Perhaps those who risk much are more likely to either win big or lose big, and then there's a huge selection bias for which we hear about later in life/history...

That's actually a very interesting question. Caesar was known for his luck throughout his life, and it certainly led to him having a larger appetite for risk than his contemporaries. But a lot of that luck was also created by him being incredibly skilled at just about anything.

Another commander known for his luck, and gambles, was Napoleon. This worked out incredibly well for him (well, for a time anyway). He equally was incredibly skilled at just about anything he set his mind to and both were tireless workers.

Gambling like that is something you can do when you are confident that whatever situation you get yourself in, you'll be able to work your way out of it. Both got into very dicey situations at times. They had the skillset and the highly disciplined troops to work themselves out of these situations (example: Napoleon losing the first battle of Marengo, realizing there was still time for a second battle and crushing the celebrating Austrians. Another example: Caesar at the battle of the Sabis managing to rally his troops after being ambushed and eventually being able to push back the Nervii. Both turned out to be crushing victories where they really should have lost.)

If you don't take great risks, you don't get great rewards.

If you don't get great rewards, someone else will gamble on them.

If someone else gambles on great rewards and wins... you may be in a position where you are no longer a great risk to them, and can be crushed/kept from successes without any great gamble.

Self-feeding, I think.