|
On December 04 2010 11:47 enzym wrote:Show nested quote +On December 02 2010 18:07 SlapMySalami wrote:On December 02 2010 16:42 enzym wrote: It's a good rule. There should be a significant drawback to being pushed into the loser's bracket. So what happens when the guy that beat you gets knocked into the loser bracket and faces you again? What is his significant drawback? Starting 2-0 against someone he already beat? Of course. He already won once. Why wouldn't that count for much of note? I think it's a very good and sensible rule.
Then everyone who comes from the winnerbracket gets to start 2-0 in a best of 7 against whoever is advancing in the lower bracket? Or only if they played before? Why does that matter?
Think about what you're saying.
German fish.
|
+ Show Spoiler +On December 03 2010 09:27 crms wrote: Someone can correct me if i'm wrong but my understanding (as a former LAN FPS player CPL 3rd wut wut!) that HALO and other FPS games compete on a LEVEL playing field. You either out play, out think, out whatever your opponent or you don't. The team that played better will win and the better team is MORE likely to win at a much higher rate. In my previous FPS experience (team fortress classic) we held a 90%+ winrate because we were simply the better players congregated into 1 team.
Now Starcraft is such a highly variable game. First, maps make a HUGE difference, getting a bad map pool for your race automatically puts you at a disadvantage. This disadvantage usually isn't insurmountable but it can be significant. This doesn't exist in (most) competitive FPS games. Not to mention how many different ways you can play SC2, build orders, cheeses, different races, etc. and the competition level at the tip top. Anyone in the top50 best players in the country can take a game or series off of one another. The best players in BW of all time have what 72% win? An insanely good gosu as fuck player may win 60% just six out of 10 games and that would make you a fucking baller. I'd wager the #50 Halo has essentially 0 chance to pull a series off the #1 Halo team. In my experience in a top FPS team I don't think we lost to anyone outside the top5 or so.
I see the argument about variables in SC, but shouldn't steps be taken to help prevent this? I realize the game is young but maybe a map pool that is more "fair" for all the races would help eliminate some of that. This is even more true when the players are so close in skill. If it's a bad map for your race and you play someone you are a lot better than well then you probably win anyways, but someone close to you almost gets a guaranteed win. All of the other variables are player generated and the nature of SC being more complex than most FPS. But I feel like map variability, at least eventually, should be far less of an issue, or at least steps should be taken to make it that way. And yes only about the top 8-10 Halo teams are on a similar level, no one outside the top even 5 could take a series off #1.
As far as the rule goes, I see arguments for both sides. I don't like how a player can be winning against another player in total games and be out, and the fact that the order itself could determine who goes home bothers me. I feel like this removes some variability in that instance.
On the other hand, both players lost a series to get there, the only difference being the player that won didn't have to go through numerous rounds in losers, which in itself can be an advantage. I still don't see it as arbitrarily giving them a lead though, they already played and had a result. I'm really on the fence about it.
I think the rule can go, but other rules put in place. One would be to never play the same map twice, don't know if this is already done, I would think it is though. Whatever maps they played the first time should be taken out the next time through.
Also for the winner of the winners' bracket, there shouldn't be an advantage given to them in games won, just require the player from the losers' brackets to beat them in 2 Bo3s in order to win the tournament. (this is what is done already in MLG if the players meeting in the grand final haven't played yet)
|
On December 02 2010 23:52 Malloy wrote: A lot of people argue that the extended series adds an advantage on the person that won the original best of 3, but in actuallity, not doing the extended series adds an advantage on the person that lost the original best of 3.
If Y plays X and wins the original series 2-0, he has demonstrated that he is better than X.
If Y meets X in the losers bracket, X now needs to demonstrate that the first series was merely bad luck and he is indeed better; thus, he must win 4 games and avoid losing an additional 2.
Okay. Chill's example again: Suppose Y suddenly loses twice and X wins the rest of the tournament coming up from losers bracket into grand finals, then wins 2 sets against the winner bracket finalist. Are you still saying Y is better than X? Should Y then play X because he beat X earlier so X can prove he's better than Y? No, you don't. You don't need to prove you're better than another person to show you're the champion. You just have to WIN.
Also, because of this rule, X might have to play an extra 2 games for NO REASON. If he doesn't ever meet Y again, he doesn't need to win 4 games, he only needs to win 2 a series.
If the extended series didn't exist, and Y meets X in the losers bracket and a fresh best of 3 occurs. X wins two games and moves on when X has only actually demonstrated that he is AS GOOD AS Y (2-2 extended).
