Nov 15 2007
Upgrade Criteria
Someone started a thread in the Cross Forum about upgrading and tewills chimed in with a couple ideas on the upgrade process. According to Candi Murray, OBRA currently has the same 3 top 3’s / 5 top 5’s criteria for cross that it does for mtb racing. It isn’t the stricter, more elaborate set that road racing has probably due to having less races to judge a rider by in a season. However, the field sizes are growing. I don’t think it’s the ideal criteria but failing a better suggestion, it’s the best we have.
Caveat - I am currently a B+35 racer. My season results (5th, 1st, 3rd) put me right at, but just below, the upgrade limits. I will not say I should or shouldn’t cat up although I am sure that 99% of any comments will say that I should. If you do think I should - I am genuinely interested in hearing why.
Anyway, I was thinking about the process this year, particularly after Brooke Hoyer began posting lap time information on his blog and the overlap that happens between categories. Brooke put the Barton lap times here. Comparing the A+ to B+, the average B+ lap time for 1st, 5th , and 10th places (about 8:10) was right at the 10th place A+ lap time (also 8:10…assuming I’m reading the data correctly). You could make an argument that those top 10 B+ riders would do reasonably fine in the A+’s. Comparing the open B’s to A’s, the difference is even less noticeable. But by current criteria, only the top 5 really get scrutinized.
So is placement a realistic indicator of performance? I’m beginning to think that maybe it isn’t. Perhaps lap times would be a more accurate representation of ability…and category placement. Instead of 3 top 3’s, maybe 3 races where your lap times are better than the 10th or 20th percentile of the next category up is sufficient. This may be the top 3 racers or the top 10. Aside from deliberately slowing down for an entire race, sandbagging would be much harder to do. Simply ‘bagging it to drop a spot or two down into the 6th-10th places that fly under the upgrade radar might actually be counterproductive.
The reality of it though is that implementing this criteria would mean significantly more work for the race officials. In addition to the time they spend getting placement right, they would now also have to track finish times and # of laps to get average lap times. The race chips would be a potential solution but they’re expensive and are reputed to be somewhat unreliable.
Obviously this is just a thought. But if we’re looking to have a category system that feels fair, maybe lap times are a better indicator of performance than placement.
Okay, about that chart. I took all the lap times for each category and grouped them. The chart then displays the range of the lap times and average lap time for each category. The top of the bar is the max lap time and the bottom of the bar is the min lap time with the diamond the average lap time.
In the post I also included a histogram of the A lap times. If you notice, there is a single lap time of 8:20 way out on the edge — perhaps a flat or crash - with the next slowest around 7:40.
The best chart to look at is the one that plots one standard deviation above and below the mean. That will give you a better idea of how the fields stack up. About 65-70 percent of the lap times for each field fall within that range. That removes any outliers like the 8:20 lap in the As.
Yeah, that makes more sense. I was under the impression that the times you got from Candi were 1st, 5th, and 10th place only which left 60-odd slower riders out there. I didn’t think you had more samples from the whole field.
The Std Dev chart does tidy up the differences between the categories. Still, there’s 15-20% on each end of those ranges. I’m thinking that those outside of that range can be bumped up or down.
They are only the lap times from the places 1, 5, and 10. For example, I got 8 lap times from each A+ rider. That means I had 24 lap times for that category. The plots show the max, min, mean, and standard deviation of those 24 lap time.
Sure, it’s only for the top of the fields, but it does show that there is some separation between the fields.
For another race (can’t recall offhand which it was), I saw some data that included the 20th placed rider. Even then, the 1st placed rider in the lower category wasn’t overlapping with him. I think it was Cs and Bs at Ranier.
I’d say it’s reasonable that the top couple of riders would slot into the middle of the pack in the next cat up. And that’s generally what happens except for a few dudes who are obviously in the wrong category at the start of the season.
After all my analysis — I wish I had more lap times to do a more comprehensive study — I think that much of the talk to fix sandbagging is a solution in search of a problem.
Each year there are some guys that obviously need to cat up and they do. Then there are the guys that worked hard and win some races and they earn the upgrades. And then there are the problem guys who sneak around the rules by soft pedaling to poorer results to avoid the upgrade. There seems to be one or two a season. Still, not a problem.
I’ve raced in C, B, B+, and A+. I’ve never had a real problem with the distribution of the talent in any of those fields.