Well, yes I think anyone taking time to analyze what happens in each goal is going to beat at any formulaic algorithm for what constitutes a goal that is easy or hard to stop. I mean, I was a software algorithm designer for predicting network bottlenecks and rerouting over SDWan using statistics and a bad beer leaguegoalie in my previous life so I kinda know some of the shortcomings of both, lol. (I know people talk a lot of shit but I'll dm you my LinkedIn if you don't believe me. It won't have my personal goalie stats tho.
But over a whole season, or career, most stats can be useful when they aren't lying. I also know what bad algorithms that don't reflect reality look like. I get it's the best we got, but I much prefer
@Bleedred 's take even if I don't always agree.