DERB

[TimM]

Does he chat at the tables? I tried Googling for his name + poker and found nothing in English.

Pretty soon that search will lead to this forum.

[rigoletto]

Took me to long to edit this some I'm reposting the edit portion:

With an SD of 30BB/100 hands (on the high side, but we'll roll with it) his confidence interval is 2.8BB/100 at a 99.7% confidence. This brings his true winrate to be between 0.2BB/100 and 5.8BB/100. Now 0.3% of the time he could still be outside this inteval, but this goes both to the winning and losing sides. But applying a little common sense we know that it is more likely to be to the loosing side.

So the math tells us that he is likely a winner, but it's within the scope to say that he might just be a small winner running good and then of course there is the 0.3% lottery.

Now if I was to give you even money on a bet that DERB is a winning player, would you take that bet! I'm not discounting the anecdotal evidence, but at the poker table we make bets all the time based on probabilities and in the face of statistics it'll take more than some bad beat stories and a few hand histories before I'll write DERB of as a fluke.

[flub]

[ QUOTE ]

We are not talking confidence interval between a population of players, but confidence interval for a 100K sample of hands where each member is 100 hands!!!!!!!!!!!!

[/ QUOTE ]

If you have 100,000 hands and use SD/100 is your sample size 100,000 or 1,000?

Is SD/100 the players SD divided by 100, or is it the players average SD per 100 hands?

-f

[rigoletto]

Confidence interval

[J_V]

99.7% confidence

When we are talking about a specific player pulled out of 100,000's, this doesn't mean very much. It's not as if I pulled this lucky Slovenian out of a hat. Plus, these stats mean nothing is there is some kind of foul play at hand - and that is not ruled out of my mind. Of course, I wouldn't dare accuse someone of that without proper evidence, but I am protecting myself. First thing is order is to learn a few Slovenian curse words.

[ QUOTE ]
Now if I was to give you even money on a bet that DERB is a winning player, would you take that bet!

[/ QUOTE ]


Tough one. If he were to continue playing full games at the party 30-60 when there were only six games and the same group of players played, I would bet that he is a losing player. However, assuming there is a chance that there are things about this game that I don't have complete grasp over, I would not be overly excited about my position [img]/images/graemlins/smile.gif[/img].

[Justin A]

[ QUOTE ]
Took me to long to edit this some I'm reposting the edit portion:

With an SD of 30BB/100 hands (on the high side, but we'll roll with it) his confidence interval is 2.8BB/100 at a 99.7% confidence. This brings his true winrate to be between 0.2BB/100 and 5.8BB/100. Now 0.3% of the time he could still be outside this inteval, but this goes both to the winning and losing sides. But applying a little common sense we know that it is more likely to be to the loosing side.

So the math tells us that he is likely a winner, but it's within the scope to say that he might just be a small winner running good and then of course there is the 0.3% lottery.

Now if I was to give you even money on a bet that DERB is a winning player, would you take that bet! I'm not discounting the anecdotal evidence, but at the poker table we make bets all the time based on probabilities and in the face of statistics it'll take more than some bad beat stories and a few hand histories before I'll write DERB of as a fluke.

[/ QUOTE ]

Yes, I understand all of that, but it doesn't work in this case, because we're not talking about a random sample of 100k hands. This guy is an outlier in a larger sample of a large number of players. The fact that there's a thread about him shows that we're looking at his numbers because of his high winrate.

I'm trying to think of an analogy here. Suppose you decide to test a coin to see if it is fair. To do this lets say you toss it 1,000 times, record the results, and repeat many times. Now you should have a set of numbers of the times out of 1,000 that the coin flips heads. The mean of these numbers should be 500 for a fair coin. So lets say it is in fact a fair coint and the mean of the tests is 500. Now suppose that we find that one of these tests gave us a heads count of 570. This test is clearly an outlier, but if we take that test by itself, we will find that our conclusion is that with 99.7% certainty, the probability of the coin being heads on any given toss is between .523 and .617. However, this confidence interval is meaningless because we chose this set of 1,000 tosses because of the results it gave.

