A thread in April discussed whether the GSL group format was flawed or not and at the time 82% people judged that the format was fine. Since then many popular tournaments have adopted it as it is shorter, easier to run and provides more drama. As more groups were played this way we've seen more games leaving that awkward feeling when there's a rematch and the player qualified isn't the one with more wins, such as the WCS finals when Hero had qualified over Stephano despite having a record of 2-3 against him. It happens because when 2 players meet again, the result of the first bo3 is discarded. This means to qualify one player has to beat his opponent in 2 bo3 while the other one qualifies with a 1 bo3 draw.
Since the GSL format feels fine otherwise I've decided to count the games where a rematch happened. I didn't count groups in bo1 such as older Code S, neither groups where a player forfeited. Everything was counted by hand. 2012 GSTL season 2 and 3 were included.
2012 Global StarCraft II League Season 5: Code S 2012 Global StarCraft II League Season 4: Code S 2012 Global StarCraft II League Season 3: Code S 2012 Global StarCraft II League Season 2: Code S 2012 Global StarCraft II League Season 1: Code S 2011 Global StarCraft II League November: Code S 2012 Global StarCraft II Team League Season 3 2012 Global StarCraft II Team League Season 2 Numericable M-House Cup MSI Pro Cup Worldwide MSI Pro Cup Worldwide: European Qualifier MSI Pro Cup Worldwide: American Qualifier MSI Pro Cup Worldwide: Asian Qualifier 2012 Ritmix Russian StarCraft II League Season 2 2012 Battle.net World Championship ASUS ROG Summer 2012 ESET UK Masters 2012 FXOpen Invitational Series 6 Arena of Legends/The King of Kongs ESL Pro Series Poland: Season V IGAOpen Summer Challenge Stim to the Win Tournament Homestory Cup (Stage 2 not yet included) Iron Squid Chapter II (Group G not yet included)
Groups not included because tournaments were running : Homestory Cup (Stage 2) Ironsquid 2 (Group G) 2013 IGN Pro Team League Season 1
Here are the results compiled :
169 groups were counted. 85 groups had 2 players meet twice.
After the rematch : 41 times the qualified player won both bo3. 44 times the loser of the 1st bo3 has qualified by winning the 2nd bo3.
Regarding games where a player qualifed by losing the first bo3 and winning the second one : 14 times the qualified player had more wins. (3-2) 20 times both players had the same number of wins. (2-0 each or 2-1 each) 10 times the qualified player had less wins. (2-3)
First observation : when you rematch players, the loser of the first bo3 has 52% chances of qualifying. It's close to even.
Second observation : when nobody won both bo3, 68% of time the GSL format is wrong about qualifying the concerned player. Either because both players were tied or because the qualified player had less wins.
Which gives us, in case of a rematch : 41+14 = 55 players qualified for good reasons (winning both bo3 or having more wins) 20+10 = 30 players qualified for bad reasons (having the same number of wins of less)
The conclusion is in the case of a rematch the GSL format gives wrong results 35% of times.
If you want explanations about why it isn't fair to discard the first bo3 you can look at the April discussion.
First, contrary to popular belief, a GSL group isn't an exact double elimation. In a double elimination the loser bracket is reversed so 2 players can't meet in the next 2 games, while it can happen in a GSL group.
Then, here's the "unfair" finished GSL style group : A 2-0 B 2-1 C 1-2 D 0-2
C won the first bo3 against B, but because of this B had an easier group than C.
Basically you give an advantage to the loser B by matching him against D, the group loser, while the winner C of the 1st bo3 gets to play A, the group winner. But it isn't actually fair either for A who had to play C and D the last 2 players of the group while C had to play A and B the first 2 of the group.
If you compare a regular round robin group with a GSL style group, in a regular round robin C wins the tie breaker against B in case of a tie, while in a GSL group format you give the loser B several advantages because he gets an easier 2nd opponent and gets the chance to redeem himself. When there's a rematch, it can be seen as a tie breaker, except one player qualifies by getting a draw 1 bo3 to 1 bo3, while the other player has to win 2 bo3.
How to fix this : First, the GSL group format is perfectly fine until two players have to meet again. This is the only case where the format should be adjusted.
-I believe the best way to fix the format is instead of rematching you play the missing games. B and C don't rematch but instead A plays B and C plays D. You end the GSL group like a regular round robin. The advantage is you count the actual results and don't make assumptions about who is better than whom. The downside are it takes more time to organize, it doesn't prevent from 3 way ties and since D is already eliminated he may not be playing at his full potential and he can influence whether his opponent is qualified or not.
-Very soft extended series : you rematch B and C in a bo5, where you give the winner of the 1st bo3 1 point to start.
-MLG style soft extended series : you rematch B and C in a bo7 while starting from the result of the first bo3.
-Double elimination style hard extended series : instead of thinking of games on their own you think of a game being an indivisible bo3. You rematch B and C in a bo3. If C wins he's qualified (he won 2 bo3). If B wins (1 bo3 to 1 bo3) you make a new bo3 where the winner is qualified.
I really dont care if the player is 2-3 and against the other guy and advances. They aren't playing a bo5. Its two 2 bo3s. I think the GSL system is by far the best out there. I'm glad more and more tournaments are using it.
1) the player who wins is better, not the better player wins 2) are you implying that losing the first game puts you at an advantage? 3) its a double elimination bracket. think about it like that if it helps you sleep better 4) its a group. not a 1v1 series. 5) isnt your solution to just play a normal group stage?
On December 21 2012 11:28 jmbthirteen wrote: I really dont care if the player is 2-3 and against the other guy and advances. They aren't playing a bo5. Its two 2 bo3s. I think the GSL system is by far the best out there. I'm glad more and more tournaments are using it.
Ya as much effort as you put in to this post... I still like the GSL format the best.
Yes the GSL style groups have a lower chance of getting the best players out of a group. That's obvious because the more games are played the higher chance you'll get the right players out and GSL style groups have the lowest amount of games possible for a 4 person group. It doesn't really matter as much as your arguing it does though. Did you ever think that the reason players win the first bo3 and then lose the second is because they are gimmicky players? They get figured out through the first bo3 so then they can't win the second.
So what you're arguing is for gsl to adopt extended series to fix this?
There are pros and cons to the system. Cons being the one you mentioned. Pros being 0 meaningless games, no need for tie breakers, map score, etc.
However, personally I do like seeing a player play everyone in their group, but gsl format is the most efficient format, so I don't really mind it either.
GSL format is the best. I hate round robin with meaningless games and players throwing or not trying in their final game. Even the cons mentioned are not cons in my book. The bottom line is the player that wins 2 matches advances.
Interesting post, but what you see as problems I don't see as problems.
Basically as I see it, the 3rd BO3 you play (rematch situation) is a much more important series than the initial one, and as such winning it holds much more value, so regardless of any results/map score etc from earlier series, it doesn't matter, because with any tournament format you need to win when it counts, and the benefits for both broadcast/smooth tournament flow (no ties) GSL format shits on round robin.
Winners vs Winners and Losers vs Losers (swiss system) is my favorite format of all time for anything. It does a great job of establishing a "pecking order" with minimum games played.
On December 21 2012 11:20 chuky500 wrote: A thread in April discussed whether the GSL group format was flawed or not and at the time 82% people judged that the format was fine. Since then many popular tournaments have adopted it as it is shorter, easier to run and provides more drama. As more groups were played this way we've seen more games leaving that awkward feeling when there's a rematch and the player qualified isn't the one with more wins, such as the WCS finals when Hero had qualified over Stephano despite having a record of 2-3 against him. It happens because when 2 players meet again, the result of the first bo3 is discarded. This means to qualify one player has to beat his opponent in 2 bo3 while the other one qualifies with a 1 bo3 draw.
Since the GSL format feels fine otherwise I've decided to count the games where a rematch happened. I didn't count groups in bo1 such as older Code S, neither groups where a player forfeited. Everything was counted by hand. 2012 GSTL season 2 and 3 were included.
2012 Global StarCraft II League Season 5: Code S 2012 Global StarCraft II League Season 4: Code S 2012 Global StarCraft II League Season 3: Code S 2012 Global StarCraft II League Season 2: Code S 2012 Global StarCraft II League Season 1: Code S 2011 Global StarCraft II League November: Code S 2012 Global StarCraft II Team League Season 3 2012 Global StarCraft II Team League Season 2 Numericable M-House Cup MSI Pro Cup Worldwide MSI Pro Cup Worldwide: European Qualifier MSI Pro Cup Worldwide: American Qualifier MSI Pro Cup Worldwide: Asian Qualifier 2012 Ritmix Russian StarCraft II League Season 2 2012 Battle.net World Championship ASUS ROG Summer 2012 ESET UK Masters 2012 FXOpen Invitational Series 6 Arena of Legends/The King of Kongs ESL Pro Series Poland: Season V IGAOpen Summer Challenge Stim to the Win Tournament Homestory Cup (Stage 2 not yet included) Iron Squid Chapter II (Group G not yet included)
Groups not included because tournaments were running : Homestory Cup (Stage 2) Ironsquid 2 (Group G) 2013 IGN Pro Team League Season 1
Here are the results compiled :
169 groups were counted. 85 groups had 2 players meet twice.
After the rematch : 41 times the qualified player won both bo3. 44 times the loser of the 1st bo3 has qualified by winning the 2nd bo3.
Regarding games where a player qualifed by losing the first bo3 and winning the second one : 14 times the qualified player had more wins. (3-2) 20 times both players had the same number of wins. (2-0 each or 2-1 each) 10 times the qualified player had less wins. (2-3)
First observation : when you rematch players, the loser of the first bo3 has 52% chances of qualifying. It's close to even.
Second observation : when nobody won both bo3, 68% of time the GSL format is wrong about qualifying the concerned player. Either because both players were tied or because the qualified player had less wins.
Which gives us, in case of a rematch : 41+14 = 55 players qualified for good reasons (winning both bo3 or having more wins) 20+10 = 30 players qualified for bad reasons (having the same number of wins of less)
The conclusion is in the case of a rematch the GSL format gives wrong results 35% of times.
If you want explanations about why it isn't fair to discard the first bo3 you can look at the April discussion.
First, contrary to popular belief, a GSL group isn't an exact double elimation. In a double elimination the loser bracket is reversed so 2 players can't meet in the next 2 games, while it can happen in a GSL group.
Then, here's the "unfair" finished GSL style group : A 2-0 B 2-1 C 1-2 D 0-2
C won the first bo3 against B, but because of this B had an easier group than C.
Basically you give an advantage to the loser B by matching him against D, the group loser, while the winner C of the 1st bo3 gets to play A, the group winner. But it isn't actually fair either for A who had to play C and D the last 2 players of the group while C had to play A and B the first 2 of the group.
If you compare a regular round robin group with a GSL style group, in a regular round robin C wins the tie breaker against B in case of a tie, while in a GSL group format you give the loser B several advantages because he gets an easier 2nd opponent and gets the chance to redeem himself. When there's a rematch, it can be seen as a tie breaker, except one player qualifies by getting a draw 1 bo3 to 1 bo3, while the other player has to win 2 bo3.
How to fix this : First, the GSL group format is perfectly fine until two players have to meet again. This is the only case where the format should be adjusted.
-I believe the best way to fix the format is instead of rematching you play the missing games. B and C don't rematch but instead A plays B and C plays D. You end the GSL group like a regular round robin. The advantage is you count the actual results and don't make assumptions about who is better than whom. The downside are it takes more time to organize, it doesn't prevent from 3 way ties and since D is already eliminated he may not be playing at his full potential and he can influence whether his opponent is qualified or not.
