(This is a post to go on for all of the worlds) 254 lost by 11 in their first match when they had a 60% chance of winning (according to statbotics)
1 Like
60% favorites lose 40% of the time, it’s really not that shocking
14 Likes
I really like Statbotics but I’ve learned to let some of these competitions breath early on.
At this point in the season a lot of these teams are who they are, but some are definitely improving, getting more reliable etc.
I guess I mean I ususally let a little bit of data be collected before I’d be super surprised byomething like that. I’ve noticed by day 2 of a competition Statbotics gets really dialed in haha
That, and we’re dealing with very, very small data sets.
Speaking as a physicist, I generally start paying vague attention to something real world (as opposed to in my physics labs) once you’ve got 250 or so data points and can start seeing what kind of distribution you’re getting (usually, but not always, normal). A team that plays six weeks in a row has a maximum of 60 qualifying matches and maybe another 20-30 eliminations matches under their belt, almost none of which is under any situation one might consider controlled–which doesn’t mean the data is bad, but it means there’s going to be a lot of jitter.
90 jittery data points doesn’t give me much confidence in predictive value beyond a vague maybe-probably.
10 Likes
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.