The Outer Line: Charting the best (and worst) WorldTour men's teams

The brief pause that is pro cycling’s off-season is the best time to crunch the men’s statistical rankings and data to examine key competitive trends of the past few years. As we discussed in an earlier piece, this exercise can reveal unconventional insights into the sport; more broadly evaluated over a period of time, it may also suggest important lessons regarding the drivers of success and failure in managing pro teams. Below, we dig into all the numbers to look at who has improved the most, who has fallen off, and what can we predict for the future.

WorldTour Victories

First, let’s look back at the figures for the number of WorldTour wins by each team for the last five seasons, where some interesting trends immediately begin to emerge.

The spaghetti-like Chart 1 (above) of total team wins over the past five years is cluttered and confusing, but we show it to demonstrate that the fortunes of many teams bounce up and down significantly on a year-to-year basis. Perhaps the most dramatic single takeaway from this chart is Deceuninck–Quick Step’s sustained dominance – that single red line standing above the rough competitive convergence of all the other men’s pro teams. The team’s sheer number of wins is very impressive, and as discussed in more detail elsewhere, seems to be a result of their “swarming” style; the team always seem to have at least one rider capable of winning, no matter the genre, profile or length of the individual race.

To parse this information more carefully, we qualitatively filtered out the results of those teams which have shown the most improvement or regression over the last five years; this makes the aggregate results shown in Chart 1 more understandable. Some tentative conclusions and hypotheses begin to emerge as a result.

Chart 2 (above) shows two teams that demonstrate the most clearly upwards-trending performance over this time period – Jumbo-Visma, and the relatively newer Bora-Hansgrohe team. While Jumbo-Visma is a longstanding WorldTour organization, it has gone through a number of different sponsors and nominal racing names, and has reinvented itself over the past few years. Bora snagged 33 wins in its inaugural WorldTour season and continues to improve, not just with superstar Peter Sagan, but also with the notable – if under-publicized – sprinters Pascal Ackermann and Sam Bennett.

Chart 3 highlights several teams that have demonstrated the most steady or “flattest” performance during the past five years. This is a slightly deceptive way to view overall results, as it includes both high and low-performing teams. For example, we once again see team Deceuninck—Quick-Step (DQS) standing relatively steady above everyone else, garnering 60 or 70 victories, year in and year out. We also see teams like Astana, Mitchelton-Scott, Movistar and AG2R-La Mondiale achieving consistent results in the middle of the peloton – generating a predictable 25 or 35 wins a year. Then there are teams like EF Education First and AG2R, which also demonstrate fairly steady performance at a lower level – good for ten to fifteen wins a year. None of these teams really significant up or down trends over the time period under discussion here.

But in Chart 4, look at the several teams that have experienced a clear-cut decline: Dimension-Data, CCC, and Katusha-Alpecin all experienced dramatic decreases over this same period, with Sunweb also in the mix. As shown, three of these teams had over 25 wins in the 2016 season while all netted fewer than 10 just three years later.

Of particular note are the dramatic slumps of CCC (formerly BMC) and Katusha. CCC went from 48 wins just two years ago to a mere six this past year; Katusha’s collapse is almost as dramatic. And even more surprising is the fact that, according to a L’Equipe report in 2016, these two teams had the second and third largest budgets of any teams in the WorldTour. On the other hand, Dimension Data and Sunweb, though lacking large budgets, were seen as scrappy and innovative upstarts which used highly-researched and nuanced personnel acquisition and training methods to build an edge.

But behind the big budgets or the supposedly innovative systems employed by these declining teams, three out of the four basically employed a strategy of relying heavily on just a few elite riders to deliver results. The BMC program had a large stable of talented riders capable of getting results, but Katusha, Dimension-Data and Sunweb all relied on a small number of star riders; more on this later.

Finally, it is important to note that not all teams fall neatly within any of the three categories summarized in this brief analysis. More insights are indicated by all the other teams that fall between the cracks. For example, Team Ineos saw its win total drop by almost half between 2018 and 2019 – 44 wins in 2018 to 26 wins in 2019. But that significant drop in wins was not reflected in top-10 or podium finishes (208 top tens in 2018 to 218 top tens in 2019, and 98 podiums in 2018 to 94 in 2019). Ineos has trended down in all three categories since 2015, but not at quite a steep or sustained enough rate to fall into our declining category. (We discuss Top 10 statistics in more detail below.)

