Some notes on STATS

[Jimbobobb]

Lately the forum seems to be overflowing with posts dedicated to personal stats generated by our much loved tool of the trade, Pokertracker. These posts are made with good intentions. Everyone wants to improve their play, plug some leaks, and get as much feedback as possible. Unfortunately, these stats just aren't the way to do it. Poker is such an incredibly complex and situational game. A page full of numbers is not going to let us know what it is that you're doing wrong, or even that anything is wrong in the first place. There are no magical 'target' stats that you should be shooting for. The level/texture/style of your game will change every time you sit at a new table, and you will adjust accordingly and play differently. Instead of stats, you should be posting hands. Hands you felt lost on. Hands you thought you played well. Hands you got outplayed on. Just hands and more hands. Read the responses to your hands and respond to those responses. Start a dialogue if you disagree with someone or want more information from a poster. Read other hands. Read the responses. Post responses to hands. Disagree with people and start writing a thesis on why effective odds are more important than immediate pot odds when you have a backdoor flush draw and an inside straight with bottom pair on the flop. THESE things and these discussions will make you a better player. If some of your stats in pokertracker are truly making you scratch your head, then feel free to post them, but include with them explanations as to why the numbers are the way they are, and why you think this is odd.

I'm just completely fed up with the:

Hands: 5,833
VPIP: 27.33%
VP$SB: 47.99
W$WSF: 28.0
BB/100: 7.24
W$SD: 52.98
PFR: 9.02
Does my VPIP and PFR look too high?

Seriously, how is anyone supposed to say *anything* that could be usefull or definitely truthfull about that block of numbers? What definites does it reveal? The answer is not a damn thing. Those numbers don't tell us anything about what you play or why, and tells us nothing about how you play post-flop, which is the *GREAT* majority of the game. Honestly if you want to know what these numbers should be, here goes....

Hands: Should be greater than zero
VPIP: Should also be greater than zero
VP$SB: Should be a real number
W$WSF: This number can indeed be imaginary, because it doesn't matter at all.
BB/100: my C*** is soooooo freakin huge. You have no idea
W$SD: Another real number here. None of that shitty 'i' stuff
PFR: This number should be less than or equal to your VPIP.

Instead try something along these lines....
my vpip is around 27%. These are the hands I typically play from early position.....mid position after a limper or two...late position after 3 limpers....I will open raise from mid and late position with...... etc. etc. etc.
This will produce much more useful feedback.

Don't just give us numbers. If you want information about your play in general, give us generalities about your game, how you play, what you play, and why. When it comes down to specifics, we want to see specific hands, not numbers. Pokertracker is a wonderful tool, and it can help you improve. The best way that it does that is by storing hands and letting you replay hands you had problems/uncertainties with. Do not become attached to the pretty numbers that it prints out. They mean nothing next to how you play. This is *not* meant as a personal attack on the people who have been posting stats lately, so please do not take it that way. I simply feel that too many people are placing much too much emphasis on these numbers, and it is not helping them to become better players in any way.

[Ajax410]

I agree with you on all counts but one.

I think that posting stats is a legitimate post if a player is concerned about moving up limits - I think often times statistics (VP$IP and PFR) are important stats to focus on when it comes to raising limits. Thus, if a player is concerned that he is too passive for 2/4 or too aggressive for 1/2 based on stats from a lower limit, I think posting these stats can be helpful. But, all in all, good post - I think we can all take something from it.

[Nottom]

You are such an ass.

[Jimbobobb]

I agree partially with this. But over a few 10's of thousands of hands, a few percentage points either way could simply be that you have really good or bad table selection and can loosen up or have been forced to tighten up accordingly. It's all just too easily skewed and the sample sizes are never large enough.

Tom eats babies.

[Ajax410]

Of course sample sizes are large enough - as bison always says - by the time you play 50k hands, only the last 15k are reflective of where you are as a player - we're all constantly learning and changing our game.

But I do think there is a place to share PT stats. Are they perfect? Of course not - that's why we post hands (and why we should all include a hand with our PT stats that we feel shows a leak in our game). But they are helpful in determining if our strategies will continue to work at higher limits, where the texture of the game has a tendency to change.

Ajax

[Jimbobobb]

[ QUOTE ]
Of course sample sizes are large enough

[/ QUOTE ]

I simply disagree.

The 10k and 20k numbers most throw out aren't enough to even narrow in on most of your numbers, and like you said yourself since we're constantly adjusting and improving our play, what we did 20k hands ago is most likely drastically different from what we're doing now.

I think strategies for playing are great to talk about and post. I don't think that the numbers show these.

[helpmeout]

I disagree here some stats are very important and can help you identify potential leaks.

Obviously you have to post hands as well but stats are definately useful.

Of course the difference between having a VPIP of 22% compared to someone elses 18% means little but if you are playing full ring games and your VPIP is 30% then this can definately help identify leaks.

Similar to preflop aggression most recommend between 6-9 slightly above is ok though. Mine was around 10 which helped me realise that I was overplaying hands like KJo.

Aggression on each street is also important. Someone with a 0.5 aggression factor on the river is obviously not betting the river enough.

Of course you have to post hands as well but stats definately don't hurt to help identify leaks.

[Jimbobobb]

Only in the most extreme cases can numbers give you any kind of insight. Nothing is every obvious or certain concerning the numbers generated by PT. That is their problem. Any amount of circumstances could account for any number it spits out. so I got dealt AK 300 times out of my last 1k hands, and my VPIP is 40% - are you going to tell me that I'm playing too many hands? The problem is people expect these numbers to point out all kinds of leaks and for everyone to have all of this wonderful feedback concerning them. They're just not that usefull or representative of how you play. I'm not saying they don't have any use at all. They're just not the end-all-be-all everyone seems to think they are. The bottom line is if you're making the correct plays at the correct time, not that you got your pre-flop raise % down to 8. When all is said and done, the people making the correct plays will be winners. Not necesarilly the ones with the prettiest numbers. Don't change the way you play because you want your numbers to fall into some crazy preconcieved norm. Changing your play should be based understanding of the game.

[Ajax410]

You and I are agreeing on almost every issue. The difference is this:

If I have played 20k hands at 1/2 and I am looking to move up to 2/4 (where the game becomes notably more aggressive and a little looser), a quick look at my stats could be indicative of any adjustments I need to make. If your PF raise is below 7, that might be something you need to look into.

The stats are not perfect - but a 20k sample size at a limit is a pretty good indication of general trends for a player. I think BB/100 is a silly stat (the more I play, the more I think this is true...I've been playing some good poker at 2/4 for the last 1500 hands and I have 10 big bets to show for it)...but I do think that, if anyone is planning on switching limits, stats can be extremely useful in identifying potential trouble points.

Ajax

[btspider]

Stats are best used to help identify which types of hands the person should be posting. I'd rather one stat post and 10 directed hand posts than no stat post and 20+ hand posts trying to figure out where they may be leaking.