But what if your beliefs were derived from facts and data? 
Well, if something doesnāt feel as good as something else, but the data says itās fasterā¦Iām not going to use it.
Iām glad you found what works for you.
Do you have a link to this data? Iād be interested to put in my stats from my last race and see what it says. But again this chart relies on static rolling resistance numbers, which just donāt exist in the real world.
Hereās a real world scenario thoughā¦
I think a lot of resistance to wide tires comes from riders whose frames canāt fit them. Everyone in this thread should be required to post which bike they have so we can determine what level of mental gymanstics are at play 
It is already posted up ^^ in the thread.
But again, that isnāt how data works. You canāt just āput in your statsā because the data is based on a specific set of criteria.
No, it doesnāt. The tester tests his setups in the real world, not on a drum.
So what is the tire of choice for 60% non-chunky gravel/40% pavement?
āIt dependsā
What are the pavement sections likeā¦how long are they? Are they just short, numerous connecting stretches, or are they longer sections with climbs?
Pavement section is relatively long and flat
Dependsā¦ But, in my experience, in most gravel races, the action tends to slow down on the paved sections (of course not alwaysā¦) and people take that time to re-fuel and, maybe, relax just a bit. This tendency makes the gravel sections matter more. For really smooth gravel and some pavement, the smallest tire I would run now would be the G One RS 45. I would still want to test both the Thunderburt 2.1 and G One RS 50 as well, because they might be faster. Iāll never run anything <45 mm now, as the G One RS 45 is fast, even on pavement.
Are we talking packed down dirt/gravel roads? How long? If itās packed down dirt/gravel and nothing chunky, probably a G-One RS pro 40, 45 or 50. (I kinda thought a conti 35 all season but these two seem the same, so might as well go wider)
+1 on āit dependsā. Some of the smoothest/fastest sections of road Iāve ever raced on were dirt. And some chip seal pavement can be as slower than fairly chunky gravel. But for your 60/40 mix example and assuming the gravel is pretty firm/packed, Iād probably be running a 38 or 42. But itās really hard to standardize this stuff when riding surfaces can range from glassy smooth to small boulders.
And you really need to consider the likely racing dynamics as well. For a given course, I might choose a very different tire if I plan to be racing in a pack all day vs. riding solo. If I plan to race in a group, Iām going to prioritize a tire that excels on the selective sections (which often means a wider tire for chunky sections). For a solo effort on the same course, it often makes more sense to optimize the tire for the surface where youāll spend the most time.
I did big sugar last year on my MTB rolling 2.2 conti race kings. Turned out great because the most selective spots were mostly loose/chunky climbs followed by sketchy downhills where my setup excelled. On the flatter/smoother sections, the extra watts I was burning with tire and bike choice were not a bother when spinning along in a group (Iām a bigger guy with watts to spare on the flats). If I was doing the Big Sugar course as a solo TT, my MTB/2.2 tire combo would not be the fastest way around that course, Iād go with a gravel bike with aero bars running 47ās.
I agree with that. Roads are chill and gravel sectors are full gas
@jkarrasch thanks for doing the testing, sharing/explaining the methodology and posting your results!
Iāve got two questions about the āpower to hold 15mphā table that you posted:
- Can you create that for pavement and Cat 1 gravel? At least for the tires youāve tested in each? I realize not every time has been tested on every surface, but for the combinations that have been tested it would be great!
- Can you comment on the precision (margin of error, plus/minus, etc) that we could interpret the power numbers with? In other words, is it fair to say that the Terra Speed 700 x 45 are always exactly 155watts for the rider CdA and weight that youāve estimated? Or does the Chung method also provide a range of resistance values that you can use to calculate something like a margin of error for those power estimates?
Again, this is really great. While I realize it may not be perfect, the methodology seems sound and replicable. Iād love to see how it expands.
I have 700x45 Terra Speeds AND a wheelset with a Conti RK 650b x 2.2 front, Schwalbe TB 27.5 x 2.1. Iām guessing the RR numbers are similar out the same as your tests, while the aero drag numbers might be slightly better for the RK/TB combo. If thatās the case, it would be worth understanding if Iām just better off running the 650b wheelset all the timeā¦
Hiya folksā¦ any of yāall still into this tire stuff might want to check out the most recent pod episode with Ronan Mc Laughlin of Escape Collective (Performance Process) on off road mtb and gravel tire testing.
Just sayin.
Sure I can model whatever. The CRR values are accurate to less than 1 watt. Same for CDA.
Regarding margin of error itās pretty small since so many data points ābuildā a virtual elevation profile.
is that ep9?
Episode 39. Free one out today I think. Here ya go
https://open.spotify.com/episode/5U7PJ9FLAqUpwHkZjEw8qr?si=pJAlpci7TgCt5s5RjbC-9g
