LAG results in SSNL

[APatterson]

Given the recent post on LAG play at higher limits. My PT stats for all the "slightly LAG preflop, Aggr post flop" auto ratings show these to be the biggest winners. They win bigger than the TAGs.

My problem though is my database isn't that big yet. I've been playing $1/2 on Full Tilt recently.

What do others with Poker Tracker see. Do the TAGs beat the LAGs/Aggr-post-flop in your game, for most money won?

p.s. wouldn't it be good for PT to show stats for each rating to allow us to asses the different styles? I instead resort to SQL and MSAccess to get these.

Also, I use the auto-rating settings linked to on this forum.

[dtbog]

[ QUOTE ]
Given the recent post on LAG play at higher limits. My PT stats for all the "slightly LAG preflop, Aggr post flop" auto ratings show these to be the biggest winners. They win bigger than the TAGs.


[/ QUOTE ]

This statement needs clarification; it may be meaningless.

For example: if the top 10 winners in your database are "slightly LAG preflop; Aggr-post-flop", then that's good for them; but if the bottom 20 losers also fit this description, then all you've discovered is that a LAG style introduces more variance. This is pretty obvious.

LAGs will win 'bigger' than TAGs, because they'll be the ones flopping two pair with ATo or flopping straight-flush draws with 76s... but many LAGs will also lose 'bigger' when they make one miscue that costs them a stack, or see a lot of flops and then miss a lot of draws.

Am I making sense?

[APatterson]

Sorry, I should have been clearer.
I meant the $ profit (total wins - total losses) was greater for "slightly LAG/Aggr-post flop" than for TAG. But I have a smallish database (<10k hands)

I was asking if others have calculated the total profit (total $ won - total $ lost) for each auto-player-category of PokerTracker. I think it would be interesting to evaluate the various styles on a large database of hands. It might help us all determine what seems to be the most profitable style for the games we play in, and the worst styles.

[wtfsvi]

</font><blockquote><font class="small">Svar på:</font><hr />
I think it would be interesting to evaluate the various styles on a large database of hands. It might help us all determine what seems to be the most profitable style for the games we play in, and the worst styles.

[/ QUOTE ] There's a huge pile of reasons why these results would not be trustable.

EDIT: Trustable? That's probably not a word /images/graemlins/tongue.gif I'm a foreigner.

[APatterson]

They are trustable as far as we use pockertracker stats for anything. I agree they're not definitive by any means, but we use them for our own study, and often on this forum to define the likely playing styles of our opponents.

Maybe it wouldn't be worth much, but we often talk of TAG versus LAG play, there was a recent post asking do you need to be LAG in the higher limits.

It may not be much, but we can at least start by looking at how profitable the LAGs versus TAGs versus how unprofitable the fish stats are. Our existing databases provide one means of defining what we mean by these and how profitable the players who have these stats are. It will not define exactly how they play - it can only lump all the players together by their stats.

Don't you think?

[elus2]

do this analysis when you have 100k+ hands. personally i think it is much more critical to figure out what situations call for LAG and what situations call for TAG play. adjusting your hand range to exploit your oppponents mistakes and using proper game selection will benefit your winrate a lot more than finding the holy grail of vpip/pfr/agg factor.

[wtfsvi]

Well, some reasons off the top of my head why the results would mean as good as nothing for deciding what style is "best":

One of the groups can consist of better players on average than the other. Or, perhaps less likely, of players with better game selection skills than the other.

A TAG that's running well will appear as a semi-lag in our limited sample size, and a cold decked semi-lag will appear as a TAG. Thus unrightfully adding "LAGs" to the winners and "TAGs" to the losers list.

The hands will be played against different oponents even if they are played at the same limit. This can somewhat skew the results since different play will be optimal against different oponents.

There are probably more. I don't think this would tell us much.

[excession]

Why not just read the analysis of AP10's 120,000 hand DB I did when I did my SSNL autorates?

http://www.andymcnish.btinternet.co.uk/newauto.htm

the short answer is yes - table for table sLAG's win about 30-50% more than TAG's

Of course they find playing 3-4+ tables harder without losing PTBB/100..

It seems that about 25% VP$IP is optimum on the Party $50NL tables..this is for full ring - a priori you would expect a higher Vp$iP to be optimum for 6 max..


That said once you get up towards the $200 tables, things may well change..