Keith Brandt 1:1 Coaching - 1 hour 
-
The Perfect MTT Player
Here is a question intended to generate discussion.
A hypothetical android multitables every available online MTT with good enough equipment and online speed to never time out. He never passes up a re-entry, rebuy, or add-on chance.
Our android plays perfect GTO, ICM, and FGS strategy, and never deviates from the equilibrium strategy.
Is our android a profitable MTT player? Why or why not?
-
11 Replies
-
When you play a strict GTO strategy you will break even against another player playing a strict GTO strategy. Since we know that it is impossible for a human to play a strict GTO strategy the android would win in the long run. The android will win because of the mistakes the human players make. The android will not win as much as someone exploiting these mistakes but will be a winner none the less.
-
The android would certainly be a cash game winner enough to overcome the rake, though not as big as if he were exploiting human opponents.
He’d probably be a winner playing SNG’s and probably satellites for value.
He would probably get more than his share of min cashes in MTT’s.
But can he get enough deep runs and championships to be a winning MTT player, without exploiting opponents?
-
I think the answer is an unequivocal yes. Exploitation is dependent on the ability to know how your opponent is deviating from optimal play and tournaments are particularly tough because the rising blinds and antes and shifting tables will decrease your ability to target a particular player to exploit their particular approach (not eliminated, just decreased compared to an 8-hour cash game). IF you have someone with a perfect read on the errors of everyone at the table and can consistently play the exploit correctly you can likely make MORE money than the GTO bot, but I don’t think that is realistic. Maybe at low stakes, but not as you move up. Thus, the GTO approach will make money at MTTs over the long term by being better at being unexploitable and at the same time exploiting by their approach whatever deviations any player makes from “perfect play” over the long term. Since it is ICM/FGS aware (not as dangerous as becoming SELF-aware so I have heard), it will have a sufficiently secure strategy to navigate the bubble and maximize its return in the later stages of the tournament.
If you could create the “iExploitbot 3600” which is a machine that is able to divine the error tendencies of each player it plays at each stage of the tournament, that Android would clean up. However, I think that such an android would hardly limit their activities to poker and we would soon be an enslaved race ruing the day we decided that to create our eventual overlord just so we could win a little extra money at poker…
-
LOLOLOLOL 😅
Maybe the most dreadful thing our winning android might do if he becomes self aware is to simply disappear… on a quest for the most perfect chocolate malt shake in the galaxy!
-
(Post without the link…mea culpa)
We already know the answer to that quest. It is Kopp’s on Port Washington Road in Glendale. (A suburb of Milwaukee, for the uninitiated).
-
-
You never cease to impress me. At least 42 times so far.
-
-
-
Have enjoyed many a Kopp’s chocolate malted over the years!
-
I have no doubt this bot would win in the long run. It might miss some a big exploit spot or two, but it would capture so much value based on mistakes in the general MTT field.
The other thing I’d note is that I think we sometimes equate GTO or non-exploitive with being boring or nitty or not aggressive enough. This is just not true. When you look at solver results, especially big trees where you give the solver lots of options, it often takes really big spots and puts a ton of pressure on its opponent. It also finds some bluff spots where most humans do not.
What it doesn’t do is adjust to someone who is tilted or drunk or what have you and play them differently. So it can lose some value there. But I have no doubt this bot would be a winning “player” in large MTT fields over time.
Log in to reply.