-Very soft extended series : you rematch B and C in a bo5, where you give the winner of the 1st bo3 1 point to start.
-MLG style soft extended series : you rematch B and C in a bo7 while starting from the result of the first bo3.
-Double elimination style hard extended series : instead of thinking of games on their own you think of a game being an indivisible bo3. You rematch B and C in a bo3. If C wins he's qualified (he won 2 bo3). If B wins (1 bo3 to 1 bo3) you make a new bo3 where the winner is qualified.
Actually, you are wrong about it not being a double elimination bracket just because the players can play each other after 1 game. Your definition of a double elimination bracket (not playing the same player until after 2 games) is incorrect. What double elim does is to TRY to put players that have play against each other farthest alway from the brackets. However, since a bracket of 4 is so small, you end up playing them right away. If brackets are bigger (32, 64, 128), players can be seperate do it take 4,5,6 series for them to meet after playing their first series..
The group stage is to determine who can get the wins that matter. If you lose to someone the second time around then too bad. You lost in a simple bo3. If you lose 2 bo3's in a normal group you are out anyway and it makes the 3rd bo3 pointless. GSL style prevents pointless matches and it definitely the best.
It helps if you think of the GSL format as its own Bo3. You have to win 2 out of 3 Bo3s to advance. Who you play in those is completely irrelevent provided you win 2 of the 3 Bo3s.
Winning the second best of 3 vs. the same person, in a strategy game, should count much more than the first, in my opinion, as you've had time to adjust and learn from your first encounter. So I don't think this is a problem.
You can't adjust the format only in the case where b v c play first then again in the final match with full round robin because it offers d a chance at redemption no other group's 4th player gets. A beats d, c beats b. A beats c, b beats d. Round robin says a now plays b and c plays d. What happens if d beats c and a beats b? Three 1-2 players in a tiebreak but if the group had d beating b (who would now be 4th) b would be getting screwed out of a shot via redemption(1st place also would also have this except he might miss via tiebreakers) . You can't just say d's game doesn't matter yo him because then you have a game where d gains nothing and c has everything to gain (imagine the controversy if they happened to be from the same team).
Sure it leaves a bad taste in your mouth to see someone get 2-0d then win 2-1 to the same person to advance over them but it provides everyone with as fair of a bracket as you can get short of some crazy format where the losers and winners games are played in g1 and g2 then the losers winner g1 plays winners loser g2 for the spot but in a league that revolves around preparation to say "prepare against these three, but you might have to pay against one of these four other people" really goes against the spirit of GSL.
Just take it from this angle: When you lose in GSL style groups, you are punished by having to play another Bo3, which is much more important then other ones.
I've often had doubts about the GSL style group format, but a full round robin system doesn't seem that much better considering the games that don't count.
On December 21 2012 12:09 SolidMoose wrote: And this is the reason why MLG had extended series. You can't be against extended series AND against this format.
Thousand times this.
Either you decide on a system that takes into account results from the past or you don't. You can't have it both ways.
The thing that's always bothered me about the GSL group stage is that something like this can easily happen (and probably has happened).
Four players: A - Clear favorite B and C - Pretty similar in skill. Will often trade games against one another D - Clear weakest link of the group
A advances by going 2-0. B advances by going 2-1. He goes 1-1 against C, but then 1-0 against D. C does not advance by going 1-2. He goes 1-1 against B, but then 0-1 against A. D does not advance by going 0-2.
I know people have different opinions on this and I do respect that, but it has not seemed fair to me that B gets to advance by beating the weakest player in the group while C gets stomped by the best player.
If you win one BO3, lose one, than lose another, you are out. It doesn't matter who the BO3's were against, the rules are simple. Don't lose two BO3's, and you advance. People need to get X player vs Y player out of their minds, and think of it as individual series, because that is what it is. In most peoples opinions(based on the poll you referenced), it is the best format. Perhaps it is not flawless, but no format is flawless.
If anything, I find it very impressive when a person manages to come back and win a bo3 after going 2-0 against someone. It shows a very good ability to analyze games and adjust within a short amount of time. Round Robin formats not only create pointless games, but almost always end up feeling very nasty for multiple players.
I think all this highlights is that starcraft skill is very evenly distributed. The fact that the losers often win the rematch means that at most players are something like 60-40 favored against each other (other than really good or really bad players). At least that's my takeaway. This also lines up with how volatile the ratings are at http://aligulac.com/
-I believe the best way to fix the format is instead of rematching you play the missing games. B and C don't rematch but instead A plays B and C plays D. You end the GSL group like a regular round robin. The advantage is you count the actual results and don't make assumptions about who is better than whom. The downside are it takes more time to organize, it doesn't prevent from 3 way ties and since D is already eliminated he may not be playing at his full potential and he can influence whether his opponent is qualified or not.
The downsides are all the reasons that you aren't playing a round-robin in the first place.
The issue at hand is that its possible for the player with the smaller head-to-head score to advance, but the format isn't about who can win more head-to-head, but who can avoid losing two series in total, regardless of facing the same opponent twice. Think of it as the quarter-finals of a tournament with a normal double-elimination bracket, but they stop before playing the final match
Personally I'm not a fan of someone winning or placing top 4 or even 8 in gsl one season. Getting knocked down to code a the next season having to battle out of code a the 3rd season and being able to play in code s again the 4th season.
On December 21 2012 12:09 SolidMoose wrote: And this is the reason why MLG had extended series. You can't be against extended series AND against this format.
Thousand times this.
Either you decide on a system that takes into account results from the past or you don't. You can't have it both ways.
You can argue for and against extended series (I know nobody's arguing yet, just saying) all you like but both sides have reasonable arguments and neither is ever really going to win on a forum argument. In the end, it depends on what the tournament organiser thinks is better.
On OP: The statistic is close to 50% so its not that big of a deal, especially considering the sample size. Another thing to consider in context is that for the first set, neither player have any immediate momentum going into the game. When they meet again, the winner of first set have just come off a loss and the loser have just won his series(more recent as it is played after winner's match but doesn't really matter other than giving the opponent more time to do whatever which some may prefer while some may not). Coming into this match, the first round loser has an edge in momentum as he just came off a win compared to the first round winner. Might be why they win slightly more often.
Anyways, while your hard work is appreciated, the statistic does not mean that much when you do not consider the skill level of players involved. Unless the higher skilled players are advancing less from this format compared to players of a lower skill level, the format does not have a problem. So long as higher skilled players advance more often, it is fine. There is always going to be some uncertainty involved but until we see alot of chumps(relative to the top players) advancing over championship contenders alot over a long period, nothing is really wrong with the format. For me personally, I don't see such things happening so I think the format is fine.
Besides, round robin format is also used in some tournaments as well. From the most recent tournament that I can recall(WCG), results weren't too pretty for the favored players.
I don't even agree that this endeavour to determine a 'better player' is worthwhile or possible, let alone justifiable to the extent that one would alter a tournament structure to falicitate it.
As i cannot get my head around that, i do not consider there to be a problem. I view every match to be a one off, with the slate wiped clean. It's nice and pure, which is partly what i like about GSL more than the 2-4 day carnival foreigner events.
I can accept extended series in the carnival tournaments as it helps ensure that players don't fly there to lose a single series, and it lets us see each player play more games. I still hate that it has an influence when you get to the finals, but it is what it is.
edit: not criticising MLG etc, i love the shit out of those tournaments, i just prefer the structured GSL format more
On December 21 2012 12:09 SolidMoose wrote: And this is the reason why MLG had extended series. You can't be against extended series AND against this format.
Thousand times this.
Either you decide on a system that takes into account results from the past or you don't. You can't have it both ways.
I'm sick of seeing threads of why GSL isn't optimal and not providing a way to make the format any better, just reiterating some form of extended series is needed.
I don't think many people have a problem with GSL's format unless one of their favorites gets knocked out.
On December 21 2012 12:09 SolidMoose wrote: And this is the reason why MLG had extended series. You can't be against extended series AND against this format.
Thousand times this.
Either you decide on a system that takes into account results from the past or you don't. You can't have it both ways.
I'm sick of seeing threads of why GSL isn't optimal and not providing a way to make the format any better, just reiterating some form of extended series is needed.
I don't think many people have a problem with GSL's format unless one of their favorites gets knocked out.
Did you read the end of the post where he made a few different suggestions?
On December 21 2012 12:40 Sikly wrote: If you win one BO3, lose one, than lose another, you are out. It doesn't matter who the BO3's were against, the rules are simple. Don't lose two BO3's, and you advance. People need to get X player vs Y player out of their minds, and think of it as individual series, because that is what it is. In most peoples opinions(based on the poll you referenced), it is the best format. Perhaps it is not flawless, but no format is flawless.
If anything, I find it very impressive when a person manages to come back and win a bo3 after going 2-0 against someone. It shows a very good ability to analyze games and adjust within a short amount of time. Round Robin formats not only create pointless games, but almost always end up feeling very nasty for multiple players.
That's wrong because this is a strategy game. Because you expose your strategies in set one and give your opponent the chance to study you and capitalize on your mistakes/fix his own in the following games.
All in all i think this is a farce. I'd rather see a real double elimination format and skip these groups alltogether. Because this is like double elimination, except unfair to some players.
On December 21 2012 12:35 Sein wrote: The thing that's always bothered me about the GSL group stage is that something like this can easily happen (and probably has happened).
Four players: A - Clear favorite B and C - Pretty similar in skill. Will often trade games against one another D - Clear weakest link of the group
A advances by going 2-0. B advances by going 2-1. He goes 1-1 against C, but then 1-0 against D. C does not advance by going 1-2. He goes 1-1 against B, but then 0-1 against A. D does not advance by going 0-2.
I know people have different opinions on this and I do respect that, but it has not seemed fair to me that B gets to advance by beating the weakest player in the group while C gets stomped by the best player.
In theory, this seems to be unfair. But we have to remember that the best player get upset and you can lose to the weakest player. So by winning the first set, even if you are facing the strongest player, you still have say 20% to advance straight away. And the losing player will have a 20% of losing outright (even against the weakest player). These 2 add up to a HUGE advantage for winning the first set and provides more than enough incentive.
Your scenario only seems unfair if the strongest player has 100% win rate and the weakest have 0%.
On December 21 2012 12:58 Goibon wrote: I don't even agree that this endeavour to determine a 'better player' is worthwhile or possible, let alone justifiable to the extent that one would alter a tournament structure to falicitate it.
As i cannot get my head around that, i do not consider there to be a problem. I view every match to be a one off, with the slate wiped clean. It's nice and pure, which is partly what i like about GSL more than the 2-4 day carnival foreigner events.
I can accept extended series in the carnival tournaments as it helps ensure that players don't fly there to lose a single series, and it lets us see each player play more games. I still hate that it has an influence when you get to the finals, but it is what it is.
edit: not criticising MLG etc, i love the shit out of those tournaments, i just prefer the structured GSL format more
That's an argument for running a double-elimination tournament (instead of single-elimination), not necessarily for extended series. How (or if) you use match history in rematches or create a Winners Bracket advantage in the finals of a double-elimination tournament is another issue. Also, MLG now sets up their Championship Brackets such that the final is always a Best of 7 starting at 0-0 with players who haven't faced each other earlier in that Bracket. (The extended series and WB/LB stuff that used to happen in the Final now happens in the Semifinals.)
