|
motbob
United States12546 Posts
Game score should always be used as a tiebreaker before head-to-head. Already I can feel the angry posts being written. Let me explain why I think what I think.
![[image loading]](/staff/motbob/UKgroupA.png)
In this group, ForGG and Babyknight were tied in match score at 3-2. Because of that tie, a tiebreaker had to be used.
There were two pieces of information that the tournament organizers could have drawn on to make their decision on who to advance. The first piece of information was that Babyknight defeated ForGG 2-1. This implied that BabyKnight should have advanced over ForGG. The second piece of information was that ForGG's record against common opponents of ForGG/BabyKnight was 7-2, compared to the 5-5 record of Babyknight. The final game score of the group, derived mostly from the record of ForGG and BabyKnight against common opponents, implied that ForGG should have advanced.
So there are two conflicting pieces of information here. Depending on which piece of information is deemed more important, one of these two players should advance.
I think that the game score is a more important indicator of who the better player is than H2H because it draws on a larger sample size. Variance exists in SC2. The best player does not always win the game (or the series). As with all phenomena with variance involved, it's always better to increase your sample size to see what is really happening.
Tournaments choose the H2H tiebreaker as the one that takes priority because it "feels right." It feels like justice is served when a tie between two players is decided by the games played between those two players. But if Player A has beaten Player B, it's very difficult for Player B to overcome that loss and do better than Player A in game score. ForGG had to obtain at least a +3 advantage in game score over BabyKnight against common opponents to overtake him. If ForGG had lost 0-2, he would have had to obtain a +5 advantage! (In this way, you can see that the H2H tiebreaker is actually included in the game score tiebreaker.) If a player who has lost the H2H match has overcome that loss in subsequent play, he deserves to advance.
ForGG should have advanced over Babyknight, and players with better game scores should advance even if they have lost the H2H match which whom they are tied.
 
|
EDIT:
([21:49:47] <@motbob> please rewrite your post to be angry)
I first thought you wanted a second "tiebreaking match" between those 2 players, which in my opinion would be the best way to decide who gets to advance.
Then again the only 2 kinds of tournaments that could possibly do something like that would be GSL (1 group per day, open end) and online tournaments, seeing that those sets of rules would destroy any schedule any offline tournament would have.
|
Aotearoa39261 Posts
I actually feel the opposite, I can never understand why mapscore should be used over head to head in any instance. Starcraft is not a perfect game and there are many factors which influence the result of a game - blind countering, bad maps and so on. Players can drop a map just due to the random nature of SC2. Matches are Bo3 to eliminate this kind of variance and to encourage the 'better player' to win. Moreover, when players are eliminated they often mentally 'give up' which means that the map scores in the latter series with this player are more likely to be 2-0's - which influences the map score of player who are still in contention. This is yet another source of variation that should be taken into consideration.
Choosing map score over head to head is not inline with this. Map score says that random variation should influence who advances not who was 'the better player' during the series. Since tournaments choose Bo3s over Bo1s, and choose to eliminate as much random variation as possible - they should be choosing H2H every time.
|
Yes, I think that is how you determine the better performance overall.
I don't know any other sport uses H2H over overall game scores.
On March 24 2013 05:48 Plexa wrote:
I actually feel the opposite, I can never understand why mapscore should be used over head to head in any instance. Starcraft is not a perfect game and there are many factors which influence the result of a game - blind countering, bad maps and so on. Players can drop a map just due to the random nature of SC2. Matches are Bo3 to eliminate this kind of variance and to encourage the 'better player' to win. Moreover, when players are eliminated they often mentally 'give up' which means that the map scores in the latter series with this player are more likely to be 2-0's - which influences the map score of player who are still in contention. This is yet another source of variation that should be taken into consideration.
Choosing map score over head to head is not inline with this. Map score says that random variation should influence who advances not who was 'the better player' during the series. Since tournaments choose Bo3s over Bo1s, and choose to eliminate as much random variation as possible - they should be choosing H2H every time.
