matlittle

That's an interesting idea! I checked the solver, and the EV of call/jam is almost identical. If you nodelock in some more folds with overpairs then the solver will just jam the T6cc. So you are definitely onto something. If we also factor in that we are unlikely to be able to play the river as well as the solver will (especially OOP), then it's probably higher EV for us humans to jam than call. The one thing that leaves me hesitant to try this play still though is the MDA data that says that the average reg in my games will just have more strong hands getting to this node to start with because weak regs take this bet sizing sequence more frequently with strong hands and use smaller sizings with bluffs. I will definitely look for more of this type of play vs good aggressive regs though!

Probably means I’m meant to sqz more high card hands and less suited cons and PPs but I think I prefer that.

My database shows I lose money squeezing the SCs and lower PPs, so that's probably a decent adjustment anyway. I think the reasoning behind it is that most squeeze opportunities IP involve a recreational player flatting in a position from MP to CO, and then calling too wide vs the squeeze.

GTOWizard recommends an 11bb size here (presuming I've got the right sim). Funnily though if you increase stack size to 125bb then it goes to 12bb squeeze size, then back down to 11bb if you go to 150bb stack size which seems very odd. It's because of reasons like this that I sometimes struggle to trust the GTOWizard sims.

Hey Alexandra, nice video!
In this 4bet pot, you mentioned that even if this weren't a range bet, you might opt to range bet anyway because of how people respond. Is this based on MDA research? Do you expect people to over-fold, under-raise, or both?

With these turn scenarios do you factor in any MDA based considerations? For example with this hand, villain bet 3/4 flop, and overbet on turn. I believe in my games this bet sizing sequence would lead to villain having a stronger range than they should on average.

For this hand the solver doesn't seem to like XRing the turn with T6s in my sim and instead opts to pure call. Having said that, both options are very close in EV. You could argue that therefore it's still a decent XR candidate, but then you either have to XF with some extra flush draw combos that can no longer call, or you end up over-bluffing.

I'm a bit confused about what you said about this hand - you said that AK is supposed to be a pure fold on turn vs the turn XR. I checked this in a sim and had AK as a pure call.

Seems like an odd line here - turn looks like a clear overbet spot for IP. What do you think is going on here? Misapplication of theory? An exploit that expects you to over-fold? I see this kind of line a lot from MDA bots, so is there a possibility this player is a bot?

Nice summary of the different types of mistakes Dan. Do you actively track which types of mistake you make, e.g. with a table and a tally for each type? Or just trying to be cognizant of potential types of mistake is sufficient?

So you would just use 1 sizing on the river of all in? Presumably the easy way to blunder the river is to not include enough nutted hands in the smaller sizing, so then the thing to worry about is our opponents finding exploitative check-raise bluffs on the river?

Interesting that you want one of the board pairs to 3bet bluff the river here. Presumably the BB bluffs board pairs, so SB doesn't bluff raise them on the river, so then the BB wants to have them to 3bet bluff as they unblock the SB's bluffs?

I personally think that most players will have a donking range here. I also would guess that most players go overboard with their donking frequency when they have Ad. I would also assume that most players would really lack bluff check-raises on the river. For those reasons, I would go for a 50% river bet, although its still very close even after those considerations.

You said for this river spot you didn't like shoving because you think BB will over call. I would agree with that, because I don't think anyone would fold 2 pairs there. How would you feel about using a smaller sizing to bluff instead? The solver likes a 50%/all-in split on the river in my sim, and facing the 50% sizing its fairly difficult to defend, with BB required to defend nearly half of the 2nd/3rd pair hands against that sizing.

Think you can just range bet this board in both ante and non ante games. It's a bit closer in ante games, but in non-ante games its a clear range bet.

Here you correctly pointed out that As4 raises less often than other A4. Because BTN is cbetting almost with range, they have to call a bunch of AsXo hands e.g. AsJd. A4 has those hands dominated, plus they will often check-down after calling the flop check-raise, so the A4 is able to effectively get value from those hands, hence why it is slightly beneficial to unblock those by not having As with your 4.

