Two Plus Two Publishing LLC Two Plus Two Publishing LLC
 

Go Back   Two Plus Two Poker Forums > > >

Discussion of Poker Sites General discussion of online poker sites.

Reply
 
Thread Tools Display Modes
Old 05-09-2009, 06:26 PM   #76
Monteroy
Carpal \'Tunnel
 
Join Date: Dec 2006
Posts: 14,100
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Dire View Post
Account creation would be logical. It's obviously in the site's best interest to get as many people 'hooked' on poker as they can. It seems reasonable to expect that people who run hot early in their experience with poker will be more easily hooked as they overestimate their ability at the game and can blame future failures on bad luck.

But I don't really think it's poignant to argue why or why not one theory makes sense. We have everything we need to test even the most obscure theories, and if one person thought it a reasonable way for a site to rig the games then somebody else probably thought so as well - and you get closer to satisfying most of everybody, or proving the site rigged depending on the results.
Any of these work in a scenario where there is 0% chance a site is caught, but as you indicate, that is never the case. Thus a company needs to measure that risk against the rewards of conducting their business in a criminal manner.

Both your belief of a logical new player boost (where again, we see tons of bad sharkscope and OPRs of new players) and action hand guys belief kind of fail the common business sense test. Both sound equally nuts to me as well

I think by now you know that those who truly want to believe it is rigged will do so regardless of the quantity and quality of data presented that shows otherwise.

Anyway, you are free to believe what you like, and act (or not act) on it as you see fit.

All the best.
Monteroy is offline   Reply With Quote
Old 05-09-2009, 06:28 PM   #77
Username^^
veteran
 
Username^^'s Avatar
 
Join Date: Apr 2009
Location: Reviewing hands in my dreams
Posts: 2,784
Re: Proof that preflop all-ins are not rigged

baning online gambling makes sence now ;-). Probably when some kind of inside company tryed to audit the games of online poker site, they found them self with similar problems.
Username^^ is offline   Reply With Quote
Old 05-09-2009, 06:30 PM   #78
Username^^
veteran
 
Username^^'s Avatar
 
Join Date: Apr 2009
Location: Reviewing hands in my dreams
Posts: 2,784
Re: Proof that preflop all-ins are not rigged

For me Durr is enough to prove that full tilt is rigged.
Username^^ is offline   Reply With Quote
Old 05-09-2009, 06:35 PM   #79
chubber8
grinder
 
chubber8's Avatar
 
Join Date: Oct 2006
Posts: 481
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Username^^ View Post
For me Durr is enough to prove that full tilt is rigged.
Oh yes. It sounds like you definitely understand more than nothing about how people win millions at poker.
chubber8 is offline   Reply With Quote
Old 05-09-2009, 06:43 PM   #80
Username^^
veteran
 
Username^^'s Avatar
 
Join Date: Apr 2009
Location: Reviewing hands in my dreams
Posts: 2,784
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by chubber8 View Post
Oh yes. It sounds like you definitely understand more than nothing about how people win millions at poker.
I do, by playing mega high variance game and running above EV for 3 years.
Username^^ is offline   Reply With Quote
Old 05-09-2009, 06:47 PM   #81
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Monteroy View Post
Not what I am asking.

What I asked was if you have 40 intervals for instance, what is the odds of one of them falling outside the range.

No idea why you think I suggested it had anything to do with the the other intervals results.

Using your lottery example, if you play every week, the odds of you winning in any specific week is the same, but if you know in advance that you will be playing every week or a year, the odds that you win are certainly greater then your chance in any specific week.

Or a less extreme example. Pick a number between 1 and 100 each day for a month. What are the odds you will get the number right on one of the days (looked at in advance of any picks). Certainly better then 1%, but I am too tired to figure it out, so you can if you like :P. That is what I am asking about the interval results.
Hmmm, I'm not sure you've had any education in statistics so I'm not sure how to explain this besides repeating myself. Your ball picking problem is completely unrelated to what you want. And your changing of the lottery problem is also incorrect. What you're looking for is what would the odds be if the winning lottery numbers never changed somehow and he just got 30 years to keep guessing. These are CONDITIONAL.