But we aren't looking for the better player during the series, we are looking for the best players of the group. And because the opponents the tied players played against are the same we can think of the whole group as a BoX where all the same variables as in the H2H match alone are present. Except for players 'giving up' which should be eliminated by a prize money structure that takes this into consideration. I would go as far as say top heavy paying tournaments shouldn't hold early group stages. Also, games against players who have given up can work against your point just as well if you are beaten by a player who only ties with you by playing his last games against players who are already out of the tournament
Although I understand your point of view of reducing variables, in my mind going H2H before match scores would make the other games less important(=more random) which I don't think they necessarily are.
|
motbob
United States12546 Posts
On March 24 2013 05:48 Plexa wrote:
I actually feel the opposite, I can never understand why mapscore should be used over head to head in any instance. Starcraft is not a perfect game and there are many factors which influence the result of a game - blind countering, bad maps and so on. Players can drop a map just due to the random nature of SC2. Matches are Bo3 to eliminate this kind of variance and to encourage the 'better player' to win. Moreover, when players are eliminated they often mentally 'give up' which means that the map scores in the latter series with this player are more likely to be 2-0's - which influences the map score of player who are still in contention. This is yet another source of variation that should be taken into consideration.
Choosing map score over head to head is not inline with this. Map score says that random variation should influence who advances not who was 'the better player' during the series. Since tournaments choose Bo3s over Bo1s, and choose to eliminate as much random variation as possible - they should be choosing H2H every time.
Variance is smoothed out by larger sample sizes. It is true that the individual pairs of matches (for example, ForGG vs AndyPandy compared to BabyKnight vs AndyPandy) that make up game score have higher variance than H2H due to this "giving up" concept. Taken as a whole, though, variance is still higher for H2H than game score.
It is worth noting that in the two instances of a tie in the example group I posted above (Jrecco/Snute and ForGG/BabyKnight), there's no way that an argument could be made that "giving up" affected the result of the game scores. Look at the match results to see what I'm talking about.
|
Yup, you're right. Mapscore > head to head, or at least so I think as well. Statistically it's obvious mapscore is superior. The issue seems to be that people have an, to me, incorrect belief that if you beat someone in head to head, you're the superior player. It feels right, it feels like a direct comparison of skill, but it's not exactly so...
|
Babyknight only needed 7 wins to score 3 wins, while ForGG needed 8 wins - so ForGG was less efficient therefore he shouldn't advance.
Head-to-Head vs Map-Score (or Goals in football) is one of those things where there is no objective better or worse I think, its just a matter of preference whats considered more fair. Even in such a global and established sport like football both of them occure (World Cup uses goal difference, UEFA Euro and Champions League use head-to-head).
|
Hong Kong9153 Posts
I'm going to have to disagree for one simple reason. I'm not convinced that in every case a 2-0 win is an effective indicator of skill than a 2-1 win. As Plexa said, game score can be up to a bit of randomness and, as such, in any group stage that uses a game series instead of a Bo1, I feel that head-to-head is the way to go for first-order tiebreaks.
|
On March 24 2013 06:20 TBO wrote:
Babyknight only needed 7 wins to score 3 wins, while ForGG needed 8 wins - so ForGG was less efficient therefore he shouldn't advance.
That doesn't make sense, haha. You can't use "more wins" to get the same amount of bo3 wins. Each person had to get exactly 6 wins to get 3 won bo3s, the rest of the maps are wins in their losses, losing 1-2 vs 0-2. In this way, ForGG won more maps (Had closer series), and lost fewer maps, giving him a better map score no matter how you spin it.
EDIT: I realise you weren't serious. >_<; *facepalm* Sorry.
|
I don't agree using mapscore simply cause not all players get to play all 3 games in their bo3. Is mapscore really fair if not all players play the same amount of games?
|
your Country52797 Posts
I think all the players should have played 15 games (3rd game even if 2-0). It will at least reduce these cases.
|
On March 24 2013 06:00 motbob wrote:Show nested quote +On March 24 2013 05:48 Plexa wrote:
I actually feel the opposite, I can never understand why mapscore should be used over head to head in any instance. Starcraft is not a perfect game and there are many factors which influence the result of a game - blind countering, bad maps and so on. Players can drop a map just due to the random nature of SC2. Matches are Bo3 to eliminate this kind of variance and to encourage the 'better player' to win. Moreover, when players are eliminated they often mentally 'give up' which means that the map scores in the latter series with this player are more likely to be 2-0's - which influences the map score of player who are still in contention. This is yet another source of variation that should be taken into consideration.
