[lonelappde]

That normalizes away, because the hours required is based on relative comparison to others, not some objective measure of talent across fields.

Math has low information volume, but it is slower information. A good historical treatment of a topic may be 2000 pages, and a good math treatment of a topic may be 100pages, but both requiring the same amount of time to study and comprehend.

[SamReidHughes]

The thing with math is it's very variable depending on talent. Maybe that's the case with history too. In one grueling math class I took, the ratio in time spent on problem sets between the kids who might become math professors and the best kid in the class was about 3. And it was only that low because the problems were really dirty. For history, I don't know what that's like. For software, the training time ratio is even bigger because the bar for "expert" is so low, and the knowledge required to start creating good software is small. As for chess, I thought it required a lot of study to play at a GM level.