yea i think this is a likely scenario

Also, a lot of the hands that are posted on 2p2 are no more weird than some of the stuff u will see from well known expert bots like sonia or the GTO bots

If i remember correctly though i think i saw a pic of it folding AJ pre when it was up against AQ. How is this explainable? Programming bug and it just happened to be up against a better hand this time? Or is it cheating?
Althought it for sure would be a very dumb way to cheat for someone who has to show his cards, right?
Cause clearly, even if it was in its fish-mode it would not make a fold like that pre im pretty sure. That would be like throwing way way more money out the window than you need to for bait purposes.

i had beaten the machine at the venetian and the bellagio at 80% of sessions. It takes some freakishly weird lines 3 betting most suited hands and calling nearly all backdoor draws oop i can't believe there's a thread for this. It's like christmas a week before christmas

Eighty percent of how many?
And if you could beat the machine 80 percent of the time, why the heck are you living in the south suburbs of Chicago?

i had lived out there for 6 months and moved back. The machine is just ultra agg and misplays/overplays all mediocre hands. I'd say 20 sessions

Proved free money. I'm moving out tomorrow.

My first session ever on these machines... I found this thread afterwards because my results seemed too good to be true... Started at 1/2 with $20, played 2/4 from 200 to 1100... Then 5/10 until 2k when I decided to cash out... Total time was about 3.5 hours.

Did I just run really good? Even though I was able to pick up on the comps patterns, I must have run above expectation as these machines could cost the casinos a lot of money.

yes you ran amazingly. no offense, but the fact that you have to ask if winning hundreds of bets in a few hours is abnormal means you probably don't understand LHE math/winrates/EV/variance/etc very well

You can lrr the bot preflop with the top of your range with a high success rate, unless they've updated the algorithm. If the bot fails to raise preflop it will have a very weak range.

I understand these basic statistical terms and their meanings... I guess I should rephrase the question.

So do you guys think the bot is possibly highly exploitable because it does not seem to change gears too much? I think playing very straight forward is the key to the game... Some bluffing works, but generally I played very straight forward as the bot doesn't remember tendencies and changing gears has no benefit against the bot....

If I'm playing heads up versus an idiot/fish/newb, I expect my win rate could be quite high when I'm getting 100+hands/hour with no rake... Just a thought, could the bot be a really bad player at the low levels?

Playing straightforward has worked best for me.

But I believe it changes personalities more often and more rapidly than you may recognize. Granted, it may sometimes sit on a passive personality for a while, but it also can jump around from hand to hand.

The best time to play the machine is when you're on a long list for the game you want to play. There's much easier money in the card room.

Is there an updated list of which casinos (if any) are currently doing comps on this thing? I have played a bunch at the Venetian/Palazzo this week, and haven't seemed to earn any comps (at least through the Grazie points) and figured I might as well be playing somewhere with some "rakeback" too.

I'm not 100 percent sure, but you might want to try Wynn.
The machines there say you can accrue points, if I recall correctly, but I never play the thing long enough to matter.

This is very interesting, I have ran across this machine at the Peppermill in Reno. I have been working on a decision tree and plan on posting as soon as I get it in a computer format. I am curious if anyone has the ability to create a minimax chart based on the possible decision values that would be represented at the various stages (PF, Flop, etc.) through out this game. I simply do not have the skills to do this, but combining such information with the games limited decision tree, might help to find an exploit.

..fingers crossed.

still working on the decision tree, the numbers are.. exhausting to say the least. hopefully i will have it up for review by the end of the week

well I don't have a way to upload the graph but for any given choice there are generally 3 follow up moves check/call, bet/raise, reraise/call, there are more possible FOLD VALUES for bets than checks, therefore the comp will typically push for a bet, rather than a check. I have a sample of about 200 hands on the machine at 1/2. session 1 $20 lost. session 2 $60 win. session 3 $100 lost session 4 $260 win. session 5 $100 lost. So I won about 50 BB over 200 hands. The machine does raise roughly 75%, definitely more than 65%, pre flop, this is most likely to encourage pre flop fold errors. The machine values middle pair and up and will call to the end with the exception of straight AND flush possibilities on the board. It often folds on the flop to a paired board, but if it calls the flop will usually hold til the showdown. The machine will typically reraise your bet if it has 2 pair, however, usually only once, so if it quits capping, most likely not stronger than 2pr/3of a kind. My sample is still relatively small, but their doesn't seem to be a straight forward exploit, you'll definitely need to still play poker. I have found getting to the flop to be the key, and getting there cheap, just call its obligatory preF raise. Remember if both players see every showdown you break even, so get to your showdowns, and value top/overpairs.