Spring Statistical Analysis - A Response to Competitve Cyclist

I’ve long been a fan of Michael Lewis’ Moneyball, his book about the Billy Beane and his statistical approach to running a Major League baseball team. So naturally I was particularly excited to read the latest post on Competitive Cyclist’s "What's New" page. (If you haven’t yet bookmarked the link or added it to your feed reader, do so now.) The post takes 2 distinct approaches to summarizing the Spring Classics season: one statistical, the other emotional. Coincidentally, I had been in the process of constructing a similar overview of my own, albeit using a different points system.

You can read the full report and analysis on Competitive Cyclist here, but I’ll do my best to summarize their process below.

1. Points were assigned to Top-10 finishes in the following races:

            Milan-San Remo
            Tour of Flanders
            Paris-Roubaix
            Amstel Gold
            Liege-Bastogne-Liege
            Ghent-Wevelgem
            Brabantse Pijl
            Fleche Wallone
            E3-Harelbeke
            Dwars door Vlaanderen
            Grote Scheldeprijis
            Tour of California
            Paris-Nice
            Tirreno-Adriatico
            Pais Vasco

2. Points were assigned as follows:
            1st = 40, 2nd = 25, 3rd = 21, 4th = 7, 5th =6….10th =1

3. Points in certain races were weighted using the following multipliers:
            MSR, Flanders, Roubaix, Amstel, and Liege x5
            Stage Races x3

4. Stage wins were not counted for points.

Now I’m not here to argue with how Competitive Cyclist constructed its system. As they admitted in the post, statistical analysis is a particularly subjective enterprise, especially in a sport that seems to resist any statistics beyond wins and placings. However, there are some places where I think I would have done some things differently.

            1. By their own admission, the Tour of California really shouldn’t be counted.
            2. Despite the persuasive argument against counting stage wins, I think these do deserve recognition. In my own point system, I accommodated them.
            3. I added Het Nieuwsblad, Kuurne-Brussels-Kuurne, L’Eroica, the Criterium International and the 3-Days of DePanne to the list. In my opinion, any analysis of the Spring Classics is incomplete without these races.
            4. Finally, and this is perhaps my biggest deviation from the good writers at Competitive Cyclist, I weighted each race individually on what I considered to be it’s own merits. I took into consideration characteristics such as prestige, length, terrain, field quality, and importance. Here is where subjectivity plays a major role. I’m sure there are holes in my logic, but I gave it my best shot.

Here’s my list of races with their weights:

            Het Nieuwsblad x3
            Kuurne x2
            Paris-Nice x4
            L’Eroica x2
            Tirreno-Adriatico x4
            Milan-San Remo x5
            Dwars door Vlaanderen x2
            E3-Harelebeke x3
            Criterium International x3
            Brabantse Pijl x2
            3-Days of Depanne x3
            Ronde van Vlaanderen x5
            Ghent-Wevelgem x4
            Paris-Roubaix x5
            Grote Scheldeprijis x2
            Pais Vasco x4
            Amstel x4
            Fleche Wallone x4
            Liege-Bastogne-Liege x5

And the points:
            1st through 10th Place: 40 – 25 – 21 – 7 – 6 – 5 – 4 – 3 – 2 – 1 points
            Stage Wins: 5 points (split stages in DePanne and Criterium International share points)

Some notes on my weighting and point systems:
            1. I weighed each of the one-day races differently; I think most make sense. I wanted to make L’Eroica’s multiplier 3, but it suffered a bit this year without Cancellara and Ballan. As for Milan-San Remo, I’m really struggling giving this race the same weight as Flanders, Roubaix, and Liege. As I said earlier—and to the agreement of Competitive Cyclist—it’s just not as difficult as other Classics. In the end, it gets the 5 for its prestige, its history, and by virtue of the fact that so many riders base their entire late-February and early-March upon doing well there. And Amstel? Maybe I underrated it, but I feel that several riders look past it to Liege or plan to peak earlier for Flanders and Roubaix. Overall, I’m content with the fact that Amstel and MSR effectively average one another out.

            2. As for the stage races, there’s not too much to say. Some might argue that Pais Vasco is much harder than Tirreno and Paris-Nice. Others might argue that the Criterium International is weighted too high. I welcome your feedback. Is there a better way?

