It does go to show how much AIEV is actually a pretty small part of luck, like there is no way Galfond is bad enough at poker to be a 44bb/100 loser, which is what the actual results are suggesting, yet we look at AIEV and it’s like “running really well, actually”. I would assume AIEV evens out at roughly the same rate as those other luck factors do.
Ah right, so the blue and the red equal the dark green.
Yeah blue and red are just components of the green (won at showdown/won without showdown, or in this case “lost” in both instances) and the yellow line is an adjusted green line.
1 Like
It’s simple: if I’m running below AIEV, AIEV is a legitimate metric that accurately measures the impact of luck in the game of poker. If I’m running above EV, AIEV is actually a dogshit metric and there are many other ways to quantify luck and, if you’d just take a look, you’d see that I’m actually very unlucky.
8 Likes
How do we know this?
Win rates like that are not possible against competent players. Even a strong AI wouldn’t be able to win at that rate. In the Libratus HUNL challenge, Libratus beat its human professional opponents at 14.7 bb/100 (which might be why Galfond chose the 15 bb/100 number). There’s no reason attainable winrates should be triple that at HUPLO.
But hasn’t Phil not been playing much lately while all the top players are using solvers?
Sure, but solvers are overrated imo and exactly why Phil is down so much as well. If you plug these PLO spots into a solver they will spit out % of what you “should” do vs the “reasonable” ranges of the villains. Our villain could be at the top or bottom of this and you make the “mistake” of doing what you should be doing 70% of the time.
Galfond is the new KLF
1 Like
Yeah, this.
Phil Galfond is really good at poker and there’s an unholy amount of variance in PLO. Swongs like this for the record are why I never found PLO appealing. NLHE had more than enough variance for me lol.
2 Likes
durrrr got to be part of Team Full Tilt for a few months before that whole house of cards collapsed, so that was pretty cool.
1 Like
It’s almost like PLO is a random walk.
1 Like
Haven’t been paying attention but he’s cheating now right?
Not sure if kidding.
He was getting absolutely destroyed so bad that he had to take a break and now he’s winning on his own site? Seems suspicious.
Welcome to PLO. This was an insane stretch I had last year where I lost 70 buyins in about 8k hands playing PLO 50 to 200, then almost instantly won back 100 buyins over the next 18k hands (which would be roughly the same # of hands as their challenge). This is all at 6-max playing a fairly nitty TAG style.
I can’t imagine the swings with 2 aggro beasts going at it HU. Phil was down what like 45 buyins and is now back up 35 or so? That’s nothing.
4 Likes
I just want Phil to get close enough that Venividi has to choose between actually playing out the match (and risk the loss) or blinding off to lock up the sidebet. The last couple hundred hands could be very interesting from an ego vs ev perspective.
2 Likes
Booo…Jman is a great guy. Whether the other guy was cheating and his cheat code stopped working, I dunno.