Team Diversity

Another insightful way to dissect the wins data is to ask what percentage of a team’s victories come from either a single rider or a small handful of key riders. Data representing the percentage of victories attributable to a single rider are shown above in Chart 5. There are several interesting takeaways here. First, note Team Deceuninck-Quick Step on the far left side; although its number of total wins far outstrips any of the other teams, its reliance on a single rider is amongst the lowest in the peloton. In other words, Team DQS is one of the most diversified in the peloton – structured with a larger number of riders who, as mentioned, can win races around a philosophy of “win everything.” It is not dependent on a single rider or two to churn out the wins.

Jumbo-Visma, has relied on sprinter Dylan Groenewegen for a steady stream of wins, but he only accounted for 29 percent of the team’s wins in the 2019 season. Bora-Hansgrohe, home to three-time World Champion Peter Sagan, has also used a win-by-committee strategy to become a dominant force. While Sagan drove the team’s performance in their first WorldTour year of 2017, in the past two seasons it has been either Pascal Ackerman and Sam Bennett who have led the team in wins. More impressively, no one rider accounted for more than 30 percent of the team’s victories. Most impressively, the team won sprint stages in all three grand tours this year with three different riders.

On the other hand, UAE Team-Emirates, Lotto-Soudal, AG2R or CCC are teams that rely heavily on a single rider for team victories. Alexander Kristoff delivered 42 percent of UAE’s wins in 2018, although the team spread the wealth around a bit better this year, with newcomer Tadej Pogačar accounting for 28 percent of the team’s first-place finishes. But for some of the more unidimensional teams, the story is different. Lotto-Soudal relied on Caleb Ewan for 43 percent of its victories and the relatively unknown Benoît Cosnefroy accounted for 36 percent of AG2R’s wins.

In 2015 the BMC team won 33 races, with Rohan Dennis taking 18 percent of those events. But by 2019 this had devolved to a total of just six wins, with Greg van Avermaet responsible for a whopping 50 percent of them. In Katusha’s 2015 heyday, they relied on Alexander Kristoff for 50 percent of its wins – and this share actually increased in subsequent years, with the Norwegian taking 52 percent of their total wins in 2016 and 53 percent in 2017.

Mark Cavendish accounted for 31 percent of Dimension Data’s first WorldTour season victories in 2016, and Edvald Boasson Hagen accounted for 40 percent the following season. The departure and/or decline of these key riders has been devastating for their teams and seems to demonstrate the strategic risks of relying too heavily on a single star rider – particularly if and when those older riders start to lose top-end fitness to age and injuries.

These data points suggest that there is a correlation between a team’s dependence on a single rider and their general future prospects. In general, if a team’s win share for a single rider is above 40 percent, a warning light should be coming on, and if it reaches or surpasses 50 percent, alarm bells should definitely be ringing. Dimension Data had 40 percent of their wins coming from a single rider in 2017 before seeing its wins and top tens fall off the cliff in 2018 and 2019. Katusha-Alpecin was counting on Alexander Kristoff for more than 50 percent of their wins in 2017 before seeing a dramatic performance drop-off in the past two seasons. In short, a rising win share by a single rider is a worrisome sign of a team lacing diversity – and one that may be about to experience a steep decline.

On the other hand, a low individual rider win share figure isn’t necessarily a good thing either. EF-Education First hasn’t had a single rider claim more than 33 percent of its win share since 2015 and no Katusha rider won more than one race in 2019. Being less dependent on a single racer doesn’t necessarily mean that a team is more successful and healthy from a competitive perspective – it may simply mean that they don’t have many riders who can win races. In other words, the flip side of being too dependent on an individual rider seems obvious: every team needs at least a couple of strong performers or it simply won’t ever going to win any races.

Sunweb provides an interesting perspective here and suggests the importance of other techniques in evaluating these data. The team has traditionally been home to a strong and balanced roster, but the injury to Tom Dumoulin in the 2019 Giro d’Italia severely hurt the team’s ability to net victories. In addition, the team’s key sprinter, Michael Matthews, while often present at the front end of the race, has not been able to convert this into wins very frequently. The example of Matthews suggests that counting just race victories is not necessarily the best or certainly the only appropriate measure of overall team success.

Top 10 Finishes

First and second place is often determined by less than the width of a tire; first place may be separated from tenth place by just a second or two. So, it makes sense that we should extend our analysis of the men’s WorldTour teams, and also look at the statistics for top-10 finishes by team. This approach paints a somewhat different picture.