No. He's proven that he's better than Y when the moment counted. If X and Y meet in every other tournament and Y has beaten X and Y beats X, but the two meet in the grand finals, should they play a super-duper-ultra-extraordinary extended series of best-of-57 with X needing to win 25 to prove he is INDEED the better player? No. He shouldn't.
Conclusion; extended series removes an unfair advatage that is gained by the loser of the first round.
There never was an advantage gained by the loser. The loser has to play more games. They start on EVEN grounds when they play their new set. It's not an unfair advantage that the later series are worth more, that's just how tournaments work. Otherwise, we'd have a round robin, not an elimination tournament. Grand finals is worth more than a pools set, like it or not. Just like a series in the loser series at a later time for all the marbles is worth more than an earlier set in the winners series. It's not an unfair advantage if they start even; in fact, it's fair.
|
Nony put it best when he said would you rather a best of 7 or two best of 3s to determine who is better between two players? Obviously the best of 7 which is what extended series delivers.
|
On December 04 2010 14:50 samuraibael wrote: Nony put it best when he said would you rather a best of 7 or two best of 3s to determine who is better between two players? Obviously the best of 7 which is what extended series delivers.
Why does it matter? Who cares about who has the better record between two players? We care about who can win when the moment matters. Once again, if A beats B, but A loses twice and B wins the whole thing, should we them make B play A once again to see who's the better of the two players? No. We KNOW B is the better player because he won. Their personal records don't matter at all. If we cared about that kinda crap, we'd be having endless loops. Furthermore, the best of 7 brings a contradiction once again: If it's an "extended series," then shouldn't player B be back in the winners bracket again if he wins?
|
On December 02 2010 07:43 StockTheFridge wrote: I've been a part of MLG for a long time and as I see it one of the main points of the extended series is to make sure that the winning player has a positive record head-to-head versus every opponent that player has faced. I would also like to note that having extended scoring may seem illogical but is not unprecedented. It reminds me of aggregate scoring in professional soccer.
Neither way is necessarily right or wrong because both systems are fair as long as they are well-defined and set up in advance of the tournament. The right thing for MLG to do though IMO is to listen to the players and the fans instead of cavalierly making some decision based on what they have done at some other time/place.
I love that MLG is trying to make their tournaments as good, fair and entertaining as possible and completly ignore the norms and doing stuff one way for no other reason that its been done that way for a long time by other people.
Super good!
However!
In this case I dont agree with the extended series. And I will actually argue for theory that suggests that it is unfair, missguided and makes a champion less legit.
The main problem with the philosophy you're using if you're arguing for extended series is that it assumes that the best player in a field is a favorite over any other player within the field and that's simply not true.
The best player out of a field is going to be the player that wins most consistantly within the playerpool. Genereally the best player out of a field is going to be the player that is a underdog to the least ammount of other players within the field.
Almost never are you going to have a champion that's a favorite against all of the players in the field. There are going to be cycles like A > B > C > A especially within Starcraft because of the different matchups.
I want you to imagine a bracket of 64 players. Now replace all the playernames with a precentage that would represent the chances of player to be the true teoretical champion within the playerpool.
For a 64 player bracket this would leave all players with 1.5625% chance of beeing the champion before any games are played. As games are starting to get played you will see these precentages increase and decrease depending on how they do against their opponents.
Lets follow player A in this example that will prove a few points.
Player A beats two players in WB before he is beaten in WB round 3 and knocked down to LB. Now once in LB he is going to have X% chance of beeing the champion. X is always going to be the same number in this possition and the same holds true for any slot in the bracket.
Now you can realize that what changes your precentage is dependant on if you win or lose, and what precentage you and you're opponent is holding.
Let's call the player that Player A beat in WB round 1, Player B.
Now imagine Player B fights his way through LB and gets to player against Player A again when he got knocked down by losing to another player in WB round 3.
Now based on WB round 1 results where Player A beat Player B 2-0, MLG would extend that series to a best of 7 starting at 2-0 in favor for Player A.
In WB round 1 actually. Player A beat random opponent with 1.5625% of beeing the true champion in theory. Player B lost to random opponent with 1.5625% of beeing the true champion in theory.
So why would you extend series ONLY if the actuall player behind the precentages happens to be the same? Why wouldn't any WB round 1 winner be granted a advantage against any WB round 1 loser once in LB?
The truth is that the actuall player behind the precentage does not matter. And the harm you're doing by using extended series is that you are enchancing individuall player collisions and making each runthrough of the bracket more dependant on random factors. If you use extended series you would actually need many more runthroughs of a bracket in order to be able to determine players true skill.
The champion of you'r bracket will end up with a smaller precentage of beeing the true champion.
