The principle the author is describing is very similar to the Optimism in the Face of Uncertainty (OFU) principle in online learning, which was proposed by Auer et Al in the context of the multi-armed bandit problem. The basic idea is that when you are sequentially making decisions in the presence of uncertainty, you should make decisions which interpret the data you have in the most optimistic light. The intuition here is that if your optimism turns out to be correct, you can capture most of the upside, but if your optimism is misguided, you can quickly gather information and reassess.
As somebody who was top ~2% in Starcraft I got heavily triggered until this part<p>>The key insight from Starcraft is that you need to be constantly testing and adjusting your hypothesis by scouting your opponent, all while continually executing whatever the best strategy you know of at any given point in time.<p>This is one of actually a good description of what is to be (in part) learned from Starcraft. However the level to which you can opt to commit to your idea without accurate information is larger than most people can suspect, and a lot of work goes into making the unwanted adjustments seem a non-issue.
Hi David, thanks for the article. Maybe poker, basketball, and Starcraft are highly correlated hobbies among software engineers in the Bay area, but I find it incredible that we share such similar interests. Starcraft not as much for me anymore, despite my HN username (though I still play maybe once a month or so).