Choosing map score over head to head is not inline with this. Map score says that random variation should influence who advances not who was 'the better player' during the series. Since tournaments choose Bo3s over Bo1s, and choose to eliminate as much random variation as possible - they should be choosing H2H every time. Variance is smoothed out by larger sample sizes. It is true that the individual pairs of matches (for example, ForGG vs AndyPandy compared to BabyKnight vs AndyPandy) that make up game score have higher variance than H2H due to this "giving up" concept. Taken as a whole, though, variance is still higher for H2H than game score. It is worth noting that in the two instances of a tie in the example group I posted above (Jrecco/Snute and ForGG/BabyKnight), there's no way that an argument could be made that "giving up" affected the result of the game scores. Look at the match results to see what I'm talking about.
Considering the "giving up" thing: In our case, ForGG only faced one ex-opponent of Babyknight after their set and the same holds vice versa, so I think that issue does not apply to our specific case.
Talking in general, if "giving up" were an issue, then head to head would favor the player, who has played more games before the in hindsight deciding showdown.
edit:
On March 24 2013 06:29 RaiKageRyu wrote:
I don't agree using mapscore simply cause not all players get to play all 3 games in their bo3. Is mapscore really fair if not all players play the same amount of games?
actually that is a valid point. anybody still remember the pokerstrategy league, where in a group stage everybody had to play everybody in a fixed number of maps?
|
On March 24 2013 06:23 itsjustatank wrote:
I'm going to have to disagree for one simple reason. I'm not convinced that in every case a 2-0 win is an effective indicator of skill than a 2-1 win. As Plexa said, game score can be up to a bit of randomness and, as such, in any group stage that uses a game series instead of a Bo1, I feel that head-to-head is the way to go for first-order tiebreaks.
Agreed. TBO also raises a good point. H2H versus game score really depends on a degree of subjectivity.
In response to motbob's attempt to rebut Plexa's point about randomness: variance is a quantifiable number. Its impact is indeed mitigated by larger sample sizes but the declaration that randomness is greater for h2h than game score because of larger sample sizes is a peculiar one. The inherent randomness in h2h versus game score are both, to my knowledge, not genuinely quantifiable. His claim is that larger sample size reduces the statistical impact of variance and game score has larger sample size--therefore game score is ultimately less random than h2h. However, one cannot categorically conclude that game score is ultimately less random (and thus a better indicator of the better player and/or the player who should advance) simply because it posesses an advantage in a single instance.
|
On March 24 2013 06:29 RaiKageRyu wrote:
I don't agree using mapscore simply cause not all players get to play all 3 games in their bo3. Is mapscore really fair if not all players play the same amount of games?
Yeah, this is a fair point to me. Playing all 3 games no matter the score. I do like that format personally.
|
First tiebreaker should be head to head IMO. It feels right, and map scores can be inflated by players who have no chance of moving on giving up.
|
Hong Kong9153 Posts
On March 24 2013 06:36 Kasaraki wrote:Show nested quote +On March 24 2013 06:29 RaiKageRyu wrote:
I don't agree using mapscore simply cause not all players get to play all 3 games in their bo3. Is mapscore really fair if not all players play the same amount of games? Yeah, this is a fair point to me. Playing all 3 games no matter the score. I do like that format personally.
You still run into the issues of 'meaningless games' that can inflate map score inappropriately. However, if it truly was a format of "play X number of games against each opponent no matter what," instead of BoX, map score would definitely be the first-order tiebreaker to use.
|
On March 24 2013 06:36 Kasaraki wrote:Show nested quote +On March 24 2013 06:29 RaiKageRyu wrote:
I don't agree using mapscore simply cause not all players get to play all 3 games in their bo3. Is mapscore really fair if not all players play the same amount of games? Yeah, this is a fair point to me. Playing all 3 games no matter the score. I do like that format personally.
