curious what the additional benefit is seen to be here
Possibly rapid placement without the requirement to have several races. Since Elo slides up and down in increments riders who don’t race often on Zwift would have less accurate placement than those who race weekly. In theory you should be able to ride 2-3 “qualifier” races and have a competitive placement.
TrueSkill is basically multiplayer Elo which makes more sense with cycling not being a 1v1 sport.
mortirolo.netlify.app/post/elo-rating-method/ gives a pretty good overview of Cycling Elo method
Sorry if this is a dumb question crew, just jumping into this new autocat system. I’m signed up for one of the classics tomorrow and have a question. I’m In the C1 pen and it looks like there are 3 B’s, me steady A, and 3 A+’s in the pen. Is this the way it’s supposed to shake out? I guess I thought it would be different. Thanks all!
How different?
They use a different metric than w/kg. So it should be different. I assume there is still a few that are not in the correct cat but it should workout out over time.
It looks to me like there is no “protection” for lightweight (low absolute watts) riders so that explains the Bs. As for you…good luck, it doesn’t look that great to me but neither is the old cat system either. Harrogate is all about 2 min power anyway, FTP isn’t a great discriminator, especially when it’s just based on what you happen to have done for an arbitrary 20 min interval in recent races.
My son - just turned 11 and has been placed in C4, this look very high when his normal ZP cat is D. His zwift power profile is 469161
Error, required tweaking or working as it should?
You should contact WTRL.
Finishing in the top 3 in a C category race sound like C4 is about the right spot.
Hi Alan, I’m entered in the 18:48 race and, looking at the entry list, one rider stood out as the only “D” rider in the race. They’re a child, the shortest rider by around 20-30cm, the lightest by about 30kg, and they’re less than half the weight of most competitors. If your son’s the same person (or in a similar situation) then I think his feedback will be especially valuable to Zwift, so I recommend he takes part, pushes hard and you feed back whatever happens, because they have less data on how to rank children accurately.
I could worry that I’ve got the joint-lowest power figures in the race over 20 minutes, but at least I know there are no “A” riders signed up, and that’s an improvement over other races I’ve been in. I’m looking forward to it!
I like it no more sandbagin:)
Yes, it was pretty darn marvelous. I was way off this morning due to an unfortunate set of circumstances but was able to hang with the lead pack for much longer than I would’ve in a regular race. The lack of sandbaggers made a huge difference. I’m thinking there were a coupla riders who should’ve been in C3 but overall the experience was much better. Thank you for taking a swing at this…
It’ll be interesting to see how the categorisation calculation itself works out over more events (I’m sure there will be plenty of people who don’t agree with the category they are put in), but just having the participants’ pen predetermined by the system and not completely open to abuse is a big step forward.
I just did the Western Australia C4 race and I’d say the categorisation seemed pretty good.
Personally I went into the race ranked 3rd of the 50 finishers but was very tired from a long ride at the weekend and a race yesterday. I finished 14th which I was happy with, losing a few places as I had nothing left to give in the last 500m.
