Statistical Convergence.

Flintoff · #1 11-01-2006, 08:08 PM

Obviously stats like VPIP and PFR converage very quickly but a change in my playing style has led me to wonder the following.
How many hands do you think it would take before you got a reasonable figure for your 'WtSD' and 'W$aSD'?

Harv72b · #2 11-02-2006, 01:07 AM

You might be able to get a mathematical answer from one of the walking computers on the Probability forum. I suspect that the true answer lies within that paradox, where by the time you've accumulated enough hands on an opponent for the numbers to converge, his playing style will have changed enough that the early hands don't apply anymore.

For your own stats, I don't really pay much attention to my WtSD, so long as the W$SD stays in the 52-55% range over any significant sample. But those numbers are mainly carryovers from my full ring game, so I don't know if they're correct for short-handed.

Flintoff · #3 11-02-2006, 01:19 AM

Thanks Harv. My current numbers are way out from the norm since I moved up to 10/20 but I have changed my playing style.

However, it's only 9000 hands. I'll get a few more hands under my belt before I mention it again.

Solid_p · #4 11-02-2006, 05:51 AM

[ QUOTE ]
Thanks Harv. My current numbers are way out from the norm since I moved up to 10/20 but I have changed my playing style.

However, it's only 9000 hands. I'll get a few more hands under my belt before I mention it again.

[/ QUOTE ]

How has your stats changed? Much higher wtsd?!

Flintoff · #5 11-02-2006, 07:16 AM

[ QUOTE ]
[ QUOTE ]
Thanks Harv. My current numbers are way out from the norm since I moved up to 10/20 but I have changed my playing style.

However, it's only 9000 hands. I'll get a few more hands under my belt before I mention it again.

[/ QUOTE ]

How has your stats changed? Much higher wtsd?!

[/ QUOTE ]

Yeah - much higher on both, G.
Im sure its partially to do with checking the turn more than I normally do.

TomBrooks · #6 11-02-2006, 03:52 PM

[ QUOTE ]
However, it's only 9000 hands.

[/ QUOTE ]
I second the suggestion to ask the folks on the Probability Forum. I'd be interested in the answer to this question too, so please post it here if you find out.

Just guessing, I think with 9000 hands your stats could be off quite a bit from what you will eventually settle into, after which time the stats will probably change slowly depending on how you play and how you run. I'll guess as much as 5% off.

Flintoff · #7 11-07-2006, 11:37 PM

No answer from the probability forum.

I did ask Stoxtrader though, and he said he usually runs around 36/57 or so.

Maybe my numbers aint that far out after all.

latefordinner · #8 11-08-2006, 01:02 AM

36/57?

57 seems way high. but I guess that's why stox makes more per hand than i do per month.

Flintoff · #9 11-08-2006, 04:07 AM

Yep - thats what he said!

ajm36 · #10 11-08-2006, 05:48 AM

The W$ASD is going to vary according to game conditions. So, someone playing at higher limits may win more often than someone at lower limits. This is mainly an artifact of the # of players at any given showdown. At lower limits (generally looser games), more players are at any given showdown on average. However, the pot size on average will be larger (in BB's). I think you can expect your W$ASD to be lower in low limit games than at higher limit games. However, because of pot size, your win rate can be the same or higher than a good mid or high limit player. Something to keep in mind.