MLG tried that once at an Arena. I don't remember the exact details but it felt pretty terrible iirc.
|
thinking about it... some of the matchups in Starcraft 2 are way more volatile than others. I don't find the thread right now but I remember someone doing an analysis showing that when a better player played vs a weaker player (using TLPD Elo differences) the chance to win differed vastly depending on the matchup, TvT being the least volatile and PvP being the most volatile. Going by map score would therefore somewhat benefit Terrans and hurt Z and P (all 3 matchups of T were pretty unvolatile)
|
I fully agree, it is a simple case of mathematics. Round robin is about who did better in the group. Beating someone does not mean you did better than them over multiple matches. In any sports league in the world you can beat the #1 and finish last, because points are what counts. Beating a single opponent does not mean a thing, and certainly should not outweigh who did better proven by numbers.
If you want to call upon randomness in mapscore over multiple series, you can do the same for the head to head that was played. It is a non argument. Variance will always exist, and should be negated as much as possible by looking at who did better overall.
|
World Cup uses goal(~"map") score before head-to-head. Different type of game, but still the logic is pretty similar.
In the end, it forces you to play your absolute best in every game, which is good. When you are 1-0, you will still play very seriously the second game - and not just cheese it, because even if you lose, you get a game 3 - no, you will fight for 2-0.
|
I fully agree. If you want h2h to count, go for a knock-out system. Group play imo is about who fares better against a diverse set of opponents: it's not "is my PvZ better than his ZvP", but "is my PvX better than his ZvX" / "am I better against different styles than he is". You can find this in the mapscores but not in the h2h score.
Even more so because having lost the h2h means you did better against the other players. If h2h counts, the match between the tied players is basically doubled: as a normal match, which already has a huge influence on who advance, AND again as to who wins the tie. I don't like it.
|
You would never look at a playoff bracket and say ForGG deserves to advance to the Ro 8 despite BabyKnight beating him in the Ro16 because his map score is 8-4 vs 7-6.
I know group play is obviously a bit different than the playoff portion (where it's 100% H2H score - 150% in the case of extended series...), but the who point of group play is to determine who gets to the playoffs. If I have to make a determination of two tied group players, I'm taking the one that would have advanced if it was a playoff.
|
I think either system is fine as long as the players are very well aware of how it works and which it is, because the two alternatives reward slightly different styles/approaches to the series'.
But if I had to pick a side, then map-score should always taken precedence, for reasons described in the OP. Map score becomes increasingly relevant with the increasing number of matches per series, and the number of players per group.
|
Plexa convinced me, in case of round robin, the players who play last will usually have better map score due to playing already eliminated players. The sample is bigger yes, but it's biased because of this.
So I agree with Plexa, I prefer H2H. Or at least it's close enough to not be as clear cut.
|
Variance is a moot point in the first place, yeah, and I'd tend to agree with Plexa and MrCon on this. I think the fact that eliminated players will no longer have a stake in the tournament is sufficiently advantageous for late players as to make h2h better in my eyes, or at least enough to make the comparison inconclusive.
|
|
|
Since H2H is what determines who wins any tournament we should be able to agree that H2H is the best way to determine who is best. + Show Spoiler +Okay, tournaments only crown champions instead of determining "who is best" but you get the point
Because many of us would rather see our favorite players play more games, have a better chance at making it out of the first stages of play and as a way to seed competitors tournaments have adopted group play (this applies to many traditional sports as well).
Since we've agreed that H2H is the best way to determine our champion it should be how we advance from groups. In a perfect group (of 6) the final result would be 5-0, 4-1, 3-2, 2-3, 1-4, and 0-5. Unfortunately, not all groups work out so simply so something other than H2H needs be used first. I'll use the OP's example. We can look at game score (4-1, 4-1, 3-2, 3-2, 1-4, 0-5) or map score (9-3, 8-4, 8,4 7-6, 2-9, 2-10). Either way, we end up with ties and since we're still under the impression that H2H determines our champion we use it. (I'm a bit too tired to discuss why I think overall score > map score.)
|
It would. JonnyREcco undoubtedly did better than Snute in the group stage.
|
All sports leagues in the world use some equivalent of goal average (that would be map score) over head-to-head, before reaching the playoff stages. Arguing that Starcraft is so special that it should be done the other way is silly.
|
On March 24 2013 22:12 Apom wrote:
All sports leagues in the world use some equivalent of goal average (that would be map score) over head-to-head, before reaching the playoff stages. Arguing that Starcraft is so special that it should be done the other way is silly.
champions league, UEFA EURO and primera division (spain league) all use head-to-head first, same for a lot of other leagues.
|
Hong Kong9153 Posts
On March 24 2013 22:12 Apom wrote:
All sports leagues in the world use some equivalent of goal average (that would be map score) over head-to-head, before reaching the playoff stages. Arguing that Starcraft is so special that it should be done the other way is silly.
