I've been following this project for 8 years now. Even back then it was
good enough to beat--or at least play even with--the best heads-up limit
hold'em players in the world. It is better now.
They claim it's essentially perfect, and while I haven't independently
verified the claim, I have absolutely no reason to doubt it. Their
methodology seems sounds, both in how they derived the strategy and how
they test for it's distance from the optimal strategy.
None of these top limit hold 'em players I'm referring to doubts the claim,
at least not to my knowledge, and several of them explicitly endorse it.
These are smart people and this is a result 15 years in the making.
The web interface is clunky, so it's kind of academic b/c it's not feasible
to play many many hands against it.
Ed
On Fri, Apr 10, 2015 at 2:59 PM, nightoftheiguana2000@yahoo.com [vpFREE] <
vpFREE@yahoogroups.com> wrote:
>
>
> Ed wrote: "I promise you can't beat it long-term. It plays within 1 big
> blind per 1000
> hands of the optimal strategy."
>
> I think they made some calculation mistakes or they dumbed down the online
> bot for some reason. It doesn't take falls like the IGT bot to try to get
> you to keep playing but I think some of its lines become too narrow and too
> readable, a mistake the IGT bot also makes. Have you tried playing it? I'm
> not arguing that the optimal strategy would still be exploitable, I'm
> arguing that it's not quite optimal strategy yet, for some reason.
> University of Alberta hasn't published all the details, so they might be
> holding back for some reason, like maybe they hope to get some big stakes
> challenges.
>
> [Non-text portions of this message have been removed]
>
>
>
[Non-text portions of this message have been removed]
Posted by: Ed Miller <ed.miller@gmail.com>
| Reply via web post | • | Reply to sender | • | Reply to group | • | Start a New Topic | • | Messages in this topic (4) |
