Celia Alexis Wagar's CritPoints

You’ve criticized randomness in games because it gives uninteresting variation, and can be unfair, making competition not about who is best, but about who lucked out. You’ve also criticized the “solution” to this as seen in poker, where many games even out the random variance to pick the best but even in deterministic sports (or as deterministic as real life can be assumed to be) there’s a significant randomness to determining who’s best. These made me think about this: https://twitter.com/nothings/status/762769777819398145 Googled for the source but couldn’t find it. Thoughts?

The short of this seems to be that people are random therefore game’s outcomes are random. I’d personally rebut by saying that people are inconsistent, not necessarily random. We’ve gone over how people are bad RNGs before, but performance is a bit less in our control.

I’ve actually heard about streaks in sports just being clumps in random distributions, and that more or less baseballers and basketballers follow their average performance most of the time, even when you think they’re “hot” or performing well. Radiolab did an episode on this actually.

There’s two factors to the contrary here however. First is that we can improve our consistency over time. Even if we’re somewhat random, as in a worst case scenario, we become less random over time as we improve at the task. This is why good players consistently dominate over weaker players.

Second is basically what they talk about around 8:45 in this episode:
http://www.radiolab.org/story/91911-are-we-coins/
Basically, individual games or shots or successes are not statistically independent. When you’re on a hot streak, you’re being affected by your previous success, not merely odds happening to collude on that specific moment.

What I’d add beyond this is that humans are affected by a lot of factors determining if they’ll succeed or not. I wouldn’t call this strictly random. More that we change over time, and some people are more consistent and others are less consistent.

That and if things weren’t at least slightly inconsistent that games would be uninteresting in the first place. The whole point is to experience inconsistency.

Yeah, people don’t always perform the same every time. Yeah you need repeated trials to get accurate results (this is why tournament finals are best of 5), but you need a lot more repeated trials to derive accurate results if there are actual randomized in-game elements.

People won’t always perform the same, but we’re not swirling in a vortex of meaninglessness here.