What it takes to finish top-five in men’s WorldTour races
Here's what eight years of data can teach us about the pointy end of the hardest men's races.
Here's what eight years of data can teach us about the pointy end of the hardest men's races.
If you’re the sort of rider who enjoys comparing yourself to the pros, there’s never been a better time to do so. The ubiquity of power data – and the willingness of many pros to post that data on Strava – means you can often see exactly what it takes to be competitive in a given bike race.
In this case we’re not talking about one specific event; we’re taking a broader view to answer the question: what does it take to achieve a top-five finish in men’s WorldTour events?
Led by Teun van Erp from Stellenbosch University (who also works as a sports scientist at Ineos Grenadiers), the researchers started with power output data from 33 professional male riders from between 2012 and 2019. All of those riders belonged to the same team, which raced at ProTeam (then called Pro Continental) level in 2012, and on the WorldTour from 2013 onwards.
While the researchers don’t say which team their data comes from, there can be only one candidate: the team currently known as DSM (previously Argos-Shimano, Giant-Alpecin and Sunweb.) Indeed, Van Erp worked with that team as an ‘embedded scientist’ from 2012 to 2020.
The researchers’ dataset includes all power output data from all men’s WorldTour mass start races (i.e. no time trials) where the rider in question finished in the top five on that race/stage (we’ll call them races from now on).
Every race was then categorised as one of four possible race types:
The researchers then applied some additional filtering to the data. For semi-mountainous uphill finish and mountain races, any finish inside the top 10 was included (not just top five) “if the power output data were collected from a general classification contender who was fighting for the general classification on the last climb of the day but a breakaway of a select number of riders (i.e. three to five) competed for the victory.”
On the flipside, some top-five data were excluded from riders who were part of the early breakaway because “the pacing profile of such races would be substantially different compared with achieving a result from the peloton, and would therefore increase the variability within the data set.”
In the end, that left researchers with:
(You might be wondering why the researchers considered top-fives rather than winning rides. Ultimately, it’s all down to data availability – by broadening the focus, the researchers had a much large sample space to work with. On that point, Van Erp told CyclingTips: “I would expect that the differences are not substantial between victory and top five”.)
For each of the 177 top-five results, the researchers took the relevant rider’s power data and assessed their maximal mean power (MMP) output for different durations during that race: 5,10,30 and 60 seconds, and 3, 5, 10, 20 and 60 minutes. These power figures were expressed in terms of both absolute power (in watts) and relative power (in watts/kg).
For added insight, those average power outputs were also expressed as a percentage of the rider’s season-best power for that duration, to see how close to peak fitness they had to get to achieve their top-five result.
The researchers also considered the timing of the maximal power outputs. Did they always occur towards the end of the race? Or were they more spread out throughout a given race?
From there they crunched the numbers to see what they could deduce.
Many of the researchers’ findings aren’t terribly surprising. Here’s a selection:
Interestingly, the researchers found maximal five-minute power outputs weren’t significantly different between race types.
So how fit did riders have to be in order to produce those top-five rides? And how far into the races were those efforts?
For flat races, riders needed to produce 80-82% of their season’s best powers between five and 15 seconds to go top five. As you’d expect, those efforts came in the final sprint. For semi-mountainous sprint races, the requirement was basically the same – 80-83% of a rider’s season’s best between five and 15 seconds.
Those percentages are perhaps lower than we’d expect. They probably show that raw power isn’t the only factor in a successful sprint – it’s often as much about positioning and timing as it is about the power numbers you hit. Van Erp agrees.
“In races, values are influenced by a lot of factors (tactics, riders) but especially fatigue,” he told CyclingTips. “I wrote a paper about this showing that top riders are less influenced by fatigue and only lose around 5% in hard races. However sprinters seem to be less resistant to fatigue and the top sprinters lose around 8%.
“Furthermore sprint tactics, like jumping from wheel to wheel or that riders have to hold back because of another rider, will lower the number as well. [Marcel] Kittel had one sprint in the TDF of only seven seconds because of a corner and drafting, so this will lower his 10-second best. “
For semi-mountainous uphill finish races, riders needed to be closer to the top of their game to finish top five: 86-90% of their season’s best for between 60 seconds and 10 minutes. Interestingly, those numbers weren’t always at the end of a stage – they occurred at 82-92% of race time, perhaps suggesting that the decisive moments come earlier than in sprints (e.g. on the lower slopes of a final climb).
For mountain stages, riders needed 91-93% of their season’s best for everything from five minutes to 20 minutes, with those numbers coming between 82-89% of the way through a race. This might also signify the lower slopes of a final climb.
It’s interesting that a higher percentage of a rider’s best is needed for mountain stages compared to flat stages. But it makes sense – doing well uphill is much more about raw (relative) power than is the case on flat stages. Positioning and timing, while not irrelevant, are far less important when it comes to excelling on hilly stages.
But why aren’t those percentages even higher? You’d expect most riders to have to do their maximum (or very close to it) to get a top-five placing in a WorldTour race with an uphill finish.