This may be an easier way to explain this. Imagine you're sitting at a poker table and for the past 20 hands somebody has flopped a set every single time. What are your odds of flopping a set the next hand? Still 11.8%. You wouldn't go well there's been 20 people flopping sets so it's incredibly unlikely this guy flops a set next hand - because they are UNRELATED.

This may sound obvious in this context, but this is exactly what you're trying to do here. You want to say that well 39/40 of the samples were normal so it makes sense for this 40th completely unrelated sample to be a bit off. No it does not. The odds of getting a 2.8% outlier is 2.8%. Putting it in context of the other results is mathematically possible but 100% meaningless.
Dire is offline   Reply With Quote
Old 05-09-2009, 07:25 PM   #82
Monteroy
Carpal \'Tunnel
 
Join Date: Dec 2006
Posts: 14,100
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Dire View Post
Hmmm, I'm not sure you've had any education in statistics so I'm not sure how to explain this besides repeating myself. Your ball picking problem is completely unrelated to what you want. And your changing of the lottery problem is also incorrect. What you're looking for is what would the odds be if the winning lottery numbers never changed somehow and he just got 30 years to keep guessing. These are CONDITIONAL.

This may be an easier way to explain this. Imagine you're sitting at a poker table and for the past 20 hands somebody has flopped a set every single time. What are your odds of flopping a set the next hand? Still 11.8%. You wouldn't go well there's been 20 people flopping sets so it's incredibly unlikely this guy flops a set next hand - because they are UNRELATED.

This may sound obvious in this context, but this is exactly what you're trying to do here. You want to say that well 39/40 of the samples were normal so it makes sense for this 40th completely unrelated sample to be a bit off. No it does not. The odds of getting a 2.8% outlier is 2.8%. Putting it in context of the other results is mathematically possible but 100% meaningless.

Dude, I know the results of each of the intervals is unrelated and independent, and again that is not my point.

I am not saying because 39 of 40 samples are normal that the 40th will have a greater chance of being unexpected. What I am saying is that if you have 40 independent samples that you plan to do, the odds that a random one of them (does not matter which) will yield a strange 1% event is greater then 1%.

Let me try one more time


Scenario 1:

Pick a number between 1 and 100. Odds you are right are 1%. Pretty rare.

There are 40 people in a room. They each pick a number between 1 and 100. Each is completely independent of each other.

What are the odds that at least 1 will get it right. Note, I am not saying any specific person, it could be the first guy the 40th guy the 21st guy, whatever.

The odds have to be better then 1% that at least one person will get their independent result correctly when looked at as a whole.


Thus, while you focus in on the interval that got a strange result (who cares which one that does not matter), my question would be, before any of the testing was done, what were the odds that one of the 40 intervals (again, does not matter which) would yield a result like that.


The odds of any one person winning a lottery is very rare, but the odds that the lottery is won by some random unknown in advance person is significant. That sort of thing.

Quote:
Originally Posted by Dire View Post
The odds of getting a 2.8% outlier is 2.8%. Putting it in context of the other results is mathematically possible but 100% meaningless.

I you ran 40 sets of samples, what are the odds in advance that at least one of them would yield a 2.8% outlier
Monteroy is offline   Reply With Quote
Old 05-10-2009, 12:00 AM   #83
Josem
human chemical weapon
 
Josem's Avatar
 
Join Date: Jan 2007
Location: Getting Trolled
Posts: 17,957
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Monteroy View Post
I you ran 40 sets of samples, what are the odds in advance that at least one of them would yield a 2.8% outlier
1 - (1-0.028)^40 = 67.8892%
Josem is offline   Reply With Quote
Old 05-10-2009, 01:50 AM   #84
tehchipdonk
grinder
 
tehchipdonk's Avatar
 
Join Date: Dec 2008
Posts: 473
Re: Proof that preflop all-ins are not rigged

i still don't get why people waste their time trying to prove/convince the non-believers to believe that the system is not rigged, what will they achieve in the end?
tehchipdonk is offline   Reply With Quote
Old 05-10-2009, 03:13 AM   #85
HundredsOfStuff
veteran
 
HundredsOfStuff's Avatar
 
Join Date: Jan 2009
Posts: 2,180
Re: Proof that preflop all-ins are not rigged

Someone needs to do a line regression that shows the correlation between how much money you lose at online poker and how rigged/shady you think online poker is.
HundredsOfStuff is offline   Reply With Quote
Old 05-10-2009, 03:14 AM   #86
HundredsOfStuff
veteran
 
