How many hands in a sample size is sufficient to trust the stats?

VitAlvM

VitAlvM

Rising Star
Joined
Sep 9, 2021
Total posts
5
hi guys, i'd like to know how many hands do you start trusting a player's stats with
 
Nafor

Nafor

Legend
Loyaler
Joined
Dec 31, 2019
Total posts
1,040
Awards
1
Welcome CardsChat VitAlyM,

You usually need a very large sample of your opponent to make good decisions. Hundreds of hands (in my experience) has given me some clue of what the opponent might do. It is not that easy though to gather a large sample from the same person if we are playing in random games with random opponents.

Rather than relying only on stats you should make notes too. That way you can make it more clear to yourself how some player has acted in some situation.
 
F

fundiver199

Legend
Loyaler
Joined
Jun 3, 2019
Total posts
10,303
Awards
1
As KristaK say, some stats will converge faster, because they happen every hand, while others might take tousinds of hands to become meaningfull. If in dought its always a good idea to look at the number of times, something has happened, and if its only one or two, then the stat will be completely meaningless. Like someone folding to flop C-bet 100% of the time, but its two out of two opportunities. Thats not a green light for you to go ahead and C-bet like crazy.

Also stats, that are way out of line, will be meaningfull earlier. If someone has stats like VPIP 91 / PFR 9 after 11 hands, then this player is with 100% certainty a huge whale, and you can start making adjustments already. Maybe his numbers will eventually drop to for instance VPIP 58 / PFR 12, but there is no chance, that this is a good player with reasonable preflop hand selection. Whereas if someone has stats like VPIP 0 / PFR 0 after 11 hands, they might just have been card dead, and you can not yet conclude, that this player is a nit. To draw that conclusion you probably need to have at least 50 or even better 100 hands.
 
VitAlvM

VitAlvM

Rising Star
Joined
Sep 9, 2021
Total posts
5
Welcome CardsChat VitAlyM,

You usually need a very large sample of your opponent to make good decisions. Hundreds of hands (in my experience) has given me some clue of what the opponent might do. It is not that easy though to gather a large sample from the same person if we are playing in random games with random opponents.

Rather than relying only on stats you should make notes too. That way you can make it more clear to yourself how some player has acted in some situation.

Thanks for the reception,

I've been using hud for a few weeks now and it's really not easy to gather a large sampling of opponents.
To make my decisions I have taken notes and in some situations I can use statistics like Vpip, PFR that end up having some kind of frequency
 
F

fundiver199

Legend
Loyaler
Joined
Jun 3, 2019
Total posts
10,303
Awards
1
I've been using hud for a few weeks now and it's really not easy to gather a large sampling of opponents.

That will come with time, so for the meantime just use the HUD for the few things, it can be used for with limited data. Like identifying the biggest whales at the tables, like I already explained. Often that is the most usefull information anyway, since the whale is the one, you need to make the biggest adjustments against. He can basically be the one making it profitable for anyone else to be on the table, and since he play so many hands, you are going to be involved with him way more than the other players. If you play cash, a main advantage of the HUD is also to help you identify, if you even want to be on that table. If there is no whale, then the answer is probably "no", and you just leave. This will already increase your winrate with several BB / 100.
 
D

dregan

Legend
Joined
Nov 12, 2018
Total posts
1,039
Awards
1
I think from 100 hands and the more the better. And still it will not give 100% certainty.
 
Starting Hands - Poker Hand Nicknames Rankings - Poker Hands
Top