GomTV3 MSL Bracket Contest - Page 7
Forum Index > BW General |
Storchen
Sweden4385 Posts
| ||
Aepplet
Sweden2908 Posts
| ||
Hot_Bid
Braavos36362 Posts
On October 19 2007 00:00 Storchen wrote: lol Hot_Bid I picked Mania, Free, rA, Savior and Xellos?! checking, you may be right, it's hard going through 150+ entrants sorry | ||
Storchen
Sweden4385 Posts
On October 19 2007 00:23 Hot_Bid wrote: checking, you may be right, it's hard going through 150+ entrants sorry np. I just want my 15 minutes ^^ picked light, oov and bisu aswell.. So I guess im out next round. | ||
Hot_Bid
Braavos36362 Posts
edit: don't give up hope, oov/light/bisu could all come back^^ | ||
Storchen
Sweden4385 Posts
| ||
Hot_Bid
Braavos36362 Posts
you two are the only ones left! both of you need Oov, Bisu, and Light to all come back from down 0-1 though | ||
Vomified
Bulgaria68 Posts
| ||
quasi -QS-
United States109 Posts
| ||
ToT)SiLeNcE(
Germany590 Posts
If we assume that each player has a 50% chance to win, the odds of getting all 5 bets right are 1/32 => participants expected to make it through after 5 bets : 196 / 32 = 6. Ok now only 2 made it, but you get my point. | ||
Storchen
Sweden4385 Posts
On October 19 2007 00:48 ToT)SiLeNcE( wrote: I don't quite get how everyone is so surprised by how this bracket contest turned out. If we assume that each player has a 50% chance to win, the odds of getting all 5 bets right are 1/32 => participants expected to make it through after 5 bets : 196 / 32 = 6. Ok now only 2 made it, but you get my point. yeah well each player dont have 50% chance to win. For example considering his shape lately I'd say Free had more like 70 or 80% chance of winning over Lomo. So its really not that simple. | ||
ItchReliever
2489 Posts
| ||
ViRii
United States826 Posts
Gogo Kwanro! | ||
ToT)SiLeNcE(
Germany590 Posts
On October 19 2007 00:54 Storchen wrote: yeah well each player dont have 50% chance to win. For example considering his shape lately I'd say Free had more like 70 or 80% chance of winning over Lomo. So its really not that simple. Yes I agree, but a players chance to win is generally a LOT closer to 50/50 then people realize. Progamers are very very close to one another nowadays. An objective observer cannot fail to notice the amount of "upsets" in modern pro starcraft. But a lot of people get blinded by their fanship. Basically the current proleague liquibet is more of a lottery than anything else. Of course you can always say "oh he predicted it right, he must have awesome prediction skills" _AFTERWARDS_ but statistically someone HAS to make the right predictions. I really think statistics are something that the current progaming community fails to be aware of. I'm sure that, if someone went through the work to do some illustrative plots of statistical expectation vs. results for pro games and took a large enough number of games into account it would totally back up what I'm saying. Also, in the RO16 there were quite a few matches in which there was no clear favourite. | ||
zer0das
United States8519 Posts
| ||
NonY
8716 Posts
On October 19 2007 04:49 ToT)SiLeNcE( wrote: Yes I agree, but a players chance to win is generally a LOT closer to 50/50 then people realize. Progamers are very very close to one another nowadays. An objective observer cannot fail to notice the amount of "upsets" in modern pro starcraft. But a lot of people get blinded by their fanship. Basically the current proleague liquibet is more of a lottery than anything else. Of course you can always say "oh he predicted it right, he must have awesome prediction skills" _AFTERWARDS_ but statistically someone HAS to make the right predictions. I really think statistics are something that the current progaming community fails to be aware of. I'm sure that, if someone went through the work to do some illustrative plots of statistical expectation vs. results for pro games and took a large enough number of games into account it would totally back up what I'm saying. Also, in the RO16 there were quite a few matches in which there was no clear favourite. Statistics are not enough to make predictions. There are other things to take into account like how many leagues are the players currently playing in, do the players have hot streaks or cold streaks, what practice partners did he have before the match, what are the styles of the players, what styles does the map encourage, etc. This info combined with a good look at relevant statistics can give enough info for good predictions. We can't ever know the actual percent chance a player had to win a match because we can't run the match over and over again. Looking at a bunch of different previous matches in order to predict the result of one match will always be significantly inaccurate, although it can be somewhat helpful as I said above. Top liquibetters get 60-65% of the points. And I doubt anyone even researches matches as much as they possibly could. If there was big money involved, I'm sure people could hit 70%. But even hitting 60-65% would be impossible to maintain if players rarely ever have a >60% chance of winning. Anyway, my main point is that predicting matches definitely isn't a useless exercise. There are still enough lopsided progaming matches and enough info available to predict them. | ||
isbunk
Sweden1017 Posts
BIIIIIIIIIIIIIIIIIIIIIIIIIIIIZU, VAAAZZZZZIN, PUUUZAN!!1111 | ||
Hot_Bid
Braavos36362 Posts
thanks for playing and good luck next time | ||
thedeadhaji
39472 Posts
| ||
PePe QuiCoSE
Argentina1204 Posts
well, i got decently close with my wild predicitions (damn oov too). How many other people got only 2 wrong? | ||
| ||