We talk a lot about Zwift bike frame and wheel choice here on Zwift Insider. Many Zwifters find it useful that our frame and wheel performance charts present very precise data about how each item in the Drop Shop performs relative to other items in game, helping us all make smart choices for races.
Note: some Zwifters spend lots of time obsessing over their bike choices, while others will tell you it really doesn’t matter. The truth is actually somewhere in between, as explained in “Does Bike Choice Matter On Zwift?”
One topic we haven’t unpacked much here on Zwift Insider is bike choice for sprinting. How much does bike choice matter at sprint speeds? And if it does matter, how should this impact the bikes we choose for races?
Let’s dig in, and ride smarter.
Sprint Test Results
We ran several popular frame+wheel setups through repeated steady-wattage solo sprints on Watopia’s classic Forward Sprint, then averaged the results. Our test bot held 1050W steady (14 w/kg) heading into and throughout the sprint. Results for each particular setup only varied by 2 hundredths of a second or less, meaning test results were precise, with very little variance.
Here are the results:
- Specialized Venge S-Works + DT Swiss Disc: 19.88s
- Tron: 19.95s
- Scott Addict RC + ENVE 7.8: 20.03s
- Canyon Aeroad 2021 + DT Swiss 62: 20.07s
- Specialized Aethos + Lightweights: 20.31s
So you can see that the difference between the two most popular setups for flat/rolling races (Specialized Venge and Tron) is 0.07s, while the difference between the fastest aero setup vs the fastest climber (Specialized Venge vs Specialized Aethos) was 0.43s.
Aero performance in our steady-state flat tests at 300W is an accurate predictor of sprint performance. That is, the most aero bikes at 300W are also the most aero bikes at 1050W.
We did some test sprints at lower power as well (600W, or 6 w/kg). Time gaps between each setup grew a bit due to slower overall speeds, with the Specialized Venge and Tron being 0.09 seconds apart with sprint times of 26.61 and 26.7 respectively. The math is fairly simple here: if sprint length increases by X%, the time gap between setups also increases by X%.
Sprint Time Gaps
Now that we know the basics of time gaps between bike setups, we need to know how much those time gaps matter. Or to put it another way: how much does bike choice factor into winning or losing crucial race sprints?
First, we looked at sprint segment data from a recent set of ZRL races. The Alley Sprint Reverse segment was analyzed since it is similar in length to Watopia’s forward sprint. We looked at the time gap between 1st and 5th place in the segment times for all four categories in Division 1 of three different timezones. Here are the results:
| 1st Place | 5th Place | AVG Gap | Race | |
| A | 25.07 | 26.473 | 0.35075 | EMEA W D1 |
| B | 24.493 | 25.544 | 0.26275 | EMEA W D1 |
| C | 26.372 | 28.244 | 0.468 | EMEA W D1 |
| D | 29.974 | 31.441 | 0.36675 | EMEA W D1 |
| A | 24.842 | 26.55 | 0.427 | AMERICAS CENTRAL D1 |
| B | 25.878 | 27.091 | 0.30325 | AMERICAS CENTRAL D1 |
| C | 27.208 | 28.93 | 0.4305 | AMERICAS CENTRAL D1 |
| D | 30.8 | 33.741 | 0.73525 | AMERICAS CENTRAL D1 |
| A | 24.596 | 25.876 | 0.32 | EMEA S D1 |
| B | 23.889 | 24.966 | 0.26925 | EMEA S D1 |
| C | 27.591 | 28.359 | 0.192 | EMEA S D1 |
| D | 29.86 | 31.839 | 0.49475 | EMEA S D1 |
The average gap between each place for all races and categories above was 0.38s.
Next, we picked a handful of scratch races that ended in pack sprints, and analyzed the average gaps between the top 5 finishers in each race. Here are those results:
| AVG Gap | Race | |
| 0.085 | ZwiftPower Link | |
| 0.186 | ZwiftPower Link | |
| 0.3355 | ZwiftPower Link | |
| 0.1475 | ZwiftPower Link | |
| 0.24575 | ZwiftPower Link |
The average gap between each place for the scratch races was 0.2 seconds.
Losing Sprints Due To Bike Choice
Based on sprint speed test results and our admittedly lightweight analysis of ZwiftPower data, we can say that, on average, bike choice is only going to cost you at most one place in a sprint ranking.
That is, assuming you have a decently fast race bike (not a pure climber), the amount of time lost with a slower race bike compared to a faster one is less than the average gap between each place in a sprint finish.
And while one place may not seem like much, keep in mind in races like ZRL, the difference between places can be worth up to 5 points!
Of course, this is only talking about averages, since this is the only way we can sensibly talk about the data. In actuality, sprints can be won or lost by 1/1000th of a second, and thus it’s possible that running a slower bike setup could cost you more than 1 place. Of course, the opposite is also true: you may lose a sprint by several seconds, meaning a faster bike wouldn’t have helped you one bit.
