[The Bryce]

Actually, you're correct on that point, I apologize about that. An easy example would be to say that player A plays a GTO game where all his frequencies are perfectly balanced for the highest possible rate of return based on the hypothetical assumption that his opponents will always make the best possible action themselves. If he plays against player B, who is a tough but non-GTO player then player B cannot win money but can still lose some money. If he plays against player C, who will fold every single hand, he will win more money than if playing against player B. Interesting point.

Also, since this always comes up: note that while player B cannot beat player A he can show a larger edge over other opponents who do not play like player A. What you often see as a result of this in higher-stakes bleeder games is that there is usually a long feeling-out period wherein many players may play very conservatively and closer to a player A style until they have a lot of data about their opponent's actions / frequencies and can assume with a high degree of confidence that their opponent will do something with a high degree of probability, at which point they may begin to shift more towards a "player B: style. Note that in some areas, like folding on the river, many players may elect to take a more conservative and proactive "player A" style with the knowledge that once they attempt to take a more exploitative "player B" style (folding with above GTO frequency to bets or raises in "tough" spots) tough opponents at these limits may become exploitative themselves in this area and begin to bluff more.

Many players do tend to "cheat" in some of these spots as well, meaning that they still do make many "hopeless" calls in these spots but, often because they assume that their opponent is not that tough or that dynamic, still do fold more than the GTO frequency. Some players do stuff like this at a subconscious or intuitive level, others do it at a mathematical level.