Chart 6 is another incomprehensible mess of spaghetti, but here we are looking at the number of top-10 finishes for each team rather than outright victories. The main thing to notice is that, in comparison to Chart 1’s total wins, the amplitude of the trend-lines here is lower – they are slightly flatter or less “wavy.” This is a good indication that total top 10 finishes might be a steadier or more predictable way of assessing overall team performance. In other words, if we isolate the top 10 finishes, team performances don’t jump all over the place year to year quite so much. And other differences emerge when we drill down further into the team trend-lines.

Chart 7 shows three teams (instead of two) that arguably increased their top 10 finishes most significantly over this period. Viewed from this perspective, Bora-Hansgrohe and Jumbo-Visma once again stand out in terms of growth and improvement. But now, UAE Team-Emirates also arguably joins the mix, with increasing top ten finishes since 2015. According to the 2016 L’Equipe piece, all of these teams had budgets that were in the mid-to-low range compared to the rest of the WorldTour, but each has improved notably since then.

Chart 8 shows the stable or “flat” grouping of teams for the top 10 category. Once again DQS dominates, while teams like AG2R and EF continue to perform steadily at a lower level. A notable change in overall performance is represented by Sunweb – which appears comfortably in the “stable” category here, instead of the declining category, as mentioned earlier in the wins category. One major reason, hinted at earlier, is that while Matthews doesn’t win that many races, he has an uncanny ability to be in the finishing mix in a lot of races – allowing the team to rank more highly on this basis, as compared to victories alone.

Chart 9 shows the primary declining teams from the perspective of top-10 finishes, and mirrors the earlier conclusions. Not only are teams Dimension Data, CCC and Katusha declining in terms of victories, they are also declining in terms of more general performance.

Points Rankings

Top 10 finishes may be a more nuanced and inclusive way to view team performance as compared to outright victories. But we can extend this observation even further by looking at the overall racing points generated by each team, which should – in theory – better summarize the overall performance ranking of the teams.

(In order to do this, we took a look at both the UCI end-of-season team rankings, the actual UCI points rankings for each year, and the ProCyclingStats [PCS] points rankings for each year. We utilize the PCS data here, rather than the UCI data, primarily because (a) the UCI made major changes to its points system in 2017, making comparison over the time period difficult, and (b) the PCS rankings are widely accepted and recognized.)

The initial all-team data rendition is (once again) an indecipherable spaghetti chart, so we will jump directly to where this broader approach yields different or interesting results. First, when the teams’ relative UCI year-end WorldTour rankings are mapped over time, we see the same general trends, and generally the same teams in each of the positive, flat or negative trends. Second, in looking at the list of improving or steady teams from the perspective of PCS points, the same general conclusions are reached – Jumbo-Visma, Bora-Hansgrohe and UAE-Team Emirates again stand out in comparison to the rest of the peloton over the time period. This approach confirms the same set of “steady” performing teams we discussed earlier.

However, in terms of the declining teams, a couple of new and slightly different conclusions can be teased out. As shown in Chart 10 above, we again see the decline of Dimension-Data, and the relative collapse of teams CCC and Katusha. But when Team Sunweb is plotted on this chart, the decline suggested earlier is lessened. Sunweb’s 2017 spike was due largely to Dumoulin’s overall victory in the Giro – although he only won one stage – and the fact that Sagan was thrown out of the Tour, opening the door for Matthews to collect a number of points as well as the green jersey. Both factors elevate the team’s overall placement.

There are several important lessons here. The UCI or PCS points designated to the winner of a grand tour are obviously very substantial, but that same winning rider’s performance might net few, or maybe not even any individual stage wins. It is these kinds of subtleties, in terms of the metrics used, that can lead to variations in perceived performance. Although typically rather minor, these differences also point out the importance of evaluating and measuring success from more than just one point of view.

Predictions

So, based on the foregoing analysis, what can be predicted about the future performance of the various teams?

First, CCC, Dimension Data and Lotto-Soudal each relied on a single rider for over 43 percent of their wins in 2019, which signals they may be in danger of suffering a performance drop-off in 2020 or beyond. CCC still has an aging Van Avermaet as its marquee rider, and he should be assisted by the arrival of Matteo Trentin – a sturdy rider reaching the end of his prime, but who is still capable of racking up podium finishes across a wide range of courses.

Lotto-Soudal leaned heavily on Caleb Ewan in 2019, but with sprint wins getting harder and harder to come by in this golden age of fast finishers, it is also at risk of seeing a performance drop-off. They have added Philippe Gilbert and John Degenkolb, presumably to hedge against this risk, but these are exactly the type of aging, high-priced stars that, as we noted earlier, have been suggested as all-too-frequent reasons for a team’s decline.