HundredsOfStuff's Avatar
 
Join Date: Jan 2009
Posts: 2,180
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by tehchipdonk View Post
i still don't get why people waste their time trying to prove/convince the non-believers to believe that the system is not rigged, what will they achieve in the end?
One more fish in the pond is what they will achieve.
HundredsOfStuff is offline   Reply With Quote
Old 05-10-2009, 04:43 AM   #87
tehchipdonk
grinder
 
tehchipdonk's Avatar
 
Join Date: Dec 2008
Posts: 473
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by HundredsOfStuff View Post
One more fish in the pond is what they will achieve.
fiends will play regardless if it's rigged or not lol. like, table games at the casino are rigged to favour the house, but there are ****loads of fiends playing on them still
tehchipdonk is offline   Reply With Quote
Old 05-10-2009, 05:26 AM   #88
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Josem View Post
1 - (1-0.028)^40 = 67.8892%
*sighs*

It's amazing that this is a poker oriented site, yet most people are 100% clueless about statistics. I understand you work for Poker Stars Josem so it's in your best interest to blindly defend the site but please only state something if you have at least some clue about of what you're talking about. You've done this same thing in other threads as well.

This is not 40 independent binary events with a probability of 2.8% (actually there are only 25 events total, but this is irrelevant to the point). This exact event was measured ONCE.

What you're trying to measure here would be if you had 40 samples of explicitly the .20 - .22 equity scenarios. But you don't have that, you only have 1 so conditionally comparing it to the other completely unrelated events makes no sense. Now on the other hand, if you did have 40 samples of just the .20 - .22 equity case then yes, you would expect to see a 2.3% outlier about 1/40 times.
Dire is offline   Reply With Quote
Old 05-10-2009, 05:32 AM   #89
Josem
human chemical weapon
 
Josem's Avatar
 
Join Date: Jan 2007
Location: Getting Trolled
Posts: 17,957
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Dire View Post
*sighs*

It's amazing that this is a poker oriented site, yet most people are 100% clueless about statistics. I understand you work for Poker Stars Josem so it's in your best interest to blindly defend the site but please only state something if you have at least some clue about of what you're talking about. You've done this same thing in other threads as well.
Your personal attack is both false and irrelevant.

There was a specific question. I answered that specific question.

As far as I know, the answer was correct. I'm no expert on this, but am interested and willing to help.
Josem is offline   Reply With Quote
Old 05-10-2009, 05:38 AM   #90
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Monteroy View Post
Dude, I know the results of each of the intervals is unrelated and independent, and again that is not my point.

I am not saying because 39 of 40 samples are normal that the 40th will have a greater chance of being unexpected. What I am saying is that if you have 40 independent samples that you plan to do, the odds that a random one of them (does not matter which) will yield a strange 1% event is greater then 1%.

Let me try one more time


Scenario 1:

Pick a number between 1 and 100. Odds you are right are 1%. Pretty rare.

There are 40 people in a room. They each pick a number between 1 and 100. Each is completely independent of each other.

What are the odds that at least 1 will get it right. Note, I am not saying any specific person, it could be the first guy the 40th guy the 21st guy, whatever.

The odds have to be better then 1% that at least one person will get their independent result correctly when looked at as a whole.


Thus, while you focus in on the interval that got a strange result (who cares which one that does not matter), my question would be, before any of the testing was done, what were the odds that one of the 40 intervals (again, does not matter which) would yield a result like that.


The odds of any one person winning a lottery is very rare, but the odds that the lottery is won by some random unknown in advance person is significant. That sort of thing.




I you ran 40 sets of samples, what are the odds in advance that at least one of them would yield a 2.8% outlier
What you are trying to do in this thread is not like what you have in your analogy. A more appropriate analogy would be like if you had 5 guys who were trying to guess a number 1/5 and another where 30 guys were guessing 3/55 and another 100 guys who were on 1/1000000, etc, etc.. Just trying to blindly merge all those completely unrelated samples into one measurement makes no sense. Even if you do the math properly it tells you nothing meaningful.

