This site uses cookies to improve your browsing experience. By continuing to browse the website, you accept such cookies. For more details and to change your settings, see our Cookie Policy and Privacy Policy.

# EV of cold calling AK vs EV of 3betting vs UTG

• Black
Joined: 05.08.2007
Hi,

Would like to know if EV of cold calling AK vs EV of 3betting vs UTG, 11% open raise range. Somebody knows solution? This generally applies for any situation where we want to compare EV of cold calling and 3betting.

I think that it is nearly impossible to accurately calculate EV of cold calling because decision tree postflop is too complex. We could do database analysis
and compare winnings for cold calling AK and 3betting, and could conclude what works for us better. But what works in general for all players better? we can still make average winnings of e.g. 10000 players from different players and if e.g. 3betting AK has much higher winnings than cold calling then it is super likely that it has higher EV. Is something like this done? However it would be only experimental result which is hardly applied on other hands, open raise ranges etc.

How could we calculate EV of cold calling AK? I think we can for postflop use 2 random strategies, and calculate EV of AK on all possible boards (around 1.7 million boards) vs villain range. The EV of AK using random strategies for both players postflop, is approximately similar to EV of AK using to 2 strategies that are equally good.

In order to simulate this, we would need to iterate over all possible boards, and for each board iterate over all possible random actions postflop for both players, and for each set of random actions postflop calculate EV of AK vs villains range. Then we calculate average EV over all boards which is total EV of cold calling AK (assuming all other players fold behind us and we play HU, this is what I want to compare)
• 5 replies
• Black
Joined: 06.04.2008
Not sure this can be exactly calculated. We even don't have "perfect" method how to calculate EV. I usually 3bet hand for value if we're ahead or close to 50/50 of villain's 4betting range. 11% opener will probably 4bet KK+ and some bluffs, if not even AA only. So I'd say it's a clear call. We maybe might consider 3betting AKs, but still it's too weak IMO. Also I think we need some combos of AK in our flatting range even against 25% opener or we're in really bad shape when facing 3 barrels on A,K hi boards. Maybe I'm a nit, but I consider 3betting AKo only against 16%+ open and still will do it less than 50% of the time.
• Basic
Joined: 17.06.2014
3 betting AK UTG in a cash game would be considered as -EV due to the psychology factor of it.

For Villain to call your 3-bet @ 11%, they would be JJ+

If the flop comes and indeed hits your TPTK, then JJ+ wouldn't continue nor would QQ. KK and AA will have you dominated. So you're only win would come from AQ or possible AJ if they have that in their 11% range.

So out of 7 hands, you lose to 2, you don't get anymore from 2, you beat 2 and you chop with 1.

So you only win from 2 hands but at a flop % of hitting at 30%.

For me, I would flat call, and then c/r the c-bet if I hit.

This will cause them to think why I didn't 3-bet if I had AK.
• Bronze
Joined: 19.04.2010
I don't think calculating the EV with random postflop strategies would be good. Our edges base on the fact that our strategies aren't random but well thought and have good reasons behind them. If the randomised strategy gives a negative EV, that actually doesn't tell us anything, only positive EV would tell us our EV is positive. For example, calling 55 on the button vs UTG is +EV for us but if we play that 55 completely random on the postflop, it's probably -EV vs a non-random opponent.
• Black
Joined: 05.08.2007
Yes you are right. It would be needed to have real strategy how regs play.
• Basic
Joined: 29.06.2014
Against an 11% opening range I don't think you can 3b AKo for value. The only realistic worse hands that will call you are TT-JJ, AQs and you're unlikely to get any more money from these (unless you have a perceived wide 3b range range).