As of November 2021, the APT Forum is closed to new posts. Like with many online forums, usage has decreased in recent years. All previous posts are still available.
An opinion on pokersnowie
Sign In to comment.
Welcome
It looks like you're new here. If you want to get involved, click one of these buttons!
Comments
Pokersnowie has bots that play in a much more game-theory optimal (GTO) way than APT's bots. But in terms of training materials, they don't have much compared to what we have. What I always tell people is that if they are trying to learn to play against computers or very top level humans, Snowie is great. The strategy snowie teaches will win - it is based on game theory, so it is a strategy that is designed to win against anything your opponent can do. However, humans at 99% of poker games and tournaments I have played in don't play that way. So while Snowie's method will definitely work, it won't win the MOST against normal human players. Humans have tendencies to make predicatable errors, and our bots are designed to play more like normal humans and teach you to take advantage of those tendencies.
So basically, Snowie is a math tool for learning optimal play - and that definitely has tremendous value. Our site is a training site designed to teach you to play against typical humans and diagnose detectable patterns and leaks in your game.
If you look at Snowie's site, they are always talking about advancements in their AI - that's clearly what they work on and they are very good at it. We don't spend as much time on our AI and spend more time developing tools to help you practice specific skills and improve on specific areas of your game.
I hope that helps.
Allen
I have been using both Snowie and APT since November 2017 and I agree with most of what Allen said above. Snowie is about as good as it gets for practicing GTO-based poker strategy. However, it is unstructured. I had to go to yet another source for the instruction (Upswing Poker) and use Snowie to practice and experiment.
Also, I have discovered that attempting to play a more balanced GTO-based game vs most players I'm actually facing is not only suboptimal but in many cases flatly wrong. Vs the vast majority of players you are likely to face, a far more exploitative approach is necessary.
I think there is a spectrum of poker players, from the most inexperienced to the nearly perfect HU player. There is a maximally profitable strategy to employ vs each part of this spectrum and the trick is to choose the right approach for the right the player pool you are facing at the moment. Against the worst players, there is no need to incorporate any balance to your game - maximally exploitative play is the way to go. As the player pool you are facing gets better, more and more balance needs to be introduced to prevent these players from exploiting you. At the theoretically perfect end of the spectrum (say vs Doug Polk HU), having a solidly GTO-based approach is necessary to be competitive.
This is a long way of saying that the "better" program for anyone depends on who they are playing and what their objectives are. To play at most stakes, just having solid fundamentals and learning to exploit leaks is going to get you far. On the other hand, learning theory expands your toolbox and may give you a better understanding of exactly what leaks you are looking for in others vs optimal play. The theory is also worth learning as a purely intellectual pursuit for those so inclined.
Each has strengths and weaknesses.
Snowie teaches how to play against a table of great players.
Most of us will never sit at that table. If we do it's a mistake.
Bet sizes are limited to certain sizes which is unrealistic.
I liked that Snowie analyzes each and every decision and puts a value on each one good or bad.
APT created in 2008 and it's so 2008. Marks any lite 3bet as questionable. Any no equity bluff as questionable.
Only assesses aggression factor post-flop not preflop.
Learning APT style will help you improve enough to beat 1,2 or 1,3
Live.
Get a hand and barrel away. An exploitable style but probably not at lowest levels.
thanks for the comments on PokerSnowie
before someone pays the $230 a year, need know its basically a cash game training thing
you can't import non-cash game hands for analysis, so if you play MTT and SNG the statistics and evaluations it offers sadly arent available
when practice with the Training against the robots, the robots are all perfect players, no human tendencies at all and you can't change their level of expertise
what i do find interesting is the emphasis on bet sizing and the EV of not only fold, call, check, raise but also the EV of various bet sizes... how it can make that determination without knowing the type of players in the hand is puzzling... however bet sizing in various situations is a big negative in my game and i trying improve that with Snowies help... i get mad it keeps tells me "WRONG" hahaha
with repetition you can get an idea better of sizes of bets in situations and streets
but sadly it no lets you accept their advice and continue with the hand... your poor decision stays hahaha
also the SNG Coach thing on Icmizer is GREAT... you see a zillion types of situations, and relates to stack sizes ... and i try keep playing them till i consistently get the 4 right smiles and it times you
oh i had SitnGo Wizard but returned it for a refund, i think its awful unless you have Holdem