If you want to measure the chances of the 2.8% being an outlier in 40 samples then you'd need 40 samples of exactly what it was measuring - the .20 to .22 equity scenario.
Dire is offline   Reply With Quote
Old 05-10-2009, 05:59 AM   #91
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Josem View Post
Your personal attack is both false and irrelevant.

There was a specific question. I answered that specific question.

As far as I know, the answer was correct. I'm no expert on this, but am interested and willing to help.
His question was misleading, improper and irrelevant. He's not asking what he thinks he's asking.

It would be like if I said: I think this coin has a 50% chance of flipping heads or tails.

He responds. Oh yeah? I just flipped it ten times and got heads 8 times in a row. What are the odds of that if the coin was fair?

You immediately jump in and say .4%.

It's true, meaningless and misleading. Somebody ignorant of math could then infer that that must mean there's a 99.6% chance that the coin isn't fair. You should expect as much when posting in the zoo. So I'm left to believe either you're ignorant of the math here and felt the question was meaningful, or you are are not ignorant of math but intentionally decided to be misleading. I can't see your post in any positive light.

Last edited by Dire; 05-10-2009 at 06:09 AM.
Dire is offline   Reply With Quote
Old 05-10-2009, 06:09 AM   #92
Josem
human chemical weapon
 
Josem's Avatar
 
Join Date: Jan 2007
Location: Getting Trolled
Posts: 17,957
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Dire View Post
...It's true, meaningless and misleading. Somebody ignorant of math could then infer that that must mean there's a 99.6% chance that the coin isn't fair. You should expect as much when posting in the zoo. So I'm left to believe either you're ignorant of the math here and felt the question was meaningful, or you are are not ignorant of math but intentionally decided to be misleading. I can't see your post in any positive light.
But the way that the .20-.22 data was selected in the first place was flawed - no one asked the question, "Do the survey and get us the results from .20-.22." Instead, the question was asked, "What are all the answers?" and then the odd outlier was selected from that.

As I said above, I'm no expert on this stuff - but I'm interested to learn and do think that I consider this stuff critically. There's no need to involve personalities here: I'm inclined to help you, because I believe that the results will prove my pre-existing belief: that there is no systematic bias.

Back to the underlying issue, it seems obvious to me that if you get 25 results, some of the results will be spot on the expected result, and some will be on the various extremities. I don't know the precise way to communicate or express or calculate this mathematically, so I'm happy to be guided by the experts.

I happen to think that Monteroy's question was reasonable, so I answered it to the best of my ability.
Josem is offline   Reply With Quote
Old 05-10-2009, 06:13 AM   #93
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Actually that's not even that great an example. Maybe a better one would be:

Me: Interesting. I just flipped this coin 100 times and got heads 50 times. I'd guess it's probably a fair coin.

Him: Oh yeah? If the coin was fair, what are the odds of you flipping heads 50 times over 100 tests?

You: 7.9%

Again, he's just not asking what he thinks he's asking and the vast majority of the people here are going to have no clue what that 7.9% means.
Dire is offline   Reply With Quote
Old 05-10-2009, 06:21 AM   #94
Josem
human chemical weapon
 
Josem's Avatar
 
Join Date: Jan 2007
Location: Getting Trolled
Posts: 17,957
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Dire View Post
Actually that's not even that great an example. Maybe a better one would be:

Me: Interesting. I just flipped this coin 100 times and got heads 50 times. I'd guess it's probably a fair coin.

Him: Oh yeah? If the coin was fair, what are the odds of you flipping heads 50 times over 100 tests?

You: 7.9%

Again, he's just not asking what he thinks he's asking and the vast majority of the people here are going to have no clue what that 7.9% means.
I think an example that captures my thinking (and I suspect, Monteroy's thinking) is as follows:

Flip a coin 100 times, and record the number. Then repeat that process 100 times.

What is the chance that at least one of those samples is a 2.8% outlier?

(obv it's not precise and exact, but I think that this represents my thinking)
Josem is offline   Reply With Quote
Old 05-10-2009, 06:28 AM   #95
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Josem View Post
But the way that the .20-.22 data was selected in the first place was flawed - no one asked the question, "Do the survey and get us the results from .20-.22." Instead, the question was asked, "What are all the answers?" and then the odd outlier was selected from that.