A 3-0 victory over an opponent who is bad and wasn't trying shouldn't be counted as 'more skillful' than a 3-2 victory over an opponent who was at an equal or better level.
|
If head-to-head is the tiebreaker, then should overall records even be displayed on liquipedia here in this manner? It just confuses people passing through. But there are only two columns: Bo3s and overall games. So I would say to myself, okay, Bo3s are obviously the goal here, and then overall games must separate those who share same Bo3 records. But then I see ForGG under BK. Now I'm wondering what the hell is going on. If anything, in tourney like this should display 3 columns for the groupstate standings: column 1 is Bo3s; column two is just called "owns tiebreaker" where the player's column only gets a star if they beat a person with the same Bo3 record head to head; the third column being overall record, which is just something people are overall curious and should be provided, and is useful in case we have a 3-way tie, each player beating the other (A>B>C>A) and THEN you refer to overall records to separate those people. And if those match obviously we need some round robin tiebreaker with just those players. That way it follows logically: Bo3s --> head-to-head --> Overall.
|
I personally don't care which tiebreak method is used if it is communicated clearly. I remember a tournament (IEM? can't remember) where even the commentators gave out wrong information during the Ret vs Nightend game. Thus, I think it's rather pointless arguing which method is better - what we should be talking about is tournaments making it absolutely clear which method is used.
My personal preference is:
1. Map score difference.
2. Head-to-head.
I don't like "maps won" because I feel that 2 scores of 2-1 and 1-2 should be the same as 2-0 and 0-2. In football (soccer), "goals scored" is legitimate because it encourages attacking play.
In the case of a 3-way (or more) tie, my preference is:
1. Map score difference of the entire pool.
2. Head-to-head of the 3 players.
3. Head-to-head map score of the 3 players.
|
Hong Kong9153 Posts
On March 25 2013 02:33 MountainDewJunkie wrote:
If head-to-head is the tiebreaker, then should overall records even be displayed on liquipedia here in this manner? It just confuses people passing through. But there are only two columns: Bo3s and overall games. So I would say to myself, okay, Bo3s are obviously the goal here, and then overall games must separate those who share same Bo3 records. But then I see ForGG under BK. Now I'm wondering what the hell is going on. If anything, in tourney like this should display 3 columns for the groupstate standings: column 1 is Bo3s; column two is just called "owns tiebreaker" where the player's column only gets a star if they beat a person with the same Bo3 record head to head; the third column being overall record, which is just something people are overall curious and should be provided, and is useful in case we have a 3-way tie, each player beating the other (A>B>C>A) and THEN you refer to overall records to separate those people. And if those match obviously we need some round robin tiebreaker with just those players. That way it follows logically: Bo3s --> head-to-head --> Overall.
Liquipedia states the tournament rules for group stages (when available) before showing results.
|
A group stage with best of 3s in Starcraft is only descriptively a round robin. Really, it's a way of compressing a bracket. The way "round robin" works best is in a game like chess where the only thing that matters is your score after a bunch of games. There is a reason that chess tournaments don't have players play a best-of in the round robin stage, and time is only one of those reasons. Chess does not suffer from variance the same way Starcraft does. Starcraft requires this best-of series.
The goal of a group stage is to compress a bracket and do your seeding for the later brackets at the same time. The winner of a regular bracket is the person who beat the player who beat the player who beat the player. Deciding a group stage based on head-to-head is more consistent in that way. It efficiently says "If both of these tied players advanced and played in our regular bracket, this player wins. So let's skip a step." That's the nature of an elimination tournament, and that is how the group stage should operate in an elimination tournament.
