Using PT reads on opponents when you only have a small sample to go by

[imported_AlphaGun]

I have read a lot of posts lately where people are basically saying that when you only have stats for say 20-30 hands on an opponent you should ignore the read and pretend they are the average player for your limit/site. Although I agree that looking at stats for 20-30 hands is clearly not statistically significant on a hand by hand basis, consider this:

Player A and Player B are identical in every respect (site, limit, playing ability, reading hand ability, etc.) and are 4-8 tabling, hence they have very limited knowledge about their opponents, except for PT stats.

Player A and Player B both play 100,000 hands. Every time Player A comes accross the situation where they have no specific read on an opponent except for 20-30 hands in PT, he/she assumes they are the typical player for their site/limit. Conversely, everytime Player B comes accross the situation where they only have a read based on 20-30 hands in PT, they go by that read. Correct me if I am wrong, but Player B should come out ahead in the LR. This is becuase, any read, even if it is over a small number hands should have a slight incremental positive expected value above using no read at all. Aggregated over many hands this expected value could mean a lot.

Thus, when responding to posts it is better to go by the read given then to assume it is useless because the number of hands the read is based on is small. Clearly a read is always more accurate after a larger sample of hands, but in my opinion, we should always respond to posts based on the read given and not assume it is useless. For that specific hand it may be close to useless, but it is a mistake to ignore it in the LR.

Let me know what you guys think about this.

[JoshuaD]

I always take the read, and then give it as much value as number of hands I have. A read with 5 hands is pretty much useless, but it might tell me he's come in for a raise every time he's come in so far. That's at least something.

[Nick Royale]

[ QUOTE ]
Player A and Player B both play 100,000 hands. Every time Player A comes accross the situation where they have no specific read on an opponent except for 20-30 hands in PT, he/she assumes they are the typical player for their site/limit. Conversely, everytime Player B comes accross the situation where they only have a read based on 20-30 hands in PT, they go by that read. Correct me if I am wrong, but Player B should come out ahead in the LR. This is becuase, any read, even if it is over a small number hands should have a slight incremental positive expected value above using no read at all.

[/ QUOTE ]
If player B is assuming this read is 100% correct he will not win as much money as player A. But if player B realise that this read tells him very, very little about his opponent and adjust his game a tiny, tiny, TINY bit he will come out ahead of player A in the long run.

A player like me (vpip 14% PFR 11%) could have stats like vpip 30% and PFR 0% after 20 hands, and even worse than that. I could look like a complete LAG raising 30% of my hands and playing 40%. This will happen quite a lot and if player B consider me a LAG and re-raises me with a wide range of hands he will lose TONS of money. But if he adjust his game by reraising me with hands like AQo/AJ/88 (hands you'll otherwise would have thrown against a UTG raise) he will show a little bit more profit.

[imported_AlphaGun]

Okay, I believe I just figured out why my logic is wrong. When considering the specific read of the player, we must also consider the distribution of typical players that we know play at a particular site/limit. I guess before we know more hands we must give more weight to that typical distribution of player than we give to the player specific read.

I hope I did not mislead anyone in this post. I now know my logic was flawed. But at least I learned something.

[TStoneMBD]

this really isnt true. 20-30 hands can be plenty enough to gather enormous information about a player.

if the guy has a 60VPIP 5PFR after those 30 hands, hes obviously not a tag. hes a loose passive.

if the guy has 16VPIP 10PFR after 30 hands, you know you are dealing with someone who isnt loose. whether he is tag or weak tight is dependable, but at least you know not to isolate this guy with junk.

if the guy is 40/25 after 30 hands, you know hes a major lag unless hes just catching an enormous amount of good cards. but most cases he will just be very laggy.

[JoshuaD]

[ QUOTE ]
if the guy has a 60VPIP 5PFR after those 30 hands, hes obviously not a tag. hes a loose passive.


[/ QUOTE ]

Every tag has a couple consecutive 30 hand samples where he'll appear a 60/5. It's pretty unlikely, but it's not definite or obvious.