Poker System

[Guest]

Hi All A while ago I posted an idea for a poker stats / tracker system. I'm now doing some testing and need some help to analyse poker player styles. I'm looking to get an individual players style from the number of calls / raises / flops seen / showdowns etc... I have all the data for each player but am unsure as to how to evaluate it?? I think the best way route is to catogarize players into the Loose/Tight/Passive/Aggressive slots but not sure how to take it forward. Anyone any ideas??? Thanks in advance. Mike

[Masterplan]

Re: Poker System tough issue ... my degree in computer science doesnt help, go seek a statistician

[GaF]

Re: Poker System It is a tough issue...... May want to look at www.bet-the-pot.com - PokerTracker Part IIa - this section has their categories (and criteria) for categorising players in poker tracker.......

[Guest]

Re: Poker System

I have defined 3 categories of pre-flop characteristics based on the percentage of times the player voluntarily put money into the pot (VP$IP).

VP$IP	Rating
> 40%	Loose
25% -> 40%	Slightly Loose
	Tight

Pre-flop, I’ve defined aggression by the pre-flop raise percentage (PFR%).

PFR%	Rating
>= 7%	Aggressive
	Passive

Post-flop, I’ve used the total aggressiveness (with the ignore pre-flop aggression stats checked).

AF-TOT	Rating
> 1.2	Aggressive
	Passive

Nice one.. just what i needed but i'm not sure what the post-flop 1.2 means ??? Cheers for the link. Mike

[gazabroad]

Re: Poker System I think it says less than or equal to, doesn´t it?

[GaF]

Re: Poker System There's a stat used in Pokertracker for Post Flop aggression - off the top of my head - it's the number of times a player bets or raises divided by the number of times a player checks or calls - so a post flop aggression factor of 1.2 would be something like the player bets post flop on 6 out of 11 hands (calculated by 6 / 5) Big weakness in this stat is that it takes no consideration of the size of the bet - there are a lot of poor players who will frequently bet the minimum every time (say $0.30) even into a large pot (say £20) - to say these players are aggressive post flop is ridiculous.......

[GaF]

Re: Poker System I'm also not keen on the preflop aggressive/passive measure - someone with 6.99% is passive - someone with 7.01% is aggressive ....... there needs to be some buffer/"middle ground". With poker tracker, a weakness is that you can only have 14 categories (I think it's 14 :unsure) of player - so the stats above were formulated with this restriction in mind and they couldn't split players up as much as they wanted to....... the combination of above gives 3x2x2 = 12 categories.....

[mcgin]

Re: Poker System Fairly difficult task if your doing all this yourself without the aid of poker tracker etc. But my advice would be calculate each individual statistic, i.e. # Raises PF, # Calls on flop, Voluntarily put money in the pot % etc for all the players. Once thats done I would take each statistic in isolation (e.g. PreFlop Raise %), so say you have 100 players for which you have calculated the statistic for, take those 100 values and calculate the mean and the standard deviation of them (I'm assuming the values will be normally distributed which is likely). You can then assign a value for each player under that statistic using their difference in SD's from the mean. E.g. Player X sees the flop 40% of the time. Lets pretend the mean value for this stat is 25 and the standard deviation is 10, therefore this guys looseness preflop is +1.5, or very loose. Similarly for player B who sees the flop only 15% of the time would have a value of -1.0, or quite tight. Repeat for all statistics you use, and then combine each to come up with profiles of your players. Thats obviously a fairly simplified approach as it does not take into consideration the number of people in each hand etc., but theres no reason you cannot extend it some how to include that if you fancy it