On December 21 2012 12:40 Sikly wrote: If you win one BO3, lose one, than lose another, you are out. It doesn't matter who the BO3's were against, the rules are simple. Don't lose two BO3's, and you advance. People need to get X player vs Y player out of their minds, and think of it as individual series, because that is what it is. In most peoples opinions(based on the poll you referenced), it is the best format. Perhaps it is not flawless, but no format is flawless.
If anything, I find it very impressive when a person manages to come back and win a bo3 after going 2-0 against someone. It shows a very good ability to analyze games and adjust within a short amount of time. Round Robin formats not only create pointless games, but almost always end up feeling very nasty for multiple players.
That's wrong because this is a strategy game. Because you expose your strategies in set one and give your opponent the chance to study you and capitalize on your mistakes/fix his own in the following games.
All in all i think this is a farce. I'd rather see a real double elimination format and skip these groups alltogether. Because this is like double elimination, except unfair to some players.
It's unfair to no one. Want to advance? Win 2 series.
Much better than some BS second-chance system to protect favorites or worse, extended series.
On December 21 2012 12:40 Sikly wrote: If you win one BO3, lose one, than lose another, you are out. It doesn't matter who the BO3's were against, the rules are simple. Don't lose two BO3's, and you advance. People need to get X player vs Y player out of their minds, and think of it as individual series, because that is what it is. In most peoples opinions(based on the poll you referenced), it is the best format. Perhaps it is not flawless, but no format is flawless.
If anything, I find it very impressive when a person manages to come back and win a bo3 after going 2-0 against someone. It shows a very good ability to analyze games and adjust within a short amount of time. Round Robin formats not only create pointless games, but almost always end up feeling very nasty for multiple players.
That's wrong because this is a strategy game. Because you expose your strategies in set one and give your opponent the chance to study you and capitalize on your mistakes/fix his own in the following games.
All in all i think this is a farce. I'd rather see a real double elimination format and skip these groups alltogether. Because this is like double elimination, except unfair to some players.
Its not unfair if you failed to win 2 Bo3s then you are out. Also a person playing a second Bo3 against somone of same race never mind same opponent either has a different strategy or is giong to face same problem.
Ill explain this by pretending for example I am DRG and I am in a group with Taeja MKP and Polt. I first face Polt and win 2-0 and then play MKP and lose 2-1. What will happen next is I will face either Taeja or Polt again. No matter which player it is my style for that day is either exposed or I came prepared with extra builds so its safe.
On December 21 2012 12:09 SolidMoose wrote: And this is the reason why MLG had extended series. You can't be against extended series AND against this format.
Thousand times this.
Either you decide on a system that takes into account results from the past or you don't. You can't have it both ways.
I'm sick of seeing threads of why GSL isn't optimal and not providing a way to make the format any better, just reiterating some form of extended series is needed.
I don't think many people have a problem with GSL's format unless one of their favorites gets knocked out.
For the record I don't really care about the format, I'm fine with either one. I'm just preempting anyone who is going to argue against GSL's format but also say extended series is bad.
The complaint is similar to complaining that if Germany and Brazil meet in a World Cup final, Brazil doesn't benefit from having defeated Germany in their group stage match. "Old" results at earlier stages in a tournament should ideally count less. And it is the loser's fault if he can't win the last BO3. He technically comes into that last game with an slight advantage. He has the mental upper ground because he already defeated the opponent once. He also had a rest period where he got to observe both potential opponents play (without mentally having to prepare for a different game first).
On December 21 2012 11:42 Snorkle wrote: Winners vs Winners and Losers vs Losers (swiss system) is my favorite format of all time for anything. It does a great job of establishing a "pecking order" with minimum games played.
I agree. There is never a meaningless game. The system never seems unfair to me and provides the most exciting system for spectators. I don't think its broken, so no need to fix it.
On December 21 2012 13:53 RaiKageRyu wrote: Fair formats are often the most dull formats since you start putting the players in a more scientifically correct method of determining who advances.
Contrary to popular notion, tournaments are not set up to find the best player, it is produced to entertain the spectators.
The simplest thing to do if you don't want meaningless games is just to assign a tie result and not have players play them. That'd work in other formats.
The disadvantage of round robin is you can have a situation where there is no clear distinction between players and you have an awkward tie situation.
The disadvantage here is that in a rematch situation, the first series played between two players is deemed completely pointless. Extended series became silly when it incorporated group stage results (the one stage where you can't get eliminated...) but it was nice that it took into account, that the match up had occurred before. To me, there is no reason a player could go 2-3 against another and still go ahead of him.
If you do a full round robin you have the same "problem". The top player might have a losing record against someone who's ranked lower. But that's not unfair. One player might have done better in the head-to-head matches, but the other did better in all the other matches. Those other matches are also relevant to determine who is better. (Maybe the player who edged out the head-to-head matches is only good in one matchup for example.) No one complains about it in traditional sports. The statistics in the OP ignore the fact that the advancing player won a whole other series while the non-advancing player lost one.
On December 21 2012 12:35 Sein wrote: The thing that's always bothered me about the GSL group stage is that something like this can easily happen (and probably has happened).
Four players: A - Clear favorite B and C - Pretty similar in skill. Will often trade games against one another D - Clear weakest link of the group
A advances by going 2-0. B advances by going 2-1. He goes 1-1 against C, but then 1-0 against D. C does not advance by going 1-2. He goes 1-1 against B, but then 0-1 against A. D does not advance by going 0-2.
I know people have different opinions on this and I do respect that, but it has not seemed fair to me that B gets to advance by beating the weakest player in the group while C gets stomped by the best player.
In a round-robin group with these four, the winner of the one match between B and C advances outright at 2-1, and both of them get to be stomped by A and stomp D. It sort of ends up the same: yes, in GSL format there's a possibility for B and C to play twice, and only the second match "counts", but as a player presumably you know this going in, so you can plan accordingly.
The situation the GSL-style groups avoid is one where you have three players of similar skill and a player who is either much better or much worse than all three, creating a 3-way tie which necessitates either looking at map score or running a bunch of tiebreaker matches (and in that case they all play again anyway, which is the same issue this thread is trying to solve). Even a map score tiebreaker has its problems from a spectator standpoint: if you know your map score is important, you're less likely to do a risky strategy which might be more interesting to watch.
The GSL format also avoids "wasted" matches: there is never a match involving a player who was eliminated or who is already qualified. (In a group of more than four players, it's not difficult to construct a situation where a match matters to NEITHER player, but matters to some third player.)
Statistical analysis is rather irrelevant here and is actually used to obfuscate the point. Statistics can not determine what is fair and what isn't. People do that.
The OP decides what a "good" and "bad" way to qualify is and what the "right" and "wrong" results are, and then makes a chart that says "here's how many people qualified for the wrong reasons" to reinforce his point, but there is no strong argument as to why it should be considered a "wrong" result in the first place. The way I see it, 100% of people qualified for the right reasons,.
It all comes down to the same argument of resolving double-elimination formats that has been rehashed for years. Whether you consider it to be fair or unfair or right or wrong depends entirely on how you choose to look at it, and is ultimately a matter of personal preference.
On December 21 2012 13:53 RaiKageRyu wrote: Contrary to popular notion, tournaments are not set up to find the best player, it is produced to entertain the spectators.
I bet you've found Naniwa's 7 probe rush quite entertaining.
Second observation : when nobody won both bo3, 68% of time the GSL format is wrong about qualifying the concerned player. Either because both players were tied or because the qualified player had less wins.
This is a faulty observation that uses statistics to try to make your point. You could easily say 88% (48% winner wins both + 77% of the 52% where the loser wins) of the time they get it right. When it is tied after both matches, who is the say the person that wins 2nd is less deserving? The most important game is the last game.
I actually think this is a really good point. It really does give certain advantage to lose the first match, as occasionally you will have a SUBSTANTIALLY easier group overall.
I disagree with the assessment that this group style is flawed and that it's a bad thing when you 'win' 3-2 and are eliminated. The players know in advance that the first BO3 doesn't count anymore and should play and prepare accordingly. If they can't win when it really counts, then maybe they didn't deserve to advance in the first place.
Furthermore, the suggestion from the OP is also flawed:
How to fix this : First, the GSL group format is perfectly fine until two players have to meet again. This is the only case where the format should be adjusted.
-I believe the best way to fix the format is instead of rematching you play the missing games. B and C don't rematch but instead A plays B and C plays D. You end the GSL group like a regular round robin. The advantage is you count the actual results and don't make assumptions about who is better than whom. The downside are it takes more time to organize, it doesn't prevent from 3 way ties and since D is already eliminated he may not be playing at his full potential and he can influence whether his opponent is qualified or not.
In this special case, where it's between B and C, the players A and D have already advanced or been eliminated. They have nothing at stake in the games except to influence who their opponent later in the tournament may be. So they may do another 6 probe rush or similar 'I just wanna go home' strategy. Also, playing out the rest of the round robin can lead to a three-way tie, which really sucks in a four player group, as it can take a while to resolve, which can lead to a conflict in the schedule (not in GSL with one group a day, but in other tournaments, where there are more rounds after the group round in the same or the next day).
On December 21 2012 13:53 RaiKageRyu wrote: Contrary to popular notion, tournaments are not set up to find the best player, it is produced to entertain the spectators.
I bet you've found Naniwa's 7 probe rush quite entertaining.
It was hilarious. Also, we're still talking about that legendary match.
player A won 2bo3 on the 3 he played, player B won 1, you're saying to me that player B deserves to advance? Being good at 1 match up, winning more maps against 1 players or whatever are your criteria doesn't make you a "better" player systematically. Feast is fundamentaly better than Stephano right ?
I don't know why all the fuss about it. Each group is a double elimination bracket, which is the best format there is. No meaningless games -> no throwing games. Period.
Honestly... I don't even see the point of doing group stage where 1/2 of the participants advance (and there's no 'bye' awarded for the top place). If 1/2 the players advance why not just do a standard bracket?
Off the top of my head, I can't think of any major sport where group stages are used and the winners / losers are split 50/50. (I'm sure there are some.)
However... I think the Swiss Danish-system (what GSL is doing) is pretty good (since I really loathe extended series). Unless there are pre-tournament seeding there should be no reason to ever adjust a match because of "who" an opponent was earlier in the tournament.
but its still the way best format ! everything else include "meh i am trough i dont care" or "meh i am out i dont care" games and this is the only way to prevent it 100% !
On December 21 2012 16:43 y0su wrote: Honestly... I don't even see the point of doing group stage where 1/2 of the participants advance (and there's no 'bye' awarded for the top place). If 1/2 the players advance why not just do a standard bracket?
Off the top of my head, I can't think of any major sport where group stages are used and the winners / losers are split 50/50. (I'm sure there are some.)
However... I think the Swiss Danish-system (what GSL is doing) is pretty good (since I really loathe extended series). Unless there are pre-tournament seeding there should be no reason to ever adjust a match because of "who" an opponent was earlier in the tournament.
because in a group you have a 2/4 chance to win so you can lose once and still go trough
that means 2 very strong player in the same group go trough but having stephano vs Taejja in the 1st round would be in normal mode th k.o of the looser ... or a stupid looserbracket where you have even alot of the same problems in
also it means in normal mode you see your favoritve player perhaps max 1 time, in group he have at least 2 games
so many reasons for groups ...
