RERUN POTD #103 A Real Tricky Hand in Madrid

A Canadian, a Hungarian, and a Dutchman play poker in Spain

At the conclusion of the post there will be a #onemorething and a video of me analyzing the hand that I shared with Premium Subscribers in the Discord channel. There are actually two versions of me discussing this hand, but the first one used a busted sim. Premium Subscriber Brian Hastings noted that I had copy and pasted the wrong ranges and I created a whole new sim and video for those subscribers. If you’d like to become a Premium Subscriber you can do so here or you can pay for an hour of private coaching with me and contact me at any of the methods outlined here.

I was legitimately thrilled that I got to write POTD #88 about a hand featuring pocket eights. I thought that once I cracked triple digits, I would no longer be able to play this matching game, but I neglected that #102-#109 can also be read as 10-2 or 10-9. Of course, like any reasonable person, when writing about poker hands I stylize “10” or “ten” as a T, but for this week and next, I will make an exception. There were no Doyle Brunson hands to write about and no “ten four good buddy” hands to write about. I wasted a T7s hand yesterday, but in the process found another T7s hand that I will write about next Tuesday. However, I did have a T3s hand that I’ve been meaning to write about for quite some time, but I struggled to find the exact HH or event that I played the hand.

Fortunately for me, the more tilted I am about a hand, the more specific I am when asking for advice from friends. If a hand is close and interesting, but I am making conversation on break, I might start off with “We’re 25-30 deep, I raise in middle position, I think the LJ.” If I’m steaming mad about a hand, you’re getting “26.5bbs deep I raise to 2.2x in the HJ, the BU has 15bbs, but everyone else covers me.” Well, I was steaming mad about this hand, so while I had to guess what blind level we were in, I am confident every other detail about the hand is correct. The other bit of kismet here is that I had yet to match the Dara O’Kearney hand I wrote about in POTD #41 with a hand of my own. That hand was a three-way pot where Dara had T3s; today’s hand is a three-way pot where I had T3s. What a great hand for POTD #103; just ignore the steam coming out of my ears as I recount this three-year-old hand that I’m still hopping mad about.

Triton Madrid € 30,000 No Limit Hold'em - 7-Handed (Event #2)
(2.5k/5k/5k) (SB/BB/BBA) Starting Stack 200k

Andras “Probirs” Nemeth (125k) makes it 11k UTG7, it folds to Tom Vogelsang (500k) in the CO who calls, it folds to me (400k) in the BB and I call T♥️3♥️

Flop (40k) T♣️3♣️2♦️: I check, Andras checks, Tom bets 13k, I call, Andras calls

Turn (79k) K♥️: I check, Andras checks, Tom checks

River (79k) 4♣️: I check, Andras bets 60k, Tom calls, I fold.

Andras has AK (I forget if he has a backdoor, I was too mad to notice). Tom mucks (or shows, I was too mad to notice, but I believe he said or showed K♣️J♦️)

What I Was Thinking

I played zero raise with range on the flop with Andras behind. I felt Andras would raise a lot of his continues, including sometimes playing shoves, which would trap Tom Vogelsang in the middle and allow me to potentially win a massive pot versus him. If Andras folds on the flop, my range looks quite weak, with a lot of draws, second and third pairs, and weak top pair, so I thought Tom would put a lot of pressure on me if we got heads up to the turn.

I figured I would lead some turns, but mostly 2x/3x and cards that filled flush and straight draws. An offsuit king seemed like one of the worst cards for my range, so I checked and it checked through. On the river, it seems like one of me, Andras, or Tom has a flush or a straight a lot. I’d expect Tom to bet the turn with top pair whenever he had it on the turn and thought it was hard for Andras to be bluffing or value betting worse than T3. So I folded. Andras had AK and Tom had KJ and I was shell-shocked that I had folded the winner.

What I Got Wrong

Technically, not a whole lot. It appears that I can split off a raising range on the flop with some two pair, sets, flush draws and gutshots, but just calling on the flop is fine. I pure check the offsuit king turn. On the river, I contemplated blocking, but it seems like a lot of what I am trying to get called by is worse Tx, which I block, and I’d rather check and see what Andras and Tom do before putting any money in the pot myself. Once Andras bets and Tom calls, in solver land I need the nearly perfect parlay of Andras bluffing with AQ/AJ/QJ with a club and Tom hero calling with 55-99 with a club or a ten. My hand is supposed to have 5% equity, which is significantly less than the ~21% I need to breakeven.

Andras is not supposed to value bet AK, and Tom has a reasonable hand to bluff catch with on the river, but he’s never supposed to check back the turn with top pair. Given what they showed up with on the river, I ranged my opponents incorrectly, but the question is, was I so off in how I ranged them that I should have the extra 16% I need to call? I think the answer is a resounding yes.

There is a simple Bayesian answer to this hand, which is they both had hands worse than mine and the river went bet-call. Something that is supposed to happen ~5% of the time happened in our 1/1 sample, which means it probably happens more than 1/20 times. If Probirs were bluffing, I’d feel more confident in my fold, but since he was value betting, we don’t just get a Bayesian look into frequencies, but also into how he thinks about this river spot. If Probirs thinks he can value bet AK here, he thinks Tom Vogelsang and I will often call with worse; perhaps he can value bet KQ and KJ he floats the flop with, and perhaps he’s floating some more unpaired stuff than I gave him credit for. All of these move the needle in one direction; with their river actions and showdowns, my opponents have given me a lot of information about their river strategy. If Probirs were bluffing AdQc and Tom called with AcTd, I could argue that I ran into one of his few bluff combos and a natural hero-calling combination that Tom should often play like that, but when I run into two hands that are worse than mine that the solver *never* plays like that, it’s hard to justify my play.

Types of Error
Poker is played by humans, not computers.

Grade

There’s a large chasm between computer poker and real-life poker. In today’s hand, I was “supposed to” have the best hand somewhere around 5% of the time. When I was shown two worse hands, it was clear that I had the best hand a lot more than 5% of the time. When I asked friends about the hand on break, they almost all told me I had to call versus the given opponents, and they were all right. I could have figured it out as others did, but I did not, and it cost me.

C-

POTD #103 ⁠onemorething

It’s easy to say in hindsight, but the tell-tale sign for me on the river should have been Andras’s bet size, which really doesn’t make sense. If there was ever a spot where it seemed like your value betting range needed to be a flush or air, this seemed like it, so why would he bet non all-in? In game I thought his size was one he’d often pick with a straight. It makes sense that he’d just call the flop with a bare gutshot and it makes sense that he’d feel compelled to value bet once he hit on the river, but he did not want to overbet all-in. It’s also possible he thought our ranges were capped and he chose a size designed to induce a hero call from second pair or worse, which are the bulk of Tom and my ranges. In my notes, I wrote that Tom Vogelsang called the river quickly, which at the time, I thought was weighted towards him also having a straight because it’s a hand that wouldn’t consider shoving or folding. When I need to be good ~20% of the time on the river, I can heavily discount flushes from my opponent’s range because they didn’t take a more aggressive action, they only have between 3-6 combos of straights between the two of them, and I’d be truly shocked if they ever had two pair or a set, folding should be off the table, even if it’s good vs two solver players.