Thoughts on All-in EV graphs

    • redrawstump
      Joined: 09.05.2012 Posts: 244
      Hey boys and girls,

      I've recently been thinking about the all-in adj line on Poker Tracker 4. I know lots of players treat this as the "truth" and judge this vs. winrate to determine whether they are running well or poorly. I've just lost a pot with KK vs JTo, as below:

      BTN: 101.4 BB (VPIP: 0.00, PFR: 0.00, 3Bet Preflop: 0.00, Hands: 4)
      Hero (SB): 176.2 BB

      Hero posts SB 0.4 BB, BB posts BB 1 BB

      Pre Flop: (pot: 1.4 BB) Hero has K:diamond: K:heart:

      fold, fold, fold, BTN raises to 2.4 BB, Hero raises to 7 BB, fold, BTN calls 4.6 BB

      Flop: (15 BB, 2 players) J:spade: 3:club: 6:club:
      Hero bets 9 BB, BTN calls 9 BB

      Turn: (33 BB, 2 players) 2:club:
      Hero bets 17 BB, BTN calls 17 BB

      River: (67 BB, 2 players) A:club:
      Hero checks, BTN checks

      Hero shows K:diamond: K:heart: (One Pair, Kings) (Pre 84%, Flop 78%, Turn 70%)
      BTN shows T:club: J:heart: (Flush, Ace High) (Pre 16%, Flop 22%, Turn 30%)
      BTN wins 64.2 BB

      Now I know shit happens but my PT4 graph green and gold lines both take a dive on this hand by the same amount. Even though I put in all of the money when I was ahead. By rights I should have (14*84% + 18*78% + 34*70%) in BB as my Sklansky bucks for the hand. I've fiddled around with PT4 and you can change the way All-in EV is calculated to include only all-in situations but that seems simplistic. I know that it's impossible to calculate EV for no showdown (working on presumed ranges is waaaay out of the question for the software currently) but I think that the current calculation doesn't give a very accurate reading.

      I guess it works the other way round if you suck out on someone. But given that a successful poker player should be trying to extract value from worse hands when playing fish, who are the main revenue stream, then I don't think this will balance out over any sample size.

      Just my thoughts, I'm open to correction and discussion.

      keep it foolish :pokerface:
  • 3 replies
    • ColorFlusher
      Joined: 19.01.2008 Posts: 842
      It balances out when your opponent doesn't hit his outs. Now the river was unlucky, but winning the pot would've been lucky in terms of ev. Most importantly there was still action left on the river so comparing your ev and the result in this spot is irrelevant and I wouldn't recommend using time on it. There's no way to calculate ev on any other spot than all in situations.
    • redrawstump
      Joined: 09.05.2012 Posts: 244
      Thanks. I was forgetting that we do better than ev when winning the pot, so it all evens out in the end.

      I still think there may be some mileage in calculating our equity share on each street as a better short-term indicator of how we are running. Not that it matters in the end...
    • Kyyberi
      Joined: 09.07.2010 Posts: 10,467
      As it was said, it's impossible to calculate EV when there are still actions to come. No application or formula can predict your actions. If you manage to calculate your EV by the equities on each streets, it doesn't understand that you can actually call without the raw equity and bluff the pot on the next street. So by raw equity you are making -EV play, which is +EV in reality.

      EV adjusted winnings graph isn't the truth about your EV. Neither is winnings. But that EV adjusted winnings is a little closer to the truth.