PS: its the fuckn same in every sport ^^ look football world championchip, you have also groups with 4 and the top2 advances (pps: i dont mean this american handegg no idea what they doing there)
On December 21 2012 16:46 CoR wrote: but its still the way best format ! everything else include "meh i am trough i dont care" or "meh i am out i dont care" games and this is the only way to prevent it 100% !
On December 21 2012 16:43 y0su wrote: Honestly... I don't even see the point of doing group stage where 1/2 of the participants advance (and there's no 'bye' awarded for the top place). If 1/2 the players advance why not just do a standard bracket?
Off the top of my head, I can't think of any major sport where group stages are used and the winners / losers are split 50/50. (I'm sure there are some.)
However... I think the Swiss Danish-system (what GSL is doing) is pretty good (since I really loathe extended series). Unless there are pre-tournament seeding there should be no reason to ever adjust a match because of "who" an opponent was earlier in the tournament.
PS: its the fuckn same in every sport ^^ look football world championchip, you have also groups with 4 and the top2 advances (pps: i dont mean this american handegg no idea what they doing there)
Pfft. Our Handegg does it right. Keeps things exciting homie. Your Football is so aggravatingly boring that you get groups of people from either side to beat the shit out of each other after every match.
The OP is very stupid and the amount of people who don't immediately see why the OP is so stupid baffles me. And OP, if you don't understand enough about mathematics to see why your post is so bad you should make less data collecting and do some thinking instead.
The winner of the first BO3 already has a huge advantage. He can lose the next BO3 and still have a chance while the loser has to win two BO3s in a row. That gives the winner of the first BO3 a 75% chance of advancing and the loser only a 25% chance. That already was the advantage and it is a huge one, so any additional advantages on top of having a 3x higher chance of advancing would be unfair. The only exception is this:
On December 21 2012 12:35 Sein wrote: The thing that's always bothered me about the GSL group stage is that something like this can easily happen (and probably has happened).
Four players: A - Clear favorite B and C - Pretty similar in skill. Will often trade games against one another D - Clear weakest link of the group
A advances by going 2-0. B advances by going 2-1. He goes 1-1 against C, but then 1-0 against D. C does not advance by going 1-2. He goes 1-1 against B, but then 0-1 against A. D does not advance by going 0-2.
I know people have different opinions on this and I do respect that, but it has not seemed fair to me that B gets to advance by beating the weakest player in the group while C gets stomped by the best player.
)
but it is only a corner case and really nothing to worry about because a) such groups are rare in general and b) those groups are even more rare in SC2 where usually everyone can lose to everyone because of BO losses / cheese. B/C will always have a shot at taking out A and they will also always have the risk to lose to D.
The system is basically flawless. You go 2-0 or 2-1 = you advance. You go 0-2 or 1-2 = you are out. Go GSL and OP you should feel bad.
so tired of this argument, seriously, it's 2 different best of 3s, it's just like double elimination brackets, it actually IS double elimination brackets. are you complaining that in the grand finals the loser bracket has to win 2 bo3s? cuz that's happened too... this is fair, you win twice, you advance, you lose twice you get eliminated. that's it.
I think the system is fair, but sort of anti-climatic...I think the winners match should happen after the losers match, instead of before. I also don't understand why GSL wouldn't consider a round-robin best of 3 format, then if needed a best of one tie-breaker; seems more exciting to me, because the winners match would happen last always....
I really like the GSL format (which actually is MSL format, and god knows where they got it from) for SC2. In general I find Round Robin systems far superior. But in SC2 tournaments, they produced too many horrible scenarios for my liking.
I think you have to win when it counts. The argument that a player with a worse map score advances half the time is not even an argument for me. Noone should be interested in map-scores, a map is like a set in tennis. You gotta win the deciding match. At the end of the day, one guy went 1-2 matches and the other 2-1.
Actually, you are wrong about it not being a double elimination bracket just because the players can play each other after 1 game. Your definition of a double elimination bracket (not playing the same player until after 2 games) is incorrect. What double elim does is to TRY to put players that have play against each other farthest alway from the brackets. However, since a bracket of 4 is so small, you end up playing them right away. If brackets are bigger (32, 64, 128), players can be seperate do it take 4,5,6 series for them to meet after playing their first series..
Don't do that, really. You could start with "OP, you are wrong about it not being [....]" instead.
gsl format in group makes alot more sense than round robin (i think its called, where everyone plays everyone)
a big reason for that is because in gsl im completely in charge qualifying or not while in normal group it always end up feeling completely random
this is speaking as a player myself in a group but also observing i like gsl format more because its easier to comprehend what results need to happen for a specific outcome while in a standard group its close to impossible to know whats going on until the very end after reading rules about map score head to head etc
On December 21 2012 11:20 chuky500 wrote: Then, here's the "unfair" finished GSL style group : A 2-0 B 2-1 C 1-2 D 0-2
C won the first bo3 against B, but because of this B had an easier group than C.
Basically you give an advantage to the loser B by matching him against D, the group loser, while the winner C of the 1st bo3 gets to play A, the group winner. But it isn't actually fair either for A who had to play C and D the last 2 players of the group while C had to play A and B the first 2 of the group.
If you compare a regular round robin group with a GSL style group, in a regular round robin C wins the tie breaker against B in case of a tie, while in a GSL group format you give the loser B several advantages because he gets an easier 2nd opponent and gets the chance to redeem himself. When there's a rematch, it can be seen as a tie breaker, except one player qualifies by getting a draw 1 bo3 to 1 bo3, while the other player has to win 2 bo3.
How to fix this : First, the GSL group format is perfectly fine until two players have to meet again. This is the only case where the format should be adjusted.
-I believe the best way to fix the format is instead of rematching you play the missing games. B and C don't rematch but instead A plays B and C plays D. You end the GSL group like a regular round robin. The advantage is you count the actual results and don't make assumptions about who is better than whom. The downside are it takes more time to organize, it doesn't prevent from 3 way ties and since D is already eliminated he may not be playing at his full potential and he can influence whether his opponent is qualified or not.
-Very soft extended series : you rematch B and C in a bo5, where you give the winner of the 1st bo3 1 point to start.
-MLG style soft extended series : you rematch B and C in a bo7 while starting from the result of the first bo3.
-Double elimination style hard extended series : instead of thinking of games on their own you think of a game being an indivisible bo3. You rematch B and C in a bo3. If C wins he's qualified (he won 2 bo3). If B wins (1 bo3 to 1 bo3) you make a new bo3 where the winner is qualified.
Bullshit! Your statistics are nearly 50-50 and you complain? The GSL format is fine, if no players meet twice, it's more or less the same as round robin, but more efficient. If players meet again, another important aspect of professional sports comes into play: bringing your best performance on point. Winning tournaments is not just about skill and it will never be. The ability to bring your best game right when it matters the most is another important factor. And I see absolutely no reason why this shouldn't be the case!
The mathematical goal of group/league play is not to finish above those you beat, it is to do better overall through playing multiple people/teams. If you win two bo3s and your opponent wins one bo3 it means you have done better. Simple math. The fact that you are 1-1 or at a losing record with him is mathematically entirely irrelevant.
Think of larger leagues which operate by the same principle you can lose to a football team twice in a single season, and finish #1 in your league while they finish #10. Nobody complains about that because mathematically the right person finishes ahead, not the one who beats the other. Group play is designed to have the right person advance mathematically. This is also why extended series don't make sense mathematically.
On December 21 2012 12:09 SolidMoose wrote: And this is the reason why MLG had extended series. You can't be against extended series AND against this format.
That's not true at all. It would take more time to type than I have right now, but the proper format is not one or the other.
As well as this:
On December 21 2012 19:40 Liquid`Nazgul wrote: The mathematical goal of group/league play is not to finish above those you beat, it is to do better overall through playing multiple people/teams. If you win two bo3s and your opponent wins one bo3 it means you have done better. Simple math. The fact that you are 1-1 or at a losing record with him is mathematically entirely irrelevant.
Think of larger leagues which operate by the same principle you can lose to a football team twice in a single season, and finish #1 in your league while they finish #10. Nobody complains about that because mathematically the right person finishes ahead, not the one who beats the other. Group play is designed to have the right person advance mathematically. This is also why extended series don't make sense.
Interesting, but I don't accept you "correct" results. They're playing 2 bo3's, not 1 bo5. This isn't MLG extended series, this is GSL group play. I would say there isn't a problem, at least until you put forward a solution that is actually a better option. The way it looks, only MLG extended series would be the solution to the "problem" you posit.
I think the main problem of this system is that it favours zerg quite heavily. Not because of imbalances or shit but because zerg is supposed to be the more reactive race and therefore it is easier for the zerg to play a second time against a protoss or terran opponent then the other way round.
Just look at recent HSC: ret loses to Hasu, then beats him. Stephano loses to Feast then beats him. DIMAGA loses to Bling then beats him.
I think that is the real issue and I would really like to see statistics about it. The "problem" in the OP is not really present in my oppinion.
I don't really see the "good reason" or the "bad reason" you are talking about... You are basically approaching the model with a different metric it was designed upon and then proclaim that the model is bad in that other metric.
On December 21 2012 17:04 Fenrax wrote: The OP is very stupid and the amount of people who don't immediately see why the OP is so stupid baffles me. And OP, if you don't understand enough about mathematics to see why your post is so bad you should make less data collecting and do some thinking instead.
The winner of the first BO3 already has a huge advantage. He can lose the next BO3 and still have a chance while the loser has to win two BO3s in a row. That gives the winner of the first BO3 a 75% chance of advancing and the loser only a 25% chance. That already was the advantage and it is a huge one, so any additional advantages on top of having a 3x higher chance of advancing would be unfair. The only exception is this:
On December 21 2012 12:35 Sein wrote: The thing that's always bothered me about the GSL group stage is that something like this can easily happen (and probably has happened).
Four players: A - Clear favorite B and C - Pretty similar in skill. Will often trade games against one another D - Clear weakest link of the group
A advances by going 2-0. B advances by going 2-1. He goes 1-1 against C, but then 1-0 against D. C does not advance by going 1-2. He goes 1-1 against B, but then 0-1 against A. D does not advance by going 0-2.
I know people have different opinions on this and I do respect that, but it has not seemed fair to me that B gets to advance by beating the weakest player in the group while C gets stomped by the best player.
)
but it is only a corner case and really nothing to worry about because a) such groups are rare in general and b) those groups are even more rare in SC2 where usually everyone can lose to everyone because of BO losses / cheese. B/C will always have a shot at taking out A and they will also always have the risk to lose to D.
The system is basically flawless. You go 2-0 or 2-1 = you advance. You go 0-2 or 1-2 = you are out. Go GSL and OP you should feel bad.
I agree that even the corner case isn't really a flaw, since both players (B and C) know beforehand that their second match has a higher value, so they can prepare accordingly. It still is worth it to win the first match (so they get at least shot of trying to beat A, as you said), but they should save their best builds for last. This is assuming all players know their expected chances before the group. Even if they don't, it still doesn't really matter. It's not the end of the world to have a second match have more value than the first, as long as the first isn't completely worthless.
The flaw with OP's logic and MLG extended series logic is that they assume that B or C would always beat D, but it's not the case in a game with variance like starcraft. This makes extended series a horribly unfair system in addition to an awful viewer experience. In GSL system, it's fair game for B and C when that second match is about to start.
GSL system is close to perfect and clearly the best system I've seen in any game or sport, given the constraints.
