How I created Draw Solver, made a million dollars, and helped to take down Phil Ivey

This is interesting, unless I’m missing a sarcastic joke. Why face up vs down like a blackjack shoe? Solver knowing more cards doesn’t help? Btw I know nothing about a solver lol.

Gotta be a top 3 winner, maybe even biggest winner (not including private online games which I’ve personally seen 3-6k HU PLO with a 1.2m buy in played + a casual 2m flip at the end!) Antonius made a boatload tho in the early 2000’s then also cleaned house on FTP. In terms of tracked results he’s probably number 1. Trust me tho insane amounts of money are exchanged on private apps

It’s not a joke.
If you deal two dead cards face-up at the start of the deal, the composition of the deck is different every hand, and all the players know the composition of the deck is different.

I’m sure we’ve all been the position where a card is exposed (say the 9d) and then we get dealt 8d7d. We need to this about whether or not to play this hand differently knowing that one of out most important outs is gone.

What I’m suggesting is that you do this on every hand, but make it two cards, so people will have to adjust from standard strategy for every hand based on what’s out of the deck.

Understood, because seeing a dead Ace changes a solver accepted “optimal” strategy of a hand like AK in a given position. Unless I am way off here this also, even if incrementally, increases the profitability of a bluff which makes the game less statistical and more nuanced.

Fun playing against the solver. Now if I only had 14K to purchase it.

You played from the link on Oleg’s site?

Why not just burn a card like normal before dealing the first two (or four, or five)? And changing anything is probably a terrible idea at the low limits.

Maybe burning one card would be enough, but the key is burning face-up. Perhaps all burn cards on every street should be face-up.

I think people don't understand why what you are saying is important. I'll try another way - imagine you have a solver chart memorized for some common preflop spot, well if you want to "prepare" for NickMPK's variation, you would need to have the solver run the spot with every combination of 2 dead cards, and then memorize all those. Its something humans are pretty good at if you play poker (the example with 9d and you hold 87dd) but to a solver, this is 52choose2 "games" instead of one.

I understand it makes it more difficult preflop, but after flop and turn doesn’t knowing more cards make it easier to deduce potential holdings as well as having betting information. My reasoning is like the end of a round of gin….seeing more discards gets me closer to the makeup of your hand, therefore optimal play could be easier.

It might make it more effective to use a solver if you could actually plug the dead cards into the solver and get solutions in real-time.

But it would make it much more difficult to use a solver to study situations to prepare for a game, as there would be exponentially more situations to potentially try to study.

(That's just my intuition FWIW.)

Ok, I agree and I realize now you are more so speaking of the solver student bringing memorized scenarios to the live table and having an advantage due to that knowledge particularly preflop. I am actually considering an experienced and capable live player now seeing more cards and being able to remember them (akin to a good stud8 player) now essentially being given an ever greater advantage. I think it makes it extremely more difficult to chart a hand strategy preflop but I can’t be sure thats the case post flop/turn/river.

And yes seeing dead cards real time online would make it much easier on the computer, we’ve already seen that with cheating accusations.

However what you have stated would be an interesting addition to AIOF tourneys….

I do think it’s likely that skilled players would have slightly a greater advantage over recreational players using this variant. But the avantage would come from intuitition and experience, rather than memorizing a bunch of solver outputs.

And I think this sort of “intuition and experience” skill is part of the allure of poker to most players, including recreational players, while the idea that the most important skills might be the ability to run cutting edge software and memorize it’s results is deeply offputting.

But I don’t really use solvers, so maybe I’m completely off-base here and people who do use them might claim this sort of dead card info could be easily incorporated into their GTO-based strategy.

Yeah, this is what most people said when I proposed this variant a while back. As a result, I never really thought about it again.

But the OP made me think it might be on to something. I do want to read the entire article OP wrote, as it looks very interesting. But he mentions that (1) his solver was only able to account for one dead card; and (2) even accounting for one dead card made the solver vastly superior to those that don't account for dead cards.

This made me think that perhaps two dead cards might not only significantly change optimal decisions at many points, but that these changes might be beyond what players are able to meaningfully simulate and memorize. So maybe it would in fact substantially change the role of solvers and memorization in poker, similar to the proposal for Random Chess (though it would change the game at the recreational level much less than Random Chess).

I wonder how this solver is in context of the Story of the heros seb86 and alexonmoon in the famous poker docu "nosebleed". The were crushing the 2-7 games 2010-2014 kind a pre this solver. Iirc seb86 quitted high stakes 2-7 when suddenly a group of Super crushing germans showed up (raulg is german too). The docu seemed to be at a point in their highstakes mixed games carriere where they knew it Was kind of over. Since I think their story is one of the most outstanding in the history of (online) poker I Would be super intersting how this solver is in context to All of that. Anyone (maybe op) has some info if and when they were aware of the solver?

I want to know more about this!

after rewatching the NOSEBLEED poker documentation on youtube I discovered the following conversation at min 53:

friend:
"did you get any action online (while beeing in london for a week for a break of vegas/wsop end of june 24)?"
seb86:
"... some 2/4 deuce against trueteller and san1ker (raulg) ..."
friend:
"how do they play by the way?"
seb86:
"they both decent now"
alexonmoon (from back of the car)
"its getting tougher"

Burn the flop & turn face up. Maybe a couple of 4pm NLHE & PLO tournaments at WSOP with that format? Or before first two draws at TD?

never thought about this ... but i get the feeling it might be a genius idea.
which is also the reason all the big sites will never allow it and/or lobby hard against it, bc they need the bots and solvers, they make up a ton of their traffic

Didn't he make those accusations in FL O8 and FL O8 only?

I'm quite sure they were not aware of our solver.

We had a spreadsheet with edge estimates versus each opponent, and we updated it every half-year. The numbers were used as input for Indirect share calculation, so we were serious about them. I just checked, and Alexonmoon was one of the very few opponents against whom we estimated our edge at 0 (zero) bb/100. Which is very impressive. Seb86 was at 0.5

op, do you mind sharing why you decided to share all this stuff

Actually, just burn one card face up right from the beginning, after the cut. Keep it simple.

Primarily because it's a personal and professional story that I'm proud of.
Also I'm still selling solvers