Using map score as the deciding factor is contradictory to using a best-of series in the first place. The whole point of a best-of series is to find who is better in a best-of series. Starcraft is ALL ABOUT making adjustments and tricking your opponent and being intelligent. A series should be taken as binary, either 1 or 0. A player who lost a first game of a series, adjusts, and comes back 2-1 should not be penalized for not all-inning in game 1. Map score decisions empower ladder play, not tournament play. I don't want to watch ladder play in MLG.
IMO, the only time a group stage should be decided based on map score is if that is the premise of the tournament. Qualifiers are a good example, and I think they should be decided on map score.
Additionally as a previous poster pointed out, deciding based on map score has problems because most tournaments do not play out a series if it starts 2-0. The times that it has been tried have poor results. It's bad for viewers, it's bad for players.
|
On March 25 2013 02:03 itsjustatank wrote:Show nested quote +On March 24 2013 22:12 Apom wrote:
All sports leagues in the world use some equivalent of goal average (that would be map score) over head-to-head, before reaching the playoff stages. Arguing that Starcraft is so special that it should be done the other way is silly. A 3-0 victory over an opponent who is bad and wasn't trying shouldn't be counted as 'more skillful' than a 3-2 victory over an opponent who was at an equal or better level.
It is round robin so you would play both opponents, and so would the guy you are tied with. A victory 3-0 victory over a bad opponent doesn't count more than a 3-2 victory over a better one. It's all about what the other tied player did against those opponents.
|
Hong Kong9153 Posts
On March 25 2013 05:39 Liquid`Nazgul wrote:Show nested quote +On March 25 2013 02:03 itsjustatank wrote:On March 24 2013 22:12 Apom wrote:
All sports leagues in the world use some equivalent of goal average (that would be map score) over head-to-head, before reaching the playoff stages. Arguing that Starcraft is so special that it should be done the other way is silly. A 3-0 victory over an opponent who is bad and wasn't trying shouldn't be counted as 'more skillful' than a 3-2 victory over an opponent who was at an equal or better level. It is round robin so you would play both opponents, and so would the guy you are tied with. A victory 3-0 victory over a bad opponent doesn't count more than a 3-2 victory over a better one. It's all about what the other tied player did against those opponents.
I understand your point, but I still feel that when and where players are within a tournament can significantly impact how they choose to play. Of course, these things depend on the individual sportsmanship and work ethic of the players involved, but there have been instances in the past where a player's cumulative progress within a round-robin group stage has directly affected how they play. Other things being equal, a player who gives another player a 3-2 victory in the first stage of round-robin might give up and give another player a 3-0 victory in the last few rounds if they are already eliminated and are thus playing a 'meaningless game.' Having the first-order tiebreak procedure be map score differential is subject to this variance.
The variance increases significantly when you start factoring in the depth of strategies available to players against other players based on map and race match-up to the point where I do not feel that, even assuming skill levels are the same, two different players can guarantee a 3-0 victory over a single 'bad' player. Intangibles begin to get in the way.
Unless the map score differential isn't really a differential and is actually only map wins, a 3-0 victory would always be counted more than a 3-2 victory (+3 versus +1) regardless of the circumstances. Having it be map wins only (+3 and +3) would lead to lots of ties and head-to-head would be inevitable as the second-order tiebreak.
In addition, I feel that I really prefer tiebreak procedures that do a minimum of intervention in terms of interpreting a player's relative skill in favor of looking at results. I actually feel that the ultimate tiebreak method should be a playoff game between the tied players. This is rarely possible, however, because of time constraints. Head-to-head looks at empirical data and shows that at some point in the round-robin, the tied players did in fact meet, and only one of them beat the other. Head-to-head, at this point, becomes the playoff game between the two, and that's why I feel that it is legitimate.
