|
The Shadow Ras
|
| Ulrich Ulriksen |
Posted on 30-04-2024 03:41
|
Directeur Sportif
Posts: 3300
Joined: 02-11-2010
PCM$: 300.00
|
In the name of trying to fix sprinting I decided to rerun the Ras using some different parcours. Two ideas were floated in another thread. (1) Go to 2 lane roads and put a turn in the last 2k and (2) make the sprints finish on a downhill. Since it is only a three stage race it is easy to run tests.
I ran the race 7 more times. Once with the original parcours (so two total with KB's actual results, set 1), twice with 2 lane roads and a turn (set 2), three times with 2 lane roads a turn and a downhill finish (set 3, I did one extra here as I was playing with the degree of DH angle) and once with 4 lane roads and a downhill finish.
I put the average results below on GC below. The correlations are averages across each test with how well the finish rank correlates with two calculations of rider quality. First 50/50 SP/AC (used in the rank column) and second 40/30/30 with FL/SP/AC. Any result of 20th or worse is shown as 20th.
| Rank | Set | 1 | 2 | 3 | | 1 | Fernando Gaviria | 12.5 | 2.0 | 1.3 | | 2 | Caleb Ewan | 9.0 | 1.5 | 2.7 | | 3 | Eduard Grosu | 14.5 | 6.5 | 3.0 | | 4 | Arnaud Demare | 16.5 | 20.0 | 15.3 | | 5 | Alberto Dainese | 1.5 | 5.5 | 7.0 | | 6 | Mustafa Carsi | 14.0 | 8.0 | 15.3 | | 7 | Kristoffer Halvorsen | 8.0 | 12.0 | 10.7 | | 8 | Matti Manninen | 3.0 | 8.5 | 9.0 | | 9 | Jasper Philipsen | 6.5 | 13.0 | 6.7 | | 10 | Daniel Vesely | 12.0 | 14.0 | 6.7 | | 11 | Gabriel Cullaigh | 6.5 | 9.5 | 8.7 | | 12 | Rick Zabel | 13.5 | 4.0 | 9.0 | | 13 | Salem Kemboi | 6.5 | 9.0 | 18.0 | | 14 | Sondre Holst Enger | 17.5 | 16.0 | 16.0 | | 15 | Marco Haller | 20.0 | 20.0 | 20.0 | | 16 | Stanislaw Aniolkowski | 8.0 | 12.5 | 13.0 | | 17 | Asbjorn Kragh Andersen | 8.0 | 13.5 | 15.7 | | 18 | Itamar Einhorn | 11.0 | 20.0 | 17.3 | | 19 | Dylan Kennett | 15.5 | 20.0 | 20.0 | | 20 | Adrien Petit | 19.0 | 20.0 | 14.3 | | 21 | Bram Welten | 20.0 | 20.0 | 20.0 | | Correl SP/AC | (0.319) | (0.642) | (0.630) | | Correl FL/SP/AC | (0.485) | (0.614) | (0.594) |
So the good news is the correlations and the optics look a lot better using either of the alternatives, so I feel like we can improve the results next year in PCM22. Demare still gets screwed (there is still something with the 4th best sprinter) but generally the top guys were much more consistent.
I don't think downhill finishes are necessary. Downhill didn't improve the results much and:
1. Putting a DH finish in is harder and it puts a lot of discretion on the stage maker to judge the right downhill. Whereas putting in 2 lanes and a turn is a pretty binary decision and also easy to implement.
2. The downhill finishes get really crowded. see example below
I abandoned the DH finish and 4 lanes because the first test just felt ugly and the results were no better. I think the 4 lane finishes also contribute to the pack splits we see because the sprinters can spread out over the road and then there is nobody connecting to the peloton.
If this makes sense to people we should be able to change all the sprint races according to this next year, will gradually try to eliminate 4 lane finishes from other races but that might take a while.
Man Game: McCormick Pro Cycling |
| |
|
|
| jandal7 |
Posted on 30-04-2024 05:04
|
World Champion
Posts: 11406
Joined: 17-12-2014
PCM$: 1120.00
|
Did we go for any breakaways in any of them? 
Cool work UU 
24/02/21 - kandesbunzler said “I don't drink famous people."
15/08/22 - SotD said "Your [jandal's] humour is overrated"
11/06/24 - knockout said "Winning is fine I guess. Truth be told this felt completely unimportant."
