Operators Beat Founders at Repeated Games

A founder’s edge is the first move. They see a thing nobody else is pricing, they commit before the crowd, and the payoff comes from being early and being right once. That is a single draw. The distribution has a fat right tail and most of the mass sits at zero, which is exactly the shape you want when you only have to win one time.

Repeated games are a different problem. When the same decision comes back every week, the rare brilliant call matters less than the median call made a thousand times. The edge stops being vision and starts being execution: position sizing, error correction, refusing to press when the read is thin, doing the unglamorous thing on schedule. Variance per decision falls, and the operator who is consistently a little better compounds past the founder who was spectacularly right once.

I think this is why the people who are good at starting things are often not the people who are good at running them, and why that is not an insult to either. They are sampling from different games. The founder is paid for the first draw. The operator is paid for the law of large numbers.

The trap is admiring the founder’s move and trying to import it into a repeated game, where it reads as recklessness. Once you are playing the same hand over and over, the question is no longer “what is the boldest right answer” but “what is the answer I can make correctly, again, when I am tired and the read is unclear.” Boldness is for the first draw. After that, it is just unforced variance.

[McKinley: insert the specific repeated-decision example here — the call you have had to make the same way, many times over.]