|
Ulrich's Projections
|
| Ulrich Ulriksen |
Posted on 17-09-2025 03:25
|
Directeur Sportif
Posts: 3914
Joined: 02-11-2010
PCM$: 300.00
|
Back with my quick projections. These are based on my point prediction model which were built to project points in the PCT. So no thought to team structure and just rankings but no point projections for CT and PT since the points are normalized to PCT and the model wasn't really built for those divisions anyway.
First a quick look at last year. Teams are ordered by finishing sequence and then the number indicators the number of positions (higher) or lower the model had them finishing. So for example in PT Polar finished 4th but was projected for 12
Spoiler | PT (Abs Ave = 5) | | | Gazelle | 3 | | Tinkoff Team - La Datcha | 3 | | King Power | -2 | | Polar | 8 | | Evonik - ELKO | -2 | | Aker - MOT | 11 | | Xero Racing | 12 | | Team Puma - SAP | -2 | | Carlsberg - Danske Bank | 0 | | Fastned | 10 | | cycleYorkshire | -1 | | EA Vesuvio | -10 | | Jura GIANTS | -2 | | Grieg-Maersk | 4 | | Rabobank | -7 | | Cedevita | 5 | | Oxxo - Frisby | -10 | | Moser - Sygic | -4 | | MOL Cycling Team | -4 | | ZARA - Irizar | -7 | | Lidl Cycling | -5 | | Assa Abloy | 0 |
| PCT (Abs Ave = 3.3) | | ELCO - ABEA | 0 | | ISA - Hexacta | 1 | | Sony - Force India | 2 | | Lierse SK - Pizza Ullo PCTeam | -2 | | Team UBS | 7 | | Indosat Ooredoo | 1 | | Benetton Bimex Cycling | 9 | | Specialized | 3 | | Tryg - Eni | -5 | | Minions | -1 | | McCormick Pro Cycling | -1 | | Ekoi - Le Creuset | 9 | | DK Zalgiris | -7 | | Bralirwa - Cegeka | 1 | | Lotto-Caloi | -2 | | Sauber Petronas Racing | 1 | | Everesting | 3 | | Team Popo4Ever p/b Morshynska | -4 | | JEWA TIROL | 3 | | Manada Coyote | -2 | | Tafjord Kraft | -2 | | Kraftwerk Man Machine | -14 | | Llapi-Vita | 0 | | Podium Ambition | 0 |
| CT (Abs Ave = 2.6 ) | | | Newton Foundation p/b Zwift | 0 | | Gjensidige Pro Cycling Team | 0 | | Trans Looney Tunes | 0 | | Simba Cement - Tanga Fresh | 1 | | Duvel-Tsingtao | -1 | | Team Würth MODYF | 2 | | Genii Hyundai N Cycling | 6 | | Spark Team NZ | -1 | | Glanbia | 7 | | SEE Turtles | 7 | | Ethiopian Airlines | -1 | | Euskadi-Murias | 0 | | Euskotren - Pays Basque | -4 | | Cervelo Test Team | -8 | | Zain - Omantel | -4 | | Babymetal | -1 | | Hilcona Racing Team | -3 | | Air New Zealand-Alfa Romeo | 0 |
On the whole not bad. The top 5 predictions were pretty solid with 4 out of 5 in both PCT and PT and 5 out of 5 with only 4th and 5th reversed in CT. Absolute average miss was highest in PT and there were a lot of big misses. This doesn't surprise me because I think the scoring dynamics are a lot different in PT than PCT and CT and the algorithm is tuned to PCT.
Looking at some bigger misses a couple of patterns emerge. The algorithm probably overvalues puncheurs and this got Kraftwerk and Versuvio with Turgis and Mohoric and Cervelo with the whole team. Kraftwerk and Vesuvio also suffered from under performing sprinters in Zabel and Van der Lijke.
I had a theory that Kraftwerk might have suffered from the sprint lead out issue more than any other team, and Zabel's 300 point underperformance to my algorithm reinforces that, given no other top sprinter under-performed by nearly that much. I wonder if Dave's conservative planning may have burned him here as PCT sprinters tended to do well in PTHC races where they weren't top dogs but my impression is Kraftwerk tended to focus on lower level races with Zabel, where he was the favorite (could be completely wrong about this).
