4-of-a-Kind
Join Date: Aug 2007
Posts: 2,875
|
|
Quote:
|
Originally Posted by daven
hmmm, if the reason for prop-bet losing is being sick then iou $9.
|
Nope - I had missed two by the time I got sick and the third was meh. I've got an idea for September I'll post in the other thread soon.
Robb's FTC Theorem (1st draft): When a player bets or raises, he is committed to a rather narrow band of fold-to-continue ratios (FTC's). A player whose FTC is too wide is easily exploitable by raises.
Corollary 1: A player whose FTC is too narrow is playing cards face up.
Corollary 2: Hero avoids unbalanced lines with proper FTC's for his aggression levels.
I guess the most obvious result I could share is something I've been thinking about since Ben said it almost a year ago: "a player who cbet's more than ~70% is probably exploitable." Without equations and such, here's the brief sketch: a player who cbets ~70% REQUIRES that Hero start raising light: I'm looking for boards that shouldn't hit him and leave me at least some kind of draw. When he cbets ~80%, I don't need the draw any more. Ten high will do. And if Hero doesn't adjust, Villain's horribly unbalanced line profits, damn him. So Hero has a certain looseness forced into his game by the unbalanced aggression of others.
The problem for the 80% cbetting villain is that there is a ton of dead money in the pot, and it's pretty cheap (relative to the pot) for me to raise the flop. The dead money and cheap raising price mean that Villain has to fold less than half of his cbets to raises or he's toasty-cooked. And with a decent PFR, say 10% or more, it's almost impossible to continue often enough to sustain an 80% cbet rate. When that PFR gets wider, the situation just deteriorates - for Villain - Hero doesn't mind  .
I should probably mention when I say dead money, I mean anything already committed, even by the active players. Those chips are worth a stab if there's decent fold equity, and Villain's wide FTC's are the way I'm finding good places to leverage FE.
Another FTC situation is steals. Most TAGG's have really awkward FTC's for their 35% or greater BTN steals. Since the 3b FTC range is pretty wide, it's easy to attack them. I know most folks KNOW this already. I'm just saying that the maths I've done back it up. Where I'm drawing conclusions based on math, the outcomes seem reasonable based on the theory of adjustments we all know. So now I'm using my FTC's at the tables, searching for exploitable villains and lines.
I've been working on a range of playable FTC's for each aggressive action: 3b's, 4b's, squeezes, cbets, steals, resteals and so forth. I'm finding some interesting results.
First, 3bets have this big wide playable range of FTC's, which is probably why about any 3b% between 4 and 10 works OK given that the player is reasonably skilled.
Second, 4b'ing is extremely brittle. You have to have just the right mixture of 4b bluffs, if any. I guess the best way to say it is that the gap between "cards face up" 4betting and "exploitably loose" 4betting is pretty thin. There's no wiggle room.
Third, as the dead money in the pot increases, FTC's are surprisingly narrow, meaning that many TAGG's will have exploitably unbalanced lines.
I don't know if I'll ever post it anywhere but here. The more I think about it, run the numbers and work with it at the tables, the more I realize it's just my limited understanding of ISF's and Renton's theorems. I do have the maths to back it up, but it's pages and pages of spreadsheets with ranges, equity, combos. Pretty boring stuff.
The problem is that I'm not sure I could teach anyone else to "use" my theorem. I'm looking at a half dozen HUD stats, making some estimates based on stats and any actions I've noted, guessing at ranges and then applying my knowledge of FTC's. The cool thing is that it works pretty well. I can tell when I'm in "thin value" situations (if re-bluffs can be called value) and when I'm home free.
|