Although, I do also think that if there is no BoX procedure, and the group stage is, in fact, conducted as "play this many games against each player no matter what," map scores are the only way to go.
|
On March 25 2013 03:51 itsjustatank wrote:Show nested quote +On March 25 2013 02:33 MountainDewJunkie wrote:
If head-to-head is the tiebreaker, then should overall records even be displayed on liquipedia here in this manner? It just confuses people passing through. But there are only two columns: Bo3s and overall games. So I would say to myself, okay, Bo3s are obviously the goal here, and then overall games must separate those who share same Bo3 records. But then I see ForGG under BK. Now I'm wondering what the hell is going on. If anything, in tourney like this should display 3 columns for the groupstate standings: column 1 is Bo3s; column two is just called "owns tiebreaker" where the player's column only gets a star if they beat a person with the same Bo3 record head to head; the third column being overall record, which is just something people are overall curious and should be provided, and is useful in case we have a 3-way tie, each player beating the other (A>B>C>A) and THEN you refer to overall records to separate those people. And if those match obviously we need some round robin tiebreaker with just those players. That way it follows logically: Bo3s --> head-to-head --> Overall. Liquipedia states the tournament rules for group stages (when available) before showing results.
Sure, but it's still a little bit clunky to display the results that way and just a little bit counterintuitive. The liquipedia guys put in enough effort as it were, though, so I can understand why they'd not want to spend the extra effort to address comments from non-contributing people when the current layout is ostensibly sufficient.
|
Counter-example. Not entirely a different scenario.
Two chess players score 5 points in a tournament of 8 games. Draws are worth 0.5, wins are worth 1. Losses are worth nothing. Each player plays black four times and white four times.
Player A scored 5-3-0 Win/loss/draw
Player B scored 4-1-2 Win/loss/draw
Player B beat Player A in their game. However, Player A scored 2 points as black and Player B only scored 1.5, and Player B was white in his game against Player A.
Should Player A advance because he performed better as black? Should Player B advance because he beat Player A? Should Player A advance because he won more games and Player B drew for 1 of his points? Making the decision based on head-to-head simplifies all of these issues. Weighing different match wins differently is a bad system.
The fact is, in a group stage, players are at different stages within the tournament even within the group stage. The player who's in last place, no matter how hard you try to believe he is, is not playing at his full strength and does not have the same amount of enthusiasm in his play. He's out of the tournament, he just still has to play his games. If Player A goes 2-1 against Player C in the first round of the tournament, is that less impressive than if 4 hours later and Player C is 0-4 and demoralized, Player B goes 2-0 against him?
|
On March 25 2013 06:41 -Kaiser- wrote:+ Show Spoiler +Counter-example. Not entirely a different scenario.
Two chess players score 5 points in a tournament of 8 games. Draws are worth 0.5, wins are worth 1. Losses are worth nothing. Each player plays black four times and white four times.
Player A scored 5-3-0 Win/loss/draw
Player B scored 4-1-2 Win/loss/draw
Player B beat Player A in their game. However, Player A scored 2 points as black and Player B only scored 1.5, and Player B was white in his game against Player A.
Should Player A advance because he performed better as black? Should Player B advance because he beat Player A? Should Player A advance because he won more games and Player B drew for 1 of his points? Making the decision based on head-to-head simplifies all of these issues. Weighing different match wins differently is a bad system.
The fact is, in a group stage, players are at different stages within the tournament even within the group stage. The player who's in last place, no matter how hard you try to believe he is, is not playing at his full strength and does not have the same amount of enthusiasm in his play. He's out of the tournament, he just still has to play his games. If Player A goes 2-1 against Player C in the first round of the tournament, is that less impressive than if 4 hours later and Player C is 0-4 and demoralized, Player B goes 2-0 against him?
I think this is a really interesting and important thing to consider, but is also entirely impossible to judge in a way that is mathematically sound and fair I think.
|
Katowice25012 Posts
This is a good way of looking at it, I agree map score makes a lot more sense here in nearly all cases.
|
I dont agree. It makes no sense for the map score to outweight the head to head result, because what it matters in case of a tie is who is the better of the two players, not who is the better of the all the players in the group. The most objective way to determine this is the H2H result. Not to mention that map score gives a lot of room of fixing - for example if two players of the same team are playing in the last round. H2H score is used in almost all sports tournament to determine the winner in case of a tie.
|
Why argue over technical decisions when everyone can argue it is the tournament's responsibilities to make sure every player (and the public) are aware of the rules. i48 organizers didn't and it's a damn shame. With all the talk about mindset and how head-to-head is better because eliminated players have less of an influence, anyone who watch the last game between JonnyREcco and ForGG could sense that the latter thought he was through.
