Cepheus calls and raises (#1 of 3) with AsAh
uhbw calls (10) with Js2s
10♠6♠8♠ are revealed.
uhbw checks
Cepheus bets
uhbw calls (10) i've been k/r'ing my strong hands here so mixed it up on this one
7♥ is revealed.
uhbw checks
Cepheus bets
uhbw calls and raises (#2 of 4)
Cepheus calls and raises (#3 of 4) why would cepheus 3b here? is it expecting AT or KK or JJ/QQ to k/r this board? this seems like a losing play here
uhbw calls and raises (#4 of 4)
Cepheus calls (20)
K♦ is revealed.
uhbw bets
Cepheus calls (20)
uhbw shows jack, ten, eight, six, two flush
Cepheus shows pair of aces with king, ten, and eight
uhbw wins 260, bringing their balance to 105

I'd either donk the flop hoping to 3 bet, or I'd check raise the flop hoping to 4 bet.

Once you don't do either, your hand is way underrepped and thus AA with the flushdraw is still strong on the turn enough to be a value 3 bet. Also since Cepheus has the Ace of spades blocking your nut combos, you will rarely be 4 betting thus it can go for thin value.

good point. flop underplay can invite overplay which is what i was going for.

this i think is false. HU every flush 4bets here given the play of the hand. having the As does NOT mean that i'm rarely 4betting.

Sorry my wording was imprecise.

You will 4 bet relatively less than you would if the Ace of spades in your range.

There are only 55 flush combos possible when Cepheus doesn't have the Ace of spades. Take away the Ace of spades from your range and that number reduces to 47. This must have some effect on the bot's value range.

i started playing around with this: http://poker.srv.ualberta.ca/strategy#

turns out those 8 combos make ALL the difference:



cepheus calls with all aces except when it has the As. so it calls w/ 1/2 its aces

Can’t the answer be that he raised the turn as a semi bluff and was planning to check back the river and just decided to **** it cal on the river

Nitpick: There's only 45 flush combos when Cepheus doesn't hold the As and 36 flush combos when Cepheus does (10C2 = 10!/(8!*2!) = (10*9)/2 = 45, 9C2 = (9*8)/2 = 36).

This being said, I'm still not sure why Cepheus elects to 3 bet here. Even if it expects you to mostly fastplay flushes, you should have a ton of 9x or 2 pair hands that are beating AsA (maybe it assumes you wouldn't xr a lot of 2 pair).

Also of note is that Cepheus will 3 bet AsK here.

Should start the questions backward by asking what am I doing wrong. Since Cepheus is assuming it is playing an GTO opponent/itself anything that derive from it is -ev. Before you even go any further to asking what you are doing wrong. Ask yourself what is the best way to learn how to play head up. I guess Cepheus website is free information but seem like a bad way to build up mental map structure since you are only getting incomplete pieces of small picture. Since you said you been mixing it up with strong flushes and decided to c/c with this one. What exactly mix % are you using and is it correct %. Second you ask do you think AT/KK/QQ/TT play like this and cepheus would never think you play like this.

Maybe I will post longer answer to how to solve this but short cut to why would AsA raise instead of call and the answer is the difference in ev is about +.1084.

Wow DonJuan, thanks for this. Please let me know if you ever write a book or publish a video on poker, GTO, or HUHU since I rarely play now but your posts are phenomenally informative.

With that said, how do you aporoach what the exact mix % 'Should' be between strong and weak flushes in real time HUHU, assuming you aren't multi tabling and playing vs one villain (Cepheus in this case)?

My guess is that EV calculations outside of playing done in the past should provide what the exact mix % is +EV / optimal play in similar spots vs a villain.

Specifically, whether it be villian playing GTO or close to it (Cepheus)

OR

vs a particular villain that may not be playing GTO (LAGTAG, TAG, LAG, Loose Passive, Tight Passive, etc)

Either way, studying EV calculations away from table provides estimate of what 'Should' be +EV with high probability if EV calculations are done often enough away from table vs particular types of opponents with various hand range assumptions in buckets based on how they play.

I don't really want to give too much information since it will hurt a lot of my win rate atm. There are software out there that will help with % raising or call with draws. Different board texture depending on ranges the % change quite a bit. However it seem like the +ev difference between calling or raising in a bunch of hands like draws are small. What important having a "balance" mix and when I mean balance I mean Cepheus are able to see the whole strategy/ranges while the best us human can do is mimic some kind of mix strategy.

Also I would start from playing pre flop as close to prefect then flop etc Should think of it as a game tree from top to bottom as you go lower the map get a lot bigger. People are trying to figure out the middle and bottom of the game tree when the top of their strategy are mess up.

I think this is true vs non adjusting players, but vs players that know how to exploit the ev difference should increase in delta.

cephus is probably trying to pay his rent.

Cepheus 3 bets the BB and then bets/calls down on a board of K8c6c5c3 with Ax9c

? The only hand it beats is 910o

Haven't sim your hand but mostly you raised on the flop which makes a lot of sense why it call down more.

Think of all the hands it can have on the river. Given the price its getting, it should only be folding ~1/9. That likely means, instead, your range doesn't have enough bluffs.

It sounds like you're giving up on the K86tt texture too much if Cepheus chooses to play this way. Having the 9c is also very relevant on this texture for the bot.