Among over performers I think Buchmann was treated like a hills rider by the algorithm but performed like a stage racer driving UBS, while nearly every rider on Bennetton, except Powless, over-performed the projection - so good planning by Adler maybe.
In the CT I think the TT teams did well, that is an area where roster construction definitely matters, a mediocre TTer on his own is pretty worthless but with 4 of his friends they all become more valuable.
PCT projection below, PT and CT will come later this week.
Edited by Ulrich Ulriksen on 17-09-2025 03:59
Man Game: McCormick Pro Cycling |
| |
|
|
| Ulrich Ulriksen |
Posted on 17-09-2025 03:29
|
Directeur Sportif
Posts: 3914
Joined: 02-11-2010
PCM$: 300.00
|
Going to start with PCT since that is what the algorithm was meant for. I did tweak it a little this year and it is not tuned to the new game version. So past performance is no guarantee of future results, as they say.
| 1 | MOL Cycling Team | 4,014 | | 2 | Hunter Valley Cycling | 3,986 | | 3 | Assa Abloy | 3,560 | | 4 | Trans Looney Tunes | 3,272 | | 5 | DK Zalgiris | 3,188 | | 6 | Specialized | 3,116 | | 7 | Minions | 3,024 | | 8 | Gjensidige Pro Cycling Team | 2,969 | | 9 | Bolt - Eesti | 2,910 | | 10 | McCormick Pro Cycling | 2,877 | | 11 | Everesting | 2,835 | | 12 | JEWA TIROL | 2,834 | | 13 | Ekoi - Le Creuset | 2,765 | | 14 | Simba Cement - Tanga Fresh | 2,688 | | 15 | Zwift - Newton Foundation | 2,486 | | 16 | Bralirwa - Cegeka | 2,482 | | 17 | Duvel-Tsingtao | 2,445 | | 18 | Tryg - Eni | 2,346 | | 19 | Lotto-Caloi | 2,299 | | 20 | Sauber Petronas Racing | 2,237 | | 21 | Spark Team NZ | 2,221 | | 22 | IESE ProCycling Team | 2,070 | | 23 | Colombini Kometa | 1,842 | | 24 | DeNA Roadstars | 665 |
A fair amount of consistency with other systems here. And as others have said, very evenly matched in the middle. The gap between MOL and Zalgiris in 1st and 5th is as big as Zalgiries and Duvel in 17th. A lot of love for Trans with Padun predicted to be a solid scorer but also good points from Zabel (the algorithm has no memory) and Govekar plus solid scoring throughout the roster.
Bolt-Eesti has now covered the full range with projections of 3rd, 9th, 14th and 20th. While Sauber has pretty concentrated projections - 13th, 15th, 18th, 20th, the average of the four is . . . 16.5. We just need someone to come in at 14 and we are there.
According to my algorithm Adler will need to bring out the planning magic again. Or maybe the machines just don't like Adler.
Edited by Ulrich Ulriksen on 17-09-2025 03:49
Man Game: McCormick Pro Cycling |
| |
|
|
| Ulrich Ulriksen |
Posted on 17-09-2025 03:29
|
Directeur Sportif
Posts: 3914
Joined: 02-11-2010
PCM$: 300.00
|
CT Rankings
| 1 | Kraftwerk Man Machine | 2,214 | | 2 | Podium Ambition | 1,421 | | 3 | Bianchi Enel | 1,351 | | 4 | Euskadi-Murias | 1,333 | | 5 | SEE Turtles | 1,272 | | 6 | Peugeot - Bancolombia | 1,233 | | 7 | Ethiopian Airlines | 1,220 | | 8 | IncaLine Pro Cycling | 1,120 | | 9 | Parmalat - Ecopetrol | 990 | | 10 | Tafjord Kraft | 977 | | 11 | Glanbia | 860 | | 12 | Euskotren - Pays Basque | 768 | | 13 | THE ONE - Fox God | 641 | | 14 | Kone-Nordea | 620 | | 15 | Air New Zealand-Prada | 568 | | 16 | Team Zappes Kölsch | 520 |
Points are fuzzy here, more an indication of relative strength than a projection of scoring.
