| Forum Index > SC2 General |
|
c0ldfusion
United States8293 Posts
On May 07 2015 22:11 Luolis wrote: Please go away with your witch hunt and leave actually intelligent people to deal with this stuff  you can't be that naive... | ||
|
Estancia
Korea (South)335 Posts
On May 07 2015 23:28 Penev wrote: He doesn't mention "Korean" anywhere? Pinnacle voided 5 bets (and another site at least one IIRC) Seems to be a confusion over this line. It refers to people who actually pay brokers to set up the match fixing. | ||
|
The_Masked_Shrimp
425 Posts
On May 07 2015 21:59 Jarree wrote: Was it 5 or 6 voided bets? I might have been wrong thinking 6, but anyway: 0,5^5 = 3,125% 0,5^6 = 1,5625% That's assuming 50-50 odds, while Creator for example was pretty big underdog. How is that math incorrect? wtf did I just read? O_o I'll spoiler my angry statistics defender part + Show Spoiler + I am not statistics professor but I am a PhD student in statistics and I did TA for introductory statistical physics for 2nd year students. You want to know how is your math incoorect? I'll give you that it's true that 0,5^5 = 3,125 and 0,5^6 = 1,5625. That's about all that is correct about it. Anything you might conclude from it is otherwise ass pulled bullshit. When medical tests are done to determine the chance of a vaccine to work or not, that is also bullshit. But at least that is not ass-pulled bullshit and people doing such "statistics" are aware of their enormous assumptions and the flaws of doing numerical analysis based on this. Statistics are only good at predicting the behaviour of deterministic microscopic objects; they're not so great with molecules and best for subatomic particles or photon behaviour even though we still make large assumptions in those cases and then results are still subject to interpretation. In sports and other fields people use "statistics" to determine trends and the risk involved with some decisions based on past results, none of those is any good to predict the outcome of anything because it's humans involved. Let's say SC had only 1 race and player A had a better winrate than player B. This does not mean player A is more likely to win if he plays against player B. It doesn't work that way because there are an undetermined number of intangible variables involved, and no one can say for sure if the winrate is even a variable involved in the outcome of that interaction. If you think you found a way to numerically determine the outcome of an interaction between two human beings, you should get a nobel prize. Don't try using "statistics" as proof of anything involving people, I maintain the only thing that can tell you if a player matchfixed is either a confession or records of communications and money transactions. | ||
|
SinO[Ob]
France897 Posts
| ||
|
viasacra89
United States134 Posts
| ||
|
ThomasjServo
15244 Posts
| ||
|
Swoopae
Australia339 Posts
Also worth noting in the cases that the betting steam went the wrong way (or case, I can only think of Byul/Terror) the odds weren't skewed anywhere near as much; Terror went from being a 2-1 underdog to very slight favourite; and this was before Byul's current run of form and Terror did beat him the previous time they played. It also could have been match fixers trying to mess with Pinnacle's algorithm so future bets on matches that are actually fixed don't get cancelled If you ask any sportsbettor; the odds for Creator to be a huge favourite in the first map and huge underdog for the series vs Soulkey are pretty much impossible without foul play; the map 1 betting line is many, many standard deviations away from what it should be based on the series line; there's a good chance it's statistically more likely for Earth to get hit by a civilisation ending meteor today (I'll actually do the math if someone insists; but it's a valid comparison) Soulkey was a -219 favourite for the series (he was -300 until close to game time; this small move makes sense if only map 1 is fixed) Creator was something like a +150 to +200 underdog for map 1 minutes before the game, then when Pinnacle raises the max betting limit as they do just before matches someone smashes the Creator map 1 line over and over until Creator is a -290 favourite and it would have gone further if Pinnacle didn't take the line down and cancel bets before the match. Steam to this degree without a real reason literally doesn't happen in sportsbetting. Ever... it only makes sense if Map 1 and only Map 1 is fixed from a statistical perspective. | ||
|
Jarree
Finland1004 Posts
On May 07 2015 23:49 The_Masked_Shrimp wrote: wtf did I just read? O_o I'll spoiler my angry statistics defender part + Show Spoiler + I am not statistics professor but I am a PhD student in statistics and I did TA for introductory statistical physics for 2nd year students. You want to know how is your math incoorect? I'll give you that it's true that 0,5^5 = 3,125 and 0,5^6 = 1,5625. That's about all that is correct about it. Anything you might conclude from it is otherwise ass pulled bullshit. When medical tests are done to determine the chance of a vaccine to work or not, that is also bullshit. But at least that is not ass-pulled bullshit and people doing such "statistics" are aware of their enormous assumptions and the flaws of doing numerical analysis based on this. Statistics are only good at predicting the behaviour of deterministic microscopic objects; they're not so great with molecules and best for subatomic particles or photon behaviour even though we still make large assumptions in those cases and then results are still subject to interpretation. In sports and other fields people use "statistics" to determine trends and the risk involved with some decisions based on past results, none