Some thoughts on game selection

[Supwithbates]

What are your thoughts on utilizing players/flop?

[bozzer]

nice post.

i similarly found that avg pot size is of very little use. i generally use players per flop, and hands per hour as my guidelines, along with making sure there are some decent stacks on my right.

i've used the open a table strat a couple of times, and agree that it does fill up quickly, and is likely to attract poor players. i'll start using it more often now when i can't immediately find a soft table.

[shyturtle27]

Nice post. I don't pay much attention to average pot. What I do is go through the FT lobby and find the most players/flop where one seat is open. Then I'll highlight the table and if there's three or more people sitting with a full stack or more then I'll open it and join. It's really worked for me.

[Fiksdal]

[ QUOTE ]
What are your thoughts on utilizing players/flop?

[/ QUOTE ]

It works alright. Players who see a lot of flops are likely to be pretty bad postflop too. If you datamine for an hour or something before you play and then sit down at the tables with the highes VPIP, then obviously that's gonna work. The problem though, is that usually there will be a waiting list and stuff, because other thinking players do the same thing. The tables with like > 60% players per flop usually have a couple of people on the waiting list. So by the time you sit down, if you are lucky enough to have the fish still sitting with their money in front of them, you're gonna have to sit with at least a couple of TAGs too.

[HBomb]

This is a terrible post, post fails to deliver [img]/images/graemlins/laugh.gif[/img]

[DMBFan23]

just wanted to bump this and say that I've seen a lot of 12/10 types and 18/16 types sitting at tables with each other (FT $25 NL 6max) way too often, and some of them probably post here. pick good tables instead of autopilot 4-tabling!

[Ender Wiggin]

bump

[xd40]

eventually u just know most of the players and where to play at

[clowntable]

The HU portion of it can actually work in your advantage, too. It's a lot easier to tilt someone playing HU because once you outplay them or suck out on them egos tend to get involved.
If all goes according to plan you can have around +25BB, and a half-tilting opponent once the first people join.
Since we are micro players, I think it's pretty good to get in some HU practice for cheap, too [img]/images/graemlins/smile.gif[/img]

Another little bit of table selection I use is a variation on stack size. I usually like to sit with position on someone who has at least doubled their stack. One could argue that the more skilled players are more likely to double their stack and thus you'd prefer position on someone else however I have noticed an interesting fact (once again tilting). If you manage to stack someone who has doubled up before then their remaining stack is likely to go your way pretty soon, too. I guess people get frustrated because they had mentally booked that 1 BI as profit and then start to chase their winnings like they would chase losses.

----

I'd be interested to study table selection on a more formal basis, too. Let me propse a simple model and some thoughts. Note that I'm not mainly looking for directly applicable advice but more for a way of thinking about this (as can be seen by the model):
1) Hero can only play one table and is restricted to 2 hours of play per session
2) There will be exactly 100 hands per hour played at the table
3) A fishier table leads to a higher winrate
4) If a seat opens up it will be an average player leaving

Now the questions:
a) How would Hero determine the influence of the table composition on his winrate? Can we create some sort of heuristic like (#fish*1.2*avg stack of fish pool) - (#tags*avg stack of tag pool)
b) How long would Hero be willing to wait in the "Waiting List" to join a table vs directly joining another table were he estimates a lower winrate. How would different hands/hour (say same table at UB vs Stars) influence the decision
c) When would it be more profitable to directly leave the table to join a jucier one vs. playing the "free" hands that are left before having to post the blinds again?

etc.

This might be total nonsense but it just popped into my head and I thought I'd share it.
I guess the most interesting would be to come up with some sort of "juice factor" heuristic for tables (and then write some code to autojoin those tables etc :P).