Poker tracker/PAhud- Thoughts on the Statistics

[Longy]

This has been asked for in another thread and I appear to have volunteered myself for it. This is by no means a complete guide and is more my thoughts on the subject. All my views are based on the game I play 6 max NLHE.

I use the statistics from poker tracker a great deal to make reads on people and get an idea of how my opponents play. I would emphasise this is guide and stats can't tell you everything, along with the numbers i take notes on how players play hands which the numbers can't tell you.

I have 10 stats displayed in game when I'm playing using PAHUD (which overlays the stats on the table from poker tracker), I will go through the main 4 and if people find stuff useful I will write stuff about the other stats to be found in poker tracker.

VPIP- Is the stat that tells you how often a player enters a pot voluntarily (checking your bb doesn't count). This is the first and most useful number when facing an unknown, it tells you if they are tight or loose and therefore we can estimate the range of hand they are playing preflop. It is also a reasonable guide to how "good" someone is and there is a correlation between winrate and vpip more than any other stat.

0-15% Anyone in this range is tight and plays few hands, they are in the main not great players to have at your table. They are however exploitable, they are easy to read as you can put them on a hand easier and therefore you can make easier reads postflop given flop texture, plus you can bully them as they will rarely play back at you.

15-30% This can be a right mixture of players from a good TAG/LAG players to weak passive players. In my opinion this is the optimal range and most of the good winning players are somewhere in there. There is no perfect number, good TAGS and LAGS can be either end of this range and be solid winners. The advantages of playing this style is that while you aren't playing junk, you are playing enough hands to make your preflop range hard to read, all pairs, big cards, suited connectors etc are all in your preflop range. These are players I most like to avoid.

30%+ There are very few players (less than 1%) that can play profitably with this VPIP. These are the players you want at your table, find them and sit with them. If you can get position on them even better, don't be afraid to get in pots with them. They are the weak players and generally play poorly postflop ,often because of their loose starting hand range. I like to isolate these players in pots meaning i will raise preflop with a wide range of hands to get the flop heads up and play from there.


PFraise%- Is how often someone raises before the flop, this gives you guide to how aggressive they are. I would look at this stat and how close it is ratio wise to vpip. It is a contentious issue but to play a close to unexploitable preflop strategy in 6 max, your pfr% should be close to your vpip, i never open limp in 6 max for example. The brief reasoning for this is that it makes you hard to read as opponents can't decipher between the hands you limp and the hands you raise, also you will take the blinds more which is never a bad thing.

This reasoning should be applied to your opponents as well, if the pfr% is more than 1/2 (I would recommend that at least 3/4 is optimal) we can reason they are aggressive and in turn are generally a reasonable player. If it is less than 1/2, these players tend to be passive and when they do raise you can put them on a strong hand in relation to their range.

Aggression Factor- Aggression factor is worked out by dividing amount of raises or bets by amount of calls. This is your best guide to how aggressive/passive someone is postflop.

0-1 These players are passive they rarely bet without the goods and unless you have a strong hand when they do, get out of the way. However they do like to call, so when playing a hand even if it is something like top pair medium kicker, value betting is key. Don't be afraid to make bets against these players, they will call you with some amazingly weak hands.

1-2.5 This is the middling range and you can assume that they play relatively straightforward, bet their made hands, strong draws and only call with marginal type hands.

2.5+ Is aggressive and these players tend to bet made hands and non made hands. Rarely call, if they are calling there is normally a reason they are doing this.

These are based on total aggression factor which is across the streets (Flop/Turn/River), there are stats in PT that breakdown the stats by each individual street. Generally good players will have an aggression factor where flop>turn> river and weaker players will almost go the other way.

Total Number of Hands

Sample size is crucial when assessing stats, I would say 100 hands is the minimum you want to be drawing any real conclusions from anything. The first two stats and vpip start to converge at this point to give you a good general guide but they probably won't become stable until you reach 1000+ hands.

Aggression factor i would say don't pay great attention until you get to 500 hands as a guide, you can get situations where a run of cards means a player is calling more than often or raising more than often.


Ok and breathe i was going to write more but I think this is enough. I will follow this up with the more obscure of the stats if people find it useful. Any comments are welcome on what I'm covering, whether you disagree with me or just to say that i have the writing skills of a 6 year old ,English has never been an academic strength.

[kinboshi]

Fantastic stuff Longy.  Makes sense to me and is very helpful.  When I've read it again and had chance to digest it and think it through, I'll be back with some questions.

[Tractor]

Cheers, will be back with some questions later.

