Discussion: Rolling Rankings

The rolling rankings have been updated for Red, Green, and Blue. The other courses will follow in the next couple of days.

Individual scores are available also. This is the chance to see how your ranking score compared to others at the same meet, and how that great/lousy race affected actually affected your ranking.

If you have any questions, let me know.

Some names are underlined. What does the underlining indicate?

:) The underlines are a mystery, even to me.

I think they are leftover from a underline formatting experiment in Excel that I can't see with the naked eye. I didn't notice the lines until I opened the file later with a browser other than IE.

If someone could look at the source and tell me what to delete, I would love to do it.

I figured it out. Didn't it make you feel special, Brooke? :)

Ha ha -- yes, temporarily special. ;-)

Valerie, this is great!

However, looking at the individual races (on Blue) it is evident that all sprints produced extremely low scores for winners. Also some low key races gave more than 110 pts to winners.

Do you see any ways how to improve these discrepancies?

May be some kind of weighting for championship and sprint races would fix the problem?

Your hard work is highly appreciated. You and your team put rankings onto new level.

Thanks a lot!

Given the way the rankings are calculated, a discussion on how to change them to make things more equal would likely look much the way the current IS scoring debate looks. I'm going to get some popcorn.

We have had a discussion about ranking sprints in the past (in the context of the IOF scheme, but the same principles are true - in fact, more so - for the USOF scheme). Loosely, the advantage of the best over the worst is less in sprints (in the proportional sense that matters for rankings). The most technical races of the year (eg the middle at the NAOC) produce the highest high scores because the slowest runners in these races are extremely slow because of errors. Also, to make it worse, sprints on Blue and Red attract larger fields of slower runners who don't run that slowly in them, which exacerbates this effect.

I think you would need to come up with a scheme that uses more moments than the current one (not just the mean and variance, loosely, but the skewness also). Alternatively, as you say, you could just rank as now and then report the results with sprint points multiplied by some small factor. Or even better, we could rank sprints, middle, and long courses separately...

I don't see any race high scores that look particularly problematic on Blue. Oli Johnson has high scores at Batona, but this is because he won by a lot. Any other race you are thinking of?

I commit now not to respond any further to this thread before tomorrow at the earliest.

More specifically, this is about Blue sprint races ##2, 12, 23, and 26.

I am taking back my "tirade" on some low key races - winners were known international athletes.

However, John's 123 pts score at NAOC middle (race #27) raises eyebrows :) He is great athlete and is definetely 100 pts runner but the score is way too high. Even my score of 101pts with almost 4 minutes lost on errors makes my heart to pound hard.

Can you check calculations for this race?

I am joining feet :)