|
On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter.
The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage".
The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far.
|
I've actually thought of this, and the main reason I don't think it's good is that you can't make it an extended best of 5 because then one player is only 1 game away from winning, and you can't make it an extended best of 7 because then one player may have to win 4 games in a row, or it might take 5 games to finish the best of 7.
I think the dual tournament format is fair enough as is.
|
Absolutely. The excuse "the loser's bracket player had to play more games so it's fair" doesn't even apply here.
|
How can you judge who the better player is? The better player is who wins. Regardless of anything else.
|
Actually, a round robin format would be better. It's only one more Bo3 and there won't be ties because you can break them by set score.
|
On April 15 2012 08:19 Fubi wrote:The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far.
The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example:
Player A, B, C and D are playing in the groups
C won against D and B
D lost against C and A
A lost against B but won against D
B won against A but lost against C
So the group standings are now:
C 2-0
A 1-1
B 1-1
D 0-2
At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better.
There are a few ways of doing this:
The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content.
The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group.
The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3.
From the player standpoint this should be the favored option because it rewards the overall better player.
As a viewer you get the 2-5 more games which means likely more content.
|
The point of a tournament is to determine who wins, not who is the best player. If you want to take into account previous series between the two players why don't you just forget having them play altogether and compare their records against each other to see who advances?
|
No get that crap away from all tournaments.
In something like GSL just 1win advantage alone would be incredible huge and more unfair to the other player then otherwise.
Can you imagine a season final where 1 of the finalists starts with 2 sets more since he won in group stage?
It basically becomes a bo3 vs a bo7, great finals you got there, ruined before it began.
|
On April 15 2012 09:28 Bazinga wrote:Show nested quote +On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content.
False analogy: extended series would also result if, say, B played against D, instead of A.
|
On April 15 2012 09:32 phantaxx wrote:
The point of a tournament is to determine who wins, not who is the best player. If you want to take into account previous series between the two players why don't you just forget having them play altogether and compare their records against each other to see who advances?
The point of a tournament is to determine who is the best player in the tournament, if it was only about winning regardless of the players skill you could go back to cointossing.
|
On April 15 2012 09:28 Bazinga wrote:Show nested quote +On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content.
You don't discard anything. The fact that both players are at 1-1 means that they are both exactly even in the group, period. Therefore, there is zero reason to give anyone of them an extra advantage at that point. Again, it's irrelevant who they have won or lost against thus far in the group. They both lost and won once, so they are facing each other again to determine who will advance with a 2-1 score.
|
On April 15 2012 09:34 Shiori wrote:Show nested quote +On April 15 2012 09:28 Bazinga wrote:On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content. False analogy: extended series would also result if, say, B played against D, instead of A.
No it would not because there was no game of B vs D before
|
On April 15 2012 09:35 Fubi wrote:Show nested quote +On April 15 2012 09:28 Bazinga wrote:On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content. You don't discard anything. The fact that both players are at 1-1 means that they are both exactly even in the group, period. Therefore, there is zero reason to give anyone of them an extra advantage at that point. Again, it's irrelevant who they have won or lost against thus far in the group. They both lost and won once, so they are facing each other again to determine who will advance with a 2-1 score.
Saying they are both even at 1-1 means that you are discarding their first encounter because at that point you say that both players roads to the 1-1 was equal while it wasn't. B had to play the harder opponent because B won A vs B Part 1.
|
There's no need to mix past matches into the result of a current match. If that was the case, we'd just go by the players history against one another and include EVERY past match to decide "who's better"
When one of the players is 2-1, and the other is 1-2, I don't find an extended series necessary. The players are preparing for a Bo3 series, and this would change it into a possible Bo5. We won't see players doing any creative strategies when they're up 1-0, because they'd be worried it'll come back to haunt them later that night.
|
I dont get why you guys are so focussed on A vs B rematch having to be fair the second time for the first winner, even if A was the winner the first time around he doesnt need to be the better player overall.
Its so clean currently, you perform bo3's in a group of 4, the first 2 to make it to 2 wins are the more successfull players, simply. Your just counting Bo3's wins.
|
On April 15 2012 09:41 Bazinga wrote:Show nested quote +On April 15 2012 09:35 Fubi wrote:On April 15 2012 09:28 Bazinga wrote:On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content. You don't discard anything. The fact that both players are at 1-1 means that they are both exactly even in the group, period. Therefore, there is zero reason to give anyone of them an extra advantage at that point. Again, it's irrelevant who they have won or lost against thus far in the group. They both lost and won once, so they are facing each other again to determine who will advance with a 2-1 score. Saying they are both even at 1-1 means that you are discarding their first encounter because at that point you say that both players roads to the 1-1 was equal while it wasn't. B had to play the harder opponent because B won A vs B Part 1.
B already got an advantage tough from winning the A vs B part 1. He only needs to win 1 more match and even can lose one before his tournament ends.
A on the other end has play with his back against the wall the whole time and 1 mistake can cost him the tournament.
So A already got a disadvantage in group stage, why should he get another one in another part of the tournament?
It is not just about A and B, there is a whole tournament going on.
|
On April 15 2012 09:41 Bazinga wrote:Show nested quote +On April 15 2012 09:35 Fubi wrote:On April 15 2012 09:28 Bazinga wrote:On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content. You don't discard anything. The fact that both players are at 1-1 means that they are both exactly even in the group, period. Therefore, there is zero reason to give anyone of them an extra advantage at that point. Again, it's irrelevant who they have won or lost against thus far in the group. They both lost and won once, so they are facing each other again to determine who will advance with a 2-1 score. Saying they are both even at 1-1 means that you are discarding their first encounter because at that point you say that both players roads to the 1-1 was equal while it wasn't. B had to play the harder opponent because B won A vs B Part 1.
