With regards to Texas Hold'em, Conterfactual Regret Optimisation [1][2] was already able to beat humans in limit games, while it seems that this new approach is able to do the same in no-limit games. My naive understanding is that no-limit games are harder to crack because there are more possible outcomes, thus the available information (poker is an "imperfect information" game) becomes "even more imperfect".

[1] https://arxiv.org/abs/1407.5042 (yes, the author is the same Oskari Tammelin who created Jeskola Buzz, a popular piece of music software).

[2] http://www.nature.com/news/game-theorists-crack-poker-1.1668...