[kinboshi]

PFraise%- Is how often someone raises before the flop, this gives you guide to how aggressive they are. I would look at this stat and how close it is ratio wise to vpip. It is a contentious issue but to play a close to unexploitable preflop strategy in 6 max, your pfr% should be close to your vpip, i never open limp in 6 max for example. The brief reasoning for this is that it makes you hard to read as opponents can't decipher between the hands you limp and the hands you raise, also you will take the blinds more which is never a bad thing.

One leak that had crept into my game now successfully plugged.  Wasn't even aware I was doing it.  Noticed that my VPIP was a lot higher than the PFR - and I'm already feeling more confident that I'm playing better.

One beer owed  .

[rossfourfive]

Thanks for that, i've never known exactly what they all mean.

One other thing, when someone says a player is 20/2/2 or any combination of three numbers what stats is this referring to?

Also what do the auto rate descriptions actually mean in profitability terms at the table? Which players should be avoided and which should be farmed?

Thanks

[Longy]

Thanks for that, i've never known exactly what they all mean.

One other thing, when someone says a player is 20/2/2 or any combination of three numbers what stats is this referring to?

Also what do the auto rate descriptions actually mean in profitability terms at the table? Which players should be avoided and which should be farmed?

Thanks

If someone is 20/2/2 it means they have:20% vpip/2% PFraise/2 Total agression factor. I think this notation came about as this is how GT+ (game time+ a forerunner of pokerace hud) displayed the stats on the screen.

The auto rate description I have never been a fan of and have never found any that have been particularly useful to me, though of course you can tailor your own in Poker Tracker.

In terms of whether a player is profitable VPIP is easily the biggest factor. When playing around poker tracker filter one day that players with less 30 vpip in my 6 max database were beating the rake on average. Whereas players above this were losing around 15 big blinds every 100 hands. Of course this is not anywhere near the answer to being profitable but its the best guide among the stats imo.

[Ecosse]

Quick question,

How usefull is poker tracker playing mostly STT/MMT tournaments rather than cash ?

Dougie.

[Longy]

Quick question,

How usefull is poker tracker playing mostly STT/MMT tournaments rather than cash ?

Dougie.

No doubt its less useful than it is for cash. I still found it useful when playing stt's basically everyone who is semi-comptent in sngs play really nittish early and this leads to them having certain stats, so you can pick out the maniacs from the good players straight away.

If i was purely an online tourney player I could take it or leave it, as for cash I think its an essential tool.

[kinboshi]

Very useful on Stars (but not iPoker) for revealing the mucked cards at showdown on the table (using PA HUD with PT), and saving the hassle of having to view the instant hand history.

However, this is probably not a good enough reason to invest in both pieces of software!

[rossfourfive]

Very useful on Stars (but not iPoker) for revealing the mucked cards at showdown on the table (using PA HUD with PT), and saving the hassle of having to view the instant hand history.

However, this is probably not a good enough reason to invest in both pieces of software!

How do you do that?

Edit: Got it sorted!

[cooker3]

BTW, Anyone who uses PT2. I would strongly suggest downloading beta of holdem manager. It is far better then PT2 in every way.
PT3 beta should be out in next few weeks as well so keep an eye out on that as well

[ACE2M]

great post longy, look forward to the rest.

[Tractor]

What would  you say is a good win rate per bb/100 on Poker tracker for 6 max NL.
Mine is in the green but not by a lot and i only have a small sample size of hands at the moment (10k)
Would 100k hands give accurate readings for this stat?

Cheers

edit - ive just realised mine is set to Big Bets (limit) not big blind, so in theory im running twice as good as i thought ...i think???....am i?

[Longy]

What would  you say is a good win rate per bb/100 on Poker tracker for 6 max NL.
Mine is in the green but not by a lot and i only have a small sample size of hands at the moment (10k)
Would 100k hands give accurate readings for this stat?

Cheers

edit - ive just realised mine is set to Big Bets (limit) not big blind, so in theory im running twice as good as i thought ...i think???....am i?

100k hand is probably the benchmark for sample size. A good winrate all depends on limits your playing etc. A simplistic guide I would say anything in the green is "good" (remember only about 10% of online players are winners long term), over 2PTBB/100 is pretty solid and anything over 5PTBB/100 is crushing the game.

These are all in PTBB/100 which is actually 2BB/100 as poker tracker takes its reasoning from limit poker where its all worked out in big bets.

[Tractor]

Cheers Longy,
I thought that was what it meant, well im over 5ptbb/100 at the moment (25/50cNL) for 10k hands.
Doesnt seem much but at least my roll is still going up!
Ive done that 10k in a week so a few months to go untill i know where i am and then hopefully move up a level.

thanks again.