Crutial miscalculation in W34s3l's video river play 101?

ConstantineSRConstantineSR BudapestRed Chipper Posts: 6 ✭✭
Starting at 16:30, when he states that OOP value bet with less than 50% equity to IP's calling range could be the optimal play, and
tries to prove this via CREW he compares two scenarios:

1:OOP value bets, IP can only call (no raising option added)
2:OOP checks, IP can bet or fold

So in the 1. option if we actually give the IP player the option to raise with a balanced range, wouldn't that lower OOP's betting expectation, and so disprove the whole concept?

Comments

  • LeChiffreLeChiffre NetherlandsRed Chipper Posts: 641 ✭✭✭
    You posted this twice for some reason.

    I haven't seen the video so I can't really say if Adam made a mistake, but if it helps I can show how it can be highest EV to bet expecting to have less than 50% equity when called.

    Suppose we have 99 on AK332 no flushes possible.
    Pot is 100 and effective stacks are 100.
    Suppose villain's range is 44, 55, 66, AQ and AJ.

    If we bet pot we get called by his entire range and he never raises, so we beat 18 combos and lose to 24. Cleary we don't have >50% equity when called.

    However, suppose that if we check he bets 66, AQ and AJ for pot. Now we don't have enough equity to call so we have to fold.

    Let's compare the EV of checking versus betting.

    If we check, we will get to showdown against 44 and 55 12 out of 42 times and fold the remaining 30 out of 42 times. The EV of a check is hence (12/42)*100 = 28.6.

    If we bet, we always get called, and we have around 43% equity. So we will get 43% of the final pot, i.e. 0.43 * 300 = 129. We pay 100 for this, so the EV is 29.

    Not much of a difference in EV but this is just an off-the-top-of-my-head example, and you could probably find better ones where the EV difference is bigger.

    Anyway, note how villain never raises my bet just like in Adam's demonstration. That seems to be key here.

  • ConstantineSRConstantineSR BudapestRed Chipper Posts: 6 ✭✭
    Hi,

    I realized later that I put it in the wrong topic so I reposted.

    As I see in your example, if villain has many hands which we beat and would call a bet, but wouldn't bet himself, our value bet can be better ev than to chk/fold.

    I suppose if we do this kind of bet with the purpose of folding to a raise (if there was money left) we could call this a blocking bet but with no stack left it's pretty weird.

    Either way it proves the point better than the example in the video, thanks.
  • TheGameKatTheGameKat Posts: 3,669 -
    edited December 2019
    To address the question in the title, no, there isn't a "crucial miscalculation." By definition there cannot be when carrying out such modeling unless one types something in wrong.

    Moderation In Moderation
  • ConstantineSRConstantineSR BudapestRed Chipper Posts: 6 ✭✭
    Apologies for the title. Better would have been minor missing info :)

    I didn't really get the point at that time since in the CREV model after the bet of 50 they both had 50 left in their stack, but IP had no tree branch to use it. Hence OOP's bet in fact is a blocking bet which wasn't really the point to prove that a blocikng bet can be more profitable than checking.

    But if we treat it like an all-in scenario it proves the point so that's what was missing for me until LeChiffre posted another example.

    Anyway I'm a huge fan of W34Z3L's videos, just needed this missing peace.
  • TheGameKatTheGameKat Posts: 3,669 -
    One sign for me of the value of w34z3l's videos is I frequently pause them so I can reflect on what he's just said.
    Moderation In Moderation
  • LeChiffreLeChiffre NetherlandsRed Chipper Posts: 641 ✭✭✭
    I didn't really get the point at that time since in the CREV model after the bet of 50 they both had 50 left in their stack, but IP had no tree branch to use it.

    You could also consider this as part of a sick read on the IP player, that he would never raise in that spot and only call or fold. In other words you could create this branch but assign no range to the raising node. Then you'll end up at the same conclusion (that betting while getting <50% equity when called could be optimal).
  • TheGameKatTheGameKat Posts: 3,669 -
    LeChiffre wrote: »
    I didn't really get the point at that time since in the CREV model after the bet of 50 they both had 50 left in their stack, but IP had no tree branch to use it.

    You could also consider this as part of a sick read on the IP player, that he would never raise in that spot and only call or fold. In other words you could create this branch but assign no range to the raising node. Then you'll end up at the same conclusion (that betting while getting <50% equity when called could be optimal).

    Months ago I played around with this and found there's a semi-plausible scenario, in which the bet-raise line has few enough combos that the bet is still profitable leaving this channel open.

    My takeaway from the video was that river play IP and OOP are very different beasts, and that with the possible exception of the c-bet, we should stop giving names to bets.
    Moderation In Moderation
  • LeChiffreLeChiffre NetherlandsRed Chipper Posts: 641 ✭✭✭
    TheGameKat wrote: »
    Months ago I played around with this and found there's a semi-plausible scenario, in which the bet-raise line has few enough combos that the bet is still profitable leaving this channel open.

    Yes, good point, IP's raising range doesn't even have to be empty.

Leave a Comment

BoldItalicStrikethroughOrdered listUnordered list
Emoji
Image
Align leftAlign centerAlign rightToggle HTML viewToggle full pageToggle lights
Drop image/file