Cycling
There is currently 1 person viewing this thread.
17 Mar 17
16:31
Joined:
Date Joined:
11 Jul 11
...Have been working on a uniform TIMEFORM rating for the WT Elite Men, since the end of the last 2016 season after playing around with it for the past few years.
The quality or merit of the rider is expressed by a number representing distance not kilograms, since a weight-for-age scale in terms of kilograms does not compute away from thoroughbred racing. Like thoroughbred racing however, I have attempted to bridge the distinction which exists between jumps horses and quality stayers, and stayers and sprinters, in terms of classics riders, cobbled-classics riders, sprinters, TT specialists, punchy climbers, and pure/GC climbers. So the distance rating is very-much scaled around 1.) the type of race or "circuit", and 2.) the type of riders involved. A lot of horses can win over a certain distance, but not many can remain (at or near) unbeaten throughout their career over the 2,040m distance, for instance, so there is inspiration in weight-for-age stakes racing, if not a clear formula which can be adopted to WT cycling on the whole.
If I can eventually figure out the applicable algorithm code to use/construct (it is presently analog) it should gain a uniform denominator which would isolate and maintain a compatible rating across all types of racing and therefore riders, instead of each type of rider and each type of race. Like comparing similar types of riders from different eras, this is not an easy thing and eventually it does become largely subjective.
I can see that if I started this project on the velodrome where the "circuit" wouldn't change so wildly, I would have a more robust foundation from which to build this uniform TIMEFORM rating, however I'll be treating it as an ongoing or developing project so please resist ridiculing me if some ratings don't balance against your view of where a rider should be against another or indeed the field, given a certain set of circumstances at a specific period of the season.
If it all goes to crap, I think I may begin from scratch with the velodrome, but hopefully it won't come to that.
Also, I'm not sure if I will get a set of selections for the main races, but I'll try.
Cheers,
SP
The quality or merit of the rider is expressed by a number representing distance not kilograms, since a weight-for-age scale in terms of kilograms does not compute away from thoroughbred racing. Like thoroughbred racing however, I have attempted to bridge the distinction which exists between jumps horses and quality stayers, and stayers and sprinters, in terms of classics riders, cobbled-classics riders, sprinters, TT specialists, punchy climbers, and pure/GC climbers. So the distance rating is very-much scaled around 1.) the type of race or "circuit", and 2.) the type of riders involved. A lot of horses can win over a certain distance, but not many can remain (at or near) unbeaten throughout their career over the 2,040m distance, for instance, so there is inspiration in weight-for-age stakes racing, if not a clear formula which can be adopted to WT cycling on the whole.
If I can eventually figure out the applicable algorithm code to use/construct (it is presently analog) it should gain a uniform denominator which would isolate and maintain a compatible rating across all types of racing and therefore riders, instead of each type of rider and each type of race. Like comparing similar types of riders from different eras, this is not an easy thing and eventually it does become largely subjective.
I can see that if I started this project on the velodrome where the "circuit" wouldn't change so wildly, I would have a more robust foundation from which to build this uniform TIMEFORM rating, however I'll be treating it as an ongoing or developing project so please resist ridiculing me if some ratings don't balance against your view of where a rider should be against another or indeed the field, given a certain set of circumstances at a specific period of the season.
If it all goes to crap, I think I may begin from scratch with the velodrome, but hopefully it won't come to that.
Also, I'm not sure if I will get a set of selections for the main races, but I'll try.
Cheers,
SP
sort by:
Show
per page
Replies: 2
|
By:
Will endeavour to commit to -- at least -- the following races:
MSR: completed. E3 Harelbeke Gent-Wevelgem in Flanders Fields Tour of Flanders Paris-Roubaix Amstel Gold Race Worlds La Fleche Wallonne Liege-Bastogne-Liege Clasica Ciclista San Sebastian Bretagne Classic Grand Prix Cycliste de Quebec Grand Prix Cycliste de Montreal Il Lombardia |
|
By:
I didn't include the ratings for my selections for RVV because I now realise that my best approach is a complete tear-down rebuild, so-to-speak. I've realised that the rating itself is too similar to a percentage rating, and that it is not the best approach given that implied probability is a percentage figure inherent in the price.
Much better to provide a key/guide, I think. |