Another vote for Kraftwerk from me.
A little surprised by Peugeot, when I glanced at this earlier I thought I had them in first. My system doesn't like Galta much and doesn't acknowledge the value of pure PRL guys like Kopfauf (which is a flaw). But 2nd through 7th are pretty tightly clustered in the point values so not sure I see a lot of difference there.
Last year we underscored SEE Turtles so whatever they did made my math much happier. Glanbia not showing the same improvement, Meintjes scores well but the algorithm is going to tend to pull outlier values back to the mean so it doesn't credit him with a Jorgenson-like season even if that is a possibility. Not a lot of other scorers on Glanbia after Meintjes and O'Loughlin.
Edited by Ulrich Ulriksen on 19-09-2025 03:52
Man Game: McCormick Pro Cycling |
| |
|
|
| Ulrich Ulriksen |
Posted on 17-09-2025 03:29
|
Directeur Sportif
Posts: 3914
Joined: 02-11-2010
PCM$: 300.00
|
And here is PT, not a lot of faith in the algorithm here and point estimates are so fuzzy that I didn't include them. Overall rankings are similar to others with a few outliers. Some of have called Lierse a PCT team, maybe my algorithm being a PCT algorithm explains the relative love there. And ELCO in 4th is also an unexpected result.
| 1 | Evonik - ELKO | | 2 | King Power | | 3 | EA Vesuvio | | 4 | ELCO - ABEA | | 5 | Team Puma - SAP | | 6 | Indosat Ooredoo | | 7 | Tinkoff Team - La Datcha | | 8 | Aker - MOT | | 9 | Carlsberg - Danske Bank | | 10 | Polar | | 11 | Rabobank | | 12 | Cedevita | | 13 | cycleYorkshire | | 14 | Xero Racing | | 15 | Fastned | | 16 | Gazelle | | 17 | Lierse SK - Pizza Ullo PCTeam | | 18 | Sony - Force India | | 19 | Jura GIANTS | | 20 | Team UBS | | 21 | Los Pollos Hermanos | | 22 | Grieg-Maersk |
Edited by Ulrich Ulriksen on 19-09-2025 03:56
Man Game: McCormick Pro Cycling |
| |
|
|
| Heine |
Posted on 17-09-2025 06:12
|
Grand Tour Specialist
Posts: 4461
Joined: 08-04-2007
PCM$: 200.00
|
Very interesting to see, and both to see the similiarities with other predictions, and the difference (if you compare to mine, the placement of Spark as an example)
Looking forward to seeing CT and PT!
|
| |
|
|
| DaveTwoBob |
Posted on 17-09-2025 09:19
|
Small Tour Specialist
Posts: 2539
Joined: 07-07-2013
PCM$: 200.00
|
I wonder if Dave's conservative planning may have burned him here as PCT sprinters tended to do well in PTHC races where they weren't top dogs but my impression is Kraftwerk tended to focus on lower level races with Zabel, where he was the favorite (could be completely wrong about this).
Spot on, a favourite sprinter with poor energy stats was a curse. This year I have followed the same conservative planning but this season I have a favourite sprinter with good energy stats. I have everything crossed for a better outcome.
|
| |
|
|
| Ad Bot |
Posted on 05-12-2025 11:01
|
Bot Agent
Posts: Countless
Joined: 23.11.09
|
|
| IP: None |
|
|
| Booker |
Posted on 17-09-2025 12:51
|
Protected Rider
Posts: 1032
Joined: 03-07-2013
PCM$: 200.00
|
Right in the thick of it, thanks for sharing Ulrich! Based on your summary, perhaps our hills depth rather than bonafide leader approach helps us slightly in the model...
|
| |
|
|
| kandesbunzler26 |
Posted on 17-09-2025 19:56
|
Sprinter
Posts: 1804
Joined: 07-07-2013
PCM$: 200.00
|
Back in the middle of nowhere. We'd take it, though 10th would be even nicer 
|
| |
|
|
| Ulrich Ulriksen |
Posted on 19-09-2025 03:56
|
Directeur Sportif
Posts: 3914
Joined: 02-11-2010
PCM$: 300.00
|
Updated with CT and PT.
