Comment | d2themfi commented on deep MATH software, which one to take CardRunner EV or Propokertoolsoracle?
assuming you want to spend all the time to learn how to use it, CardrunnersEV, its not close
March 8, 2013 | 3:34 p.m.
Teddy, what you can do is choose 1 turn betsize, 1 turn check raise size, 1 river betsize, 1 river checkraise size. Then you input one player's strategy (guess), and maximally exploit it (you'll want to get detailed on the board texture notation for accuracy for unknown river). You see where its deficient, and make adjustments so its not as exploitable as it just was. Rinse and repeat about a thousand times, adjusting for card removal factors, and you have a near equilbirum for two set ranges on the turn, with specific betsizes. Not going to share anything atm, maybe in the future
March 4, 2013 | 3:27 a.m.
dont use checkdowns unless you are just looking for something really quick that doesnt need to be very accurate to help you. This isnt the case for most preflop decisions in NLHE. At the very least I would simulate the flop with some board texture conditions. Ideally you could do all 3 postflop streets but just doing the flop is probably the best time/reward ratio, especially if you have experience doing all 3 streets and have a feel for what % of equity different hand types will realize in different scenarios (SPR, range composition, position etc)
Feb. 28, 2013 | 6:25 a.m.
you cant just pick a 3bet range, figure out what range you need to 5bet and assume thats GTO or even balanced. I think in general its better to think a bit exploitatively with preflop decisions because so much of the game tree is unknown at that point in the hand that you have to make potentially very inaccurate estimates about what is optimal postflop for both players, or worse yet make guesses at what % of a hands equity it realizes in various postflop scenarios. When we are playing we almost always have a decent sample for preflop stats, and if we have some postflop stats even better. For 3betting, obviously theres things like open %+fold to 3bet%+4bet% and his fold/raise cbet stats in 3bet pots, but you also have to factor in how he plays in SRP, since that can change your thresholds for which hands you want to flat and 3bet. Add in figuring out thresholds b/w 5betting and flatting 4bets, and you've got a complex problem, that you only get an acceptable answer to by modeling
Feb. 28, 2013 | 6:18 a.m.
Its not terribly difficult in CREV, esp with 40bb since the postflop modeling is a lot easier. It would take several hours though to get something close to an equilibrium unless you get lucky and guess very close to it at the beginning, since you will have to do a lot of back and forth with the two players strategies
Feb. 28, 2013 | 6:05 a.m.
You are somewhat close. Betsizes determine the opponents range that continues, and you can base you v-bet range based off that. If theres more than 1 bet left, and your opponent can checkraise hands for value, you'll need more than 50% equity vs his non-folding range to bet IP. And ofcourse you can find the range he can checkraise for value(and corresponding bluff combos) with the same method. With some adjustments back and forth you can find the equilibrium given 2 set ranges, and 1 discrete betsize for each player (well more than 1 if you have time)
Feb. 28, 2013 | 6:01 a.m.
theres too many factors to give exact advice to apply generally. Off the top of my head the main things that matter- range composition of both players (both overall equity and nuttiness/polarization or lack thereof), nature of the board-how closely do equities run on that specific board, betsize, and expected action on future streets.. all of which is impossible to model together except with CREV. You can solve for pretty accurate solutions with the software though
Feb. 28, 2013 | 5:55 a.m.