GSL means no pointless matches, so players are always trying to win. It means in order to progress you have to win two Bo3s, it doesn't matter who against, and if you play the same guy twice; beating him means you figured him out and/or correct your mistakes from the first series. So gz to you!
I don't understand the problem. What does it matter if a player does not win against a specific player both times? There will always be two players who wins two matches, and two that does not. Why shouldn't it be fair that the players who win two matches will go through?
When the same players meet again, and the same win I feel cheated. This player had only one player of the group to beat in order to advance.
However when they meet again and the other win, it means that these two players have approx the same level. However, the one coming from the looser match has won against someone else in the group whereas the one coming from the winner match has not.
Well you can't just put everything on the rematch. Think about since the first match one player managed to lose a bo3, while the other won o ne, so they both earned their way back to the rematch. No reason to call that "unfair".
On December 21 2012 19:40 Liquid`Nazgul wrote: The mathematical goal of group/league play is not to finish above those you beat, it is to do better overall through playing multiple people/teams. If you win two bo3s and your opponent wins one bo3 it means you have done better. Simple math. The fact that you are 1-1 or at a losing record with him is mathematically entirely irrelevant.
Think of larger leagues which operate by the same principle you can lose to a football team twice in a single season, and finish #1 in your league while they finish #10. Nobody complains about that because mathematically the right person finishes ahead, not the one who beats the other. Group play is designed to have the right person advance mathematically. This is also why extended series don't make sense mathematically.
But what are you arguing? In the GSL-format you actually finish in front of the people you beat, and behind the people you lose to (and in front/behind people you don't play).
As for the general discussion:
The problems with the GSL-system are:
1. Since race match-ups have a huge impact on the game, getting a lucky draw for the order of match-ups can have a big impact on the results. By meeting only two of three players in the group, you create a more random system than round robin.
2. It is possible to beat your opponent for second place in maps won, but still not advance. Outside of hand-waving about being "clutch" (which is what rolling a dice can look like from a first-person perspective), looking at it from a matter of probability, it is a problematic system that rates bo3 > bo5 (0 - 2; 2 - 1).
The GSL-system is a more random system than the round robin system. There is no way around it.
It is down to a matter of taste whether the lack of incentives in some situations in the RR-format is enough to prefer the GSL system. For my taste it is a no. I think you can work around the problems of incentives by having teammates always meet in the first match (which is uncontroversial in a RR format, but problematic in a GSL-system) and by incentivizing 3rd and 4th place (you win money/ranking).
I think having the better system for determining the best players should take precedence.
One of the problems with the GSL system is the momentum going into the rematch. A vs B + C vs D -> A wins C wins, So now its : A vs C+ B vs D-> A loses B wins-> Momentum swings into b's favor because he just won his last series where A just came of a loss.
My take on all this: GSL format would be fine if they would make the momentum swing less impactfull.
Even if player A advances with a 2-3 record against player B, he will still have a winning record against player C while player B has a losing record against player D.
I like the format, and I think it's the best one atm. All formats have flaws, but how the pariings stack up etc is much bigger than just this imo.
On December 21 2012 19:53 o)_Saurus wrote: I think the main problem of this system is that it favours zerg quite heavily. Not because of imbalances or shit but because zerg is supposed to be the more reactive race and therefore it is easier for the zerg to play a second time against a protoss or terran opponent then the other way round.
Just look at recent HSC: ret loses to Hasu, then beats him. Stephano loses to Feast then beats him. DIMAGA loses to Bling then beats him.
I think that is the real issue and I would really like to see statistics about it. The "problem" in the OP is not really present in my oppinion.
Don't think this is true at all. While they might be the reactive race, if zerg players would blindly react to what they've seen in earlier games, they would get abused like crazy.
If you look at the results you posted, I think people are more surprised at the fact that those protoss players beat the zerg players in the first place. Stephano, DIMA and ret are generally regarded as stronger.
On December 21 2012 11:31 courtpanda wrote: 1) the player who wins is better
You would have a very, very hard time convincing anyone that this is the case in SC2.
Actually the winner picked the better build and won the mindgames, so he is better, the looser might be a god in practice, but when he got "outthinked", then it´s legit to say the winning player is better at all the SC2 sorrounding aspects
The GSL system groups are better to me for two big reasosn: -When the only possible scores are 2-0 and 2-1 it means its way too likely to have awkward 3-way ties in round robin groups which either results in Bo1 tie breakers or some guy going home because he lost one extra map which I doubt the players like. Its also not fun for the viewer to constanly be presented with the wrong/contradictory information about who is going through depending on what results. -Considering lots of players play on same team or are just friends with each other plus that once you're out you only have yourself to play for (or not play for) it means you create incentive to sometimes not do your best in round robin groups.
On December 21 2012 19:40 Liquid`Nazgul wrote: The mathematical goal of group/league play is not to finish above those you beat, it is to do better overall through playing multiple people/teams. If you win two bo3s and your opponent wins one bo3 it means you have done better. Simple math. The fact that you are 1-1 or at a losing record with him is mathematically entirely irrelevant.
Think of larger leagues which operate by the same principle you can lose to a football team twice in a single season, and finish #1 in your league while they finish #10. Nobody complains about that because mathematically the right person finishes ahead, not the one who beats the other. Group play is designed to have the right person advance mathematically. This is also why extended series don't make sense mathematically.
But what are you arguing? In the GSL-format you actually finish in front of the people you beat, and behind the people you lose to (and in front/behind people you don't play).
As for the general discussion:
The problems with the GSL-system are:
1. Since race match-ups have a huge impact on the game, getting a lucky draw for the order of match-ups can have a big impact on the results. By meeting only two of three players in the group, you create a more random system than round robin.
2. It is possible to beat your opponent for second place in maps won, but still not advance. Outside of hand-waving about being "clutch" (which is what rolling a dice can look like from a first-person perspective), looking at it from a matter of probability, it is a problematic system that rates bo3 > bo5 (0 - 2; 2 - 1).
The GSL-system is a more random system than the round robin system. There is no way around it.
It is down to a matter of taste whether the lack of incentives in some situations in the RR-format is enough to prefer the GSL system. For my taste it is a no. I think you can work around the problems of incentives by having teammates always meet in the first match (which is uncontroversial in a RR format, but problematic in a GSL-system) and by incentivizing 3rd and 4th place (you win money/ranking).
I think having the better system for determining the best players should take precedence.
The GSL-system is not more random than Round Robin. In fact, every Bo3 in the GSL-system has meaning whereas Round Robin will have meaningless games when top players have already guaranteed advancement. Why did you think the Naniwa-probe rush incident happened in the first place?
You're putting too much stock on who is playing who and that's entirely irrelevant. The fact of the matter is with the GSL format, you need to win two Bo3s. Plain and simple. Doesn't matter against who or what race.
I've noticed the same as OP, but I've never thought the system particularly good or bad in the first place, just average at best. I don't think of it as unfair if someone goes through with a 2-3 record against the same player (since it's groupbased/double elim, and you know what you're getting into to begin with), but I can understand that the GSL system is perhaps more harsh on players who rely more on preperation.
If you 'blew' your two prepared builds on the wins and then get the same player again later and lose 1-2, it's probably extra frustrating/doesn't feel right. Every tournament system has it's pros and cons.
since D is already eliminated he may not be playing at his full potential and he can influence whether his opponent is qualified or not
This is the main reason why GSL-style is, without argument, better than round-robin. There's been so many drama around this thing, both in e-sports and traditional sports, and GSL system provides a very good workaround.
Also, don't write "MLG" or "extended series" in a thread that's supposed to be serious
On December 21 2012 19:40 Liquid`Nazgul wrote: The mathematical goal of group/league play is not to finish above those you beat, it is to do better overall through playing multiple people/teams. If you win two bo3s and your opponent wins one bo3 it means you have done better. Simple math. The fact that you are 1-1 or at a losing record with him is mathematically entirely irrelevant.
Think of larger leagues which operate by the same principle you can lose to a football team twice in a single season, and finish #1 in your league while they finish #10. Nobody complains about that because mathematically the right person finishes ahead, not the one who beats the other. Group play is designed to have the right person advance mathematically. This is also why extended series don't make sense mathematically.
But what are you arguing? In the GSL-format you actually finish in front of the people you beat, and behind the people you lose to (and in front/behind people you don't play).
As for the general discussion:
The problems with the GSL-system are:
1. Since race match-ups have a huge impact on the game, getting a lucky draw for the order of match-ups can have a big impact on the results. By meeting only two of three players in the group, you create a more random system than round robin.
2. It is possible to beat your opponent for second place in maps won, but still not advance. Outside of hand-waving about being "clutch" (which is what rolling a dice can look like from a first-person perspective), looking at it from a matter of probability, it is a problematic system that rates bo3 > bo5 (0 - 2; 2 - 1).
The GSL-system is a more random system than the round robin system. There is no way around it.
It is down to a matter of taste whether the lack of incentives in some situations in the RR-format is enough to prefer the GSL system. For my taste it is a no. I think you can work around the problems of incentives by having teammates always meet in the first match (which is uncontroversial in a RR format, but problematic in a GSL-system) and by incentivizing 3rd and 4th place (you win money/ranking).
I think having the better system for determining the best players should take precedence.
The GSL-system is not more random than Round Robin. In fact, every Bo3 in the GSL-system has meaning whereas Round Robin will have meaningless games when top players have already guaranteed advancement. Why did you think the Naniwa-probe rush incident happened in the first place?
You're putting too much stock on who is playing who and that's entirely irrelevant. The fact of the matter is with the GSL format, you need to win two Bo3s. Plain and simple. Doesn't matter against who or what race.
Not quite sure how to respond. You either didn't read or didn't understand.
It all comes down to the question: what is more important: the good entertainment for the viewers or the perfect fairness for the players. And because the viewers pay the GSL's bills, they chose the entertainment with no meaningless games.
On December 21 2012 19:53 o)_Saurus wrote: I think the main problem of this system is that it favours zerg quite heavily. Not because of imbalances or shit but because zerg is supposed to be the more reactive race and therefore it is easier for the zerg to play a second time against a protoss or terran opponent then the other way round.
Just look at recent HSC: ret loses to Hasu, then beats him. Stephano loses to Feast then beats him. DIMAGA loses to Bling then beats him.
I think that is the real issue and I would really like to see statistics about it. The "problem" in the OP is not really present in my oppinion.
Don't think this is true at all. While they might be the reactive race, if zerg players would blindly react to what they've seen in earlier games, they would get abused like crazy.
If you look at the results you posted, I think people are more surprised at the fact that those protoss players beat the zerg players in the first place. Stephano, DIMA and ret are generally regarded as stronger.
It's not about blindly reacting but about getting knowledge of your opponents behaviour. On a "new" or uncommon map it is for example not always clear where your P opponent places his forge or stargate/twilight etc.. So it might happen that your 1 or 2 suicide overlords don't see anything. After that you watch the replay and do it better in the 2nd set. Same for attack ways or pylon placements.
Of course you can argue if it is good or bad that you get a second chance to win a match you only lost before because you had bad scouting/luck with your scouting. But generally I think there are more advantages given for a zerg to play the same player again thanfor protoss or terran against zerg.
and btw i dont think dimagas zvp is better than blings pvz and i also think hasu and ret are on equal footsteps at the moment but ofc these are just my personal oppinions
On December 21 2012 21:19 graNite wrote: It all comes down to the question: what is more important: the good entertainment for the viewers or the perfect fairness for the players. And because the viewers pay the GSL's bills, they chose the entertainment with no meaningless games.
