UCI time trials are a class race contested in the tiny fractions of a percent, tiny differences in equipment become very important. Riders practice in wind tunnels to tweak everything from clothing to pedalling style, because a shift that gives an overall gain of 0.01% in speed can result in a win.
Considering more than just the top gear, the jump from 50/11 - 50/12 (the two highest gears) and 54/12 - 54/13 is noticeably bigger (9% vs 8%). That may well explain the change, even if it results is slightly lower peak speed, since it means that on slight slopes the rider has better gear options available. Note that on downhill sections, the 54/12 rider has the 54/11 ratio available, where the 50/11 rider has no choice but to spin faster.
The difference between gearing 50/11 and 54/12 is about 1% (4.545454... vs 4.5). The larger chainring is bigger and heavier so it has more aero drag but lower mechanical losses (especially the loss difference between 11T and 12T is significant... but that loss should be less than 2% of the total power output). I suspect the human power output difference between those two at a given speed is less important than the comfort factor of being able to choose a gear that feels right.
It does not even have to translate directly into better average or total power output from the rider, as long as it gives a better time. Also possible is that it makes the rider feel faster, and that directly affects performance.
Gear ratio selection
When riding, people care about pedal rpm, which is determined by gearing. Humans have a power/speed curve. The further from their optimum cadence they are, the less efficient they are (and efficiency is what matters in a time trial). The closer the gear ratios are, the closer to the right gear they can get, so the closer to their peak efficiency point they can stay.
These tables show what's likely to be the next 4 cogs lower than the original question on the two cassettes we're discussing. The question I'm looking at here is "Going up a slight rise, what gear options are available?"
54/12 top gear
cassette cogs |
ratio to next gear |
Development (m) |
12 |
|
9.05 |
13 |
92.31% |
8.35 |
14 |
92.86% |
7.76 |
15 |
93.33% |
7.24 |
16 |
93.75% |
6.79 |
Average |
93.06% |
|
50/11 top gear
cassette cogs |
ratio to next gear |
Development (m) |
11 |
|
9.14 |
12 |
91.67% |
8.38 |
13 |
92.31% |
7.73 |
14 |
92.86% |
7.18 |
15 |
93.33% |
6.70 |
Average |
92.54% |
|
The immediate answer is that the 54T rider can make a slightly smaller shift. Ratios let us ignore the actual gear and focus on the size of the shift. The 54T rider shifts to a gear 92.3% of their current one, the 50T rider drops to 91.67%. And that happens every time—on average the next lower gear is 93% for the 54T rider and 92.54% for the 50T.
(It's possible that the 54/12 rider will keep the 11T small cog so they have a downhill gear, but we can ignore that for this comparison because in that case the 50/11 rider is out of options.)
That sounds really minor, but remember that those riders are fighting it out in the fractions of a percentage point.