of those is any good to predict the outcome of anything because it's humans involved. Let's say SC had only 1 race and player A had a better winrate than player B. This does not mean player A is more likely to win if he plays against player B. It doesn't work that way because there are an undetermined number of intangible variables involved, and no one can say for sure if the winrate is even a variable involved in the outcome of that interaction. If you think you found a way to numerically determine the outcome of an interaction between two human beings, you should get a nobel prize. Don't try using "statistics" as proof of anything involving people, I maintain the only thing that can tell you if a player matchfixed is either a confession or records of communications and money transactions. Yea I didn't do my M.sc in english, so statistics might have been wrong word to use. I don't really know the proper english terms. Probabilities? Anything else you wrote makes no sense, read my earlier post where I illustrated it. | ||
|
maGicc
Finland134 Posts
Don't try using "statistics" as proof of anything involving people Should've said that from the start so people can know how useless it is to argue with you | ||
|
Swoopae
Australia339 Posts
What we should be doing is asking all of the followup questions I posted in my first post in this thread. Without additional details; Kespa's statement, while a step in the right direction, doesn't actually provide us with any relevant information on how they reached their conclusion that no matches were fixed. For us to take their statement seriously, they need to provide the evidence that they used to reach said conclusion, elaborate on how the investigation was conducted, etc. | ||
|
necrosexy
451 Posts
| ||
|
Penev
28500 Posts
| ||
|
OtherWorld
France17333 Posts
On May 07 2015 23:57 Swoopae wrote: Math doesn't lie; while it's possible no matches were fixed it's extremely unlikely based on the betting patterns we witnessed. What we should be doing is asking all of the followup questions I posted in my first post in this thread. Without additional details; Kespa's statement, while a step in the right direction, doesn't actually provide us with any relevant information on how they reached their conclusion that no matches were fixed. For us to take their statement seriously, they need to provide the evidence that they used to reach said conclusion, elaborate on how the investigation was conducted, etc. Asking for transparency from KeSPA is like asking for a post-victory striptease from INnoVation. Unnatural and not gonna happen. | ||
|
sagi
Finland346 Posts
On May 07 2015 22:22 maGicc wrote: Any statement like "players were not involved" just sounds like officer's Barbrady's "Okay move along, move along people, there's nothing to see here!" Atleast it's not Chief Wiggum: "Okay folks show’s over. Nothing to see here, show’s o-oh my God, Altho I still fear KeSPA might have to do just that eventually. | ||
|
TBone-
United States2309 Posts
| ||
|
geokilla
Canada8244 Posts
| ||
|
maGicc
Finland134 Posts
On May 08 2015 00:08 TBone- wrote: I'm really thankful we can put this behind us and that KeSPa officially investigated. Yes i am glad everything got resolved. All the bad shady gambler-investors are behind bars and they wont temp the innocent hardworking players no more. | ||
|
Jarree
Finland1004 Posts
On May 07 2015 23:49 The_Masked_Shrimp wrote: If you think you found a way to numerically determine the outcome of an interaction between two human beings, you should get a nobel prize. Had to still comment on this one. I wouldn't get a Nobel. It's already been done millions of times. Wisdom of the crowds (also known as collective intelligence) does it every day in every sports booker. If you want to read more about it, for example James Surowiecki covers it pretty well in his 2005 book The Wisdom of Crowds: Why the Many Are Smarter Than the Few. | ||
|
The_Masked_Shrimp
425 Posts
On May 08 2015 00:13 Jarree wrote: Had to still comment on this one. I wouldn't get a Nobel. It's already been done millions of times. Wisdom of the crowds (also known as collective intelligence) does it every day in every sports booker. If you want to read more about it, for example James Surowiecki covers it pretty well in his 2005 book The Wisdom of Crowds: Why the Many Are Smarter Than the Few. My comment refered to the outcome of a match between two people. Not a crowd. And what a crowd think of the outcome of a match has no impact on said match. Also crowd stats are not so well developed yet and are on the same reliability scale as medical stats. Some people making money out of it does not make it more reliable btw. | ||
|
StarStruck
25339 Posts
On May 08 2015 00:02 OtherWorld wrote: Asking for transparency from KeSPA is like asking for a post-victory striptease from INnoVation. Unnatural and not gonna happen. Considering how KeSPA handled the WC3 and BW scandals and the domino effect it had on the scene over there (ofc there were other variables too). Perhaps KeSPA does want to sweep such things under the rug. I'd love to look at the reports of this police investigation and see what they looked at and what proof they found against the people they jailed. So I guess they were jailed for mischief. They would need significant proof and see where money was going or written exchanges to do such things. | ||
| ||
StarCraft 2 StarCraft: Brood War Super Smash Bros Other Games |
|
Korean StarCraft League
CranKy Ducklings
IPSL
dxtr13 vs OldBoy
Napoleon vs Doodle
LAN Event
BSL 21
Gosudark vs Kyrie
Gypsy vs Sterling
UltrA vs Radley
Dandy vs Ptak
Replay Cast
Sparkling Tuna Cup
WardiTV Korean Royale
IPSL
JDConan vs WIZARD
WolFix vs Cross
LAN Event
[ Show More ] |
|
|
EPT