Man Game: McCormick Pro Cycling |
| |
|
|
| Heine |
Posted on 19-09-2025 06:50
|
Grand Tour Specialist
Posts: 4461
Joined: 08-04-2007
PCM$: 200.00
|
While I hope to do better than 6th I have to respect it, I would also guess me getting so many "weak" guys with loanins and talents hurts in general, and if your system dislikes Galta its natural. I just hope I can beat this 
|
| |
|
|
| Luis Leon Sanchez |
Posted on 19-09-2025 06:55
|
Team Leader
Posts: 5568
Joined: 12-06-2013
PCM$: 500.00
|
Pretty happy to see us sitting in 2nd and so close also to a MOL team that others are predicting could run away with the division title. It certainly puts some pressure on us to finish in the automatic promotion spots but glad to see the faith too!
Thanks for organising and sorting this one Ulrich 
|
| |
|
|
| sammyt93 |
Posted on 19-09-2025 11:43
|
Classics Specialist
Posts: 3800
Joined: 03-07-2012
PCM$: 300.00
|
Ulrich Ulriksen wrote:
While Sauber has pretty concentrated projections - 13th, 15th, 18th, 20th, the average of the four is . . . 16.5. We just need someone to come in at 14 and we are there.
The fact nobody has gone lol Sauber 16th is a little surprising, Starting to think we are the team the Division is balanced around
|
| |
|
|
| SotD |
Posted on 19-09-2025 14:48
|
World Champion
Posts: 12630
Joined: 29-11-2006
PCM$: 2980.00
|
4th PT would be a crazy achievement… 
|
| |
|
|
| Ollfardh |
Posted on 19-09-2025 19:41
|
Tour de France Champion
Posts: 15024
Joined: 08-08-2011
PCM$: 11500.00
|
It's official, we're not going down!
Changed my sig, this was getting absurd.
|
| |
|
|
| baseballlover312 |
Posted on 19-09-2025 21:46
|
Tour de France Champion
Posts: 16552
Joined: 27-07-2011
PCM$: 13638.70
|
Trying to prevent these predictions from going to my head too much. Glad to see another top 10 though!
RIP Exxon Duke, David Veilleux, Double Feature, and Monster Energy |
| |
|
|
| Caspi |
Posted on 20-09-2025 10:16
|
Domestique
Posts: 705
Joined: 22-09-2022
PCM$: 15500.00
|
Pretty solid predictions last year for CT and PCT. Hope we get close to your prediction again this year and secure a top 15
|
| |
|
|
| seancoll |
Posted on 20-09-2025 13:31
|
Breakaway Specialist
Posts: 763
Joined: 20-12-2022
PCM$: 500.00
|
Not sure how I missed this ranking over the past few days, but I am very happy with my CT ranking here and fits pretty much what I'm expecting. Especially seeing Kraftwerk so far ahead of the group, it's a battle for the remaining four promotion spots in my mind.
|
| |
|
|
| Gustavovskiy |
Posted on 21-09-2025 14:24
|
Team Leader
Posts: 6174
Joined: 20-07-2008
PCM$: 200.00
|
Nice to be predicted in 11th, but as I've said before I think it's flattering. Funny to see Bolt being placed so differently indeed. One of the more interesting teams to follow for sure. Thanks for sharing your prediction!
|
| |
|
|
| TheManxMissile |
Posted on 28-09-2025 16:12
|
Tour de France Champion
Posts: 17833
Joined: 12-05-2012
PCM$: 0.00
|
I don't think 2nd-8th will be quite that ttight, but the lower promotion places usually are there or there abouts. I would happily take the top side of tight! I just worry, based on history, about underperforming. Any vote of confidence is cool, can I somehow turn that into stat boosts?
|
| |
|