            3. To try and allow for a point of comparison, I used Competitive Cyclist’s point system. Changing the points would have made such a comparison difficult if not impossible.

            4. I think 5 points is a fair method to recognize stage wins without creating an imbalance. Ignoring them is simply too easy. Does the winner on Montelupone really deserve nothing for his exploits? Does Pozatto deserve no credit for the 1st stage of DePanne?

So let’s get on with it!

Here’s how things shaped-up over at Competitive Cyclist:

            Quick Step 659
            Cervelo Test Team 574
            Saxo Bank 558
            Katusha 528
            Columbia 394
            Astana 360
            Silence-Lotto 276
            Caisse d'Epargne 267
            Rabobank 185
            Diquigiovanni 175
            Garmin-Slpistream 125
            Liquigas 112
            Euskatel 100
            Lampre 86
            Acqua Sapone 85
            LPR 70
            Francaise des Jeux 64
            Milram 51
            Skil-Shimano 34
            An Post 25
            Vacansoleil 22
            Ceramica Flaminia 15
            Ag2R 11
            BMC 7
            BBox 6
            Elk Haus 5
            Landbouwkredit 5
            Andalucia 3
            Topsport Vlaanderen 1
            Verandas Willems 1

Here’s how they turned-out for me:

            Quick Step 946
            Cervelo Test Team 838
            Saxo Bank 822.5
            Columbia 789
            Katusha 610
            Diquigiovanni 450
            Rabobank 422
            Astana 360
            Caisse d'Epargne 340
            Liquigas 334
            Silence-Lotto 319
            Euskatel 179
            Milram 165
            Lampre 148
            Francaise des Jeux 139
            Garmin-Slipstream 136.5
            LPR 115
            Acqua Sapone 110
            Agritubel 85
            Skil-Shimano 80
            Ag2R 71
            An Post 50
            Vacansoleil 44
            Landbouwkredit 20
            BBox 19
            Ceramica Flaminia 15
            BMC 14 (tie)
            Cofidis 14 (tie)
            Elk Haus 10 (tie)
            ISD 10 (tie)
            Andalucia 6
            Besson-Sojasun 5
            CSF Group-Navigare 3
            Topsport Vlaanderen 2
            Verandas Willems 2
            Fuji-Servetto 0

Some things I noticed in my analysis:
            1. Quick Step romped to the win in both my and the Competitive Cyclist analysis. The most interesting thing though is that they did it without scoring ANY points after Paris-Roubaix. Truly, this is a team built for cobbles.

            2. In my system, Silence-Lotto suffered the most and Diquigiovanni benefitted the most (almost entirely due to Scarponi and Rebellin).

            3. Interestingly, Astana scored the same number of points in both systems. Contador’s stage wins obviously compensated for the lack of Levi’s ToC.

            4. Lampre scored zero points prior to Pais Vasco. Clearly Ballan’s absence hurt them. And while the same can’t be said for Saxo Bank as a team, one can’t help but wonder what effect an on-form Cancellara would have had on the final tally—of several teams.

            5. Whatever happened to the French? FDJ aside, they were largely absent—as teams—from the majority of races. Those that did score points did so largely with riders from other countries. Individuals that did well did it while riding for foreign teams. It’s a shame we no longer have a French contender for Roubaix riding for a French team. Maybe someone will soon offer the big Euro’s to Chavanel to come home?

            6. Should BMC feel better or Cofidis worse?

            7. And finally, poor, poor Fuji-Servetto. Are they getting a raw deal?

All in all, this was a terrific exercise; it truly lent a new perspective to the races and the spring season as a whole. I did it all freehand—my Excel skills are negligible—which I think almost made it more enjoyable. However, I would welcome anyone with the time and talent to create a spreadsheet, particularly so that we can use it again next year.

In fact, I enjoyed it so much that I’m hoping to post a ranking of individual riders and maybe even countries—just to prolong the fun and opportunities for debate.

Finally, I would be remiss if I did not extend one final salute to the folks at Competitive Cyclist. Not only are they a quality retailer, but they also offer some of the web’s greatest journalism and food for thought. Thanks!

Now, let the comments begin!