What the computer says:

3/4 POT is not viabale RFI sizing

Seems reasonable, though NL can be quite hard to simulate. You need to take shoving into account, which I'm pretty sure is the highest EV action with many hands.

Yes, yes.

Impossible to know if this is useful without knowing the full tree / what the players behind are allowed to do. Impossible to imagine 3/4 pot as only open size is good but as an exercise to see the range it wants to play, sure, seems vaguely useful

8-max NLO8!!!!!!!!!

I'm not familiar with this program. Why are KsJd8d4d selected in the box on the right?

Ignore the selection to the left its an artifact.

When it comes to making preflop ranges the amount of Memory required is enormous. That's why all bets are usually explicitly stated. All ranges are approximations

Where I come from, NLO8 is mostly played as a once per orbit dealer button so there can be a 9-handed NL08 game, but full ring NLO8 is played on the weekends. Sometimes there are NL Big O games.

NLO8 is a money siphon from the NLHE specialists to me!

any game where the correct strategy is to play 16% of hands is a terrible game.

i.e. close to every raked 6-maxed poker game.

probably true, unfortunately. what would you think it would be for PLO high?

I have a 50bb sim. It's 16% with kinda low rake.

There are solutions to this issue like antes, 4-max games etc. However, does it really even matter what the optimal is? Poker works because people are given opportunities to make mistakes.

No matter which game is being played, when it is a big bet, community card game the amount of hands that can be played in each position tend to be around the same % no matter which game is being played. To me this makes sense, since it has to do with bets relative to stack size and the amount of money currently in the pot. I mostly try to model 100bb stacks.

Like amok said, it's about the people who are playing the game with us. They certainly are not computer opponents, and they likely have desires other than squeezing out all the theoretically possible EV.

ante or 3 blind games are probably better theoretically to force more action. I guess in most small stakes live environments, it's not necessary.

Yeah, after completing any sim the EV of each GTO player is a % of what was in the pot pre-flop. The only players that are losing money are the Blinds, because they are forced to enter the pot without seeing their cards first. The GTO players would never make a play that was -EV so once the game is solved all the expected winnings are just the Blinds (less rake) divided up amongst the players. Really depressing! I agree that if we want more action we just force more money in the pot pre-flop. I guess if I wrote a sim where there was a $1,000,000 in the pot and everyone had $100 stacks we would see almost every hand being played to the River.

Luckily, where I play people will voluntarily add a straddle of $10 in a 1/2 game in almost every hand. In GTO land they are the biggest losers, but in the real world most players make so many mistakes that some of these Straddlers are still winning players.

in gto land an ante will make everyone play looser, a third blind will make everyone play tighter.

I know you were being hyperbolic but even if you put $100 in the pot with $100 stacks you would see in the sim every player probably all in every hand pre. At a $1M in the pot I'll throw my $100 in with everything that isn't small quads

Good point on Straddle vs Ante!

The UTG+1 range is tighter and is doing lots of shoving.

Below is what happens when someone straddles for 5BB UTG in an NLO8 game. They actually turn the SB into a small winner! (1000 mchip = 1 Small Blind -----> 1 mchip = 0.001 Small Blind)

Just for a fun. Here is what we should play UTG if someone throws $200 in the middle of our 1/2NLO8 game. I'm actually surprised it's not looser.




The reason we don't shove all AA is because the filter brings in AAAx off suit, and those we should fold.

There also is a phenomena that as players fold we should consider being tighter in late position. I don't have time to prove it, but I believe that it has to do with card removal. I ran a sim with $1000 in the pot of a 1/2 100BB game and UTG (96% open all-in) is looser than the Button (78% open all-in).

This may be because the chance of it folding to button is so low that that branch of the game tree is not really getting iterated. Its close to a non existent node so the results get real weird / screwy.