There's no perfect fairness. With meaningless games, you can also have unfair situations when you depend on the victory of a player who is already eliminated to settle a tie situation with another player.
I don't think the GSL group stage is unfair if you account for the different weight of each matches. You have to remember that there is not necessarily one bad player who loses 2 matches and in recent seasons, you often have 4 very good players in one group. When you lose your first match, you're not even certain that your opponent is going to be easier since neither you or your previous opponent played against him. So it is not a good strategy to lose on purpose your first match. On the other hand, the player who went through the "loser bracket" has to play one more elimination match which is harder to play mentally. The player from the "winner bracket" has had a longer time to rest, has the first win psychological ascendant on his opponent and had no risk of getting out the tournament until now. If you consider these advantages, it is easier to see a player move on with a 2-3 score.
Finally, I think that if there is actually an unfair aspect about GSL group stages, it's the order in which you play. It changes the whole dynamic of the group, allowing you for example to avoid to play the best player and keep your confidence high. It is important, especially for the player who tends to choke. To settle this discussion, GSL has a seeding rule who act differently for each round of group stages. First you place the highest seeded players against the lowest. It gives a harder time to new opponents but it seems to work fine. And during the second round, you have the players picking their opponents one by one. I think it is a fine way to settle those problems, a little bit advantageous for players that have been around for a longer time but it always worked fine for me and create a lot of great head to head.
I feel like GSL format is the best format overall. I see your points but in the end, the person who wins the first Bo3 out of the 2 similarly skilled people still gets a shot at playing a Bo3 for direct qualification. Even if you presume that the group loser is a free win and the group winner is a clear favorite then the winner of the first Bo3 still has a chance, even if it's 30% or less, for direct qualification, which is still a bonus. Not a perfect format, but the best we've got. Extended series would make this a mess I feel.
I cant believe theres still people who argues against double elimination format which has been used for more than a decade (probably way more than that) in esports and real sports and TV shows.
Yeah in case ppl doesnt know, GOM has been using the most basic double elimination style. Yup this so-called GSL style is D.O.U.B.L.E E.L.I.M.I.N.A.T.I.O.N style. GOM is probably the 395946812th organizer who adopts this. God damn it
On December 21 2012 21:53 Arceus wrote: I cant believe theres still people who argues against double elimination format which has been used for more than a decade (probably way more than that) in esports and real sports and TV shows.
Yeah in case ppl doesnt know, GOM has been using the most basic double elimination style. Yup this so-called GSL style is D.O.U.B.L.E E.L.I.M.I.N.A.T.I.O.N style. GOM is probably the 395946812th organizer who adopts this. God damn it
It's because how liquipedia format GSL group stage some people don't realize it's a double elimination bracket with 4 players, where both upper bracket winner and lower bracket winner advance.
It's good to see the pros and cons of every system. The gsl format of course has its cons but for me, from a spectator point of view I much prefer it. It's short and sweet and has direct consequences from winning/losing a series instead of it resting on map difference, which can be an anticlimactic finish. The gsl format is much much better for viewing.
This is a terribly stupid post. The entire premise that a player who has better head to head results should be ranked higher or qualify is stupid and any conclusion coming from that is silly.
In any competition with more than 2 players you frequently encounter situations where A is worse than B head to head but A is ranked higher because of better results against the rest. This is normal and not a problem of any system whatsoever.
The player in the GSL who advances is 2-1 the loser 1-2, completely fair.. You could make a point that in a hypothetical group where player A wins 100% and player D loses 100% the first match between player B and C is pointless, since they are garanteed to meet in the last match anyway. This situation is completely hypothetical though and never happens, in reality the winner of the first bo3 had a chance to go 2-0 while the loser had a chance to go 0-2.
Head to head results should not be made more important and at most used as tiebreakers in any system. Carrying through results like MLG does is completely rediculous and changes a fair situation into an unfair one, the loser's bracket becomes unequal as it matters who you lost against which it shouldn't really.
Someone else put it well by the way. Tournaments are not to determine who the best player is in a fair way, they are first and foremost to create a cool event which spectators want to watch. If you want to determine fair rankings of players you would never go for a knockout type format or have any groups at all, you would just play a swiss style competition. Knockout style however is by far the most interesting in a game like sc2 where draws can't occur (knockout is not so great in games like football where that does happen frequently). Round robin is just a terrible format for small groups. The only advantage is that all matches are known beforehand which makes it easier to schedule and advertise matches, which is pretty important for sports like football. In team sports players don't tend to throw games either as you are always fighting for your spot on the team. In sc2 however scheduling is not a problem at all if you play all matches on the same day and place anyway, plus match throwing or not trying your best is a serious problem with countless matches to prove so. GSL format is the best I think, double elimination with groups so later on players are on equal footing unlike events like MLG where the finals is often not as exciting because one player starts with an advantage.
The GSL group format (burrowed from MSL if I remember correctly) has more advantages than disadvantages: It has a fixed lenght (no more ties than the one which is created for the fifth match.) It also has only games which are meaningful for both players.
While a regame in the last match can result in a situation where a player has overall less wins versus the other and still gets out of the group, it is still fair as the overall performance of both players brought them there.
On December 21 2012 11:20 chuky500 wrote: Which gives us, in case of a rematch : 41+14 = 55 players qualified for good reasons (winning both bo3 or having more wins) 20+10 = 30 players qualified for bad reasons (having the same number of wins of less)
The conclusion is in the case of a rematch the GSL format gives wrong results 35% of times.
You miss one VERY important point right there: The Player who advance from the group won TWO Best-of-3s. The Player who didn't advance won only ONE Best-of-3.
Despite the map score between them, the one with more WINS advance. That is not a BAD REASON.
IMHO: GSL format is not perfect, but its the best one there is for group play.
EDIT: BTW in normal round robin lets say we have players A, B, C, D. Here's the scores: A 2:1 B C 2:1 D A 0:2 C B 2:0 D A 0:2 D B 2:0 C
Final Standing: B: 5:2 C: 4:3 D: 3:4 A: 2:5
And in this situation will You say that B advances over A for bad reason, because A won vs B??
To those saying the GSL format is fair because if you get 2 wins you're qualified if you get 2 loss you're eliminated, I thought of another way of running the group. It would go like this : Winner plays Other Loser, and Other Winner plays Loser. Every game counts. Players have to win 2 games. No rematch. This way the Winner gets the advantage of playing the potential last of the group, while in a GSL group the Loser of the first match is getting this advantage.
Depending on the results you have 3 possibilities : If both Winners win, they're qualified and the group stage is over. If exactly 1 player won both games, the last game will be 1st Winner vs 2nd Winner, the winner of this game qualifies and -the group stage is over. If neither 1st winners wins again, you then play Winner vs Winner and Loser vs Loser.
Such a group can be shorter than a GSL group, provide the drama and as precise as round robin to determine the winner.
Also I agree with JJH777 and o)_Saurus when they say the gimmicky player usually loses the 2nd bo3 and that it's most often a zerg. I believe it's even truer for PvZ/ZvP as it's a more gimmicky matchup where both players usually win by hiding buildings or running by. I actually started to count the race of players but then I realised I'd have to weight in the amount of players per race and since I was doing that by hand it would take too long, maybe look irrelevant due to the small sample size and probably look like trolling.
edit : actually the group idea is wrong as if exactly 1 player wins both 2 games, you have a rematch
On December 21 2012 23:24 Ch3rry wrote: You miss one VERY important point right there: The Player who advance from the group won TWO Best-of-3s. The Player who didn't advance won only ONE Best-of-3.
Despite the map score between them, the one with more WINS advance. That is not a BAD REASON.
IMHO: GSL format is not perfect, but its the best one there is for group play.
This, this and this.
For people who think the GSL group format is unfair and that round robin group play is the way to go, I have one question: What do you think is the lesser evil, a group format which might be unfair in regards with that the player who won less maps (not sets) might advance, or a group format in which an already eliminated player has to play against his friend with the fact in mind that if he loses, his friend will advance, and if he wins, his friend is eliminated as well?
In fact, here's a poll:
Poll: Which is the lesser evil:
[double elimination] Player with fewer maps won advances (6)
86%
[round robin] Player plays a friend; if he wins friend is eliminated, if he loses friend advances (1)
14%
7 total votes
Your vote: Which is the lesser evil:
(Vote): [double elimination] Player with fewer maps won advances (Vote): [round robin] Player plays a friend; if he wins friend is eliminated, if he loses friend advances
I don't understand the whole pretense of this post. Why are you assuming that it's bad if someone goes 2-3 against someone and still advances? It's a group, not a head to head... if you want to avoid this you could just do single elim but as everybody knows that is even worse.
As said in the OP, the loser who won 2 bo3 had the chance to play the last of the group while the winner had to play the group winner. You can't assume the winner of the 1st bo3 wouldn't have beaten the last of the group as well.
On December 21 2012 23:29 chuky500 wrote: To those saying the GSL format is fair because if you get 2 wins you're qualified if you get 2 loss you're eliminated, I thought of another way of running the group. It would go like this : Winner plays Other Loser, and Other Winner plays Loser. Every game counts. Players have to win 2 games. No rematch. This way the Winner gets the advantage of playing the potential last of the group, while in a GSL group the Loser of the first match is getting this advantage.
Depending on the results you have 3 possibilities : If both Winners win, they're qualified and the group stage is over. If exactly 1 player won both games, the last game will be 1st Winner vs 2nd Winner, the winner of this game qualifies and -the group stage is over. If neither 1st winners wins again, you then play Winner vs Winner and Loser vs Loser.
Such a group can be shorter than a GSL group, provide the drama and as precise as round robin to determine the winner.
Also I agree with JJH777 and o)_Saurus when they say the gimmicky player usually loses the 2nd bo3 and that it's most often a zerg. I believe it's even truer for PvZ/ZvP as it's a more gimmicky matchup where both players usually win by hiding buildings or running by. I actually started to count the race of players but then I realised I'd have to weight in the amount of players per race and since I was doing that by hand it would take too long, maybe look irrelevant due to the small sample size and probably look like trolling.
Maybe you should show your new format with player, ABCD. Because I don't think it works at all.
So why don't make another poll. I don't think that the results would change. GSL format is the quickest and fairest way to clearly determine 4 placements out of 4 players.
On December 21 2012 23:40 Pokebunny wrote: I don't understand the whole pretense of this post. Why are you assuming that it's bad if someone goes 2-3 against someone and still advances? It's a group, not a head to head... if you want to avoid this you could just do single elim but as everybody knows that is even worse.
It's this, meaningless games is always worse than someone winning a best-of-3 advancing.
There was a talk about the MLG format at TL some time ago(if curious, go ahead and find a link) compared to GSL format.
Majority of votes were against the extended series format finals that is currently used by MLG and IPL. GSL group stage is simple. You just need to win two Bo3's. That's all there is to it.
Statistics you showed based on less than a hundred examples are just misleading. The method can't be aplied to such small number of occasions. I really don't like terms "a good reason to qualify" and "bad reason to qualify" you use. It just reminds me of "he did such a retarded thing, so I lost".
On December 21 2012 23:29 chuky500 wrote: To those saying the GSL format is fair because if you get 2 wins you're qualified if you get 2 loss you're eliminated, I thought of another way of running the group. It would go like this : Winner plays Other Loser, and Other Winner plays Loser. Every game counts. Players have to win 2 games. No rematch. This way the Winner gets the advantage of playing the potential last of the group, while in a GSL group the Loser of the first match is getting this advantage.