For this hand, you mentioned that you felt the squeeze size was too small. What would be a good size here in your opinion? Are your squeeze sizes derived from preflop sims?

Hello Callum,
Solid play as always and thorough analysis throughout. One thing I noticed though was that the tables are slightly blurred in the video. I'm assuming it's an issue with recording CoinPoker tables, because I think I saw the same problem in another video played on Coin Poker.

On this board, you were investigating whether higher turn cards might cause us to add a block sizing in the probing strategy for BB. From my research in probing (in 6 max), higher board cards lead to larger probe sizes. On high boards protection is less important, so checking is less punitive to paired hands, and probing is more polar. Conversely, if the board cards are low, paired hands concede more equity by checking, so probing pairs for a small sizing can be more attractive. What you suggested seemed to be the opposite of this trend, although the solver did not agree with you. So I'm wondering whether the mechanics of probing are the same in HU or not in general?

When you see this massive raise size that only reps a hand like 33, do you think that this player just has 33 the vast majority of the time and therefore this kind of spot is very under-bluffed?

Just saw the review in the solver and the solver liked the bigger raise sizing on the 2-tone board too. Is it a heuristic in general though that we should size down often on 2-tone boards compared to the respective rainbow board?

Enjoyed the video as usual! For this QT6t board, you said that on the 2-tone board you will use this 50% check-raise sizing, and on a rainbow board you will use a bigger one. Is this because the 2-tone board is more dynamic, causing our flop value hands to become devalued more often, so we opt to invest less in the pot before the river?

Hi Paul, nice video!
How are you finding the CoinPoker games? Are there enough tables running there to play 200nl at most times of the day? How soft is the player pool compared to GG for example?

Congrats on the results Gary and good luck at 500nl+. If you played 1m hands at RNC this year, how many hands would you aim for playing regular tables next year?

Ah right I see, if you plan to call down unimproved then I can see the logic! Must be a fun player to have at the table!

This hand I would have snap folded the flop, but you mentioned that this player might be spewy so you want to call it. My thoughts are that they will bet the turn way too often in general and especially after this flop sizing given that bigger bet sizings from recs are associated with higher turn follow through. If you pick up a heart and face a bet, you are going to want to shove, but then this player type will call the shove way too light as well which also decreases the EV of calling flop. Usually a player being spewy adds to our EV, but in this instance I'm struggling to see that being the case, unless you can convince me otherwise.

The hand in the video was SB vs CO, not SB vs MP. There was a player sat out at the table which may have made it look like it was SB vs MP, which is probably the source of the confusion here.

You mentioned in game that this kind of call doesn't tilt you anymore like it might have done previously. This is the exact kind of scenario that would send me on tilt in an instant, despite knowing that in the long-run my opponents mistakes are good for my winrate. How did you get to the point where this kind of thing doesn't really affect you anymore?

For this hand it looked as though you didn't use the RNG for 3betting this hand and opted to pure call it preflop? Can I infer from that that vs recreational players you will 3bet linear here rather than polar? If so, what do you think are the borderline hands that are good enough to 3bet here vs a recreational player? I am using a linear range vs recs but never quite sure exactly how strong a hand should be to 3bet here. Also I'm not entirely sure whether I want hands like 65s in there too at a small frequency vs a rec.

For this multiway hand at 16.55 you decided to just pure give up on river with bluffs, because you felt that J0hnMcclean wouldn't bet the flop with much air in a 3way pot. Would it also be true that he wouldn't bet the flop with many middling hands either, e.g. A5s or JJ? So then after he checks back the turn card, my guess would have been that he would have lots of backdoor draw hands that bricked turn, plus some TX?

Does the recreational player being in the pot sway you way more to giving up too? Would you expect them to call too often on river in multiway pots in general because they don't adapt enough to ranges being tighter?

(It was 3BP BTN vs UTG 15.05)