[ICL] Santos-Euskadi | [PT]  Xero Racing
5x  x5
2x  x2
2x  x2 |
| |
|
|
| Laurens147 |
Posted on 30-04-2024 07:21
|
Protected Rider
Posts: 1139
Joined: 26-04-2020
PCM$: 900.00
|
Great work Ulrich, thanks for the insights
MG - Lotto - Caloi
[MG] New Manager of the Year - PCM.daily Awards 2022
|
| |
|
|
| MacC |
Posted on 30-04-2024 08:13
|
Sprinter
Posts: 1598
Joined: 15-07-2008
PCM$: 700.00
|
Not quite sure I follow but I trust Ulrich completely!
One thing from playing is that I find (anecdotally) that late turns make it very hard to catch break away and secondly its almost like they act as randomisers. You can go into a turn in a great position and next you know twenty riders haver passed you ..,..maybe that's a graphics thing but I really don't like it
Edited by MacC on 30-04-2024 08:53
|
| |
|
|
| KaiserAdler |
Posted on 30-04-2024 08:26
|
Domestique
Posts: 570
Joined: 13-07-2022
PCM$: 5800.00
|
Insiders have been asking around developers about the 4th sprinter bug. From what I have heard one of the developers has stated that it ‘should’ be fixed in PCM24. So if we ever make the jump to a Mid Mountain version I’d say 24 is the best bet in the near future if it does get fixed.
|
| |
|
|
| Nemolito |
Posted on 30-04-2024 08:42
|
Classics Specialist
Posts: 3386
Joined: 20-04-2020
PCM$: 425.00
|
Same!
|
| |
|
|
| Ad Bot |
Posted on 21-12-2024 12:21
|
Bot Agent
Posts: Countless
Joined: 23.11.09
|
|
| IP: None |
|
|
| redordead |
Posted on 30-04-2024 10:02
|
Grand Tour Specialist
Posts: 4910
Joined: 18-10-2017
PCM$: 200.00
|
Ulrich Ulriksen wrote:
I think the 4 lane finishes also contribute to the pack splits we see because the sprinters can spread out over the road and then there is nobody connecting to the peloton.
I think we need these kind of narrow roads with a turn in TT races also, so we don't get what happened in Deutschland.
Awesome work Ulrich!! 
"I am a cyclist, I may not be the best, but that is what I strive to be. I may never get there, but I will never quit trying." - Tadej Pogačar
|
| |
|
|
| quadsas |
Posted on 30-04-2024 13:00
|
Small Tour Specialist
Posts: 2520
Joined: 18-01-2013
PCM$: 300.00
|
4 lane epidemic is finally over, as long as I don't have to read constant garbage in race threads I am cool (and we don't go back to inferior versions of the game)
|
| |
|
|
| baseballlover312 |
Posted on 30-04-2024 16:23
|
Tour de France Champion
Posts: 16451
Joined: 27-07-2011
PCM$: 10538.70
|
Thanks for the testing Ulrich! I trust you completely! I guess it does seems to me like Test 3 led to better results than Test 2, but yeah, the downhill aspect is probably too much to implement for mild benefits.
It looks like the changes would hurt AKA (rightfully), but it would allow me to simply find weaker fields like I used to, so no complaints there. That's how it's supposed to work.
RIP Exxon Duke, David Veilleux, Double Feature, and Monster Energy |
| |
|
|
| Ulrich Ulriksen |
Posted on 01-05-2024 02:52
|
Directeur Sportif
Posts: 3300
Joined: 02-11-2010
PCM$: 300.00
|
Thanks for the thoughts. A couple more points. There is by no means a big enough sample here to be absolutely sure, Cunego has said he might try and do some more which would be good. Also, I don't think this is a total solution Grosu still did miserably in one test in set 2.
@red - yes the TT races would be second priority after the sprint races. Definitely Deutchland. If people have particular races that stick with you as problem let me know.
@Mac - agree adding turns causes some randomness. I think this is compensated for by two things (1) it helps to have a good lead out because you are probably better positioned to avoid the shuffle in the corner (which is realistic) which balances some of the disadvantages of a lead out and (2) it slows things down which helps the front runners.
@BBl - my calculation was the same as yours that the hills might be a little better but that it wasn't enough for the added cost to stagemaking. If cunego gets some time to test we might be able to refine that.
@Adler - When I first tested 22 I really thought the 4th sprinter bug was fixed but having seems more results it seems I was wrong (so much for trusting Ulrich). I do think it has changed in 22 and is less obvious but does still seem to exist.
Man Game: McCormick Pro Cycling |
| |
|