|
On March 24 2013 05:38 motbob wrote:Game score should always be used as a tiebreaker before head-to-head. Already I can feel the angry posts being written. Let me explain why I think what I think. In this group, ForGG and Babyknight were tied in match score at 3-2. Because of that tie, a tiebreaker had to be used. There were two pieces of information that the tournament organizers could have drawn on to make their decision on who to advance. The first piece of information is that Babyknight defeated ForGG 2-1. This implied that BabyKnight should have advanced over ForGG. The second piece of information was that ForGG's record against common opponents of ForGG/BabyKnight was 7-2. Compare that record to the 5-5 record of Babyknight. The final game score of the group, derived mostly from the record of ForGG and BabyKnight against common opponents, implied that ForGG should have advanced. So there are two conflicting pieces of information here. Depending on which piece of information is deemed more important, one of these two players should advance. I think that the game score is a more important indicator of who the better players is because it draws on a larger sample size. Variance exists in SC2. The best player does not always win the game, or the series. As with all phenomena with variance involved, it's always better to increase your sample size to see what is really happening. Tournaments choose the H2H tiebreaker as the one that takes priority because it "feels right." It feels just that a tie between two players should be decided by the games played between those two players. But if Player A has beaten Player B, it's very difficult for Player B to overcome that loss and do better than Player A in game score. ForGG had to obtain at least a +3 advantage in game score over BabyKnight against common opponents to overtake him. If ForGG had lost 0-2, he would have had to obtain a +5 advantage! (In this way, you can see that the H2H tiebreaker is actually included in the game score tiebreaker.) If a player who has lose the H2H match has overcome that loss in subsequent play, he deserves to advance. ForGG should have advanced over Babyknight, and players with better game scores should advance even if they have lost the H2H match which whom they are tied.
Good post explaining the concept.
The real key here - the real essence of the problem - is that given these two possibilities (H2H match or group score), you want to choose the best one. Neither method is a bad method - the issue is that one is better than the other. The fundamental problem with choosing H2H is that you are choosing to throw out so much additional pertinent information. Mathematically, it's very analogous to using an ELO-like system vs. using a "true" ladder. A true ladder will result in a new top player every time someone beats the top player, but it will give a poor indication (a single moment in time) of how close or far the top few players' performance is from each other.
To look at it the opposite way, if H2H was the best tiebreaker, then round robins or any group format except elimination are inferior. If we could get more information about who the better player is from H2H, then having every group stage mechanism be a H2H elimination (loser leaves the tournament) would be the best possible system. If we take this logic even further, we might even could argue that the best thing to do to determine player performance would be to have an entire tournament of nothing but single elimination (1-0). The fundamental issue is that there is more information and less variance across more matches.
A lot of people are bringing up the point that random circumstance can lead to one player being artificially favored by a group score. That's definitely 100% true. But does random circumstance have more variance in one match or over many? There are a large number of factors that can affect a single match, and singling out one of them is intellectually dishonest for the purpose of determining a system to find the best player. We have to trust in the group system because our only other option is to trust in an overall inferior system.
|
On March 24 2013 05:48 Plexa wrote:
I actually feel the opposite, I can never understand why mapscore should be used over head to head in any instance. Starcraft is not a perfect game and there are many factors which influence the result of a game - blind countering, bad maps and so on. Players can drop a map just due to the random nature of SC2. Matches are Bo3 to eliminate this kind of variance and to encourage the 'better player' to win. Moreover, when players are eliminated they often mentally 'give up' which means that the map scores in the latter series with this player are more likely to be 2-0's - which influences the map score of player who are still in contention. This is yet another source of variation that should be taken into consideration.
Choosing map score over head to head is not inline with this. Map score says that random variation should influence who advances not who was 'the better player' during the series. Since tournaments choose Bo3s over Bo1s, and choose to eliminate as much random variation as possible - they should be choosing H2H every time.
I think you counter your own argument. H2H is 3 games, overall game score was potentially 10+. How is that not significantly better at eliminating the "randomness" you're talking about.
|
|
|
|
|
|
EPT
