Does Flat Speed Translate to Uphill Speed?

One of the most interesting aspects of uphill specific run training is the amount of variability in training methods in order to reach the same goal. I have a real passion for identifying the "optimal" uphill running training philosophy (the question of whether an "optimal" training style exists in any sport is debatable, but that doesn't mean it's not worth trying to identify). For example, some international trail runners may exceed 25,000-30,000 feet of elevation gain per week, while others will rarely exceed more than 10,000 feet. In addition, some may opt for full specificity in their training and complete all endurance training hours through running, while others may choose to include many hours of cross-training (through cycling or different types of skiing) to supplement their running volume. However, what I'm most interested in is the translation of flat running speed into uphill running speed. The extent of this translation could help aid training decision making for those aiming to optimise their uphill running performance.

Vertical Nasego, image by WMRA.

In order to gauge the extent of this translation, I looked at numerous high-level uphill races (either international races or trial races to gain selection for international honours). The races selected are the 2023 World Trail and Mountain Running Championship (vertical race) , the 2024 British senior Home Countries International event (at Glas Tulaichean, Glenshee) and the Skiddaw uphill race (which acted as a trial race to represent Great Britain at the European Championships in 2024). I looked at every male participant in these races and calculated their race pace in metres covered per minute. I then searched for each athlete's road times (via Power of 10, World Athletics or Strava) and noted any road races completed within a year of the uphill race. Naturally, some runners completed 5 kilometre races, while others ran races in excess of 20 kilometres, which made it difficult to compare performances. In order to compare performances of varying distances more accurately, I gave every runner a VDOT score based on their fastest road time within a year of the uphill race. For now, a scatter graph with metres covered per minute on the Y axis and VDOT score on the X axis seems to be the simplest and most accurate way of measuring the extent of which flat running speed translates to uphill running performance. The sample sizes used within each graph are not substantial (between 13 and 22 participants for each race), but I feel they're large enough to draw judgements from. The smaller sample sizes are due to some athletes' road times not being available online, or some athletes simply not running a road race within a year of the uphill race.

Results

Here's the first scatter plot, analysing the VDOT score and metres per minute covered during the vertical race at the 2023 World Championships in Innsbruck. There's a clear correlation between VDOT score (which is used to quantify road speed) and pace at the vertical race, but the correlation isn't extreme. In addition, there are numerous cases where runners with a lower VDOT score covered more metres per minute at the vertical race. We can see that the fastest runner at the vertical race (in metres covered per minute) was far rom the fastest flat runner in the group.

Extremely similar data is seen when analysing the numbers from the Skiddaw Uphill race, which acted as a trial race for the 2024 European Off-Road running Championships in Annecy. Although this wasn't an official international race, the standard was identical to that of an international event and featured many athletes who've represented their country in mountain running. Here, the correlation is clear, but similarly to the World Championships vertical race, there are occasions where slower flat runners massively outperform faster flat runners. Again, the fastest uphill runner was far from the fastest flat runner (in this chart alone, there are nine athletes with a larger VDOT score than the fastest uphill runner in this event). In this race, a runner with a VDOT score of 70 manages to cover 36 more metres per minute than a faster flat runner (who has a VDOT score of just over 72).

The data from the Senior Home Countries International (a race featuring English, Welsh, Scottish and Northern Irish international teams) tells a slightly different story to the World Championships vertical race. The fastest flat runners generally perform better than slower flat runners in uphill only events. However, this correlation is slightly more extreme than the previous two, which is likely down to the less steep gradient of the race. For reference, the average gradient of both the World Championships uphill race and the Skiddaw Uphill Trial was just over 14%, while the gradient of the Senior Home Countries event was 9.3%. Naturally, the less steep the event, the more flat specific attributes will be advantageous. This explains why the correlation between flat speed and uphill speed is stronger in less steep races. In this chart, the fastest uphill runner is the second fastest flat runner, which is a pretty substantial contrast to the previous charts where the fastest uphill runner may be far from the fastest flat runner.

Now it's pretty clear that flat speed correlates with uphill speed on a macro-scale (in terms of the top 10% of a vertical race will almost certainly have more fat speed than the top 30-40%), the impact of flat speed at the absolute world-class level differs slightly. To measure this, I simply zoomed-in on the top 5/10 finishers at each race, to see if the correlation was still clear.

Albeit an extremely small sample size, there's either an extremely small or zero correlation between flat speed and uphill speed at the absolute highest level. In all four races selected, the fastest uphill runner was never the fastest flat runner, showcasing the importance of uphill specific qualities even at the highest level of competition. Perhaps the outstanding uphill specific quality is muscular-mechanical related. This could be muscular endurance (although this is often associated with eccentric muscle contractions that are most useful in steep/fast downhill running) or more related to force production abilities when elastic energy from (mainly) the achilles tendon is compromised or non-existent, as is often the case in very steep uphill races.

As expected, the whole data suggests the importance of flat speed qualities in uphill running performance. The clear correlations between flat and uphill speed on the macro-scale suggest that in order to even be competitive at uphill races, flat speed is imperative. However, it seems the importance of flat speed attributes decreases as the standard of competition increases. Additionally, the importance of flat speed decreases with the increase of average gradient, which explains the stronger correlation at Glas Tulaichean compared to a weaker one at the World Championships in 2023. There seems to be a "threshold" of flat speed that the vertical athlete must surpass, but once that threshold has been surpassed, the importance of flat speed is replaced by either muscular endurance, non-elastic force production or both.