Implications for Bike Choice
How should these sprint findings impact our Zwift bike choice? I would summarize my thoughts this way: if you think the race is going to come down to a sprint, and you’re going to be in the mix, then you want a bike and wheelset that sprints well.
When I choose an aero bike for a typical drafting road race on Zwift, it’s usually not because of the watts it saves me while sitting in the pack’s draft for 95% of the race. I choose an aero bike because of the edge it gives me in sprints.
On the other hand, if I’m in a race that includes tough climbs, I need to balance the importance of climb performance vs sprint (aero) performance. And that’s a very personal, situational choice.
Conclusions
There are many factors that combine to determine the outcome of a race sprint. You’ve got your own power, of course, and even that is being read from a power meter that probably has a margin of error of 1-2%. But also how you time that power matters, as well as what sort of powerup (if any) you use, other riders’ behavior, and your choice of bike frame and wheels.
Based on the numbers above, I would say bike frame and wheelset selection are definitely a factor in sprint finishes, but they are one of the smallest factors. I would rank the key factors in order of importance thus:
- Your raw power
- Your timing
- Other riders’ behavior
- Your powerups
- Your frame and wheelset choice
Your Thoughts
Have you considered sprint performance in bike choice on Zwift? Do you agree with my ranking of sprint performance factors? Share below!
Great piece! I have some extra questions:
I would also be interested in ramp up speed for various bikes. I mentally feel like the Venge holds speed better, but kick is slower then Tron
Yeah, I’ve tried sprinting with a faster combo than the TRON, but it hasn’t worked out. Maybe just bad timing 🙂 But I’d be curious how the frames compare accelerating. I feel like the TRON has a significantly faster kick, especially if it’s at all uphill (~1%)
I also have always felt like the TRON snapped quicker than the venge. I also have always felt the TRON received a better draft
Very good question. I’ve noticed this as well. Tron seems snappier to me, but I seem to gain ground quicker ans slice through the pack on the Venge in a longer sprint.
This is the big question Here. @eric Schlange. What about accelerations before reaching top Speed?
Everyone always wants to know about accelerations. We’ve tried testing them, but can’t get granular enough timing for it to be meaningful.
My hunch is, the top race frames all accelerate essentially identically, because their weights are pretty close to the same.
For these speed tests, we did a long run in, so the entry speed was basically “top speed” on all tests. No acceleration figured into it.
Bike choice in a draft does make a big difference. Follow one of the pace partners with a MTB and see what your wkg is compared to those that you’re drafting.
Or be a bit more realistic, and pick the Zwift road bike and see.
There’s a % saved when drafting. The more riders you draft, the higher the % saved will be. But that % is applied to what you would have to do without draft.
So even the fastest non-aero road bike will have to push harder than someone with an all aero setup.
What is that % though? Let’s take the most aero bike (Venge) with a flat time of 3077 (from https://zwiftinsider.com/charts-frames/ ) and the fastest climber (Aethos) with a flat time of 3093. If one was sitting in with a group doing 250 w on the Venge, one would expect to how to put out 3093/3077 * 250 w = 251.3 w to sit on with the Aethos, or 0.52% more watts.
Nope, because power and speed aren’t linear. The faster you go, the more resistance there is, and therefore even more power is needed to go +x kmh faster.
And that 3093/3077 is for a specific weight and power. Change either of those variables for both riders, and you will get a different % of separation between the two bikes.
In the example you gave, the time was calculated using 300W and not 250W. So the total time and the difference between the two would be different
Okay, please do the calculation using whatever specific weights speeds you like and give me the percentage then. Until then I’m the only one who’s actually provided some useful info, and you are just saying whatever you want. And yes I know that power/speed is not linear I was doing an approximation, and we are talking 0.52% difference, so it can’t be that far off.
Your info isnt useful because its wrong. You might not like this but it true
If we go with 300w as you suggest then the calculation becomes 3093/3077 * 300 w = 301.56 w, which is again 0.52% difference. So 1.56w instead of 1.3w, so I guess that’s a HUGE difference and you’ve invalidated my whole argument then, eh?
Sorry if I somehow upset you. And of course it’ll still be 0.52%. You’re dividing the same two numbers. At 250W there would be a difference between the times, and would have a different percentage. And you agreed that it’s not linear, but are using, linear (applying the same % to two different wattages). Two examples of power and speed. On another page Eric mentions that at 300W two riders will do 42kmh. At 375W then would do 45kmh. That’s +75W for an extra 3kmh. From https://www.gribble.org/cycling/power_v_speed.html A 185lbs rider on a 15lb bike (with a frontal of 5.4788 and… Read more »
Just wondered where you got that 0.4mph estimate from? (extra marks for showing your working)
Thanks Eric, Excellent analysis and explanation. I usually ride a Tron just to play it safe but when your raw power is as low as mine, bike/wheel choice and the other factors never seem to matter much, lol.
I hear that! I’ve never been able to muster much more than 10wkg over 15 seconds, even on a good day with super fresh legs. I’m convince my legs have zero fast twitch fibers.
Eric and in case of punchy hill sprint?
For example next zrl stage will finish with a punchy hill sprint.