Depending on the results you have 3 possibilities : If both Winners win, they're qualified and the group stage is over. If exactly 1 player won both games, the last game will be 1st Winner vs 2nd Winner, the winner of this game qualifies and -the group stage is over. If neither 1st winners wins again, you then play Winner vs Winner and Loser vs Loser.
Such a group can be shorter than a GSL group, provide the drama and as precise as round robin to determine the winner.
Also I agree with JJH777 and o)_Saurus when they say the gimmicky player usually loses the 2nd bo3 and that it's most often a zerg. I believe it's even truer for PvZ/ZvP as it's a more gimmicky matchup where both players usually win by hiding buildings or running by. I actually started to count the race of players but then I realised I'd have to weight in the amount of players per race and since I was doing that by hand it would take too long, maybe look irrelevant due to the small sample size and probably look like trolling.
Maybe you should show your new format with player, ABCD. Because I don't think it works at all.
You're right actually the 2nd option where exactly 1 player wins both game forces a rematch :/
On December 21 2012 11:20 chuky500 wrote: Which gives us, in case of a rematch : good reasons (winning both bo3 or having more wins) bad reasons (having the same number of wins of less)
BTW in normal round robin lets say we have players A, B, C, D. Here's the scores: A 2:1 B C 2:1 D A 0:2 C B 2:0 D A 0:2 D B 2:0 C
Final Standing: B: 5:2 C: 4:3 D: 3:4 A: 2:5
And in this situation will You say that B advances over A for bad reason, because A won vs B??
No, he will say it is good that B advanced, since B has more wins. (B is 2-1 in Bo3s, A is 1-2 in Bo3s)
On December 21 2012 23:45 TigerKarl wrote: So why don't make another poll. I don't think that the results would change. GSL format is the quickest and fairest way to clearly determine 4 placements out of 4 players.
Well, according to the OP the fairest way: a) the 5th game is not played when the same player meets again (first match winner advances) (doesn't sound fair to me); b) the 5th match when the same player meets again is Bo5 starting from first match score (maybe more fair than a) but player who lost 0:2 in first game has very slim chances of advancing).
On December 21 2012 11:20 chuky500 wrote: Which gives us, in case of a rematch : good reasons (winning both bo3 or having more wins) bad reasons (having the same number of wins of less)
BTW in normal round robin lets say we have players A, B, C, D. Here's the scores: A 2:1 B C 2:1 D A 0:2 C B 2:0 D A 0:2 D B 2:0 C
Final Standing: B: 5:2 C: 4:3 D: 3:4 A: 2:5
And in this situation will You say that B advances over A for bad reason, because A won vs B??
No, he will say it is good that B advanced, since B has more wins. (B is 2-1 in Bo3s, A is 1-2 in Bo3s)
And that is exactly what happens in GSL format, which OP has problems with.
On December 21 2012 11:20 chuky500 wrote: First, contrary to popular belief, a GSL group isn't an exact double elimation. In a double elimination the loser bracket is reversed so 2 players can't meet in the next 2 games, while it can happen in a GSL group.
A > B C > D
A>C (winners final) D>B
D>C (consolidation final)
I don't really get how it's different from a double elim bracket without a grand final because 2 people qualify? I understand how the loser bracket side is flipped every other round, but that wouldn't happen with 4 players regardless? I'm at work so it's quite possible I'm missing something obvious.
On December 21 2012 11:20 chuky500 wrote: Which gives us, in case of a rematch : good reasons (winning both bo3 or having more wins) bad reasons (having the same number of wins of less)
BTW in normal round robin lets say we have players A, B, C, D. Here's the scores: A 2:1 B C 2:1 D A 0:2 C B 2:0 D A 0:2 D B 2:0 C
Final Standing: B: 5:2 C: 4:3 D: 3:4 A: 2:5
And in this situation will You say that B advances over A for bad reason, because A won vs B??
No, he will say it is good that B advanced, since B has more wins. (B is 2-1 in Bo3s, A is 1-2 in Bo3s)
And that is exactly what happens in GSL format, which OP has problems with.
EDIT: sry for double post
OP only has a problem with the specific scenario where there is a rematch in the final game of GSL format, and the two final players head-to-head scores go the other way from who advances.
In round robin, that will never happen. (ie: a tie in map score and a worse head to head score should always lose the tiebreaker rules)
If OP is against player with better head to head score advance, does that mean in a round robin format, players who do not advance can only win against other players who don't advance??!
Really, head to head score means very little besides its effect on overall score in a group.
I guess OP just want a giant single elimination bracket for GSL
On December 21 2012 21:53 Arceus wrote: I cant believe theres still people who argues against double elimination format which has been used for more than a decade (probably way more than that) in esports and real sports and TV shows.
Yeah in case ppl doesnt know, GOM has been using the most basic double elimination style. Yup this so-called GSL style is D.O.U.B.L.E E.L.I.M.I.N.A.T.I.O.N style. GOM is probably the 395946812th organizer who adopts this. God damn it
It's because how liquipedia format GSL group stage some people don't realize it's a double elimination bracket with 4 players, where both upper bracket winner and lower bracket winner advance.
I thought people actually watches the GSL, instead of looking at liquidpedia, no?
The statistical analysis is flawed in the opening post. Sometimes the worse player wins the first BO3. Say this happens 30% of the time. We are more likely to see a rematch when the result of the first match was "wrong" skillwise.
If the better player won the first game he has a 44% chance of losing the winners match*. If the worse one won he'll lose the winners match 64% of the time*.
If the worse player lost match one he'll win match two 44%. If the better player lost match one he'll win match two 64% of the time.
So when the better player won the first match we'll have a rematch 0.44*0.44=0.1936 or 19% of the time.
When the worse player won we'll have a rematch 0.64*0.64=0.4096 or 41% of the time.
Rescaling for the fact that the better player wins match one 70% we get the result
0.70*0.1936 = 0.13552 (52.24% of total) 0.30*0.4096 = 0.12288 (47.76% of total)
What is the interpretation of these numbers? Under our simplified model, when 2 players meet in the final match the winner of the first match is actually the worse player 48% of the time. So we should expect the loser of the first match to win very often, exactly because he is the better player very often.
* To understand where these numbers come from we need to look at the results of the other bracket. I made the same assumption there: the better player wins 70% of the time. If 2 of the better players play in match two both have 50% to advance. Basically there are two types of players in this model: "better" and "worse". Players of the same type score 50% against each other and better beats worse 70% of the time.
On December 21 2012 21:53 Arceus wrote: I cant believe theres still people who argues against double elimination format which has been used for more than a decade (probably way more than that) in esports and real sports and TV shows.
Yeah in case ppl doesnt know, GOM has been using the most basic double elimination style. Yup this so-called GSL style is D.O.U.B.L.E E.L.I.M.I.N.A.T.I.O.N style. GOM is probably the 395946812th organizer who adopts this. God damn it
It's because how liquipedia format GSL group stage some people don't realize it's a double elimination bracket with 4 players, where both upper bracket winner and lower bracket winner advance.
I thought people actually watches the GSL, instead of looking at liquidpedia, no?
Not that many people in US can watch GSL with ungodly hour
On December 21 2012 11:28 jmbthirteen wrote: I really dont care if the player is 2-3 and against the other guy and advances. They aren't playing a bo5. Its two 2 bo3s. I think the GSL system is by far the best out there. I'm glad more and more tournaments are using it.
I agree with this. I hate each other style of group play, especially those that end up in endless tie breaker matches or when casters need to make themselves look like fools in front of cameras counting maps wins, opposed wins and other bullshit tie breaker mechanics.
I would even implement GSL system into other sport group plays (like football). It would make those sports more interesting as playing to tie would be useless, each match would need to have a winner.
I think it is fine to think that it is not fair that a worse player may win in the Finals of a group stage, but this is how the system works. It is not a gentlemen's club on the digital battlefield; you either get him, or he gets you and that's all there is to it.
To be honest, I wish they kept the Bo1 matches like in the MSL Survivor group stage, but people argue that *that* is not fair. Apart from Starleague seedings, imagine how much more mystique surrounds players who can consistently crack Ro16 CodeS in a Bo1 format.
Unless he has a tendancy for gimmicky or cheesy play like Kwanro or Shine
well, the player that wins the first match has a 50% shot at advancing first place, while the player that loses first match has a 50% shot at going down to last place (important in code S ro32 becuz 4th placers go straight to ro48 code A), i say its fair the way it is
On December 21 2012 12:40 Sikly wrote: If you win one BO3, lose one, than lose another, you are out. It doesn't matter who the BO3's were against, the rules are simple. Don't lose two BO3's, and you advance. People need to get X player vs Y player out of their minds, and think of it as individual series, because that is what it is. In most peoples opinions(based on the poll you referenced), it is the best format. Perhaps it is not flawless, but no format is flawless.
If anything, I find it very impressive when a person manages to come back and win a bo3 after going 2-0 against someone. It shows a very good ability to analyze games and adjust within a short amount of time. Round Robin formats not only create pointless games, but almost always end up feeling very nasty for multiple players.
That's wrong because this is a strategy game. Because you expose your strategies in set one and give your opponent the chance to study you and capitalize on your mistakes/fix his own in the following games.
All in all i think this is a farce. I'd rather see a real double elimination format and skip these groups alltogether. Because this is like double elimination, except unfair to some players.
Its not unfair if you failed to win 2 Bo3s then you are out. Also a person playing a second Bo3 against somone of same race never mind same opponent either has a different strategy or is giong to face same problem.
Ill explain this by pretending for example I am DRG and I am in a group with Taeja MKP and Polt. I first face Polt and win 2-0 and then play MKP and lose 2-1. What will happen next is I will face either Taeja or Polt again. No matter which player it is my style for that day is either exposed or I came prepared with extra builds so its safe.
Hit the nail right on the head. This is why GSL is so prestigious. You can't just walk in with one style and a couple of builds. You need like a playbook of builds so that after you expose said build to your opponents you can just fall back on something else that hits at a different timing or something wonky. So if you don't prepare enough and can't win 2 Bo3's because you were too lazy to study enough to come up with multiple builds against every player you don't deserve to advance.
OP made good suggestions, but I think it would be best to just bite the bullet here and accept that the player with fewer wins would advance. I wouldn't mind an extended series, but I don't think GSL will do that.
Who advances from the GSL-groups depend on a larger degree on chance than round robin does. Saying otherwise is simply wrong. You are still allowed to prefer the GSL-system, but the system comes with downsides that needs to be acknowledged.
In the GSL-system, bo3 > bo5 in deciding who advances. That is a flaw inherent in the system.
You get a better idea of the strength of each player by having each play each-other. You get a worse idea of the players strength by only letting each play meet two other opponents. That is a weakness in the GSL-system compared to the round-robin format.
Saying otherwise is sticking your head in the sand.
Wait a second, so after 2 years of people bitching about extended people are now bitching about GSL format? It offers a good combination of fairness and entertainment I'm all for it.
On December 22 2012 01:13 m0ck wrote: Who advances from the GSL-groups depend on a larger degree on chance than round robin does. Saying otherwise is simply wrong. You are still allowed to prefer the GSL-system, but the system comes with downsides that needs to be acknowledged.
In the GSL-system, bo3 > bo5 in deciding who advances. That is a flaw inherent in the system.
You get a better idea of the strength of each player by having each play each-other. You get a worse idea of the players strength by only letting each play meet two other opponents. That is a weakness in the GSL-system compared to the round-robin format.
