The OVL calculations for all riders are formula based, using a rating for each rider for the categories of 'Puncheur', Stage Races', 'Sprinter', 'Time Trial', 'Cobbles' and 'Climber'.
Based on these formulas, I've calculated a category rating for each team, based on their best 3 riders for each category. A rider can only be ranked 1st for his team in one category, and then 2nd for his team in one other category - with the exception of Sprint and Cobbles.
This results in a team division rank for each category (shown below) alongside a team category rating. I've experimented with a few methods, based on the ranks and the ratings, to predict the overall PT standings.
Puncheur
1
Hugo Boss
:
Ponzi
Bilbao
Slagter
2
Strava
:
Bakelants
Pichon
Vanendert
3
Evonik - ELKO
:
Skujins
Van Stayen
Siskevicius
4
Moser - Sygic
:
Sagan
Vakoc
Velits
5
Team Puma - SAP
:
Kelderman
Demare
Felline
6
Gazelle
:
Ginanni
Furdi
Senyenov
7
Tinkoff Sport Academy
:
Trofimov
Chernetskiy
Zakarin
8
Festina - Dexia
:
Koretzky
Spilak
Coppel
9
Jayco - Cobra9
:
Bobridge
Haas
Morton
10
Aegon - Lavazza
:
De Bie
Villella
Olivier
Spoiler
11 Pendleton's
12 Ayubowan!
13 eBuddy
14 Vesuvio - Accumalux
15 Aker - MOT
16 Spotify - Haglöfs
17 Project: Africa
18 Wiesenhof - Andritz
19 Porto - Prio
20 Iberia - Team Degenkolb
21 Nemiroff - ABBYY
22 RBC Pro Cycling
Stage Races
1
Wiesenhof - Andritz
:
Taaramäe
Cattaneo
Borisavljevic
2
Festina - Dexia
:
Spilak
Lecuisinier
Novak
3
Nemiroff - ABBYY
:
Pluchkin
Nepomnyachsniy
Scarponi
4
Gazelle
:
Madrazo
Bibby
Kangert
5
Tinkoff Sport Academy
:
Kritskiy
Trofimov
Kunshin
6
Iberia - Team Degenkolb
:
Tenorio
Faiers
Izagirre
7
eBuddy
:
Gesink
Brambilla
Aru
8
Vesuvio - Accumalux
:
Schleck
Keinath
König
9
RBC Pro Cycling
:
Phinney
Brown
Rosskopf
10
Aegon - Lavazza
:
Keizer
Dekker
Olivier
Spoiler
11 Porto - Prio
12 Pendleton's
13 Jayco - Cobra9
14 Moser - Sygic
15 Project: Africa
16 Spotify - Haglöfs
17 Aker - MOT
18 Team Puma - SAP
19 Hugo Boss
20 Ayubowan!
21 Evonik - ELKO
22 Strava
Sprinter
1
Moser - Sygic
:
Bewley
Vesely
Habtom
2
Aegon - Lavazza
:
Swift
Groenewegen
Sbaragli
3
Evonik - ELKO
:
Van Stayen
Van Avermaet
Dzamastagic
4
Iberia - Team Degenkolb
:
Degenkolb
Zabel
Merino Criado
5
Strava
:
Guarnieri
Boeckmans
Fonseca
6
Spotify - Haglöfs
:
Ahlstrand
Stallaert
Reimer
7
Team Puma - SAP
:
Ciolek
Lander
Demare
8
Aker - MOT
:
Kristoff
Enger
Feillu
9
RBC Pro Cycling
:
Howard
Boivin
Houle
10
eBuddy
:
