Ahh the Thursday drumming. No weekly ride calendar would be complete without it.
Yes, yes, another absolute hammering today in Stage 2 of the September Zwift Racing series, this time on the Two Villages Loop route over on Makuri Islands.
I had planned to get on the bike earlier today, ideally around 11:40ish so as to get a proper warm up. Yesterday I ended up working an hour later than expected so was hoping to claw back that time today. Not so. I was called into a meeting at 11:50, and managed to get out with just 5 minutes to spare.
So, basically no warm up.
Not that it really matters.
This one, being only ~13km wouldn’t take too long. Just long enough for me to get dropped. As per.
It definitely would have helped to have done some route recon today. I wasn’t aware exactly how much climbing was in store, and ended up burning all my reserves at the base of the only real climby part of the lap, due to the afore mentioned lack of knowledge. Whoops.
However, even if I had been able to stick to the pack at the bottom of the climb, the only difference would have been me being dropped slightly later on. Perhaps enough to claim a few extra seconds on the lap, but probably not much more than that.
So yeah, if you are doing this one, be aware the climb comes around the 7km marker.
It’s really hard to see on the minimap, too.
It lasts for about a kilometre and gets up to, at most, around 5%. Mostly it’s around 3%, but expect the bunch to push the pace.
After that I was able to see out the remaining ~4km or so with one other rider. We took a couple of turns each, winding down the distance. However he was able to completely blitz me with better tactics, and a very crafty use of the draft power up with 200m left to go.
Soundly beaten. But a good incentive to push right to the end.
Rather than use Zwift Power today though, I came across another racing results / analysis site yesterday called ZwiftRacing.app.
The names a bit weird, as it’s not an ‘app’ in the sense of being like, well, a mobile phone app.
It’s just a website.
The interesting thing about the site is this:
And then this:
ZwiftRacing.app uses the concept of Elo, which they have craftily branded as vELO. Nice.
For those unfamiliar, Elo here’s a ChatGPT summary:
- Initialisation: Each rider starts with an initial Elo rating, which can be set to a predefined value (e.g., 1200) when they first join the competition.
- Race Outcome Prediction: Before each race, the Elo system predicts the outcome based on the rating difference between the two riders. The rider with the higher rating is expected to perform better, but the system also calculates the probability of an upset.
- Updating Ratings: After each race, the Elo ratings of both riders are updated based on the actual race results. If the lower-rated rider wins or performs better than expected, they gain more rating points. Conversely, if the higher-rated rider wins, they gain fewer points, and the lower-rated rider loses fewer points.
- K-Factor: You can set a K-factor to determine how much a rider’s rating should change after each race. This factor can be adjusted based on factors such as the importance of the race, the number of participants, and the rider’s experience level.
- Rating Decay: To account for changes in a rider’s skill level over time, you can introduce rating decay. If a rider is inactive for a certain period, their rating gradually decreases. This encourages regular participation and ensures that ratings reflect current skill levels.
By implementing the Elo rating system in online bike racing competitions, you can create a fair and competitive environment that motivates riders to improve their skills and provides a clear way to measure their progress.
And also useful:
The term “Elo” in the context of the Elo rating system doesn’t actually stand for anything specific. It’s named after its creator, Arpad Elo, who was a Hungarian-American physics professor and chess enthusiast. The system was developed by Arpad Elo in the mid-20th century to rate the skill levels of chess players, and it has since been adapted for various other competitive activities. So, “Elo” is simply a name and doesn’t represent an acronym or abbreviation.
With that in mind then, here’s some of the stats from the race:
And my rating:
What we see here then is that I’m in Platinum (the 6), and the Top 3 are in Emerald (the 3).
I’m racing against a field, on average, two categories higher than myself. And the winners are three categories higher than me.
No wonder I don’t win ****.
And for reference, Cat C:
So the winner of Cat C race is 30 points higher than I am. Remind me again how I’m not allowed to race in Cat C?
Ridiculous.
Anyway, it highlights just how broken Zwift’s racing categories are. But I already knew that. It’s just now I have some good proof.
Hopefully Zwift incorporate this kind of thing into their forthcoming, yet apparently very buggy “Race Score” feature. I won’t hold my breath.
See you next week for another drubbing.
