The Chariot or the Horse: An Analysis of VAM vs. Price

It’s an age old question in the cycling world: how much of a rider’s performance is a result of the strength and fitness of the rider, and how much derives from the quality of the bike? Are there any relative bargains to be had when choosing a good climbing bike? We’ve gathered a lot of data in Strava which could help answer these questions.

I began my research by compiling all of the KOMs (and QOMs) in the entire Strava dataset from Category 3 or harder climbs. This rendered a list of  717 climbs, along with the bikes listed for these rides. Grouping the bikes by manufacturer, I selected the most common 5 bikes chosen by Strava riders for further investigation. There were 2 questions I aimed to answer: How fast, on average, does each manufacturer’s bike go uphill? And, what’s the average cost of each bike?

To answer the first of these questions, I employed a method of comparing climbs called VAM. VAM was designed by Michele Ferrari to help compare and prepare elite cyclists; in Italian VAM is short for Velocita Ascensionale Media, in English, we call it the average climbing speed. VAM boils down to a simple equation: how many meters in elevation are gained per hour while climbing ((delta elevation(m) / delta seconds) * 3600). VAM works well when comparing climbing efforts because the vast majority of your power goes to fighting gravity, which VAM accounts for. When you’re not climbing (for example, in a draft line on a flat road), much less of your energy goes to fighting gravity and factors like road conditions and wind resistance have more influence. Also, because VAM tends to be more accurate on longer, sustained climbs, only Category 3 and harder climbs are used in factoring the results of the equation. To put VAM into perspective, the best climbers in history, Pantani, Armstrong, etc., achieve 1700+ numbers when winning an important mountain stage.

The experiment that we wish we could run would be to take the same group of riders and put them on different bikes, measuring VAM on the same climb with each bike. This would really get at the question of the impact the bike has on performance. Instead, we can pool our data across riders on different bikes, assuming that the average rider on each bike is of approximately equal ability. We recognize this is a big assumption. If bike choice among the top 5 most popular bikes is biased by biker ability, then this analysis is only going to reveal that bias and not the impact the bike has on performance.

The following chart shows the average VAM and price for each of the top 5 bike manufacturers, for KOMs on Category 3 climbs or harder, in the Strava dataset.

I’m assuming several things with this analysis — first, that the average frame by each manufacturer ridden by Strava athletes is the mid-level frame in each manufacturers’ series of racing level frames. I’m also assuming that, with 717 Category 3 or harder climbs, the average climb ridden by each frame manufacturer is on average the same length and grade. These assumptions may be wrong, and warrant further research.

Now let’s take a look at the bikes:

Cannondale Six 13	Cannondale had an impressive showing with a 1076 Average KOM VAM. A popular choice for the Cannondale line is the Six13 (shown left). It has a carbon/aluminum frame with a fully-built weight in the 16.5 lb. range (7.5 kilos), that makes it a clear top choice for climbers. The Cannondale was used by Team Barloworld in 2006, and assisted Ivan Basso to achieve numerous Mountain Stage wins in several tours. This line of bikes has been replaced by a new line of 6′s — the Carbon 3, 5, 6 and Féminine 3 and 4.
Scott Addict SL	The Scott line of bikes — specifically the Addict — is another popular choice among climbers. Mark Cavendish has shown the world that Scott bikes can be made to go fast as well. With a ridiculously light 790 gram frame, the Addict SL can be built way under the legal UCI limit (I’ve heard rumors of 13 to 14 lb. builds of this bike when set up as a pure climbing machine).
Trek Medone	Trek bikes — specifically the Medone — have been made a popular choice for all types of cyclists by Lance Armstrong, who rode Treks to all his Tour de France wins. Trek has been pushing the envelope by committing a large budget to R&D for years. There are 15 choices in the Medone line alone — its carbon frame can be grouped with components to make it either light or super-light.
Specialized Tarmac	With a monocoque carbon frame and fork, the Specialized Tarmac line is also a great choice for climbers — made evident by watching Andy Schleck pilot one of these to within a whisker of winning this year’s Tour de France. This is a bike that should be on the short list for any cyclist who has a bent on climbing.
Cervelo R3	Cervelo seems to be everywhere now. A popular choice is the R3 which can seemingly suit evey purpose. Cancellara has won with this bike on the cobbles, so it’s clearly rugged — yet it can also be built into a sub-15 lb. climbing machine.
Parlee Z3	One notable bike that isn’t in the top 5 is the Parlee Z3, which looks like a traditional ride. Built in Beverly, Massachusetts, this bike is often the choice of very discriminating cyclists. Nearly everyone who owns one raves about it. Unfortunately they don’t come cheaply.

Two other questions worth looking into are: who are the top climbers using Strava? And, what bikes are they riding?

Strava has some very impressive climbers. Here are the top 5 via VAM analysis:

1. Brent Mellen on Burke Mountain:  Bike:  Scott Addict SL, VAM: 1685 (watch this ride on YouTube here, here and here — and note the Strava kit!)
2. Rowan Dever on Noojee – Vesper:  Bike:  Specialized Tarmac SL3, VAM: 1492
3. Ben Silberfarb on Mount Ascutney:  Bike:  Parlee Z3, VAM: 1455
4. Jacob Berkman on Alba Road:  Bike:  Cervelo R3, VAM: 1424
5. Jamie McJunkin on Old La Honda:  Bike:  Cervelo R3, VAM: 1410

Brent holds the top 3 overall climbs on Strava for VAM analysis as well. Ben Silberfarb (number 3 on the list) currently owns the KOM on Old La Honda with a VAM score of 1419, which is nearly as high as his ascent up Mt. Ascutney. This climb is interesting in that he provides a conduit through which climbers from the West coast of the US can compare themselves to East coast climbers, and proves the value of VAM as a tool for comparison.

What questions have been answered by this look at bikes and their climbers?

My analysis of the Strava data shows that certain bikes tend to reach the top of climbs before others. These results may be attributed in some part to the qualities of the bike, but the results could just as easily be attributed to the fact that Strava’s strongest climbers gravitate towards certain bikes.

• Can an average climber be made great by a great bike? Probably not.
• Do bikes make a difference in overall climbing speed? Probably, but perhaps less than one might hope.
• Are there any relative bargains to be had? It appears that there are, though it’s up to you to decide what they are. The reasons for harmony between bike and rider can be complicated — price is only one ingredient.

The evidence of bike superiority is anecdotal and may very well change as more data rolls in. The data that I found most intriguing is:

• None of the top 5 climbers in the Strava database chose the bike with the highest average VAM, a Cannondale.
• 4 of the 5 top climbers in Strava chose one of the top 5 average VAM machines (with the Cervelo R3 earning the choice of 2 of Strava’s top 5 climbers).

Ultimately my initial research does seem to indicate that the strongest climbers tend to get to the top of climbs quicker than most regardless of the bike they ride. So, I’d say, “Chalk one up for the horse!”  However, I can’t ignore the data. There does seem to be the possibility of increasing one’s VAM through the right bike choice — perhaps enough to gain valuable seconds on rivals.