While the cobbles stage didn't go my way I have to be happy with the final result. A GC 4th with two more top 20's is quite good for a CT team.
Lehtinen taking the youth jersey is awesome - great to finally see him get a bit of glory for himself!
While the cobbles stage didn't go my way I have to be happy with the final result. A GC 4th with two more top 20's is quite good for a CT team.
Lehtinen taking the youth jersey is awesome - great to finally see him get a bit of glory for himself!
tsmoha wrote:
Can't really agree that Czech Republic (MO/HI issues) and this one should be compared. I can see your frustration, but it's not like all climbers gave a sh*t about the stage here. It's pcm after all and no, it's definitely no issue with the stage in this case, as then also other climbers like Roglic (similar GC position) would have ride like Kwiatkowski did. We can talk about Morton, who did not give a f**k about T-A (or P-N) this season. Not a stage issue though. Just PCM deciding to let a stage favorite doing nothing or too less.. Can't replay stages, when two or three guys did not deliver. The fact that guys like Valls, Penasa or Roglic did act like they had to, should count after all.
Are there any other cases of major favorites with 78 mo losing 10 minutes to guys with nearer to 70? The AI doesn’t suddenly let a clear, far and away top 2 favourite lose that sort of time without a crash or a stage issue, and we’ve turned off crashes this year.
Also, you say only 1 climber was affected, but realistically there were 6 top climbers in this race (iirc the rest have signifanctly lower acc but I can’t check stats right now) being Roglic, Valls, Pensada, De la Cruz, Bennett and Kwiat, whether or not they were helping others, that’s half the best climbers losing huge time and honestly the next best climber, Garby lost pretty big time on some much weaker rivals as well. I’ll check rider stats on that when I get home, but that’s probably a higher percentage than in Czech Republic afaik, so that’s a similar extent of issue in my opinion.
I guess my problem is that what if we relegate now by say, 50 points, then I have to look and go “oh, if both were replayed I’d still be up, if both weren’t I’d still be up, but because one was and one wasn’t, I’m losing Pedersen and Kwiat and all my long term plans are dead.” That would be super demotivating.
Edited by trekbmc on 15-11-2018 04:45
Great final stage for us A shame we couldn't topple GVK but he fought well and deserves that 2nd place. Absolutely stoked to be on the podium, the route suited Debesay but needless to say he exceeded expectations massively Hope he can make a habit out of it when he comes back maxed in the future. Bennett with a nice last stage, was hoping he could tussle for the stage but the Joker was too good, and that was very nice to see A jump to 12th on GC is awesome, will certainly be bringing these two along again in future. A very nice points haul but also some positive and strong racing is great to see.
The points jersey is a nice bonus, that green looks good on Debesay (what I thought yesterday when he had it during the stage) but that transfert looks hideous. Will have it changed to a dark blue one next year Thanks for the excellent reporting as ever tsmoha and congrats to Matt, bbl, Booker, whitejersey and more on their success.
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
Not the best day from Riesebeek, but 6th GC is simply very, very much above expectations. I probably just got lucky, but I can't help but think I read this race well by making sure he wouldn't get blocked. This profile screamed AI issues, so I feel like it's something you have to act on.
Thanks tsmoha for this route and the reports! I loved every minute of it!
Imo races/stages should ONLY be re-run if one of the 3 below is an issue:
- Teams/riders not correct
- Narrow roads
- Wrong information about HI/MO ratio
Poorly tested races, bad AI, weird results or stuff like that doesn't give any reason to re-run a race. For me stuff like that opens up for a lot of discussion and losely debateable causes, which is not ideel.
Whether Czech Republic should have been re-run or not, I don't know, but I see absolutely no reason why this should be. Bad AI or passive behavior from decent/good riders is always a risk - even more so in races like this where we have little reference to draw conclusions from.
I would have NEVER opted to send Lecuisinier or Spilak (when I had him) to a race like this, because the risk of the AI deciding for a different rider to be more valued was waaay too big. I have seen Spilak be 3 star favorite, only to end up in the 40-somethings quite some times... It happens to all of us.
SotD wrote:
Imo races/stages should ONLY be re-run if one of the 3 below is an issue:
- Teams/riders not correct
- Narrow roads
- Wrong information about HI/MO ratio
Poorly tested races, bad AI, weird results or stuff like that doesn't give any reason to re-run a race. For me stuff like that opens up for a lot of discussion and losely debateable causes, which is not ideel.
Whether Czech Republic should have been re-run or not, I don't know, but I see absolutely no reason why this should be. Bad AI or passive behavior from decent/good riders is always a risk - even more so in races like this where we have little reference to draw conclusions from.