Hopefully this does shed some light and if you have any questions regarding the theory behind this please send me a private message, I would be happy to explain deeper and more detailed should it be neccessary.
Last but not least I would like to say that MLG for sure has to be the best organisation we could ever have. Sure nothing starts off perfect but it's the potential that is important and it's so great that this organisation have people reading this stuff that we're discussing and actually cares and respects the viewers and players oppinions. We couldnt have a better thing happen to the Starcraft 2 scene then having MLG pick up this game.
Cheers MLG!
|
Why would they do this? Isn't beating someone in a BO3 enough to say your "better," and wouldn't the person be able to counter your previously prepared BOs then?
|
On December 04 2010 16:09 ghrur wrote:Show nested quote +On December 04 2010 14:50 samuraibael wrote: Nony put it best when he said would you rather a best of 7 or two best of 3s to determine who is better between two players? Obviously the best of 7 which is what extended series delivers.
Why does it matter? Who cares about who has the better record between two players? We care about who can win when the moment matters. Once again, if A beats B, but A loses twice and B wins the whole thing, should we them make B play A once again to see who's the better of the two players? No. We KNOW B is the better player because he won. Their personal records don't matter at all. If we cared about that kinda crap, we'd be having endless loops. Furthermore, the best of 7 brings a contradiction once again: If it's an "extended series," then shouldn't player B be back in the winners bracket again if he wins?
What Tyler is saying, if I'm not mistaken, is that by playing two best-of-3 instead of an extended best-of-7 you run the risk of having a tie in the number of series won. For example, player A beats player B in winner's bracket. Then, sometime later in loser's bracket player B beats player A. If this was two best-of-3 this means both player B and player A won once and we don't know who's really the better player. How can we then say that player B should advance because he won the second best-of-3? It makes no sense, because while player A also dropped to the loser's bracket, he was dropped by a better opponent further along in the winner's bracket (at least statistically it will be so almost every time). Thus player B, having been dropper earlier is by definition the worst player of the two right now. Therefore player B must prove that he is indeed better than player A and that the first games in which he lost did not reflect his actual, average (or "normal") skill level.
It's the same reason Blizzard uses a hidden ranking to match you on the ladder and not your actual ranking. The actual ranking is your average skill level, the hidden ranking is your skill level now. If you keep playing at your hidden ranking's skill level, your actual ranking will move up or down consequently. Any match-making system that uses past performance to choose your next opponent will have that kind of hidden ranking somewhere in the math. In the MLG, the extended series rule comes as a way to take into account this "hidden ranking" that comes from having both a winner and a loser's bracket.
For comparison, in other sports that do not use double elimination, the highest ranking (or seed) at the end of the regular season gives you the chance to play the worst team first plus have home advantage. Home advantage then goes to the team that can win over their opponent faster, thus showing that their hidden ranking is higher (if the winning team was the lower seed) or at least as good as their actual ranking (if the winning team was the higher seed). The home advantage is admittedly not great, but as this is not double elimination and there is no way for a team to play two series against each other, it does not matter much. On the other hand, playing the worst team first is extremely advantageous of course, but inasmuch as we do not want (or cannot because not all players are ranked yet) to systematically play higher seeds against lower seeds in Starcraft tournaments, the double elimination format with extended series is better. Plus, to actually give the home advantage to the better player in single elimination, you would need to give the first map choice to the player who won his series in less games so far, if equal then to the highest seed. But again, home advantage is a poor substitute I think and plays too much to map imbalances instead of simply giving a comfortable playing atmosphere like in other sports (since the playing field is always the same). So either play the best seeds against the worst seeds, when MLG begins to have a rank for every player, and change to single elimination or keep random seeds and double elimination if rankings are too volatile would seem the best options.
|
I don't think its an unfair rule, but its simply not fun to watch as a spectator. It makes things confusing, and many people will SEE it as unfair, which will just kill the whole competition for them. Because of this, i don't believe it should be used.
|
On December 01 2010 15:14 Phraxas wrote: The extended series is bollocks.
InControl: "I will f*ing smash you in an arm wrestle. Does that mean that the next time we play I need a small child punching you in the crotch whilst we have a rematch?"
Best quote of the podcast
|
In starcraft, if player a is a favorite over player b does that mean they are neceserally the better player? If player a won in the winners bracket, when they play again, they are both x-1 so they both figure to be of equal calibre so should be given an even match.
|
The only scenario I can see it not working is if there was like several days between the games. Otherwise I don't think there's any argumant that can objectively show it's a bad rule. I can't vote because I don't care whether they have this rule or not but I'm undecided =)
It's sort of rule that some people like, some don't. But the rule itself is just.
|
I'm not sure if this has been mentioned before (Read: Haven't read through all 12 pages :p), but someone has done a statistical analysis of the MLG Extended Series rule.
http://www.teamliquid.net/forum/viewmessage.php?topic_id=168345
Granted he cannot account for everything in his models, such as if player skill is transitive, or the psychological effects of the rule and tournament setting, I think he does a very good job at providing an analysis of the rule.
