I'm talking about being able to accurately assess the comparative chances of success between two (or many) startups, not to accurately predict the chances of a

I'm talking about being able to accurately assess the comparative chances of success between two (or many) startups, not to accurately predict the chances of a single startup's success. This is something that many talented developers do. Technical prowess does not mean company success, but if you are able to identify a company that with significantly higher comparative odds of success, in which your technical prowess can directly impact the success of the product, you are in a much better position to achieve a positive outcome than if you had joined a company at random where your talent didn't have significant incremental value-add. With regard to your prowess adding significant value-add, I think it is easiest to give a counter-example / example. SnapChat's success likely doesn't ultimately lie in their ability to create beautiful HTML5 landing pages for their app on the web, so their front-end web developers likely don't add huge incremental value. Conversely, Google's success largely hinged on their ability to return highly relevant search results with great speed, so early developers with talents in these fields had a much larger impact on the eventual success of the company. - This is the kind of situation you should aim to find.