I would have NEVER opted to send Lecuisinier or Spilak (when I had him) to a race like this, because the risk of the AI deciding for a different rider to be more valued was waaay too big. I have seen Spilak be 3 star favorite, only to end up in the 40-somethings quite some times... It happens to all of us.
+1
This had AI gamble written all over it and handling that was a big part of the race IMO.
The thing is with anticipating bad AI there were arguments in Czech Republic that from the profile you could tell that there would be some mountain influence, so I don't think quite fair to say that I should've anticipated the bad AI, because others could have also anticipated the need for the mountain stat in Czech Republic, iirc it was only clear that it was suited towards the hill stat and honestly, what sort of race requires you to anticipate bad AI? Races are generally tested before being implemented, as was this one from what tsmoha said.
To be completely honest, I don't think Czech Republic should've been replayed, but the truth is that it was, and (at least, as I intepreteded it) it was done on the premise that the same course of action would be followed if the events were to repeat themselves in other races, for the sake of consistancy.
Here, the only real difference I can see between Czech Republic and Cyprus, is the number of seriously affected managers, but that's only because the other managers affected (Jandal, White Jersey & to some extent BBL), had no GC aspirations for their climber and ended up with a good result from another rider, in Czech Republic, the losing managers had GC aspirations associated with their best puncheur, so there were more people who were in support of a replay and it doesn't help that in Czech Republic, all the managers with a positive result in the first run had a good result in the replay, except for my team, so it's admitedly very frustrating to be caught in this not once, but twice - and on opposite sides
And I get that most managers are happy with the (somewhat ridiculous) result of that stage, because people came to this race with low expectations (honestly, only Matt and I had reliable hopes for a GC podium), so under normal circumstances, I would just move on, take the disapointment and look to the next race, but unfortunately, I don't think the possibility of relegating over this is at all fair, so even if we don't replay, I think there should be some sort of comprimise.
The thing with the Czech replay was that the Mo Hi Ratio was set completely wrong. It was supposed to be a hilly stage (hilly looking route, hill icon -> ratio <0.35 could be expected) but iirc it was set to sth quite a bit above 0.5 so it was quite obviously a stage issue there. That's not AI reasons, that is a mistake in the stage files.
The precedent that was set is that a stage will be replayed when
a) obviously wrong mo/hi ratio
b) narrow roads
c) wrong startlist
d) not working stage mechanics (e.g. no cobble icon on cobbled races)
affect a race. Bad AI hasnt been a reason to replay a race and it should not be. Let's face it: PCM is far from perfect and AI could be a lot better and every manager is impacted by it at points.
Narrow roads and a wrong ratio could be replay-worthy reasons here but tsmoha is one of the most experienced managers and i trust him to notice narrow roads causing someone like Kwiatek to drop.
So i think the only chance for you to get a replay is when someone checks the mo hi ratio and finds out it is below 0.5. Otherwise this result will stand as it is. But i dont think there was a mistake because the results show quite a tendency that it was with a very high ratio. The top positions on the crucial stage were dominated by climbers (Valls, Triple P, Jeanesson, Roglic, ...) and amongst them Medvedev who is one of the most one dimensional climbers in the game.
So that leaves us with AI reasons for the underperformance. I dont know whether he had a -5 day or worked for a team mate or sat up once dropped for whatever reasons but i think it should rather be compared with Swift not bothering in Qatar or the 2016 USAPCC(?) where climbers couldnt match the expectations of many than with the Czech replays.
As much as i like riders like Kwiatek and as much as i like your team but i really dont think a reply should be on the cards. It's a case of bad AI and you see that so damn often with PCM that it cannot be a reason for a replay for a replay. I deaply hope that this wont make a difference for the relegation fight.
Completely agree with knockout and that's why a replay won't happen here. I know the MO/HI ratio, by the way, since I made these stages by myself just to clarify: I hated to see Kwiatkowski not bothering too much about the mountain stage, since he was in the position to attack the GC like Roglic did. Much better position than Valls, too, who actually won the stage eventually. As I said before: spent hours of testing this race with many many different settings, stage orders, cobbles kilometers and stuff. And as it is now, it offered climbers a chance to stay in GC contention and actually caring about the mountain stage then. Why Kwiatkowski didn't, I don't know. Only Cyanide may know. It sucks, yes. But at least he didn't throw away his race completely, did he? As shown in the final stage.
I guess we two, trek and myself, won't come to an agreement here. But for me it's pretty obvious, that it's not about the stage mechanics, but awful AI in Kwiatkowski's case.