Well that’s the big question, isn’t it? No simple answer here, as it depends on the length of the hill and the steepness of the climb.
Question is about next zrl stage, so basically libby hill, to give the question mor complex, with the cobble, just for the final could how much gain could you have with a gravel bike? 🙂
Absolutely none. You will easily be beaten up the hill. Gravel bike will not give you an advantage on the cobbles. And it’s heavier and the wheels aren’t as good.
I’m sure the Zwift road will beat the gravel up it.
If you’re talking something like the Temple KOM, then yes the gravel would be the definite choice. But Libby is too short (and I don’t think the cobbles have an effect on your riding)
As well, you would be left so far behind in the previous 32Kms if you went with gravel.
I’d add that before the sprint the speed ups sometimes a lot, and for that part, being on an aero bike is very important.
the decision is a strategy thing : will you attack on a climb, and finish at the top? If yes choose wisely.
in any other case aero rules : if I get away on a climb (which i can almost always do, being a very lightweight B, under 250w of 95% 20 minutes power but at 4.7 wkg), i will need to stay away, so aéro still rules!
Great advice as always, shame you didn’t post this before the “Chase the Sun” ZRL. I was messing around with bike swaps in the pen, and ended up starting on a gravel bike (thought i was on the Tron). I then produced my best 2 min effort, to fall out of front group as sprint started, and it was only when coming up to the next sprint i realised my issue. Changed to Tron, and managed a 5th FTS on 3rd Sprint (small leadout from fellow team mate), and 1st on last sprint, with no leadout, and beat 2nd by… Read more »
I’ve written before about how much slower gravel bikes are in the pack and at higher efforts. Suffice it to say, you won’t be winning competitive sprints on a gravel bike!
See “Power Required to Sit In on Flat Paved Sections” here https://zwiftinsider.com/zrl-s4-w8/
Great that you listed timing in there. I would almost call that and raw power 1a and 1b. I’m terrible with timing my sprints. I almost always see someone else to too early and think I need to chase them down to or lose out on FAL, so I go to early, blow past the first jumper at the halfway point and get overtaken myself with 10-20% of the sprint left because I spent the whole thing out in the wind while people who used my draft slip past me at lower raw W and W/kg.
Hi Eric, I’ve been wondering about this test for a while, so good to see the answer. I’m also curious if aero matters less in a pack than when riding solo as most of your tests are. In other words, if riding a light weight set up costs say 10% more speed in your solo tests, it’s a big issue. But if drag resistance is applied after draft benefits (correctly?) You may only lose 5% in a pack and it may become viable to use in a race. I guess a test for this would be more complex but I’m… Read more »
Aero still matters in the draft, but the difference is, most racers are not on the rivet if you’re sitting in the draft. In that case, if there’s a key climb in the race, it absolutely makes sense to get a good climbing bike for that climb, and have it “cost” you a bit in the flats where your effort is lower.
Any kind of up hill component before the sprint and those ZIPP disc wheels are a pig to spin up. Most sprints it is the speed you cross the start line at that is critical. NYC Outer Park Loop sprint / France sprints / Watopia Villas sprint – the lightweight wheels in those 100 meters before the line – coming off a gradient – into the flat – are vital.
As others have mentioned, I think you are missing the key factor of acceleration, lighter bikes in theory accelerate faster
Not missing it – purposely left it out. See note on first comment at top.
What about the drafting impact during the sprint? A more aero bike can benefit more, no?
Not sure you could say an aero bike will benefit more from the draft than a climbing bike. But I think the difference in draft benefit between both styles of setup is negligible.
Great article. Very interesting.
In ZRL, FTS points often come down to hundredths of a second. I think the difference matters most for those going for FTS on the sprints. For FAL and finish positions, you’re most likely going to be in the same position whether you’re on Venge/Disc or Tron, as Eric concludes above.
What about the gravel bike? (I was the idiot that attacked first in your league on Tuesday night at the first sprint. I sensed a slight lull at 500m into the race and launched only to not get anywhere, later I noticed that in doing bike swaps practice in the pen that I had started on the gravel bike!)
It’d be interesting to hear from a sprinter who wins a lot without necessarily the best power profile 😉
Got any suggestions?
I guess the final piece of the puzzle is whether these split second differences in sprint times are, in actuality, dwarfed by the effects of the draft. By this I mean your selected frame/wheels might be half a second quicker over the sprint in a non-drafting drag race, but how does that play out when for the entire duration of your sprint you are in the draft? (In most sprints I do, I hardly ever actually hit the front and go into the wind.) My guess would be that the draft effects totally dwarf the effects of frame choice, so… Read more »
In short, more power wins over bike choice.
And an aero boost power up usually helps as well.
Network connection. Long ping times and latency can be seen in normal riding when people put down the power and back off. These riders can be seen moving through the pack after efforts and quickly going backwards again. Same sort of effect as having low trainer difficulty over that terrain. I’d imagine that given the tight time gaps if you have a poor connection your visual representation of where you are adds delay to your sprint start and in receiving that increased power data.