Grosu
Van Poppel
Te Brake
Spoiler
11 Festina - Dexia
12 Project: Africa
13 Vesuvio - Accumalux
14 Pendleton's
15 Porto - Prio
16 Nemiroff - ABBYY
17 Hugo Boss
18 Wiesenhof - Andritz
19 Tinkoff Sport Academy
20 Jayco - Cobra9
21 Ayubowan!
22 Gazelle
Time Trial
1
Festina - Dexia
:
Coppel
Vlatos
Spilak
2
Porto - Prio
:
Fiedler
Zmorka
Oliveira
3
RBC Pro Cycling
:
Cataford
Phinney
Rosskopf
4
Jayco - Cobra9
:
Durbridge
Bobridge
Hepburn
5
Tinkoff Sport Academy
:
Yatsevich
Kritskiy
Malori
6
Vesuvio - Accumalux
:
Kittel
Jungels
Goos
7
Aegon - Lavazza
:
Dumoulin
Keizer
Dekker
8
Gazelle
:
Koren
Kangert
Madrazo
9
Team Puma - SAP
:
Sütterlin
Van Baarle
Bevin
10
eBuddy
:
Howson
Walsleben
Gesink
Spoiler
11 Hugo Boss
12 Nemiroff - ABBYY
13 Wiesenhof - Andritz
14 Spotify - Haglöfs
15 Pendleton's
16 Moser - Sygic
17 Project: Africa
18 Aker - MOT
19 Iberia - Team Degenkolb
20 Ayubowan!
21 Strava
22 Evonik - ELKO
Cobbles
1
Moser - Sygic
:
Bewley
Zhupa
Boroš
2
Evonik - ELKO
:
Van Avermaet
Taubel
Nooytens
3
Aker - MOT
:
Trentin
Kristoff
Ringheim
4
RBC Pro Cycling
:
Summerhill
Young
Laporte
5
Pendleton's
:
Senechal
Fenn
Abraha
6
Team Puma - SAP
:
Blythe
Declerq
Weber
7
Jayco - Cobra9
:
Vanspeybrouck
Saber
Van der Haar
8
Ayubowan!
:
Baugnies
Chatarunga
Petit
9
Wiesenhof - Andritz
:
Zepuntke
Franzoi
Bolivar
10
Iberia - Team Degenkolb
:
Maes
Kluge
Insausti
Spoiler
11 Vesuvio - Accumalux
12 Festina - Dexia
13 Spotify - Haglöfs
14 Tinkoff Sport Academy
15 eBuddy
16 Strava
17 Porto - Prio
18 Gazelle
19 Project: Africa
20 Aegon - Lavazza
21 Nemiroff - ABBYY
22 Hugo Boss
Climber
1
Wiesenhof - Andritz
:
Cattaneo
Taaramäe
Preidler
2
Pendleton's
:
Nibali
Dombrowski
Rosch
3
Festina - Dexia
:
Lecuisinier
Vasyliv
Spilak
4
Vesuvio - Accumalux
:
Keinath
Schleck
König
5
Aker - MOT
:
Guldhammer
Galta
Beltran
6
Porto - Prio
:
Reis
Machado
Vinhas
7
Project: Africa
:
Ricco
Martin
Meintjes
8
Hugo Boss
:
Carapaz
Alarcon
Slagter
9
Aegon - Lavazza
:
Dekker
Olivier
Keizer
10
Gazelle
:
Formolo
Madrazo
Sosnitskiy
Spoiler
11 Moser - Sygic
12 Team Puma - SAP
13 Spotify - Haglöfs
14 Ayubowan!
15 Nemiroff - ABBYY
16 Iberia - Team Degenkolb
17 Jayco - Cobra9
18 eBuddy
19 Evonik - ELKO
20 Tinkoff Sport Academy
21 RBC Pro Cycling
22 Strava
Summary
Led by Sam Bewley, Moser - Sygic top the tables for both Sprint and Cobbles. Meanwhile it is the combination of Taaramäe and Cattaneo that lead Wiesenhof - Andritz to the top in both Stage Races and Climber. Of the two teams, Moser back those top ranks up the strongest in the other categories. The new signing of Ponzi puts Hugo Boss as the top puncheur squad, so 3 of the 6 top ranks are coming from last season's PCT teams.
Festina - Dexia top the tables for Time Trials, and are then up there across the board. They have the best consistency, with all categories in the Top 12. Next in that count are Vesuvio - Accumalux, always in the Top 14 - and with the smallest range between ranks of 4 to 12. Pendleton's are always Top 15. But in terms of consistency, ignoring cobbles, then look no further than Aegon - Lavazza. Top 10 in all 5 other categories, albeit down the lower end of the Top 10 for a few of them.
Spotify - Haglöfs always Top 16 - although the Swedes have only one Top 10, in Sprints. They are one of only 4 teams that only has one Top 10 category. The others are Nemiroff - ABBYY, Project: Africa and Ayubowan!. In other divisions some teams don't have any Top 10s, but in PT every team has at least one Top 8.
Team Puma - SAP best rank may only be 5th, but they are one of only four teams to have four Top 10s. The other are RBC Pro Cycling. The Canadian squad is an example of a team who are strong in some areas and weak in others, another example is Evonik - ELKO, with 3 good and 3 bad.
PU
SR
SP
TT
CO
CL
Aegon - Lavazza
10
10
2
7
20
9
Aker - MOT
15
17
8
18
3
5
Ayubowan!
