Earlier this year I bought some new wheels for my commuting bike. They certainly felt like they made it faster, but did they? Since I record all my rides on Strava and go the same route between Witney and Oxford everyday I have a good dataset. In theory, changing the wheels was the major difference between 2015 and 2016 so if I compared the same period (June to Sept), it should give me a pretty good idea of how much of a difference the new Shimano Ultegra 6800 wheelset made.

First, let’s look at just after Wolvercote Roundabout to just before Cassington, a distance of 2.6 miles. It is all cycle path with no interuptions, with just a few gentle lumps in the form of two bridges. As the wind typically blows from the south-west and strengthens during the day, I usually have a light tail-wind in the morning and a stronger head-wind on the way home. In 2015 the out and return legs took took me, on average (excluding the fastest and slowest 10% of rides to remove outliers), 10 min 17 sec and 9 min 24 sec, respectively. This was a surprise, since it meant I was nearly a minute faster on the way home, despite tending to have a headwind. Perhaps I put more effort in the afternoon compared to the morning. Note that the Python code detected between 39 and 86 rides for each year/direction combination, so that is a good number of rides.

How did the wheels do? Well, the averages for each leg decreased by 24 sec and 23 sec, respectively, a reduction of around 4%, which doesn’t sound like a lot, but over a 50 min commute is, on average, 2 minutes, which is noticeable.

Wolvercote to Cassington, afternoon, tends to be a headwind

Averages though can hide a lot, so let’s take a peek at the distributions. These more clearly show that on the way home, the rides are more tightly bunched and I cover the distance in between 8 and 10 mins, whereas on the way in there is much more variation in how long it takes me to cover that segment, with most days it taking me between 8 and 12 mins. Both cases clearly show, however, that I tended to be faster in 2016 with the new wheels, than 2015. But, as both distributions overlap, there were many days when I was faster in 2015 with the old wheels than I was with the new wheels in 2016.

Cassington to Wolvercote, morning, tends to be a tail wind

In conclusion, the new wheels did speed up my journey. Perhaps more important is something I can’t capture with data: the bike “feels” faster and is more fun to ride.

Interestingly, given my previous wheels were old, fairly heavy and came with my Dawes Galaxy about 9 years ago, shifting to Shimano Ultegra should be about as big an upgrade as one could reasonably make, yet the improvement is “only” 4%. Food for thought if you are doing a more minor wheel upgrade.