The solution to this problem would be a traditional group system and not extended series.
|
On April 15 2012 09:41 Bazinga wrote:Show nested quote +On April 15 2012 09:35 Fubi wrote:On April 15 2012 09:28 Bazinga wrote:On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content. You don't discard anything. The fact that both players are at 1-1 means that they are both exactly even in the group, period. Therefore, there is zero reason to give anyone of them an extra advantage at that point. Again, it's irrelevant who they have won or lost against thus far in the group. They both lost and won once, so they are facing each other again to determine who will advance with a 2-1 score. Saying they are both even at 1-1 means that you are discarding their first encounter because at that point you say that both players roads to the 1-1 was equal while it wasn't. B had to play the harder opponent because B won A vs B Part 1.
No, that reasoning is flawed, and for few reasons
1) "harder" opponent is subjective, and it isn't always true. The recent Ryung/Naniwa/MVP/Puzzle group is good example of that. Naniwa won first set against Puzzle, and had to face Ryung as oppose to MVP. Those games clearly tells you that Ryung is definitely not a "harder" opponent than MVP.
2) You can't use subjective advantage to balance out a system: "harder opponent" and "gets more tired from playing more games" are not absolute advantages. You must balance out a system systematically.
Using your example, we're not discarding their first encounter: the fact that B won the first encounter gave him an absolute systematic advantage: A chance to advance the group; as in his next win allows him to advance, and at worst if he loses, it will simply puts him back to where he was. Now by contrast, the Loser is punished because he faces elimination if he loses, and even if he wins, he doesn't get to advance.
Therefore, their first encounter is already taken into consideration, it wasn't discarded when they faced again: The fact that they are now both 1-1, means it counted the first game, and they both have won once and lost once in the group.
*edit: TLDR version: winning the first set is always better than losing, regardless of your opponent. Because at worst, even if you lose, you'll just be put back to where you were; while the loser actually has to win his next set or face elimination.
|
Harder opponent? lol please keep that shitty categorization away from GSL.
|
On April 15 2012 09:47 Assirra wrote:Show nested quote +On April 15 2012 09:41 Bazinga wrote:On April 15 2012 09:35 Fubi wrote:On April 15 2012 09:28 Bazinga wrote:On April 15 2012 08:19 Fubi wrote:On April 15 2012 07:23 Bazinga wrote:I feel like a bo7 is better than two bo3s and because of that that extended series are good. Here are a blog entry and a thread explaining the merit of the extended series: The mathematics of Extended Series (Bo7) vs 2 Bo3sStatistical Analysis of Extended SeriesThe only reason not to use an extended series would be because it doesn't feel right from a traditional standpoint, but i do not know why that should even matter. The math is wrong. It did not consider each player's performance of their other games in the group (or the whole tournament for MLG's case); but that is the whole point of having a group stage, to find the best players in that GROUP (by extension, best players in the tournament for MLG). Therefore, you can't simply consider their head to head and be like "oh this player did better against this guy in their last meeting, lets give him an advantage". The better player of the two isn't simply which one of the two beat eachother more, it's which one of the two performed better overall in the group/tournament thus far. The point is that it considers each players performance in the groups till that point, discarding the first match between both players does not. I hope i can explain it by the following example: Player A, B, C and D are playing in the groups C won against D and BD lost against C and AA lost against B but won against DB won against A but lost against CSo the group standings are now: C 2-0 A 1-1 B 1-1 D 0-2 At this point the only players that are left to be evaluated are A and B and in order to do that you have to determine which one of both players is better. There are a few ways of doing this: The first option would be to say that B already won against A so B should advance. This is the worst option because it results in less content. The second option would be discarding the first evaluation of A vs B and starting anew, which can be done and is a valid way of determining the better of both players. The problem here is that you are not rewarding player B for his win in the first game, but you are instead punishing him because he had to play C, the tougher opponent, in his 2nd match, while A had to play D, the worst player in this group. The third option would be using an extended series format to determine the order of Player A and B more accurately than with another bo3. From the player standpoint this should be the favored option because it rewards the overall better player. As a viewer you get the 2-5 more games which means likely more content. You don't discard anything. The fact that both players are at 1-1 means that they are both exactly even in the group, period. Therefore, there is zero reason to give anyone of them an extra advantage at that point. Again, it's irrelevant who they have won or lost against thus far in the group. They both lost and won once, so they are facing each other again to determine who will advance with a 2-1 score. Saying they are both even at 1-1 means that you are discarding their first encounter because at that point you say that both players roads to the 1-1 was equal while it wasn't. B had to play the harder opponent because B won A vs B Part 1. B already got an advantage tough from winning the A vs B part 1. He only needs to win 1 more match and even can lose one before his tournament ends. A on the other end has play with his back against the wall the whole time and 1 mistake can cost him the tournament. So A already got a disadvantage in group stage, why should he get another one in another part of the tournament? It is not just about A and B, there is a whole tournament going on.
Part 2 of A vs B in this case would still be the group stage in the GSL, which is the tournament we are talking about.
That disadvantage disappears when both players meet again, because now both of them have their backs against the wall and you could argue that Player A already got his mind set from his first elimination game while Player B has to deal with a new situation AND Player B still had the tougher opponents in the given groupstage.
Edit: Harder opponent means the better performing opponent, matchups do influence this part quite a bit sure, but taking this into consideration is impossible because you would have to look at every player seperately.
|
|
|
|
|
|
EPT