12
20
21
20
8
14
eBuddy
13
7
10
10
15
18
Evonik - ELKO
3
21
3
22
2
19
Festina - Dexia
8
2
11
1
12
3
Gazelle
6
4
22
8
18
10
Hugo Boss
1
19
17
11
22
8
Iberia - Team Degenkolb
20
6
4
19
10
16
Jayco - Cobra9
9
13
20
4
7
17
Moser - Sygic
4
14
1
16
1
11
Nemiroff - ABBYY
21
3
16
12
21
15
Pendleton's
11
12
14
15
5
2
Porto - Prio
19
11
15
2
17
6
Project: Africa
17
15
12
17
19
7
RBC Pro Cycling
22
9
9
3
4
21
Spotify - Haglöfs
16
16
6
14
13
13
Strava
2
22
5
21
16
22
Team Puma - SAP
5
18
7
9
6
12
Tinkoff Sport Academy
7
5
19
5
14
20
Vesuvio - Accumalux
14
8
13
6
11
4
Wiesenhof - Andritz
18
1
18
13
9
1
Bringing it together
The next stage is bringing all of the above into a divisional standing prediction. Five different rankings have been produced here, the first two come from the ranking positions, the other 3 use the team category scores themselves - sometimes adjusted, sometimes weighted.
And the average of these produces my Slightly Mathematical PT division prediction:
1
Festina - Dexia
2
Moser - Sygic
3
Aegon - Lavazza
4
Pendleton's
5
Team Puma - SAP
6
Vesuvio - Accumalux
7
Wiesenhof - Andritz
8
RBC Pro Cycling
9
Evonik - ELKO
10
Aker - MOT
11
Jayco - Cobra9
12
Gazelle
13
Porto - Prio
14
Tinkoff Sport Academy
15
Iberia - Team Degenkolb
16
eBuddy
17
Spotify - Haglöfs
18
Hugo Boss
19
Strava
20
Project: Africa
21
Nemiroff - ABBYY
22
Ayubowan!
Festina - Dexia are predicted for the hat-trick of PT titles, but this season should be closer than last season, with Moser - Sygic predicted to be top in 2 of my 5 team rankings.
Aegon - Lavazza, Pendleton's and Vesuvio - Accumalux all take 3rd in one of the rankings, and ultimately place in the Top 7 alongside Team Puma - SAP.
One of the interesting differences between predictions is Evonik - ELKO. R3 involves the scores assigned to each team for each category for Puncheur, Stage Race and the best three of the other 4. They are punished for their weaker categories and end up in the relegation positions. R4 gives the bottom 5 in each category the rating of the 18th placed team, and repeats the calculation - and that takes Evonik right up to 3rd. Of all teams in all divisions, only they have such a difference between those two ranks. In the end, the formulas have placed them 9th.
Looking at the relegation zone, there are three teams who are consistently in the bottom 4: Project: Africa, Nemiroff - ABBYY and Ayubowan!. The final two places are less clear cut. The predictions suggest it will be two from Spotify - Haglöfs, Hugo Boss and Strava. But some of the next teams up are never too far away in the predictions.
Thanks for the preview. It gives a good overview over the teams i agree a lot with most teams.
My own team also shows the problem with formula based previews. Whoever creates the ranking can change the look of a ranking a lot with just a few adjustments to the formula. The range between R3 and R4 for my team is hilarious considering that it only compensates for ignoring one terrain. Thanks for the explanation for why there is the big difference between the rankings
Must say I feel this mathematical prediction is pretty right, at least pre-season 10th for my team sounds realistic, eventhough it's probably what we'll get with a good run, and then probably a couple of places down if we're unlucky.
Interesting to see Pendleton's that high up. They're definitely a joker for a podium if Lutsenko can build on last year's performance and maybe even do better now that he's trained. Vesuvio down in 6th with this formula isn't that surprising, eventhough I expect them to be a top four team, at the cost of Pendleton's.
Wiesenhof in 7th feels pretty high, but underlines that you get rewarded for being superior on paper for climbs/stage races, but they really need Taaramäe to deliver. Same goes for Nemiroff and Pluchkin, but they obviously don't have the same kind of high quality depth that the fellow promoters at Wiesenhof got.
Difficult to predict the teams for relegation, but I must say it feels like Hugo Boss should be able to survive, as Ponzi looks like a beast now that he's trained. Though, Strava could be direct rivals in the fight to avoid relegation, meaning every race with Ponzi and Bakelants could get an extra point of interest.
