Here’s an image to ponder:

Pts/Serve is ultimately what we care about the most. At the end of the day, we need to score points, so the servers that are scoring the most points are (barring something like an exceptionally-poor or good defensive rotation for the server) our best servers. All the statistics we take are basically attempting to say, “here’s the server that’s going to score the most points, assuming they have an equal team around them.”

We can see pretty clearly that Serve-In % has a pretty low correlation with Pts/Serve. We can see that Ace % has a better correlation, but it’s still not perfect. And we can see KO % has the best correlation.

This isn’t surprising. KO % is a more comprehensive stat (it measures Aces + Out-of-System passes and it contains a bit of Serve-In as well), so we expect it to have a higher correlation to success. What’s great about KO % is that it’s also easy to understand. It’s human scale. The best servers on this team (as tends to be the case in HS/club volleyball) are at about 67% KO, which is about, “2 out of 3.” That human scale makes it easy for players to understand and process.