Well, this read is actually pretty straightforward, and I think pretty much everyone is sensitive to it, but I find it very difficult actually to nail down a lot of the time: Who buys pots in LP?

Specifically in PLO, it seems to me VERY important because it's going to be the basis for deciding whether it's better to check-raise or bet out. Surely others besides me have had this happen: I think LP is pretty loose and will bet, hence check my top set. Then it gets checked around AND the flush hits the turn with betting from a credible player. Well... maybe he would have called my bet anyway, but it's aggravating.

Anyhow, I think the reason it's a lot harder for me to get this read rather than others (like "pursues thin draws," "slowplays," "bets only made hands," "loose/tight hand selection," etc.) is that the situation actually would arise typically only once every 9 hands even if the particular player were play every hand in LP. And if the player makes what looks like a buy attempt even twice in a row, that can also be that he just had a hand those times.

Also, I think almost everyone is willing to bet thinner values in LP, but the question is just how thin. The chronic LP buyer is one imo who is going to bet to checks 80% of the time or more over a decent number of hands.

So, my project is going to be this: With every player in my database on whom I have at least 100 hands (I might change that number a little depending on how this turns out), get an actual percentage on how often they bet to checks in LP, then given them an icon I can easily recognize based on that percentage. Around 80% I'm going to take as chronic LP buyer whom you need to check-raise on your big hands.

Something that also belongs in this thread: I'm thinking that a good percentage to have here for oneself is probably somewhere in the 50% area--maybe up to 60% if the table is pretty straightforward and on flops that aren't highly multi-way, and maybe as low as 30% (which is getting close to the value bet range but no doubt throws in a fair number of hands you wouldn't have bet in EP) on multi-way flops and/or tables with a lot of slowrollers/trappers.

Interesting idea, I agree this information could be useful. I don't think it's possible to do this using auto rate rules, but I can easily write up a database query to get this information. Let me know if you want the query text (you need either Microsoft Office Access or be familiar with postgreSQL tools). It'll give you a list of players with bet% when checked for x positions off the button, for each street, filtered by the number of hands you specify. Or I can just add it to my program if you'd prefer, in which case I can automate the changing of icons.

Hmmmm...

I do have Access and know the program pretty well, but it's been a while. I do have some trouble writing out the actual code unless it's real easy (more along the lines of those docmd.openform... things and stuff like that, which I've used more for organizing windows in a few access databases I use).

I'll PM you my email address, and maybe we can take it from there.

What I've also done before (but am pretty cautious about because I don't want to screw up my database in weird ways) is open pokertracker in Access. Anyhow, as amateur and non-programmer, I'm often just winging it on this kind of thing but can usually get it figured out if it's not too complex--but I don't really understand which programs are doing work and really how the Access code does exactly what it does (I always use the Query-builder windows, etc. and have never really understood just how you could create your database out of a bunch of SQL statements and that kind of thing).

Anyhow, thanks very much for the offer, and I'll PM you my email address in the hope that you can walk me through it enough that we can get it to work.

aggression frequency actually does a pretty good job of spotting the excessive stealers.

interestingly, from a game theory perspective, there should be a fair bit of stealing going on in pot limit games when it's wrong for you to call down light or resteal.

This is a really interesting query. Here's the results from my database (50,000 of my own hands). BetFlopLP represents the percentage of the time a player bet the flop from the CO or Button when checked to. Hands represents the number of hands the player played from the CO or button.



Looks like I outed a few BTPers too I come in at a piddly 39. What's interesting is that it doesn't always correlate with AF as Kuso claims, there are definitely some hardcore opportunistic stealers who aren't otherwise notably aggressive.

Here's the query text. I borrowed the VPIP and PFR queries from the pokertracker forum.

------------------------

SELECT players.screen_name AS ScreenName, round(SUM(game_players.fbet)/(sum(game_players.fbet) +
sum(game_players.fcheck)),2)*100 AS BetFlopLP,
iif(SUM(game_players.fcall + game_players.tcall + game_players.rcall), ROUND(SUM(game_players.fraise + game_players.fbet + game_players.traise + game_players.tbet + game_players.rraise + game_players.rbet) / SUM(game_players.fcall + game_players.tcall + game_players.rcall), 2), SUM(game_players.fraise + game_players.fbet + game_players.traise + game_players.tbet + game_players.rraise + game_players.rbet)) AS PostFlopAF,
ROUND(AVG(game_players.vol_put_money_in_pot), 4)*100 AS VPIP, ROUND(SUM(game_players.praise) / COUNT(game_players.praise),4)*100 AS PFR, COUNT(game_players.game_id) AS Hands
FROM (players INNER JOIN game_players ON players.player_id = game_players.player_id) INNER JOIN game ON game_players.game_id = game.game_id
WHERE game_players.off_the_button < 2
GROUP BY players.screen_name
having COUNT(game_players.game_id) > 100
--------------------------

copy and paste this into the sql section of the new query window in Access, then click run. I don't think it works on postgres. The important part of the query is this:

SUM(game_players.fbet)/(sum(game_players.fbet) + sum(game_players.fcheck))*100 AS BetFlopLP

this says "SUM all flop bets made by a player, then divide by (the number of times they bet + the number of times they checked)" .

Note that the game_players.fbet column refers exclusively to first in bets, otherwise it is recorded under game_players.fraise. The WHERE clause down the buttom restricts it to the CO and Button.

This was a great idea Ais, I'm going to add some advanced queries like this to my program after I get the initial release done.

hmmmm, so the player that you have the most hands on, YouGotScoopd, only bet 44/1,064 = 4.1% of the time that it was checked to him in LP on the flop? is that right? it seems a player would be betting more often than that just by value betting alone. what a rock!

edit: whoops, nevermind. it pays to actually read what you write. 44 would be the % of time bet when checked to on the flop and 1,064 would be the sample size. sorry for knocking you, YouGotScoopd. you're actually a TAG!