So, I work in AI/Pharma, and the new deep researcher tools are cool. They essentially combine a bunch of AI agents on a research team with pubmed/internet access. Of course, I had to ask it our favorite question: where does motocross rank in terms of how physically demanding it is? Here is the report:
When you consider cardio, strength, and skill, motocross belongs at the very top of the list of demanding sports—right up there with XC skiing, 2k rowing, MMA/boxing, and water polo.
Why Motocross is “S-Tier”
1. Cardiovascular load: near max the whole moto
On-track physiology studies show motocross riders sustain 95–98% of HRmax for almost the entire race. Mean oxygen uptake is ~32 mL·kg⁻¹·min⁻¹ (≈70–80% VO₂max) with post-moto blood lactate around 5 mmol·L⁻¹. In one 30-min simulation, riders spent 87% of time above 90% HRmax.
Konttinen et al., 2007, J Sports Sci
Konttinen et al., 2008, J Strength Cond Res
2. Strength & impact: repeated isometric/eccentric load
Handgrip strength drops ~16% immediately post-moto. Lower body muscles stay at ~24–45% of maximal activation just absorbing landings, bumps, and side-to-side bike control.
3. Heat strain in full gear
In real races, 33–50% of riders hit core temps ≥39.0 °C (102.2 °F) despite hydration, with 4.3–8.7 L sweat losses over an event. Protective gear reduces evaporative cooling, compounding the load.
González-Alonso et al., 2012, J Strength Cond Res
4. Vibration & forces every lap
Landings, bumps, and throttle forces expose riders to whole-body and hand-arm vibration levels that exceed ISO health-risk thresholds, accelerating fatigue.
McCauley & Kent, 2010, Occup Environ Med
5. Skill & decision-making under duress
Every lap the track changes—ruts, bumps, traction. Riders are constantly making split-second choices at speed while squatting, deadlifting, and pulling the bars to throw 100+ horsepower where it needs to go. Reviews of motorsport physiology emphasize the unique “triple threat” of mechanical + physiological + cognitive stress.
Prada et al., 2019, Front Physiol
How It Compares to Other Elite Sports
XC skiing: highest VO₂max values recorded (80–90 mL·kg⁻¹·min⁻¹).
Rowing (2k): ~36 kcal/min at race intensity.
Boxing/MMA: HR near max, lactate up to 15 mmol/L.
Ice hockey: repeated high-intensity glycolytic bursts.
Water polo: ~60–85% of match time >85% HRpeak.
Motocross combines all of these loads simultaneously: sustained max HR, continuous strength demands, high heat load, vibration/impact, and cognitive skill. That’s why many researchers and riders alike put it in the “S-tier” of the world’s most physically demanding sports.
...but you have a motor! 😂
Cool info- thanks for posting.
I don't disagree with the overall assessment, but here's yet another reason to not blindly trust AI:
Point #5 from the AI: ...and pulling the bars to throw 100+ horsepower where it needs to go...
Yes, even though “deep research” is a bit better, it’s still just a bunch of LLMs, which basically pattern complete language based on all the language out there. They don’t actually know anything. Regardless, I thought the report was helpful, it basically finds all the evidence available and summarizes it well.
Pro Soccer and Pro MX have been considered at the top since the 1970’s
The Shop
Free shipping: VITALMX
Luxon 4-Post Bar Mounts
$189.95 - $239.95
How bout an 8 minute moto on the vet track?
S tier in my book brother
"it basically finds all the evidence available and summarizes it well."
Well, yes, except for the obvious part that it got completely wrong. 🤣 How can we trust the rest of the result with such a blatant mistake included?
AI can be very useful as a tool, so long as the results are verified by someone who knows what they're doing. The problem is that it's being used more and more without verification. Look at all the posts here that people blindly copy and paste the Google AI result from a search. Our society is quickly losing the ability to think critically (what little ability we had anyway).
Yeah, you can credit me for training GPT that a 32:1 premix mixture is gas:oil, not the other way around. 😄
That said, as skeptical as I was about AI and how it's become such a buzzword, you can't deny its usefulness and possibilities.
I agree, with the caveat of "Trust, but Verify". I just feel that there's a lot more trust than there is verification these days...
Not only that, but because people continually post “articles” and “blogs” that are entirely AI written, the AI then gets in a feedback loop that sites its own garbage as a credible source.
