A post from Arty.

Sample sizes ?

Not too many about 1K hands.

I think Stuey Ungar thought that GTO was a car and yet he has the best tournament winning % of all time and is regarded as the best NL Holdem and Gin Rummy player of all time.

Sigh.

1k hands is nothing. Nothing. Everyone constantly wins or loses a few stacks over a few thousand hands. This is normal variance, not a sign of skill. You can't use short winning sessions as an argument against Snowie. Especially if the skill difference is small then obviously anything can happen over a few thousand hands.
You rarely hear from players getting crushed by Snowie over 1k hands. Because nobody would brag about that.
It's very obvious why 1k hands mean nothing. I could go into detail but in 2016 most players should know the reasons.

Please, by all means please make Snowie look bad by playing something like a 25k or 50k hand challenge against him. If you could do that the programmers would be under a lot of pressure to make improvements.
Just don't use these silly small sample sizes to prove a point.

Play 25K or 50k with play money?

If they agree to pay me if I win I would.

That's now how things work lol. You're questioning Snowies usefulness, nobody's questioning you. The developers don't carry the burden of proof, the ball is in your corner.

And who are you to make any rules here. They are selling a product that has been shown by many posters to be a scam. Read the thread. I am not selling anything, they are.

You want 100K hands? Read this.

http://forumserver.twoplustwo.com/sh...postcount=1060

http://forumserver.twoplustwo.com/sh...postcount=1058

So according to your thinking a drug company says it has a drug to cure cancer. Then the public has to take it to test it for a year and die if the drug is not effective. Dumb ass.

I'd recommend snowie for a lot of flop and OOP plays on flops. Seems to do pretty good there and I agree mostly with how it constructs its ranges. It's mostly river spots and multiway spots I don't agree with much. Also, it definitely does some wrong things, but if you don't take the time to analyze what it is suggesting you do and just attempt to mimic it, then you're using the software incorrectly to begin with.

These posts are rants, not evidence.
This is starting to get boring, you're not forcing me to play my cards better.
Funny strawman argument by the way. So not relevant to this discussion at all.

can someone shed some light on this scenario for me?



after player 5 opening, snowie has player6 responding with only 3.66% of hands total, calling 0.96%, 3betting 2.7%. Folding even TT in this spot in a full ring game. It seems strange to me.

With the stacks behind I think you're basically getting what snowie thinks is a +EV 3bet shove range . I've seen a spot before of like 40-50bb where hj opens , co 3bets and I'm told to fold JJ from sb or btn because stacks are awkward and it doesn't want to coldcall or 4b/fold or 4b/call so it just folds . Think this is similar where it just finds that stacks are too awkward so it folds a hand as strong as TT rather than 3b/f or 3b/c and probably lose more EV overall

LOL, I have seen it fold AQ preflop to Min Raise in 6 player ring all fold except min raiser. Makes no sense

It makes a lot of sense when you think about sample sizes.

I have recreated this spot with 100bb stacks behind, it has the exact same range for player6

If you call in that spot, you're very vulnerable to squeezes and overcalls. Even if someone just calls (IP or OOP), you have the worst relative position post-flop and a capped range, so it's very hard to make money.
Snowie doesn't like flatting in position very much at all. It much prefers iso-3-betting, or folding.

You get out of snowie what you put in. If you use it and try to base your game off of it your going to lose a lot of money. That IMO is being incredibly lazy and using minimal brain power. Trying to play like Snowie as if it is a cheat sheet is just dumb. Best way to use it is to construct your own ranges to play against it and think about how you play different parts of your range in different spots. I try to play as unexploitable as possible based on my range and my style of play, as I know that if I can achieve that against Snowie and take that mindset to the tables it will be quite tough for someone to exploit me. It teaches you to think about your range and not just your hand, helps to minimise tilt in spots you bluff or bluff catch and get it wrong as you know that your range is profitable and it frees up a lot of mental capacity when you know your range more and more. You can do the same in flopziIlla, crev, combonator etc, but don't get the actual playing part of it so it helps massively with the visualisation of how a range plays out. Obviously we adjust to weak players but same principals aply with our range and how we think, it just becomes purposely exploitable to exploit fish tendencies.

Anyone on here who goes off on one because it says it is gto and it isn't and conplain that they lose money playing with et level is extremely short sighted. Use it as a tool to improve your game, not transform it into a style one) we can't achieve and two) isn't maximising EV against 80%+ of the field!

If that is what you use it for would you not be better off playing a tougher bot , such as neobot, recording the session and then analyzing it later to see mistakes and improve your game.
Don't understand why reviewing what Snowie calls errors improves anything as snownie's game is terribly flawed.
Would you take coaching lessons from a coach that plays much worse than you?

I'm not looking at what Snowie is telling me to do. Like I check it to see if it's logic is better than mine or why it is telling me to take a different action so that I understand why it thinks my play is bad, but if my logic is fine for my ranges I ignore it.
I have not heard of neobot but yes would potentially do the same. Does neobot save hands and results? I tend to not look back over loads on Snowie as I just spend a lot of time on tough spots when they come up but being to look back is an advantage.
Also, although I agree Snowie is beatable (I'm beating for about 8bb/100 after 10k hands but yaknow, small sample, I am aiming to get about 20k hands in total by end of April) it still beats a lot of decent players. Sure you can trick it, but if you took an average 50nl reg and didnt tell them they were playing Snowie I think they would get crushed. Even the average 100nl reg in the same circumstances wouldn't beat it I don't think so given that it is going to be helpful to play.
I'll investigate neobot ,thanks

Neo Bot only saves your overall profit or losses. It does not save hands. What I do is video record the session then go back and edit the video to only leave hands where I made mistakes for future reference.

And how exactly do you know NeoBot is tougher than Snowie ?

LOL, that is easy. Because I have played both.
Snowie does not protect a vulnerable hand, does not adapt and bluffes all in, risking the whole stack, when a smaller bet would have worked just as well. Just to name a few of the many blunders of Snowie.

Neobot adapts to your play and has won was a top place finisher at both the 2012 and 2013Annual Computer Poker Competitions. More information at official website. www.computerpokercompetition.org

Snowie is afraid to enter the competition and play against other bots.
Results for neo.poker.lab
http://www.computerpokercompetition....2p-nolimit.pdf

http://www.computerpokercompetition....ts_2pn_tbr.pdf

Want more?

Could you please give some examples for this. I am interested in the theory not defending Snowie.

If you try to play like Snowie, when you have a pair and the board has a flush draw or straight draw possibilities, Snowie checks I presume to get more money in the pot, instead of making a large bet to protect the hand and get players to fold. Instead it takes it to the river where many times you are outdrawn. So it allows a made hand to get outdrawn.