Interestingly the rankings of riders before the race didn’t really translate to finishing positions. Here’s the list of finishers, their ranking before the race and the position they would be expected to finish if it went in order of ranking:
| Finish Position | Race Ranking (before race) | Ranking Position |
|---|---|---|
| 1 | 497.59 | 26 |
| 2 | 231.02 | 1 |
| 3 | 575 | 40 |
| 4 | #N/A | |
| 5 | 298.1 | 4 |
| 6 | 547.58 | 37 |
| 7 | 539.09 | 33 |
| 8 | 353.08 | 11 |
| 9 | 378.36 | 15 |
| 10 | 539.62 | 35 |
| 11 | 267.49 | 2 |
| 12 | 539.27 | 34 |
| 13 | 369.55 | 12 |
| 14 | 287.49 | 3 |
| 15 | 380.54 | 16 |
| 16 | #N/A | |
| 17 | #N/A | |
| 18 | 398.58 | 17 |
| 19 | 581.03 | 41 |
| 20 | #N/A | |
| 21 | #N/A | |
| 22 | 417.39 | 19 |
| 23 | 375.57 | 14 |
| 24 | 417.65 | 20 |
| 25 | 487.12 | 24 |
| 26 | 522.36 | 29 |
| 27 | 581.45 | 42 |
| 28 | 479.19 | 22 |
| 29 | 335.93 | 8 |
| 30 | 513.41 | 27 |
| 31 | 529.38 | 30 |
| 32 | 592.91 | 44 |
| 33 | 344.91 | 9 |
| 34 | 327.78 | 7 |
| 35 | #N/A | |
| 36 | 309.95 | 5 |
| 37 | 538 | 32 |
| 38 | 566.25 | 38 |
| 39 | 488.74 | 25 |
| 40 | 542.3 | 36 |
| 41 | 572.81 | 39 |
| 42 | 466.98 | 21 |
| 43 | 400.87 | 18 |
| 44 | 351.57 | 10 |
| 45 | 325.24 | 6 |
| 46 | 588.07 | 43 |
| 47 | 374.58 | 13 |
| 48 | 479.2 | 23 |
| 49 | 519.64 | 28 |
| 50 | 537.99 | 31 |
I was in that race too. Finished DFL but that wasn’t unexpected with how my life went in the last coupla days… There were actually 12 guys behind me but apparently WTRL only placed up to 50th. You had nothing to give in the last 500m; I had nothing to give in the last 15km. Nonetheless, the start wasn’t the usual bloodbath and I could see under normal circumstances I could’ve made the initial selection, something that rarely happens in “regular” Zwift races. The guy that won put down some serious power and probably should’ve been in C3.
I was cat A in the last ZRL. I could just about keep with front pack in most races, but definitely at the lower end of cat A. I’ve dropped to cat B on ZP as I haven’t raced since then. My ranking has also dropped to 560, from 213 in Feb. I’ve been placed in C3. I think this is probably about right for me, as I had Covid quite bad and have struggled a bit with recovery, but I’m guessing WTRL didn’t know that ;). It may have been right anyway, as I am better at the 20 minute end of the power curve rather than the 2-3 min.
There are 6 cat As in C3 too, the rest are Bs. Look at the various key stats (Yorkshire KOM time, w/kg, pure watts) I am about mid table. Seems like the top C1/C2/C3 covers top A, middle A, and top B/bottom A. No conclusions drawn, just interesting observations. Looking forward to the race.
Initial reports seem to indicate that racing has been improved but would be interesting to know which of the following has had the biggest impact on this -
Be interested to see what IF people end up racing at…
I would hope people are racing at .9 through to 1.1 as that would suggest the right amount of effort is being asked of people…
if they are racing at .7 or .8 then it might be to easy and need tweaking.
Personally when I race I want to on the boundary of pushing PBs but get others don’t always feel this way…
If you’re not, you’re sandbagging amirite?! 
I get that’s tongue in cheek… but it is tough thing to gauge, as the opposition can decide how hard the race is… it’s not the users fault currently… this is a slight change to that dynamic in that the total pool of racers is selected by zwift / wtrl
It depends. A traditional race is about trying to beat the opposition, otherwise we might as well just do individual time trials at 100% for the whole course and simply compare the times at the end. If you can do well without going all out for the full race that’s fine IMO. The problem comes when some in the race have a massively different ‘all out’, and we can’t force anyone to give everything (or indeed check that they did). The only thing that would truly prevent people gaming any ‘best effort’ system for their own benefit is result-based categorisation, as has been discussed at length.
Interestingly and ironically, being confident that everyone in the race is of similar ability may actually result in fewer all out races, and things being a little more tactical.
In an ideal world, this would vary massively race on race. If riders are trying to establish a breakaway, they could be riding hard, but as a sprinter you may gamble on riding easy in the pack and hoping there is no breakaway to sprint at the end. Unfortunately as the drafting is broken it rarely plays out this way.
