Poker Tracker Thought

[fnord_too]

I have been noodling around with how to use PT to help in tournies. I have not used it in any way with tournies yet, but I think I may ave hit on a good thought:

Make a 1st hour, 2nd hour, etc. database just for tournies. That is, for each hour you have a seperate database.

To date, I have relied only on notes I take durring a tourney and memory of specific players, but it would be really nice to know the stats on players that have not showed down any interesting hands. (And because I have a bad habit of multi-tabling when I am playing tournies, I don't take nearly the notes I should).

What do you all think? This should greatly mitigate the problem of having history from different tourney stages mixed together and blurring the picture.

The question that remains is what is a good breakdown? Certainly, the first parts of tournies (on PP) are pretty wild. Also, right after the bubble play seems to be overly loose in gerenral. I don't like the idea of not having breaks on the hour (since then I can download exactly the hands from that hour with no chance of error), but maybe the best break down would be something like:

1st hour
2nd hour
3d-in the money
in the money to 30 left
30 left - final table
final table

So, comments on the general idea and specific implementation would be appreciated.

[FloppedFlush]

Rather than keeping separate databases for each stage of the tournaments, you can filter the results based on the blind levels in the preferences tab in the game time window. That way you have everything in one database if you want to look at it all together. It also keep the importing a lot simpler.

For the first hour, look at blind levels 5-10 through 50-100 (or whatever they are for the site you're playing on) for the second hour, look at blind levels 75-150 through 200-400, and so on. Or even better, use a sliding scale as the tournament progresses by adding in the next blind levels and removing the older ones as appropriate based on the type of tournament and how much data you have on the opponents.

Just try not to look at pre-rebuy and post-rebuy stats together if any of the data could be from rebuy tournaments.

[gergery]

I think there’s too much statistical variation in the cards your opponents get to breakdown PT results at any level other than the aggregate. Your sample sizes would be very small, too small I think to be helpful.

I think its more helpful to use pokertracker to get an aggregate picture of how they play, and they use that to help predict how they’ll act in the future.

I’ll wait til I have 20 hands before even looking at data, and want 35+ hands before I start really relying on it.

For example, my notes will say:
“loose (33 VPIP), passive (PFR 5); will defend blinds (0% fold to steal), doesn’t steal (10%), good on flop (25% WSD, 80% wonSD), Loves chkrs (6%)”. So I conclude that he’s a decent player that likes to see flops before deciding what to do but makes good decisions on the flop. I’ll then pull up the blowout menu in the gametime tracker to see which hands he’s limping with/raising with when I actually need to make a decision in a hand against him.

If I’ve taken those notes during early part of tourney, then I just calibrate that for later on: he probably tightens up as the blinds get bigger, but I still need a better hand than usual to try to steal his blind, and need to watch out if he’s stealing mine, and never semi-bluff with position on a draw vs. him.

Note: I’m actually looking for a better way to evaluate how well an opponent plays on the flop, since going to showdown/winning showdown and Won when saw flop can be misleading. But they’re better than nothing. Ideas?

--Greg

btw, there was a player in my tourney this weekend called FnordDiscord—that you?