That's referring to Prado's Kawi after he said they finally got the motor where he needed at Ironman.
I thought that was just some shit Bruce Brown pulled out of his ass ? 😁
The problem still sits with people using it. You need to constantly make sure AI explains where it got its information and then you can check how reputable it is. The content - at first glance - used above is reputable in that it comes from peer reviewed scientific articles. Each article could have problems however, but you can ask questions of AI to figure out what these might be. As for 'pulling handlebars' that might be in the original article and be a phrase from a scientist unfamiliar with moto terminology.
I was always told that you just sit down and twist the throttle.
Yes, but soccer has the secret can of spray to heal injuries.
To be fair the intelligence community in conjunction with MSM had the circular sourcing loop perfected way before AI. More proof it can't invent new things lol.
I can't wait until we are all being treated by the next generation of doctors that earned their degrees 99% using AI.
Pit Row
This is about as spot on as it gets. Working in the chemical industry, I see people using AI for technical reference without checking the information all the time. The information given is easily wrong +75% of the time, and the more technical, the less accurate the information is. AI has a long way to go in many technical industries and markets, that is if it can ever even get there. Yes It can be a helpful tool, but actual human thinking can't be replaced unless the information is incredibly remedial.
At the pro level, 100% moto is S-tier. Motocross gets more physically demanding the faster you get as a rider. A slow rider like me won’t have the balls or skill to push my body to its physical limits on a motocross track. But I can go run a 5k and will be absolutely gassed even as a slow runner.
Indeed - AI can be extremely useful but it’s not intelligent in the sense that we are. This famous paper (famous as far as philosophy papers go) makes a compelling argument that computers can’t know or understand anything or have any cognitive states.
https://rintintin.colorado.edu/~vancecd/phil201/Searle.pdf
As far as how demanding motocross is, when bought a YZ 250 10 years ago at 43 years old and went to a sand track after not riding for 20 years, I was astonished by how physically demanding it was. I remember wondering if maybe this was a bad idea! I’d been doing kickboxing training so it’s not like I was completely sedentary but riding that YZ in the whooped out sand was another level of difficult. It’s not easy to convey that to someone who has never done it.
Unfortunately anyone who thinks racing motocross is easy because they rode a four wheeler one time at their uncle’s house isn’t going to read or believe any of this. I gave up trying to convince others that it was a tough sport 25 years ago when I was in high school.
If you could somehow feel what RC has done at like glen Helen, 35x2, almost all would have a heart-attack 15min into first moto. The phys demand is sky-high. Now, you also better be able to think right, or, well, lights out!
100% agree. I use AI at work everyday for bits of code and some data processing. I have to fix a lot of what the AI spits out because it makes a lot of mistakes, even with simple stuff. However, because I know what I'm doing, I can see the errors and fix them. With a lot of simple/routine jobs, it works out quicker to use AI and fix the mistakes than it does to start from scratch. My colleagues, who don't know how to code or analyse data, also use AI but they can't see the mistakes! They don't even know what a mistake would look like, let alone point one out. The same seems to be true with the person who started this thread. That overview of the research they posted has way to many red flags to be taken seriously. What they posted needs to be checked, reviewed, and rewritten before it can be taken seriously.
Or just sit down and smash the burger...and fries
But one thing separates this sport from many others where you have to be at a very, very high standard physically. Skill on the bike is such a big factor that relatively speaking a slightly overweight Cooper Webb can win races and even a championship.
That is why i will never put MX/SX as the most physically demanding sport in the world. It's up there obviously, but i would at least put XC skiing, roadcycling and soccer ahead of our sport.
Those calculators based on "fact" you are 100% using your fitness and strength all the time but in reality you jump stuff and flow the track so you are not anywhere near 100% all the time. Beginner can use his full stamina to get around track pro´s are riding track level their head stays clear. I also think Konttinen calculations are not 100% correct as i have seen his study for examble 5-9 L sweat loss need more track time in reality. AI can do stuff but if your master data isnt correct you gonna go outside of correct answers like 100 Horsepower and so on.
Soccer??? Are you kidding me?
Well football of course, but this is mostly an american forum so i thought i would use the american name.
Or do you disagree with it not being one of, if not the most demanding sport?
Post a reply to: Motocross is “S-tier” for overall physical demand