We're doing the same here with DERB. His stats were chosen because of the bizarre results, therefore we cannot use confidence intervals to figure out whether or not he is a winning player. I hope all that made sense.

[NLSoldier]

[ QUOTE ]
[ QUOTE ]
Took me to long to edit this some I'm reposting the edit portion:

With an SD of 30BB/100 hands (on the high side, but we'll roll with it) his confidence interval is 2.8BB/100 at a 99.7% confidence. This brings his true winrate to be between 0.2BB/100 and 5.8BB/100. Now 0.3% of the time he could still be outside this inteval, but this goes both to the winning and losing sides. But applying a little common sense we know that it is more likely to be to the loosing side.

So the math tells us that he is likely a winner, but it's within the scope to say that he might just be a small winner running good and then of course there is the 0.3% lottery.

Now if I was to give you even money on a bet that DERB is a winning player, would you take that bet! I'm not discounting the anecdotal evidence, but at the poker table we make bets all the time based on probabilities and in the face of statistics it'll take more than some bad beat stories and a few hand histories before I'll write DERB of as a fluke.

[/ QUOTE ]

Yes, I understand all of that, but it doesn't work in this case, because we're not talking about a random sample of 100k hands. This guy is an outlier in a larger sample of a large number of players. The fact that there's a thread about him shows that we're looking at his numbers because of his high winrate.

I'm trying to think of an analogy here. Suppose you decide to test a coin to see if it is fair. To do this lets say you toss it 1,000 times, record the results, and repeat many times. Now you should have a set of numbers of the times out of 1,000 that the coin flips heads. The mean of these numbers should be 500 for a fair coin. So lets say it is in fact a fair coint and the mean of the tests is 500. Now suppose that we find that one of these tests gave us a heads count of 570. This test is clearly an outlier, but if we take that test by itself, we will find that our conclusion is that with 99.7% certainty, the probability of the coin being heads on any given toss is between .523 and .617. However, this confidence interval is meaningless because we chose this set of 1,000 tosses because of the results it gave.

We're doing the same here with DERB. His stats were chosen because of the bizarre results, therefore we cannot use confidence intervals to figure out whether or not he is a winning player. I hope all that made sense.

[/ QUOTE ]

Makes sense. Thanks for the clarification.

[mikelow]

Now I hope you didn't click on any of the links. [img]/images/graemlins/smirk.gif[/img]

[theBruiser500]

jesus guys, let this thread die alreayd

[rigoletto]

[ QUOTE ]
We're doing the same here with DERB. His stats were chosen because of the bizarre results, therefore we cannot use confidence intervals to figure out whether or not he is a winning player. I hope all that made sense.

[/ QUOTE ]

I have to disagree. Your objection would be true if we had say 1,000,000 hands with DERB and picked a 100K series out of them because this series was bizarre. But the fact that we chose to isolate all the hands where DERB was at the table doesn't bias HIS stats.

You analogy is flawed because you are assuming that each player can be represented by the same coin. A more true analogy is to make 1000 cointosses with 1000 coins and the result is say 500,152/499,848. You record all tosses and what coin they where made with. Now you isolate your sample to a specific coin and find that the result for this coin is 611/389. You now have good reason to suspect that this particular coin is different from the others because the 1000 tosses with this coin can statistically be treated as a seperate event.

Let me try my own analogy: let say we randomly sample 4000 people and ask them who they would vote for as president Andyfox or Ed Miller and the result came back 50/50. Now we pick a subset, say ages 18-30, and for this subset the result is 30/70. Just because we picked a subset it doesn't mean that the sample is not random anymore; it is a random sample within the universe of 18-30 year olds. Just as hands with DERB in them is a random sample within the universe of hands DERB participated in.