As I said above, I'm no expert on this stuff - but I'm interested to learn and do think that I consider this stuff critically. There's no need to involve personalities here: I'm inclined to help you, because I believe that the results will prove my pre-existing belief: that there is no systematic bias.

Back to the underlying issue, it seems obvious to me that if you get 25 results, some of the results will be spot on the expected result, and some will be on the various extremities. I don't know the precise way to communicate or express or calculate this mathematically, so I'm happy to be guided by the experts.

I happen to think that Monteroy's question was reasonable, so I answered it to the best of my ability.
The survey was not flawed as it did not introduce bias. If there's no bias and sample sizes are sufficient to give a reasonable level of confidence then you should be able to manipulate the data any way you want and get proper results. If somebody asked for every third hand, filtered to players with an odd number of characters in their username and then split down by equity it should still be proper. If there's something wrong with the data there then it's not the survey's fault.

The problem with just randomly bunching all the results together is that these are not independent binary events, and the events are also not related. So it just doesn't mean anything. If you figure out that you data would occur exactly 17% of the time by chance then that tells you absolutely nothing more than that your data would occur exactly 17% of the time by chance. It says nothing about the reliability of any given unrelated event within it. If an event within the set is 2.3% likely, then it's 2.3% likely.
Dire is offline   Reply With Quote
Old 05-10-2009, 07:30 AM   #96
Monteroy
Carpal \'Tunnel
 
Join Date: Dec 2006
Posts: 14,100
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Josem View Post
I think an example that captures my thinking (and I suspect, Monteroy's thinking) is as follows:

Flip a coin 100 times, and record the number. Then repeat that process 100 times.

What is the chance that at least one of those samples is a 2.8% outlier?

(obv it's not precise and exact, but I think that this represents my thinking)
Correct. I really have no idea how and why Dire is off on the tangent he is, but he seems to enjoy being there, and as I say people are always entitled to their beliefs.

My overall point would be that you run a huge test broken up into dozens of pre-defined segments. The fact that a couple of those segments produced unexpected results (particularly the ones with much smaller sample sizes) is in and by itself not that shocking.

This would be like running a bunch of trials of coin flipping (with a diferent amount of attempts in each trial), and one of the trials generating a 1 in 100 event.

The next step is to do further studies on the outlier quadrants to ensure the data does not have some sort of flaw, and then do further larger studies on those quadrants to see if the same outlier results occur.

In this case the quadrant had a relatively small sample size of 2,325 hands when other quadrants had 40,000-55,000 hands so I assume the odds of an outlier event are greater with a smaller sample size.

I mean i you look at the .04-.06 quadrant of 46 hands in which 2 won to it in with a 0.043 result. Now, imagine if only 1 or 0 or 3 or 4 had won. It would have been a huge outlier, but with a sample size of 46 who cares (well, rigged fanatics would pounce on it, but they tend to be pretty easy going with whether a stat is significant).


Quote:
Originally Posted by Dire View Post
*sighs*

It's amazing that this is a poker oriented site, yet most people are 100% clueless about statistics. I understand you work for Poker Stars Josem so it's in your best interest to blindly defend the site but please only state something if you have at least some clue about of what you're talking about. You've done this same thing in other threads as well.
Paranoia, passive aggressiveness and arrogance. Sadly, those traits tend to not be as independent.

Quote:
Originally Posted by Dire View Post
What you're trying to measure here would be if you had 40 samples of explicitly the .20 - .22 equity scenarios. But you don't have that, you only have 1 so conditionally comparing it to the other completely unrelated events makes no sense. Now on the other hand, if you did have 40 samples of just the .20 - .22 equity case then yes, you would expect to see a 2.3% outlier about 1/40 times.
Again, that outlier may be rare when looked at by itself, but the fact that an outlier existed should not on the surface be that concerning. I would have assumed one or more of the segments based on how they were defined would have produced an outlier. It involved one of the smaller samples (which is better then if one of the 55,000 samples were off).


Dire, if the .06-.08 segment produced these results instead

[0.04 - 0.06): 46 4 0.086%

How much would you think this mattered overall?