Again it comes down to fatigue and tactics. But also the length of the final climb. Very rarely is a rider’s effort on a final climb exactly five minutes, 10 minutes or 20 minutes. As Van Erp notes “a climb of 18 minutes will lower the 20-minute power,” given their power would be lower for the two minutes before the climb.
Alright, so those are some generalities. Let’s put some actual numbers to these top-five performances.
To go top five in your average, flat men’s WorldTour race, you need to be able to average:
It’s worth noting that these figures are lower than the five- and 10-second power requirements for a top sprinter at the Tour de France, as reported by Van Erp, Kittel, and colleagues in previous studies. At the Tour, you’re looking at 1,610 W (17.9 W/kg) and 1,515 W (17 W/kg) for those five- and 10-second efforts.
It’s not surprising that the requirements of the Tour are greater than those of your average WorldTour race. After all, there’s no bigger race than the Tour, and no race where sprinters are more in form.
As noted above though, these power figures are still quite a bit lower than what a sprinter can produce at their absolute maximum.
In what has become a seminal table showing the power outputs of cyclists of various abilities, created by renowned exercise physiologist Dr Andy Coggan and TrainingPeaks, world-class male sprinters are said to be capable of up to 19.4 W/kg for five seconds – quite a bit higher than the 17.9 W/kg found here.
Why the difference? Again, there are various in-race factors that can affect a rider’s effort relative to a pure sprint effort in training, say.
“The numbers are slightly lower because these are measured with fatigue and all the other components that influence PO [power output] in a sprint (tactics, holding still because of other riders, etc),” he said. “Secondly in figure 1 [see below] you can also see that some sprinters [have] even higher relative PO values [than] reported by Coggan. Thirdly I have no idea where the numbers came from which are reported by Coggan.”
Here are the requirements for a top-five finish in semi-mountainous flat-finish races:
These numbers are lower than for flat races, which isn’t surprising. Semi-mountainous flat-finish races have more climbing than flat races, which means higher power for longer durations to survive the climbs, and therefore increased fatigue for most riders by the finish.
For semi-mountainous uphill finish and mountain races, absolute power isn’t as useful a measure as it is for flatter stages, given the role of gravity and a rider’s mass in determining their climbing speed. Instead it’s relative power that’s more indicative.
For semi-mountain uphill finish races, Van Erp and co found that the 20-minute relative power required for a top five was 5.1 W/kg. For mountain races, that was higher again, at 5.7 W/kg. While still well beyond the reach of all but the very best riders, these numbers are also lower than we might expect. The Coggan/Training Peaks table suggests somewhere in the vicinity of 6.3 W/kg is possible for the world’s best over 20 minutes.
Again that’s the result of a couple factors: fatigue, tactics, and the length of those final efforts relative to set efforts of five, 10 or 20 minutes.
“Also here you see in the figure [above] that riders can reach up to 6.3 W/kg for 20 minutes in an ideal situation,” Van Erp added. “In training or fresh, these numbers can even be higher as values are reported around 6.5-6.6 for the very best.”
The researchers note that semi-mountainous uphill finish races have higher requirements for both relative and absolute power over 30 and 60 seconds, compared to mountain races. Why? Because semi-mountainous uphill finish races tend to have shorter climbs than pure mountain races, and generally speaking, shorter climbs can be raced harder than longer climbs.
These numbers provide an interesting snapshot of the strength of the best male riders in the world. But we need to view these results with a certain amount of caution.
It’s important to remember that these numbers are averages and there’s significant individual variation between races. “For example,” the researchers write, “it can be clearly seen that for some of the analyzed races, riders needed to reach close to their personal best for a specific duration to achieve the top-five result while at the same time, similar results were achieved at a lower percentage of their personal best.”
Also, all riders are different. Some riders are more aero on the bike than others, for example, which will change power requirements for a given effort. Other riders are better at positioning and conserving energy, which might affect how much effort they have to expend (or have available) for a given result.
And we need to remember that, even for a single rider, there are a multitude of factors that can affect their power output for a given performance. How hard a race is in the lead-up to the final kilometres is an important factor – if it’s been super hard, power numbers might be lower in the crucial moments that follow. And again, every race is different. Different tactics, race dynamics, and even weather conditions can change how a race unfolds, and therefore how much power, and for how long, a rider must produce in order to be competitive.
And that’s to say nothing of individual race profiles as well. In this study the races are divided into four categories, but races vary dramatically within those categories. For example: “while the results of this study clearly show the importance of relative power output of long duration in [mountain races], it can still be that for one [mountain race] 10-minute relative power output is of greater importance and for the other [mountain race] 60-minute relative power output was of greater importance (e.g. due to the difference in climb length).”
And the WorldTour classification isn’t uniform either. One-day races are very different to week-long tours, and week-long tours are entirely different to three-week Grand Tours. The cumulative fatigue from days or weeks of racing can be a real factor in longer races, affecting power outputs.
In short, there are plenty of caveats here. As such it’s probably best to view the numbers above as ballpark estimates of what it takes to be competitive at the highest level, rather than concrete figures. There’s a significant margin of error on either side of those averages, but they still provide valuable insight on just how good the best riders in the world are. And, if you’re looking to get more specific, there’s always the pros’ power files from individual races.