DERB

[AceHigh]

I understand what you are saying...but the math is all on Andrew's side. DERB might "look" like a bad player, but he consistently wins. Over a long period of time for a high rate. That's my definition of a winning player. It's not just the high winrate or the long period of time, it's the combination that makes the arguement/math compelling.

It can't explain it, but it doesn't make it less true. And anecdotal evidence is notoriously weak. See Anecdotal evidence

It's a little like a trial, where the one eye witnesses say a person is the guilty, but the DNA evidence says it was someone else. The DNA evidence is so compelling and eye witness are somewhat unreliable, so we have to believe the DNA.

[James282]

[ QUOTE ]
I understand what you are saying...but the math is all on Andrew's side. DERB might "look" like a bad player, but he consistently wins. Over a long period of time for a high rate. That's my definition of a winning player. It's not just the high winrate or the long period of time, it's the combination that makes the arguement/math compelling.

It can't explain it, but it doesn't make it less true. And anecdotal evidence is notoriously weak. See Anecdotal evidence

It's a little like a trial, where the one eye witnesses say a person is the guilty, but the DNA evidence says it was someone else. The DNA evidence is so compelling and eye witness are somewhat unreliable, so we have to believe the DNA.

[/ QUOTE ]

You can't say "I understand what you are saying" and then make an analogy that clearly shows that you don't understand what I'm saying if you want me to take your opinion seriously.
-James

[AceHigh]

I'm saying I understand he doesn't look like a typical, winning player.

[James282]

[ QUOTE ]
I'm saying I understand he doesn't look like a typical, winning player.

[/ QUOTE ]

And what you don't understand is that I'm saying I understand how unlikely it is, according to the math, that he is not actually playing a winning style. That is the whole point of the discussion.

[DMBFan23]

before I look like a total [censored], does the data GUARANTEE that he is a winning player? let's say 99.99999999% sure.

[rigoletto]

[ QUOTE ]
before I look like a total [censored], does the data GUARANTEE that he is a winning player? let's say 99.99999999% sure.

[/ QUOTE ]

No according to the math it there is maybe about 5% (not trivial) chance that he is a losing player. Then we have anecdotal evidence from some well respected players that would suggest that he is in fact a losing player.

I haven't played DERB, so I can't tell one way or the other. But I have in the past seen good players write of succesfull players because they played an untraditional strategy, so I reserve my judgement.

[RandomFluke]

[ QUOTE ]
No according to the math it there is maybe about 5% (not trivial)

[/ QUOTE ]

Where does this 5% number come from? I mean, I believe james is probably right and derb is the 'lotto winner' but I think over 100k hands it is much less than 5% that a bad player could win this much. I'd say less than 1% chance. This is the flipping the coin 10000 times and it comes up heads 70% of the time type situation.

[NLSoldier]

[ QUOTE ]
[ QUOTE ]
No according to the math it there is maybe about 5% (not trivial)

[/ QUOTE ]

Where does this 5% number come from? I mean, I believe james is probably right and derb is the 'lotto winner' but I think over 100k hands it is much less than 5% that a bad player could win this much. I'd say less than 1% chance. This is the flipping the coin 10000 times and it comes up heads 70% of the time type situation.

[/ QUOTE ]

Yeah, my instincts are telling me its easily gotta be less than 1% or we would have come across more of these guys by now.

[rigoletto]

[ QUOTE ]
[ QUOTE ]
No according to the math it there is maybe about 5% (not trivial)

[/ QUOTE ]

Where does this 5% number come from? I mean, I believe james is probably right and derb is the 'lotto winner' but I think over 100k hands it is much less than 5% that a bad player could win this much. I'd say less than 1% chance. This is the flipping the coin 10000 times and it comes up heads 70% of the time type situation.

[/ QUOTE ]

You need a much larger sample than 100K to give him a winning confidence interval of 99%!

[NLSoldier]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
No according to the math it there is maybe about 5% (not trivial)

[/ QUOTE ]

Where does this 5% number come from? I mean, I believe james is probably right and derb is the 'lotto winner' but I think over 100k hands it is much less than 5% that a bad player could win this much. I'd say less than 1% chance. This is the flipping the coin 10000 times and it comes up heads 70% of the time type situation.

[/ QUOTE ]

You need a much larger sample than 100K to give him a winning confidence interval of 99%!

[/ QUOTE ]

Even considering how high his BB/100 is?

I just learned about confidence intervals in a class. I should be able to figure this out. Im gonna try anyways.

edit-ok nm its already been done, i think this is what you are talking about anyways.

[ QUOTE ]
Does anyone know from their PT DB what his Sigma/100 hands is? It has to be higher than what most of us are used to.

Since I don't have the data I can't do any serious analysis, but we can wait for Cypher for that.

Just for the fun of it I calculated the probability of a break even player running this well or better over 80k hands.
Assuming 80k hands and a winrate of 3BB/100:
Sigma/100 ---- Probability
15 ----------- 7.7e-9
20 ----------- 1.1e-5
25 ----------- 3.4e-4
30 ----------- 2.3e-3

[/ QUOTE ]