And as alexkr00 kind of points out, eBuddy is probably the team that stands out as the most underrated team in this preview. Though, they probably need another brilliant season from Gesink to make this prediction look like a fool
knockout wrote:
My own team also shows the problem with formula based previews. Whoever creates the ranking can change the look of a ranking a lot with just a few adjustments to the formula. The range between R3 and R4 for my team is hilarious considering that it only compensates for ignoring one terrain. Thanks for the explanation for why there is the big difference between the rankings
Both R3 and R4 involve taking the Puncheur and Stage Race score, and the 3 best from the other 4. The difference with R4 is that you've been given a better score in your worst 3 categories (of which 2 are still being used in the calculations). So it negates the fact that you are so much worse than the average in those categories - on the basis that the 18th best team isn't likely to actually outscore the 22nd best team by much.
And finally the most important mathematical preview
Not sure if i will publish a PT preview by myself, but this seems to be similar to what I have in my mind. Jayco clearly missing the depth in MO/HI, so while I expect Bobridge to become one of the best puncheurs with the schedule he has, I'm not surprised to see us only on 9th in this puncheurs ranking. Overall, finishing around 10th seems realistic.
High quality analysis as we've become accustomed to. The weight of history underpinning SN's mathematical approach means a lot of the kinks have been ironed out, hence a very sensible looking prediction imo. Maybe a minor quibble here and there, but in general looks on the money from pre-season perspective.
Interesting to note that 6 of the promoted teams occupy the 6 lowest placings, and look destined to squabble to avoid relegation. Demosntrates the difficulty of the step up. Alternatively Wiesnhof - Andritz look to have overcome the yo-yo effect and break the top 10 here, which again is quite hard to argue against looking realistic. Teams like Aegon which have similarly yo-yo'd a few seasons back now very established in the top tier.
Note that there isn't a 'junker' team this season that sits bottom of the heap in everything and looks destined for relegation from the get-go that we have generally seen in the last few seasons. Relegation looks to be a mighty scrap - points from day 1 will be preciously hoarded!
Thank you for putting this together SN. Right now I would totally accept another season of at least being close to the fight for the overall win. But honestly, losing to Festina would be tough for the 3rd straight year. Is the most likely scenario though.
I like to see how my puncheur depth suddenly looks so nice after all these poor years after I sold Alex Efimkin. One have to wonder if this depth can get me closer to Festina this year. But it is not only about Festina, as I surely can see all teams from top 8 to end up on the overall podium.
Agree a lot with the look of these rankings, but I see eBuddy finishing quite a bit higher, Porto on the other way a bit lower.
Nice to see your math on PT division as well, in especially when my team is rated that high.
As I said in matt`s preview, I still don`t believe it as I just lack the top (non-sprinter) scorer and not sure if 4-5 solid ~7th-12th GC riders can replace a top rider on this terrain.
But who am I to judge a math formula? You can`t beat the Science and now my Top5 PT Goal is looking rather comfortable.
knockout wrote:
My own team also shows the problem with formula based previews. Whoever creates the ranking can change the look of a ranking a lot with just a few adjustments to the formula. The range between R3 and R4 for my team is hilarious considering that it only compensates for ignoring one terrain. Thanks for the explanation for why there is the big difference between the rankings
Both R3 and R4 involve taking the Puncheur and Stage Race score, and the 3 best from the other 4. The difference with R4 is that you've been given a better score in your worst 3 categories (of which 2 are still being used in the calculations). So it negates the fact that you are so much worse than the average in those categories - on the basis that the 18th best team isn't likely to actually outscore the 22nd best team by much.
I understood that. Hence thanking you for the explanation
Right on the edge, as espected will be a tough season ahead.
This preview underestimated us last season, so hopefully we will to better this year aswell
We struggled with signing depth in cobbles section as you see by our 13th ranking there but we still hope Stallaert will do better then that.
Last year we were fighting relegation, so it's nice to be projected a solid mid-table team. Being dead last in the sprinter category is our team's main founding principle, so that will never change. But due to market conditions we failed strengthening our cobble line-up, which should indeed hurt our performance this season.
I suppose R2 puts us in 8th place because that algorithm lets you get away with having two weak categories?
If I can manage 7th I will be really happy. Everything depends on Taaramae doing what he should do, and then the others having some consistent if not fantastic scoring
10th for my team sounds realistic, eventhough it's probably what we'll get with a good run, and then probably a couple of places down if we're unlucky.