Dimension Data’s performance can’t get much weaker, and unfortunately it got little production from the talented Dane Michael Valgren in 2019. A return to form for him in 2020 could see the team get a good performance bump. Also, the team has more financial flexibility in its post-Cavendish era, and seem to have taken advantage of this by signing Max Walscheid and Victor Campenaerts in the offseason.

The rebranded Bahrain-McLaren team recently brought the high-profile British performance company on board as a co-title sponsor, but its roster construction should have alarm bells ringing. The team has built its 2020 roster around the expensive and recently under-performing Mark Cavendish and the perennially inconsistent Mikel Landa. While the squad still has consistent winners like Sonny Colbrelli and an up-and-coming star with Matej Mohorič, the Middle Eastern team seems to be employing a risky strategy by putting so many both financial and emotional resources into Cavendish and Landa.

Both of these riders have drawn a lot of attention but have also seemingly caused disruption at their respective teams in recent years, while delivering few tangible results. This can create an unstable team atmosphere and could potentially distract the rest of the team, in terms of delivering results in 2020.

Other teams to keep an eye on for potential underperformance in 2020 are Ineos and Movistar. Both have a lot invested in aging stars (Thomas, Froome, and Alejandro Valverde). While Ineos has a fantastic roster of young talent, one risk is that major opportunities may be offered up first to the aging core rather than emerging younger riders.

Expect DQS to continue their dominance. The team has become very skilled at parting with aging stars just as their performance is falling off, and shifting their resources to younger, cheaper and faster riders. And Jumbo-Visma and Bora-Hansgrohe will be well-served by their bona fide superstars; each team has a deep supporting cast of riders who are individually capable of winning races.

Summary

After wading through all of these different metrics and evaluation approaches, there are at least a few general conclusions that can be drawn. Perhaps the clearest lesson involves the over-reliance on a single star leader – i.e., the old adage about not putting all your eggs in one basket.

While it might be tempting to sign a big name, the experiences of the (now-disbanded) Katusha and Dimension Data (soon to be NTT) suggest otherwise. It’s great when your star is winning, but if he falters, you have nothing to fall back on. And everyone eventually falters. It’s exactly analogous to relying too heavily on a single big customer in any other business. Both of these teams went all-in with aging fast finishers and enjoyed some initial success, but it was short-lived. As those individual riders aged or moved on, the lack of diversity or depth on the team was quickly exposed. If the number of victories attributable to one rider is approaching 50 percent, you likely have a problem. On the other hand, if no single rider has more than 25 percent of the team’s win and your team has a relatively high win total, you are more likely to have a well-balanced and diversified team.

By contrast, teams like Ineos, Jumbo-Visma and Bora-Hansgrohe, while employing a number talented riders, have built strong organizations and deep rosters without necessarily focusing on a single star. They have instead largely relied for their success on the ability to identify and sign lesser-known riders, who have gradually turned into bigger stars under their tutelage. That kind of diversity, and spreading around of the talent, allows those teams to flourish even if one or two of their top riders hit a rough patch or decide to move on.

A second clear finding is that while the budget is clearly a critical factor, it doesn’t appear to be the definitive explanation for the success or failure of a team. Team Ineos’s WorldTour-leading budget didn’t buy them that many wins or even a particularly high WorldTour ranking (6th) in 2019. In fact, the team’s wins and top-10 finishes have actually decreased as its budget has increased, while Deceuninck-Quick Step continued to dominate all three categories with (according to limited public information) a significantly smaller payroll.

A certain threshold budget is obviously critical to achieve success at the WorldTour level, but it appears to be more table stakes than deciding factor. What is more important is access to fertile recruiting grounds and a strong organizational structure and process. (Note: It would clearly be very helpful to have better team budget data for this sort of analysis. The UCI has indicated to The Outer Line that there is on-going discussion about making team budget information publicly available, but the topic is still under consideration at this time.)

Building a successful cycling team obviously requires a number of key inputs and factors. This analysis suggests that there are three key and intertwining drivers which must somehow combine in the right manner in order for a team to really shine on the road. These include the purely quantitative size of the team’s budget; the related and less quantifiable but critical significance of the quality of the team’s riders, and: third, the least quantitative or imprecise factor of all – the team philosophy, operating structure and management style. Obviously, all three of these factors interact and are overlapping with each other, but, just as clearly, it is differences in these three attributes which separate the winners from the losers – and the teams which are improving versus those that are declining.