4

sorry for what is likely a straightforward question but when using equilab I see the following:

my hand: 8♥8⋄

board (flop+turn): 6♥8♣Q♣K⋄

If I calculate the equity of my hand versus:

  • A♣9♣ it is (me/villain): +/- 84/16
  • K♣J♣ it is (me/villain): +/- 82/18

since I can only be beaten by a flush both hands have the same number of outs so why is the equity different?

Radu Murzea
  • 3,527
  • 2
  • 24
  • 51
roel
  • 43
  • 2
  • How did you end up with this `84/16` equity? I understand the `82/18` though. Can you make it clear in the question? –  Aug 03 '15 at 21:27
  • Just looked at this quickly so correct me if I'm wrong but doesn't A9 give you 8 outs (A234579T) while KJ gives 7?? (23457TJ) – Chris Aug 05 '15 at 10:17

2 Answers2

3

I ran a simulation of around 5K hands and i found the following:

88 vs A9s (84% to win)

  • Four of a Kind, won 128 times (3%)
  • Full House, won 999 times (20%) <---
  • Three of a Kind, won 3113 times (62%)

88 vs KJs (82% to win)

  • Four of a Kind, won 116 times (2%)
  • Full House, won 900 times (18%) <---
  • Three of a Kind, won 3125 times (63%)

I think the "issue" lies to the fact that a K on the board and a K on villain's hole cards reduces the ability for the 88 to hit as many full houses against KJ than against A9 (only 2 kings left).

EDIT

Once again, i upped the sample size to confirm if there are any gaps due to small size above:

88 vs KJs (81% to win, 500k hands)

  • Four of a Kind, won 11334 times (2%)
  • Full House, won 90622 times (18%)
  • Three of a Kind, won 306948 times (61%)

88 vs A9s (84% to win, 500k hands)

  • Four of a Kind, won 11302 times (2%)
  • Full House, won 101518 times (20%)
  • Three of a Kind, won 307754 times (62%)

88 vs KJs (81% to win, 5m hands)

  • Four of a Kind, won 112983 times (2%)
  • Full House, won 909900 times (18%)
  • Three of a Kind, won 3069012 times (61%)

88 vs A9s (84% to win, 5m hands)

  • Four of a Kind, won 113070 times (2%)
  • Full House, won 1024030 times (20%)
  • Three of a Kind, won 3068647 times (61%)

I may be wrong or my code may be wrong, though the statistics are not lying; the 88 vs A9s hits around +2 % more full houses than 88 vs KJs. Additionally, if the KJ hits trips, we will hit full houses then.

  • 1
    vlzvl that sounds like the correct answer. I did not think of that, thx! May I ask what software you used for your simulations? – roel Aug 04 '15 at 15:24
  • This is wrong. The hands are 100% equivalent, and a 5000-hand simulation is just going to produce a big margin of error. Yes, against the KJ we will hit fewer full houses, but he will win with trips by exactly that same number of times more. Simulations are nice, but actual accurate math is better. – Lee Daniel Crocker Aug 04 '15 at 19:25
  • Hi Lee, it is not wrong. It is maybe not perfectly explained but he is right. In the K/J hand the villain has 8 outs while in the A/9 hand the villain has 7 outs (he loses one extra out - the Kc since that would make a full house - next to the 6c). – roel Aug 04 '15 at 20:56
  • @roel, this is a custom app, using [Cactus Kev](http://suffe.cool/poker/evaluator.html) evaluator –  Aug 05 '15 at 00:18
  • @roel, I edited the answer as well –  Aug 05 '15 at 00:19
  • Please read the question carefully. OP specifies the suits of every card. Your generic equities here are irrelevant. In the question given, each opponent has 9 outs, period. My answer below is correct. – Lee Daniel Crocker Aug 05 '15 at 01:47
  • @Lee Daniel Crocker, But the OP asked _specifically about the generic equities_ here, 84% vs 82% and i provided that margin through simulation and that's it. OP already confirmed he knows about the same number of outs, i said he had more outs? –  Aug 05 '15 at 02:09
  • That's not how I read it, but OK. Seemed to me like he was asking why the calculator gave different answers for two clearly identical cases. – Lee Daniel Crocker Aug 05 '15 at 02:16
  • While you're at it, consider switching your simulator code to ojcardlib. You'll be able to run millions of hands instead of thousands. :-) – Lee Daniel Crocker Aug 05 '15 at 02:18
  • @Lee Daniel Crocker, i'm very aware of your _excellent lib_ Lee ;) I'll take a look into yours and maybe someday raise a question benchmarking them ;) (however, mine can still simulate around 10 millions of hands in less than 45secs in i3 laptop, which is way enough than my likes, although im not fond of big sample sizes) –  Aug 05 '15 at 02:29
  • @LeeDanielCrocker Why don't you read the question carefully? You are wrong, period. With A9 the K is dead as it makes a boat. With KJ the K is not dead so it has one more out. – paparazzo Apr 22 '16 at 13:08
  • You're right. Ignore my comments here, KJ does indeed have one more out. Not sure why it took me so long to see that. – Lee Daniel Crocker Apr 22 '16 at 16:50
0

The hands are not equivalent. A flush beats trips but a boat beats a flush. The 88 can make a boat with a 6 or K. So the 6 or K on the BOARD are dead for both flush draws. Since one hand holds the K then it is not dead as it is not on the board.

52-8 = 44 cards out

On the A9 neither the 6 or K is good so there are 7 outs.
23457TJ
(44-7)/44 = 84.09%

On the KJ only the 6 is not good so the are 8 outs.
234579TA
(44-8)/44 = 81.18%

6K of clubs would have no flush block and pick up 2 outs to an over boat for 11 outs.

paparazzo
  • 6,908
  • 2
  • 21
  • 50