That is my point, that this one outlier of [0.22 - 0.24): 2325 547 0.235% is interesting and rare in and by itself, but not that shocking within the big picture, though further study (more hands, more detailed breakdown of the segment etc) would be good to see.
Monteroy is offline   Reply With Quote
Old 05-10-2009, 07:43 AM   #97
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

In the 1-(1-.028)^40 case there you are saying you have 40 events, each with a probability of exactly 97.2% of occuring. What are the odds that at least one of the events fails to occur? Do you understand how this isn't related at all? Were the odds of .30-.32 running at 3.17% exactly 97.2%? Were the odds of .26-.28 running at .261 exactly 97.2%?

EDIT: Even if you were JUST measuring lots of the .20 - .22 samples it still wouldn't make sense. If you get .21 the first time, was that 97.2% likely to occur, and .215 the second was also 97.2%, etc, etc, etc. It just a nonsensical way of interpreting the data and even if you do balance and do the math properly, it doesn't mean much of anything. It will just tell you that your whole data set would occur by chance 'x' percent of the time. Meaningless!

Last edited by Dire; 05-10-2009 at 08:01 AM.
Dire is offline   Reply With Quote
Old 05-10-2009, 07:53 AM   #98
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Josem View Post
I think an example that captures my thinking (and I suspect, Monteroy's thinking) is as follows:

Flip a coin 100 times, and record the number. Then repeat that process 100 times.

What is the chance that at least one of those samples is a 2.8% outlier?

(obv it's not precise and exact, but I think that this represents my thinking)
Well the odds of a 2.8% outlier are 2.8%!

Over 100 samples your expectation would be 2.8, so it should be extremely likely for at least one 2.8. Repeat the .20 - .22 experiment 100 times and I wouldn't bat an eye if 4 of them were off by a bit, but that's not what you have here.
Dire is offline   Reply With Quote
Old 05-10-2009, 08:02 AM   #99
Dire
banned
 
Dire's Avatar
 
Join Date: Nov 2005
Location: wasting less time
Posts: 8,581
Re: Proof that preflop all-ins are not rigged

Be back in a week.
Dire is offline   Reply With Quote
Old 05-10-2009, 08:31 AM   #100
Monteroy
Carpal \'Tunnel
 
Join Date: Dec 2006
Posts: 14,100
Re: Proof that preflop all-ins are not rigged

Quote:
Originally Posted by Dire View Post
Well the odds of a 2.8% outlier are 2.8%!

Over 100 samples your expectation would be 2.8, so it should be extremely likely for at least one 2.8. Repeat the .20 - .22 experiment 100 times and I wouldn't bat an eye if 4 of them were off by a bit, but that's not what you have here.

But you do have different sets of samples, some with as few as 40ish hands, some with 50,000+ hands based on predefined criteria (the % intervals). Sure they all came from the same overall database, but once seperated into categories they are effectively smaller sample sizes in each of the categories.

That is why if the one with 40 hands was off you probably would not care, because 40 hands does not really prove much, except to those fixated on QQ vs Q2 matchups. That 6-8% category needs more hands for it to be powerful data, regardless of where it happens to fall for now vs expectations.

In this case a segment with 2000ish hands is off by a bit. How big a concern should that be?

My overall point that you seem to continue to misinterpret is that I would assume some of the arbitrary sub categories would generate some outlier results, the more categories (ie: fewer hands in each) the greater liklihood of outliers. This is kind of "well, duh" stuff.

I mean if you go to the extreme and make every single hand their own category then they will all be off by a ton except for 100/0 scenarios.


So while you keep hammering home that the odds of this event is for sake of argument 1 in 100, my point is that really in and by itself does not matter. Yes, I get that the odds of that particular result is 1 in 100, but does a 1 in 100 sub category with 5% the hand sample size of many of the other categories mean that much other then something to look at with more data?

Odds were that one of the smaller sub categories would produce some outlier (my point all along if I say more then 1% I do not mean to link it to the 1% result that took place).

The fact it was this one in this specific data set is irrelevant. If you ran 100 separate studies of different independent similar sized data bases I would expect outliers to pop up here and there all over the place based on these sub categories.

If they kept showing up in the same intervals then I would be more concerned that something was off, either in the data manipulation/gathering or maybe in the site itself if nothing else could be proven.
Monteroy is offline   Reply With Quote

Reply
      

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


Forum Jump


All times are GMT -4. The time now is 04:42 PM.


Powered by vBulletin®
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.
Copyright © 2008-2020, Two Plus Two Interactive