Cutthroat or Litespeed for the Raymond ride? It's a fair question. Titanium gravel bike, light and 'skinny' (for a gravel bike). Or the carbon bikepacking rig that's comfier and more stable when things get chunky. I've owned both for a while and I still couldn't tell you which one is actually faster. Vibes are not data.
So I went and measured it.
The test track
There's a loop out by the bay I ride constantly — the West Dumbarton Bay Trail, clockwise, 3.36 miles of packed dirt. It is silly flat. I pulled the GPS track and the entire elevation range across the whole thing is 2.0 meters (6.5 feet). Not the climbing — the range. Highest point to lowest point. (And that may be a measurement illusion)
That matters more than it sounds. On a climb, weight dominates everything, and I am a clydesdale (that's what the industry calls people above median weight. No, seriously. the average American man weighs more than the design limit for much biking equipment). Any climb comparison is really a weight comparison wearing a costume. On something this flat, weight basically drops out of the math. Which means whatever's left is the bike, the tires, and me.
Surface is what I'd call packed moist. It's bayside, so it's rarely bone dry, and keeps some minor residual rutting. Close to 'champagne gravel' but not perfect. Remember that part, it turns out to matter. (The post cover photo is on the trail)
The protocol: ride the loop holding ~200 watts, steady. Same effort, different setup, compare the clock. Turns out holding steady power on flat ground is way harder than it sounds (you're constantly correcting, and there's no hill to hide the surges), but I got there.
February: case closed
Feb 25, Cutthroat on 2.25" (57mm) Vittoria Mezcals. 202 W. 15:46.
Feb 27, Litespeed Gravel on 40mm Vittoria Terreno Drys. 198 W. 14:19.
Four watts (2%) apart. 1:27 faster on the Litespeed. That's 8.5%, which is way outside anything I'd call noise. I wrote in the ride notes that day: "It's looking like for pure efficiency, the LSG is the way to go."
Done. Titanium bike wins. Go ride. Well, sorta. I later learned that there would be some segments where I'd like the Cutthroat so I ended up taking it.
May: well… kinda
Then, later I put 2.4" (60mm) Continental Dubnitals on the Cutthroat and ran the same test again, to see how much worse they were.
12:49.
They were not worse, but better. I actually wrote "That's madness" in the ride description. The wider tire, on the heavier bike — the Cutthroat is about 1.6 kg (3.5 lbs) heavier (not that it matters on this segment) — beat the setup I'd just concluded was the efficient one. By another 1:30.
So I did it twice more to be sure. 13:24. 13:23.
The numbers
All five at 198–202 watts. Under 2% spread in effort.
| Setup | Time | Watts | Speed |
|---|---|---|---|
| Cutthroat, 2.4" Dubnital | 12:49 | 201 | 15.7 mph |
| Cutthroat, 2.4" Dubnital | 13:24 | 199 | 15.0 |
| Cutthroat, 2.4" Dubnital | 13:23 | 198 | 15.0 |
| Litespeed, 40mm Terreno Dry | 14:19 | 198 | 14.1 |
| Cutthroat, 2.25" Mezcal | 15:46 | 202 | 12.8 |
Averaging the three Dubnital runs: the 40mm Terreno is 8.5% slower. The 2.25" Mezcal is 19.4% slower.
Nineteen percent. For the same watts. On the same bike.
But it's the low pressure, right?
That was what someone else insisted. Nope. Both wide tires run at roughly the same pressure — I'm somewhere in the 23-25 psi range on either, versus 40-50 for the 40mm. So inflation isn't the variable on the wide tires.
And width alone doesn't explain it either, because the Mezcal is wider than the Terreno (2.25" vs 40mm) and it's the slowest thing here by a mile. So it's not "wide = fast."
What's left is casing and tread. The Mezcal and Dubmitals are knobby XC race tires. The Terreno Dry has a smooth center strip. Net, net: tread and casing, at appropriate pressure. Width is the enabler, not the answer.
The part that surprised me most is how little roughness it takes. This trail is pretty smooth by dirt standards. Minor ruts, that's it. And that's apparently enough to hand a 19% spread to the tires.
The part where I get humbled by a robot
Full disclosure: I didn't do the analysis myself. I've been running my training data through an AI (I call it Claudius, long story) and I handed it the whole segment history to chew on.
It got the February result right, then talked itself out of it. It built a fancy model across every effort I'd ever logged on that loop, decided the February comparison was an outlier, and retracted the conclusion. Then, on cleaner data, it retracted the retraction and landed back where I'd been in February. Which is where I'd have been if it had just… looked at the three numbers.
To be fair, it also caught two things I'd have missed. Strava's leaderboard reports elapsed time, and that I don't always ride the trail for performance - it's nice out there. One of my efforts had 76 seconds of standing around baked into it, which is a full mph on a test where 1 mph is the whole point. It checked all five of these against moving time. Clean, all of them.
And it flagged that my fastest-ever run on the other dirt test segment I was looking at for testing was set in 10-20 mph wind, which — reading my own ride notes back — I had cheerfully admitted at the time and then completely forgotten. That number is garbage. Ignore it.
So, Raymond?
Cutthroat. On the Mezcals despite the cost.
Which is a slightly annoying conclusion, because I spent a while agonizing over two bikes when the thing that mattered cost was something else. But had I swapped tires, the heavier bike with the fatter tires would have also been the faster bike on this surface, and the more comfortable one. What's amazing is that the Mezcals and the Dubnitals are considered fast tires in their class. It wasn't comparing dissimilar tires.
Go check your tires. 😏
Comments
Post a Comment