Wednesday, August 19, 2009

Lies, Damn Lies, and Statistics

Continuing the theme from earlier, I haven't had a lot of time to post to this blog. I still think a lot about poker, but unless I have 30+ minutes set aside to write, few of those thoughts actually make it to (virtual) paper. I don't see things improving in the near term, either. Work, family, life (and poker table time, of course!) are conspiring to shove this blog down on the priority list.

Okay, that said, I do have 30 minutes today, so I figured I'd jot down a few thoughts I've been pondering recently. Specifically, I've been thinking about PT3 HUD stats, and trying to make them more useful and user-friendly so that, in the heat of battle, I can find the key statistics that will help me make the right decision. Serendipitously, today's lunchtime poker strategy meeting focused on this same topic. So, with that in mind, here are some thoughts:

PT3's default pop-up stats are, for lack of a better phrase, too damn complex. When you click on a players stats at a table, PT3 gives up a pop-up that includes five separate tables, 4-5 columns per table, and 10+ rows of data. That's something like 250 stats that you have to negotiate to find what you're looking for. With the count-down timer ticking away on the screen, pressuring me to fold, call, or raise, my mind sometimes glitches and I end up making a bad play. Now, don't get me wrong; I like all the stats, especially the positional stat pop-ups, but I really only look at those numbers maybe 1 time out of a 300. More often than not, I never even click on the turn or river stats. A custom (i.e., much simpler) pop-up table is definitely in the cards (pun intended) for me.

Okay, next let's talk about what stats I use, starting with my on-table numbers:

  • VPIP. This is the key stat upon which most decisions begin to get made. Mine is the upper left, on-table (i.e., non-popup) stat I use. Mine is color coded to give me a quick idea of whether I'm against a tight, moderate, or loose player. A VPIP<20>
  • PFR. This is the second most important number, and can really help narrow down a range. If a player has a VPIP of 50, but a PFR of only 2, I need to be really cautious when he raises preflop.
  • AFq. This number defines how "aggressive" the opp is, but I really only use it for post-flop decisions. Dunno if this is smart or not, but I basically only look at it when the board is dealt and I have to figure out what my opponent's bet is telling me. To be honest, I could probably remove this from my on-table stats and move it to the pop-up category.
  • # of Hands. This is another really important stat. If the number is less than 50, I don't put much credence in the other stats. Consequently, I have this stat color coded: less than 50 is red, for caution, and greater than 50 is green (to tell me the stats are relatively valid).
  • $ Won. This stat is a double-edged sword. On the one hand, it tells me how my opponents are doing for the session and, therefore, gives me an indication of whether they're leaning toward tilt or not. That's the good news. The bad news is I tend to use my own stat to see if I'm up or down for the sessions. This, of course, if Results Oriented Thinking (ROT), and as we all know, ROT is not something we want on our brains.
  • Attempt to Steal. Self explanatory, really. If someone has a high number, I don't give him much credit and often re-pop him, or float at least. Similarly, if someone has a low number and is raising from LP, I tend to give him/her more credit.
  • Fold BB to Steal. I have stopped using the Fold SB to Steal, but I do use this one a lot to see if I should attack or not. Coupling this with VPIP is a generally a strong indicator of whether you should steal or not against the opp.
  • C-Bet Flop. I have recently started using this stat to determine whether the opp is for real or not when he c-bets third street. Unfortunately, I'm not sure what ranges of numbers constitute loose or tight c-betting ranges. Once I figure this out, I will probably color code the stat.
  • Fold to C-Bet. The flip side of the previous C-Bet Flop coin. I.e., this number tells me whether I should continue a bluff post flop or not, but again I need more data to figure out what is tight and what is loose.

When it comes to the pop-up stats, I'm still trying to find my way through the chaff to find the wheat. Numbers like 3bet %, cold call %, and W$SD all are useful and probably should be included. Other things, like Aggression Frequency isn't so useful (i.e., because I'm already displaying Aggression Factor on my on-table numbers. ). I'm also weighing things like Fold Steal Attempt to Blind RR numbers. Guys like Raster find this useful, but the Guru thinks its a weak stat to consider. I think the real answer lies somewhere in between; i.e., the stat is probably semi-useful to call up in the heat of battle, but only if you have hundreds (or maybe even thousands) of hands on the villain prior to the stat being meaningful.

Similarly, the positional stats are intriguing, but I'm not really sure how useful any of it really is. Unless I have 500-1000+ hands on a player, the positional stats are pretty useless. In fact, I probably need more like 2K+ hands to really make this number useful.

Oops, my 30 allocated minutes are up.... Gotta run.

All-in for now...
-Bug

No comments:

Post a Comment