EV calc wrong with shortie AI multiway

[ongameplayer]

In the hand below, the guy with KK has 75% equity on the turn versus the A7h, but HM has worked out his All-in EV from the 5% he has against the A2.

HM worked out the EV$ diff to be -$86.55. Seems it should be something liek -$18 if my calcs are correct. (If you ignore the short stack AI on the flop.)


***** Hand History for Game 493564035889 *****
$50.00 USD NL Texas Hold'em - Friday, April 10, 07:42:00 ET 2009
Table Dschidda (Real Money)
Seat 2 is the button
Seat 1: highnoon88 ( $49.54 USD )
Seat 2: cateatsdog ( $53.79 USD )
Seat 8: Dirty Mendez ( $41.19 USD )
Seat 9: Palminho57 ( $10.00 USD )
Dirty Mendez posts small blind [$0.25 USD].
Palminho57 posts big blind [$0.50 USD].
** Dealing down cards **
Dealt to highnoon88 [ Kc Ks ]
highnoon88 raises [$1.50 USD]
cateatsdog folds
Dirty Mendez calls [$1.25 USD]
Palminho57 calls [$1.00 USD]
** Dealing Flop ** [ 2h, 2d, Jh ]
Dirty Mendez checks
Palminho57 checks
highnoon88 bets [$3.50 USD]
Dirty Mendez calls [$3.50 USD]
Palminho57 raises [$8.50 USD]
highnoon88 raises [$10.00 USD]
Dirty Mendez calls [$10.00 USD]
** Dealing Turn ** [ 4s ]
Dirty Mendez checks
highnoon88 bets [$34.54 USD]
Dirty Mendez calls [$26.19 USD]
** Dealing River ** [ Kh ]
highnoon88 shows [Kc, Ks ]
Dirty Mendez shows [Ah, 7h ]
Palminho57 shows [Ac, 2c ]
highnoon88 wins $99.23 USD from main pot

[morny]

What version of HM are you using?

[ongameplayer]

At the top it just says 1.07.

[morny]

Update to http://www.holdemmanager.com/downloads/Hm1.08Beta.exe and then go to Options > Fix Ev for current player

[ongameplayer]

Fixed. Thanks.

[aaltvater]

Full Tilt Poker Game #13786769429: Table moneymannc - $2/$4 - No Limit Hold'em - 5:33:27 ET - 2009/08/03
Seat 1: CoolFireX ($580.10)
Seat 2: zzyg ($797)
Seat 3: 030186 ($400)
Seat 4: edouardo63 ($352)
Seat 6: martin678 ($50)
Seat 7: smileyeeee ($86.55)
Seat 8: cacareco1 ($555.35)
Seat 9: TheChicken69 ($922.10)
030186 posts the small blind of $2
edouardo63 posts the big blind of $4
The button is in seat #2
*** HOLE CARDS ***
Dealt to martin678 [Ts Td]
martin678 raises to $10
smileyeeee folds
cacareco1 folds
TheChicken69 folds
CoolFireX folds
zzyg calls $10
030186 has 15 seconds left to act
030186 raises to $44
edouardo63 folds
martin678 raises to $50, and is all in
zzyg calls $40
030186 calls $6
*** FLOP *** [Qs 2s Ac]
030186 has 15 seconds left to act
030186 bets $110
zzyg calls $110
*** TURN *** [Qs 2s Ac] [9d]
030186 bets $240, and is all in
zzyg has 15 seconds left to act
zzyg has timed out
zzyg folds
zzyg is sitting out
030186 shows [Ad Kc]
martin678 shows [Ts Td]
Uncalled bet of $240 returned to 030186
*** RIVER *** [Qs 2s Ac 9d] [Ks]
030186 wins the side pot ($220)
030186 shows two pair, Aces and Kings
martin678 shows a pair of Tens
030186 wins the main pot ($151) with two pair, Aces and Kings
martin678 adds $80
*** SUMMARY ***
Total pot $374 Main pot $154. Side pot $220. | Rake $3
Board: [Qs 2s Ac 9d Ks]
Seat 1: CoolFireX didn't bet (folded)
Seat 2: zzyg (button) folded on the Turn
Seat 3: 030186 (small blind) showed [Ad Kc] and won ($371) with two pair, Aces and Kings
Seat 4: edouardo63 (big blind) folded before the Flop
Seat 6: martin678 showed [Ts Td] and lost with a pair of Tens
Seat 7: smileyeeee didn't bet (folded)
Seat 8: cacareco1 didn't bet (folded)
Seat 9: TheChicken69 didn't bet (folded)


hm says preflop equity 4,5% for tt could you please adjust that... i dont know how to make a good value for preflop but 4,5% just sucks

[netsrak]

Please mail the original handhistory with a link to this thread to support@holdemmanager.net

[Rvg72]

Hi, your EV is calculated from the Turn because that is the spot where everyone left in the hand is all-in. There is no way to know your all-in % from an earlier street since some players in the hand at that point did not show their cards

Roy

[aaltvater]

yeah but then you can give them directly 0,0% equity because they will only bet when they improve postflop.. so the equity is completely useless because it doesnt say that iam unlucky because 40% of the time an ace appeared on the flop which would be way over average... so for me the best approach would be to equilate AKo vs. TT vs player3 range(something like his cc range) it would be much nearer to the truth then the actual calculation