|
MLG Extended Series... Much Bitching is Le Ensued.
MLG's "extended series" format, where a Bo3 becomes a Bo7 with previous results counted, guarantees the advancing player has a winning record head-to-head. The other system that people keep proposing is the "win two Bo3" system, where the previous loser has to win 2, the previous winner only 1, Bo3. By counting in terms of series, it avoids the possibility of the "losing head to head advances" problem. In terms of games, tie records are possible, as in a 2-1, 0-2, 2-1 advance (or 0-2, 2-1, 2-1), but the series is made the main thing so that doesn't really matter.
Anyway, I thought it would be interesting to run some numbers and see how these two systems compare.
+ Show Spoiler [W/L Table] +Extended from 0-2 W W W W L W L L W W L L L W W W L L L L W W W W L L L L
Extended from 1-2
W W W L W L L W W L L L W W W L L L
2 Bo3 System
W W W W L W L L W W L L L W W W L W L W W L L L W W W W L W L L W W L L L L
Assume the two players are equally good. Then in the different situations, the loser of the first series has:
Extended Series (0-2): 18.75% chance to win 2 Bo3 System: 25.00% chance to win Extended Series (1-2): 31.25% chance to win
In a regular DE tournament, the loser of the first match would still have a 50% chance to win, so any of these situations gives the first winner an advantage. That could be good or bad: I tend to think it's a good thing overall, especially since it's a known part of the format. I prefer the MLG extended to the 2 Bo3 because it rewards a "better" loss in the first place. Series wins are good, but the point of the series is to find out who's the better player.
Other Thoughts
MLG could juggle the brackets so that rematches simply don't happen in the open bracket. It's not that hard. There are 4 OBs and 8 LBs: so split the OBs (as it were) by finalists to create 8 brackets, then rotate the losers. For example, for the first LB round OB 1.1 losers -> LB 1, then the next round OB 1.1 losers -> LB2, etc. There are then 8 brackets and only 5 seeding-downs so the rotation never goes all the way around and there are no Open rematches and everyone's happy.
The championship bracket is a different story, naturally. Even with rotation there are 6 seeds from only 4 pools so you're almost guaranteed to get a rematch at some point. It's sort of surprising there aren't more: Anaheim had only 2; Columbus had 6; Dallas had 4. But if pool were made really and truly only for seeding, then extended series rules shouldn't apply to the championship bracket.
|
... There is no discussion between 2 bo3's or extended series... The discussion is about extended series or a fresh start at 0-0 when facing each other in the loser brackets.
|
... A fresh start in mlg would be two best of 3's ZpuX. The first, when you first meet, the second in lose losers bracket. Where as extended series is akin to a best of seven, starting with the score from your last series.
|
On August 02 2011 00:53 ZpuX wrote: ... There is no discussion between 2 bo3's or extended series... The discussion is about extended series or a fresh start at 0-0 when facing each other in the loser brackets.
I'm guessing it's talking about the finals, where this actually does come up.
Edit: Looks like OP was just confused - the "2 bo3" system is exclusively being talked about for the finals, which is the only real debate... everyone outside of MLG agrees that extended series before the finals is terribad.
|
On August 02 2011 01:33 Uncultured wrote: ... A fresh start in mlg would be two best of 3's ZpuX. The first, when you first meet, the second in lose losers bracket. Where as extended series is akin to a best of seven, starting with the score from your last series. *edit* re read the OP, he is talking about winning "a bo3 consisting of bo3's". When you face each other in the LB again, the winner from WB only has to win one bo3 while the other player has to win 2 bo3's.
No, a fresh start would be as if you never played each other before. Why should someone have a harder road to the finals depending on who he got to face in the winner bracket?
Two players get knocked down to LB in the first round.
In the third round in LB, player A has to play an extended series, Which means player A has to win either 3 or 4 games to advance while player B only has to win 2 games. Does this seem fair?
Another point, why isn't this applied to every single round in the LB then? I mean it should make no difference who you lose to in WB. Playing from the LB will give you a disadvantage once you reach the final which is what should count.
|
Like ZpuX said there isn't really much of a mathematical controversy between Extended Series (ES) and "first to 2 Bo3" systems.
Both are fine when the LB and WB winners meet in the finals and both are outrageous when two players meet in the LB. Instead the "2 Bo3" vs ES discussion is of a subjective and much more debatable nature. + Show Spoiler +- People being against ES argues that it lessens suspense as the chance to come close to or to a final deciding game is much lower. (This is why many double-elimination tournaments eliminates the rule) - People for ES assert the added fairness, which your above example is proof of.
Which system is is more fair in the finals isn't too much to debate about either as it's pretty easy to deduct which one is correct for which situation: + Show Spoiler +It's always fair to use "first to 2 Bo3" with WB winner being up a Bo3 at the start. The extended series is also provably even more fair if the players has met before.
Note that the "losing head to head advances"-dilemma is illusionary for every instance except in the LB winner vs WB winner finals as it is based on false assumptions.
|
|
|
|