For those of you who don't know what snowie is, it's basically a program that's designed to find out how to play GTO by playing a ridiculous amount of hands against itself. While it's still far from finding perfect GTO strategy, it's supposedly good enough that a new player can quickly learn to break even at 500nl zoom by just following its advice.
Here's the website, http://www.pokersnowie.com/ Also Improva of DC made a huge thread about it.
Anyway, I was wondering how applicable snowie can be to low stakes live play. If it derives its optimal decisions against what it would do if it were playing against itself, how can it have any sort of data on how you should play in some sort of situation where the action leading up to your decision was clearly so far from optimal that snowie bots are never seeing that type of leading up action against themselves.
For example, Snowie bots would probably never play in such a way so that a hand is opened 4x with 3 callers and you're in the big blind wondering what to do? If snowie bots just don't get themselves into 5 handed pots, does that mean that the program won't have any data on how to play GTO in the event that you actually find yourself in a 5way pot?