Quote:
Originally Posted by ZenFish
I was mostly interested in hearing whether there was a reason for the AUTO tree builder to insert multiple sizings pretty much everywhere.
Fwiw, I have tested Monker's SB v BB equilibrium vs a benchmark Pio simulation (one geometrical bet size per street, 198 flops, 2bb/100 convergence) and found that {90 buckets, Perfect/Large/Large} and {30 buckets, Perfect/Perfect/Large} both reproduce the Pio results accurately. Frequencies same, and the shape of the ranges close to identical. In particular, the suited/offsuit polar regions in BB's 3B range are reproduced almost perfectly, which does not happen with default settings.
So we can verify that Monker reproduces accurate NLHE HU ranges when we push the accuracy settings up a bit.
Quote:
Originally Posted by gtlol
How many iters nodes you solved for this? When you think we should stop the calculation?
I ran this for 50 I/N. I've explored convergence a bit, and how far you should go depends on what ranges you're interested in.
For example, if you solve CO v BTN v SB v BB, you'll see the 4-way single raised pot branch (everybody flats behind CO) converge very slowly (BB's flatting range takes forever). But the ranges for CO open, the first flatter, and 3-bets converge quickly.
That said, the 10 I/N rule of thumb is much too loose. I'd say run everything to 30 I/N if possible. If this is impractical, try to snip away the biggest single-raised pot. For the 4-handed game, we can snip BB's range for overcalling vs CO open + BTN flat + SB flat, and we are left with only 3-handed single-raised pots. This makes sense, since 4-handed play is rare. And since we can't converge this branch well anyway (unless we run with low accuracy or for a very long time) we might as well snip it. The biggest single-raised pot in the simulation is also the biggest branch in the tree, and from what I recall, this simplification will cut the tree size roughly in half.
Multiway simulations are fun to do and very fascinating if you're into poker theory. But it's a good mindset to be a little pragmatical and not get hung up in perfection. You can chase perfection in HU sims with Pio, but for multiway work in Monker, it becomes a costly habit. If it isn't useful, leave it if it's expensive to compute.
Last edited by ZenFish; 09-13-2017 at 10:10 AM.