Saying otherwise is sticking your head in the sand.
that happens in round robin too
2-1 2-1 2-1 0-3
you don't let them only play 2 opponents all the time and if this happen they just don't play either against the best or worst player of the group which is not a problem.
I really appreciate people who go through the lengths of complying this kind of data, but I think you are throwing around terms like "qualified for good/bad reasons" way to loosely. The rules are clearly laid out, the player who advances always does so for a good reason because he won within the rules.
When analyzing this data you should also consider that in the situation where 2 players go 1-1 and meet a second time, it is more likely that they are close in skill, making an 50/50 record for the winner of the first BO3 somewhat expected and acceptable. Add to this that the first BO3 is likely the one that the players have been preparing for the most and that some players excel in matches were players can prepare and some do not, and it becomes more clear that this is just a nuance of a format with both predetermined and non-predetermined opponents, not an inherent flaw.
On December 22 2012 01:18 Nick_54 wrote: Wait a second, so after 2 years of people bitching about extended people are now bitching about GSL format? It offers a good combination of fairness and entertainment I'm all for it.
Lol, I find this amusing too. This has been discussed to death. Kind of a boring topic now.
If the game is balanced around Terran's stronger early/mid game and the stronger late game armies of P vs T, and Z vs both P and T, Terran's tricky builds are a strength that will be diluted the longer a series goes. A longer series/more games played does not necessarily determine a better player. It can also expose imbalances as one race can adapt better than another. That also is not necessarily the strongest factor, but something to keep in mind before claiming that a bo101 will always determine the better player.
Well, ok, first off...you can't make statistics stating things like "times where a player advanced for good/bad reasons". These bad reasons are completely subjective and totally your own opinion. I would argue that a player advanced for good reasons in 100% of cases, i.e., he won the damn match when qualification was on the line.
As to the bigger point, there is a very good reason why round robin is not used in the GSL, and GOM has stated it already. There will be situations where players can throw games for their friends/teammates because they have nothing to lose. In those same situations, even if the games aren't thrown, the player has no need to play with any urgency, creating a very boring viewing experience. One way to fix this might be to run round robin with money for each win, but even then, you may experience a teamkill situation.
Extended series have been discussed to death, but I think they are terrible. Your argument about the player who lost first receiving an easier draw is actually more applicable in an extended series scenario, because with an extended series, the player who wins the first Bo3 will gain a huge advantage if he happens to face his defeated opponent again later. It creates situations where a player is rewarded for something that is totally out of their control (the results of a match that they are not participating in).
Finally, I want to say that even though all matches are played during a "group stage", it does not mean all these matches need to hold equal significance. A qualifying/elimination/ace match is more important than a regular match, that's just the way it always is. If you lose the unimportant match and win the important match, good for you, you turned it on when it mattered.
On December 21 2012 11:20 chuky500 wrote: A thread in April discussed whether the GSL group format was flawed or not and at the time 82% people judged that the format was fine. Since then many popular tournaments have adopted it as it is shorter, easier to run and provides more drama. As more groups were played this way we've seen more games leaving that awkward feeling when there's a rematch and the player qualified isn't the one with more wins, such as the WCS finals when Hero had qualified over Stephano despite having a record of 2-3 against him. It happens because when 2 players meet again, the result of the first bo3 is discarded. This means to qualify one player has to beat his opponent in 2 bo3 while the other one qualifies with a 1 bo3 draw.
I completely disagree with this. The penalty for losing your first bo3 is that you go down and have to win another game to get into the 2nd place game. You should not be penalized in the 2nd place game in any way (please no extended series). The GSL format is great. The only problem with it is that there is some luck involved with who you play against first, but there is luck in any kind of draw by definition.
On December 22 2012 01:18 Nick_54 wrote: Wait a second, so after 2 years of people bitching about extended people are now bitching about GSL format? It offers a good combination of fairness and entertainment I'm all for it.
Lol, I find this amusing too. This has been discussed to death. Kind of a boring topic now.
the most "amusing" thing about this thread is people declaring every tourney should apply "GSL Style"....
The GSL format is not double elimination. Double elimination is balanced around several points. First, the winners bracket is harder than the losers bracket. To compensate that there are 2 downsides about being in the lower bracket : you can be eliminated at every round and very important you have to play twice as many games as someone in the winners bracket. And to make sure the lower bracket player plays more games, players can't meet 2 rounds after playing each other because the lower bracket is reverted. People seem to forget point 3 and 4 when comparing the GSL format and double elimination but GSL is actually double elimination without point 3 and 4. So if you consider double elimination balanced, then removing those points shifts the balance towards the lower bracket.
On December 21 2012 11:20 chuky500 wrote: A thread in April discussed whether the GSL group format was flawed or not and at the time 82% people judged that the format was fine. Since then many popular tournaments have adopted it as it is shorter, easier to run and provides more drama. As more groups were played this way we've seen more games leaving that awkward feeling when there's a rematch and the player qualified isn't the one with more wins, such as the WCS finals when Hero had qualified over Stephano despite having a record of 2-3 against him. It happens because when 2 players meet again, the result of the first bo3 is discarded. This means to qualify one player has to beat his opponent in 2 bo3 while the other one qualifies with a 1 bo3 draw.
I completely disagree with this. The penalty for losing your first bo3 is that you go down and have to win another game to get into the 2nd place game. You should not be penalized in the 2nd place game in any way (please no extended series). The GSL format is great. The only problem with it is that there is some luck involved with who you play against first, but there is luck in any kind of draw by definition.
That is the benefit of the system and it really is no different than the old BW formats made for TV.
On December 21 2012 12:10 OfficerRobert wrote: BO1 Round robin is the correct way to do groups.
Bo1 is never the correct answer in sc2
Say what you like, but the RO8/RO4 of the OSL was solid. It's only one datapoint so I won't get carried away.
I actually think that even Proleague is trash because its BO1... Terrible suggestion.
If we're talking about TV format it really isn't an option.
On December 22 2012 02:20 chuky500 wrote: The GSL format is not double elimination. Double elimination is balanced around several points. First, the winners bracket is harder than the losers bracket. To compensate that there are 2 downsides about being in the lower bracket : you can be eliminated at every round and very important you have to play twice as many games as someone in the winners bracket. And to make sure the lower bracket player plays more games, players can't meet 2 rounds after playing each other because the lower bracket is reverted. People seem to forget point 3 and 4 when comparing the GSL format and souble elimination. So GSL is actually double elimination without point 3 and 4. So if you consider double elimination balanced, then removing those points shifts the balance towards the lower bracket.
Other guy just told you what the real benefit of the round is and to be frank. If it were up to me and I had an allotted timeslot to use I wouldn't even consider using double elimination. The players are fortunate. If you win the winners good on you. If not you still have an opportunity to move on. This is actually one of our oldest systems we have and it's widely accepted.
On December 22 2012 02:20 chuky500 wrote: The GSL format is not double elimination. Double elimination is balanced around several points. First, the winners bracket is harder than the losers bracket. To compensate that there are 2 downsides about being in the lower bracket : you can be eliminated at every round and very important you have to play twice as many games as someone in the winners bracket. And to make sure the lower bracket player plays more games, players can't meet 2 rounds after playing each other because the lower bracket is reverted. People seem to forget point 3 and 4 when comparing the GSL format and double elimination but GSL is actually double elimination without point 3 and 4. So if you consider double elimination balanced, then removing those points shifts the balance towards the lower bracket.
GSL is a true double elimination bracket, with 4 players.
the true meaning of double elimination format is: - a player has two chances, which means, you could only be eliminated by losing TWO matches - one can only face the other for the second time in the grand final (however, this rule is broken quite often due to organizer messing up the seeding process in big bracket (ie 64-man)
So yeah, a double-elimination bracket for 2-man tourney is essentially a bo3 by definition.
the number of games to be played in LB is not twice. It depends on how big the bracket is you just cant determine if WB is harder than LB. Thats clueless. If any, the only measure for that is number of games, in which LB players always play more
So with 2 players advancing from a group of 4, only consolidation final is needed. The grand final of that group of 4 might be played to truly determined 1st place and 2nd place but it would be too excessive for broadcasting. And I guess it's not the problem that ppl here seems to dwell into.
On December 22 2012 02:20 chuky500 wrote: The GSL format is not double elimination. Double elimination is balanced around several points. First, the winners bracket is harder than the losers bracket. To compensate that there are 2 downsides about being in the lower bracket : you can be eliminated at every round and very important you have to play twice as many games as someone in the winners bracket. And to make sure the lower bracket player plays more games, players can't meet 2 rounds after playing each other because the lower bracket is reverted. People seem to forget point 3 and 4 when comparing the GSL format and double elimination but GSL is actually double elimination without point 3 and 4. So if you consider double elimination balanced, then removing those points shifts the balance towards the lower bracket.
Arg... NO... You cannot set your own rules about what double elimination is. The GSL format is by DEFINITION a double elimination format. Although your point 3 and 4 are valid by itself, it is not the core requirement of a double elimination format. Those things just occur when the double elimination format is bigger than 4 players.
And players do play more games in the LB. WB winner only plays 2 games while the LB winner has to play 3.
you balance your whole analysis around not even caring about the second game, which is precisely what is wrong here. what you do basically is that you consider the group winner as untouchable and the group loser as a freewin; it's exactly where you fail : let's assume it's that way : it now means that the first bo3 is meaningless too, since either way (group winner is unbeatable, group loser is a freewin) it will all go down to a last match rematch : it means everything is perfectly balanced. now, considering this claim is false, and you can actually beat the group winner and the group loser can beat you, both players come in the last match with a different path : one had a probability to qualify, while the other had a probability to drop out. thus winning the first bo3 already had an impact, it created an opportunity : positive for the one that won, negative for the one that lost. GSL format is perfectly balanced.
I like this BECAUSE it creates this dynamic. It would be silly to say a player is better than another just because of one matchup. Far more reasonable is to judge a player based off of how they do against the same exact players.
On December 22 2012 02:20 chuky500 wrote: The GSL format is not double elimination. Double elimination is balanced around several points. First, the winners bracket is harder than the losers bracket. To compensate that there are 2 downsides about being in the lower bracket : you can be eliminated at every round and very important you have to play twice as many games as someone in the winners bracket. And to make sure the lower bracket player plays more games, players can't meet 2 rounds after playing each other because the lower bracket is reverted. People seem to forget point 3 and 4 when comparing the GSL format and double elimination but GSL is actually double elimination without point 3 and 4. So if you consider double elimination balanced, then removing those points shifts the balance towards the lower bracket.
No see, all the things you listed are a consequence of a double elimination bracket. A double elimination bracket is a winners/losers bracket system where a loss in the winners bracket knocks you down into the losers bracket before being eliminated (and the lower bracket entry points flip sides every round). Yeah, this means that you can be eliminated after taking a loss and yeah, this means you have to play more matches if you're in the lower bracket.
Everything else is just a consequence of the bracket format. GSL Format is a 4 person double elimination bracket without a grand finals as 2 people qualify. No more, no less. If you make a 4 person double elimination bracket, and you take out the grand finals, you've got "GSL Format". Try making a 4 person double elimination bracket and end up with something else, I'll readily point out why it's wrong.
EPT![[image loading]](http://i.imgur.com/NaVMQ.png)
![[image loading]](http://i.imgur.com/Xc3uF.png)
![[image loading]](http://i.imgur.com/lOF5s.png)