The results and conclusion for those who are too lazy to read the thread:
+ Show Spoiler +4.4 EFFECT OF EXTENDED SERIES
We now consider the effect of the extended series in isolation. Specifically, how often is the extended series used, and how often does is "correct an injustice" from the winners' bracket?
In this case, we consider a 64-player tournament in double elimination with extended series format. In a standard double-elimination format, 127 matches will be played.
Simulation shows that, on average, 18.8 extended series will be played in a 64-player tournament. This means that 15% of matches, on average, will be rematches of players.
Similarly, of these 18.8 matches, 3.03 of them will result in "corrections". A correction is when the better player loses in the winners' bracket and wins the extended series to continue in the tournament. In 2.17 of the matches, the worse player won in the winners' bracket and won the extended series, meaning the extended series failed to "correct" the result from the winners' bracket.
The worst possible outcome is when the better play wins in the winners' bracket and loses the extended series. The extended series does well here, only introducing 0.55 such results per tournament, or 4% of the extended series.
Considering the disadvantage that the better player has when entering the extended series, it does surprisingly well at correcting these results, succeeding 58% of the time. At the same time, it only introduces bad results 4% of the time.
I am tempted to conlude that extended series is successful at letting the better player continue in the tournament, however data is missing to compare against a standard double elimination tournament. A good area of extension for this study would be measuring the outcome if a regular best-of-three were done, and comparing its correction/injustice rate to the extended series. The ratio from the extended series (58%/4%) seems pretty hard to beat -- I would expect a best-of-three to allow the better play to proceed more often, but have a much higher injustice rate.
5. CONCLUSION
Whe considering individual matches, the extended series appears to perform well to make sure the better player continues in the tournament. In this sense, it fulfills its purpose.
But when looking at the larger picture, it appears that the extended series has little effect on the outcome. While the extended series rule does slightly improve outcomes, these differences are not particularly significant compared to the overall double elimination format.
What is clear from these results is that both elimination formats leave much to be desired when compared to a round robin tournament. Although round-robin is impractical due its large number of games, other tournament formats such as swiss-style or those with rounds play deserve further consideration.
Basically, based upon his player models, the MLG Extended Series rule has the most probability besides a Round Robin of letting the "best" player win, as well as letting the "best" player advance when they meet again and must play an Extended Series.
However, the difference of who wins the entire tournament between it and a normal double elimination tournament is very small, as to not really effect the results of the tournament at all.
So basically, the extended series slightly improves the outcome of the tournament, but not to such a great extent.
|
On December 04 2010 17:56 Bidouleroux wrote:Show nested quote +On December 04 2010 16:09 ghrur wrote:On December 04 2010 14:50 samuraibael wrote: Nony put it best when he said would you rather a best of 7 or two best of 3s to determine who is better between two players? Obviously the best of 7 which is what extended series delivers.
Why does it matter? Who cares about who has the better record between two players? We care about who can win when the moment matters. Once again, if A beats B, but A loses twice and B wins the whole thing, should we them make B play A once again to see who's the better of the two players? No. We KNOW B is the better player because he won. Their personal records don't matter at all. If we cared about that kinda crap, we'd be having endless loops. Furthermore, the best of 7 brings a contradiction once again: If it's an "extended series," then shouldn't player B be back in the winners bracket again if he wins? What Tyler is saying, if I'm not mistaken, is that by playing two best-of-3 instead of an extended best-of-7 you run the risk of having a tie in the number of series won. For example, player A beats player B in winner's bracket. Then, sometime later in loser's bracket player B beats player A. If this was two best-of-3 this means both player B and player A won once and we don't know who's really the better player. How can we then say that player B should advance because he won the second best-of-3? It makes no sense, because while player A also dropped to the loser's bracket, he was dropped by a better opponent further along in the winner's bracket (at least statistically it will be so almost every time). Thus player B, having been dropper earlier is by definition the worst player of the two right now. Therefore player B must prove that he is indeed better than player A and that the first games in which he lost did not reflect his actual, average (or "normal") skill level.
