Just wanted to chime in with one minor thing to contribute to this conversation. But I'd like to preempt any detailed questions about specifics to the Ecology event. I don't know any details of how the event was run or specific team answers, nor do I have easy access to them. I'm also not likely going to share scoring information beyond what's already publicly available.Uber wrote:.......Ecology is an extremely variable event where almost every tournament has a different style and focus. Yes, the style is somewhat different from other tournaments, but the content it tested was overly basic, with some random sections mixed in. The random sections involved guessing what the proctor wanted. Since most of the test was too easy, the series of fill in the blank biodiversity and random vocab was what determined most of the final rankings, and we ended up on the wrong end. When 25% of the teams score within 3% of each other, there's something wrong with the test.
One of the numerous checks and balances we have as part of the scoring process is a cross-correlation between each events rankings and the overall team rankings. This is built into my Excel scoring system and primarily used to catch major errors like having low score wins instead of high score wins. The math involved is pretty simple, and anyone with a copy of the final score sheet that is posted on the website could generate these correlation coefficient numbers.
However, it's trivial easy for me to just share these with you for a bit of perspective (note these include ONLY the competition events, not the trial events):
Max event correlation: 0.52
Average event correlation: 0.39
Standard deviation: 0.09
Min event correlation: 0.20
Ecology had a correlation coefficient of 0.28. This means it was about 1 standard deviation below the average correlation for events. There were several events with lower correlations.
In essence, what this means is that statistically, the resulting ranks in Ecology are reasonably well aligned across all teams with the overall team ranks.