Is there any such discussion/analysis other than Doug Polk's $499 course on this exact subject?

I found a brief discussion by the human opponents of Libratus at:

http://www.pokerlistings.com/how-to-...-w/o-being-one
Starts 2/3 down the page TL;DR below

I'm interested because I'm both a poker player and an AI (machine learning) researcher. I understand
Libratus' & AlphaGo's "Reinforcement Learning" approach vs. PokerSnowie's "Neural Network" approach.

TL;DR
- Very balanced actions (similar hands with vastly different play)
- Bet-sizing was totally independent of pot-size (pot-size is just an arbitrary 'reference point' for human betting)
- Pot overbetting is the 'norm' (even with, say, middle-pair). Is this 'in general' OR only for fold-equity purposes?
- Bluff & call frequencies much higher than human players
- Bluffing and calling in situations where no human would ever consider such plays
- Libratus would always have a bluff available in any situation
- Libratus would 'tank' on the turn & river, exactly like a chess endgame (where there is always a provably correct play)

There was a lot of discussion in the huge NVG thread. http://forumserver.twoplustwo.com/29...asino-1647075/

Not a lot of it was strategically useful though, although some of the hand histories that were posted were quite entertaining. As ever with NVG, a lot of the posts were misinformed/gossipy and the signal:noise ratio was not great. I don't think much of the actual science/research has been widely circulated officially. It was more like a publicity stunt than a research project imo. CMU appeared to be less forthcoming about how its bot works than say the Alberta Uni team is with its rival programs (e.g. DeepStack AI). There's a thread about that bot too: http://forumserver.twoplustwo.com/29...00-a-1647923/?

I used search function and there are only three thread titles that mention Libratus. This seems odd to me since it seems completely intuitive that his success indicates the future death of the online game. A smart person running an AI simply will not get caught with the current system. Bots get caught sometimes but a smart AI that has been programmed correctly to mimick human pattern , including spots where time is used to mimick though, like in close equity calls will not. Unlike the stock market where people just don't care because essentially the people offering the game are in on the action, this will destroy the game of online poker eventually.

Please call 1-800-273-8255 if you find yourself struggling with reality.

Does anyone know where I can find the Libratus research paper? I didn’t have much luck googling for it

I highlighted "Libratus research paper" in your reply, right-clicked and chose "search Google for..." and it was the second-from-top answer.

I think this PDF (by Brown and Sandholm) is what you're were hunting for: https://www.cs.cmu.edu/~noamb/papers...I-Libratus.pdf

Yeah I saw that result but assumed that wasn’t the full paper. I’m thinking the complete paper is behind some academic paywall/database or something

This site has some more information about Libratus, as well as the main Algorithm that these bots used, CFR.

The GTO algorithm is more important, IMO