Why doesn't it make sense? The second series mattered more. Round of 8 matters more than round of 64. Grand finals matter more than Round of 8. What's wrong with that? The later the rounds, the more they matter. That's how a tournament works. Otherwise, how can we say the winner is indeed the winner of a tournament if he drops a series along the way? Because his later matches mattered more. Furthermore, this is saying we should take history into account. The problem is, we don't, or else we'd take all previous head to heads, and make them play an extended series throughout all MLG tournaments. But we don't care about that crap. Why should we? It's about who's better at the moment. The fact of the matter is, B won when it mattered more. If A were better, he would've won at that moment. Here's another way to look at it. Flash vs Jaehoon okay? Let's suppose Jaehoon is a practice bonjwa. He can go 20-15 vs Flash in Practice. His "hidden skill" is better. But guess what? When he gets out on stage, there's no DOUBT that Flash will crush him. If they play an OSL series and flash wins 2-0, they're only 20-17. Do we care? No. Flash won when it counted. He was indisputably better at that moment. Who cares about the past?
It's the same reason Blizzard uses a hidden ranking to match you on the ladder and not your actual ranking. The actual ranking is your average skill level, the hidden ranking is your skill level now. If you keep playing at your hidden ranking's skill level, your actual ranking will move up or down consequently. Any match-making system that uses past performance to choose your next opponent will have that kind of hidden ranking somewhere in the math. In the MLG, the extended series rule comes as a way to take into account this "hidden ranking" that comes from having both a winner and a loser's bracket.
I have no idea this happened, and I think this is idiotic for tournaments. There shouldn't be any "hidden rankings" stuff. Tournaments aren't about the past once seeding is done. They're about the moment. Who's the best player at the moment?
For comparison, in other sports that do not use double elimination, the highest ranking (or seed) at the end of the regular season gives you the chance to play the worst team first plus have home advantage. Home advantage then goes to the team that can win over their opponent faster, thus showing that their hidden ranking is higher (if the winning team was the lower seed) or at least as good as their actual ranking (if the winning team was the higher seed). The home advantage is admittedly not great, but as this is not double elimination and there is no way for a team to play two series against each other, it does not matter much. On the other hand, playing the worst team first is extremely advantageous of course, but inasmuch as we do not want (or cannot because not all players are ranked yet) to systematically play higher seeds against lower seeds in Starcraft tournaments, the double elimination format with extended series is better. Plus, to actually give the home advantage to the better player in single elimination, you would need to give the first map choice to the player who won his series in less games so far, if equal then to the highest seed. But again, home advantage is a poor substitute I think and plays too much to map imbalances instead of simply giving a comfortable playing atmosphere like in other sports (since the playing field is always the same). So either play the best seeds against the worst seeds, when MLG begins to have a rank for every player, and change to single elimination or keep random seeds and double elimination if rankings are too volatile would seem the best options.
What? Starcraft DOES use seeding. They obviously place like unknowns vs Jinro first rather than go oh, Jinro vs Ret, then Jinro vs. Select or something. No, that doesn't happen. Starcraft seeds, better players get an advantage, that should be it. In NFL games, you didn't see NE start off 1-0 vs the Giants in a best of 3 for the superbowl do you? You don't see the Lakers start off like, 4-1 vs the Mavericks do you? No, it's all about the moment.
Also, no one has answered the point about the loser winning an extended series should be placed back into the winners bracket again. CONTRADICTIONNNNNNNNNNN
|
Voted no. With the obvious exception of seeding the first round, the tournament rules should not influence the outcome of any one match.
|
it just doesn't make sense .... i say no
|
@Mr. Wiggles: Thanks didn't know about it. Basically his conclusion is pretty much the same as mine. It does no harm (better player is more likely to advance) but it doesn't help much either. That's why I don't really care much about this rule. From spectator's POV it's good to see more matches but most spectators are used to the usual double elemination and don't like it because they find it wierd.
Still, it basically looks like a rule that barely influences the tournament outcame (in a good way) but doesn't account for 'people factor' (ppl feel something is wrong with this rule but can't objectively say what is and why). The question is, is it worth the trouble? Slight help vs angry mob
|
IF they meet again it should be a BO3 or 2xBO3 if its the big final!
All the rest is stupid and doesnt deserve any more words for it!
|
If you don't use extended series and you want to use a double elimination tournament, then one player gets punished because of the arbitrary order of the games they play.
If A wins 2-0 first, and then loses 1-2 later, he is has beaten B 3 times to 2, but because of the order the games were played in, he is out, while B moves on.
I fail to see how player B is punished a second time under extended series, it just gets rid of the imbalance caused by the arbitrary order of the games they played.
This is the inherent flaw in doing a double elimination tourney (without extended series) rather than a single elimination tourney, the order the games are played in matters in determining which of two players advances.
|
|
|
|