Printable View
I read a post where someone implied that he although he had 800 hands on someone he did not have much of a read. I was confused by this. How many hands does it take before you have a solid read on someone?
And how much different is it for different stats? Obviously VPIP and preflop raise are going to be more accurate with less hands than Aggression Factor or Won $ at showdown %.
HUD stats aren't great for reads period. Looking at hands opponents have shown down and reviewing their lines is much better.
But since most of us multi-table, sometimes all we have are HUD stats. I start taking pre-flop stats (VPIP & PFR%) seriously around 100 hands. But if someone's 80/25 over 40 hands, I still take that into account. With post-flop stats (Aggression, c-bet %, folds to c-bet %, WTSD%) I like to have 500+ hands. I basically ignore these stats with less than 200 hands and start considering them between 200-500.
These are rough guidelines off the top of my head, ymmv. An eye-opener is to look at how your stats vary from session to session. I find that in 100-200 hand sessions my pre-flop stats can vary a lot... VPIP 16-26, PFR 6-15... and post-flop stats even more.
This is one reason why I like Poker-Spy, because it records the last 17 hands and displays them on the screen as your playing.
Downside, I can only handle 3 tables at a time with it, since it takes up the bottom half of your display.
The other program I started using recently is PO.
Good program for the stats, but I am missing alot of betting patterns.
Good news!!! NOT SPAM :shock:
Poker-Spy is working on a HUD for there next version. I'm looking forward to it, and I am on the beta testing list. I hate when I miss the mucked cards, so this way I can check the last 17 hands quickly and maybe get a better read.
I find my notes are also better with Poker-Spy, since the betting patterns really "jump" out at you.
I just seen some guy playing on PartyPoker 25NL, and after 100 hands had a vp$ip of 4.0/ PF raise 1/ won at showdown 100%.
With those stats I would be scared to play anything but the nuts againt this guys bet or raise, even though it is only 100 hands.
Some stats really jump out at you and some don't. Just play your cards straight forward most of the time.
Also if you want to try something really scary, try playing without the HUD after using it for a while. :twisted:
(Naked Poker?? That the way it feels :wink: )
Quote:
Originally Posted by Massimo
more than a 100 hands ,,,,i recommend you datamine if your going to use HUD stats , i dont look at stats if i dont have atleast have a 100 hands on a player ... ,
in order to get good aggro calcs you need like 1k hands
I assume you are all talking about higher limits where people mix up more and play trickier? Because at the micro limits, many of the player stats are reliable after as few as 30 hands for me.
Quote:
Originally Posted by Wyvver
Ever went 30 hands without seeing a pocket pair or AK?
I know I have.
Sure but how often does it happen? Poker is a game of probability and incomplete information, so if I have 30 hands on someone and the stats tell me he is tight passive, yes there is a small chance that he cought nothing for 30 hands. But it is much more likely he really is tight passive, so I like my chances.
If you dont have the HUD stats (for whatever reason) and get only normal reads on a table, would you ignore your reads for the first 100 hands?
And with loose players it's even more likely that the stats are correct, if someone has 60% vpip over 30 hands, he is loose...I cant ignore the stats just because there is a very small chance he cought premium hands for 30 hands.
Also, I usually check the stats again after a larger number of hands, just to see how they change and sure, there are small changes. Someone might go from tight-passive to tight-neutral over a bigger hand sample, but I cant remember the stats changing from passive to loose or the other way around for example, just because the sample went from 30 to 300 hands.
after getting like 1k hands vs an opponent you can usually look at a few things like where they draw the line for raising pp's, how they like to play sets, and other stuff like that.
Wyvver, take a few random 30 hand samples from your own PT stats and you'll see how misleading 30 hands of data can be.