Ditlev 7:19:17 vs Blummenfelt 7:21:12: who had the faster day?
The two fastest IRONMAN times ever are 1 minute 55 seconds apart. That gap hides a much bigger story. One race had an ocean current behind it. The other had cool air and a north wind. Below, we put both races on one 226 km course, then remove what came from the course and keep what came from the athlete.
Both races compared at the same distance, timing mat by timing mat. The lines show who is ahead and by how much.
The splits
| Swim 3.8 km | T1 | Bike 180 km | T2 | Run 42.2 km | Total | |
|---|---|---|---|---|---|---|
| Blummenfelt · Cozumel 2021 | 39:41 | 1:52 | 4:02:40 | 1:37 | 2:35:24 | 7:21:12 |
| Ditlev · Florida 2026 | 47:52 | 4:05 | 3:51:16 | 2:32 | 2:33:31 | 7:19:17 |
| Gap | Blum. 8:11 | Blum. 2:13 | Ditlev 11:24 | Blum. 0:55 | Ditlev 1:53 | Ditlev 1:55 |
Official splits as published after each race. Ditlev's transition times come from the race timing data in Finisher. Each split is rounded to the second, so the legs may not add up exactly to the total.
Blummenfelt won the water and the transitions by about 11 minutes. Ditlev won the bike by 11:24 and the run by 1:53. The question is how much of each gap came from the athlete.
Swim: it's a current, not the tide
Cozumel's swim is fast, but tides aren't the reason. Tides in the Caribbean around Cozumel are small. The push comes from the Yucatán Current, a permanent ocean current that flows north through the 18 km-wide channel between the island and the mainland. Oceanographers who moored instruments there measured a mean along-channel flow of about 1.1 m/s at 30 m depth (Athié et al., 2011, citing Chávez et al., 2003). Near the shore, where the race swims, the current is weaker and less steady. It still pushes the swimmers if the course runs with it, and IRONMAN describes the Cozumel swim as down-current.
Our results database shows the size of that push:
| Median swim time | Cozumel 2021 | Florida 2026 |
|---|---|---|
| Pro men | 42:49 | 51:11 |
| Age-group men | 58:24 | 1:20:50 |
| Age-group men 35-49, every full IRONMAN since 2021 | 1:16:01 | |
The whole Cozumel field swam about 23% faster than a typical full IRONMAN. Florida's swim was slower than typical. It was two laps of 1.9 km, with a run up the beach between them and another to transition, all of it counted as swim time.
The key point is that both men came out of the water with the leaders. Blummenfelt exited 2 seconds behind the first pro. Ditlev was 11 seconds behind the first man out (47:41). Neither lost or gained much on his own field. The 8:11 between them is almost all water, not swimming ability.
Two more details. Slowtwitch listed the 2021 Cozumel swim as 2.1 miles (3.4 km). IRONMAN said the course was accurate and recognised the times as world bests. Water temperature mattered too. IRONMAN rules allow pros a wetsuit only up to 21.9 °C. Neither race's pros wore one. Cozumel's water was well above the limit, and in Florida the pros also swam without wetsuits (the Open-Meteo marine model puts the Gulf near 24 °C that morning). So the 8:11 isn't a wetsuit effect either.
T1 and T2: the layout, not the hands
At Florida the athletes run up the beach and across the road to a parking lot with change tents. Cozumel's transition sits next to the swim exit. Again, the field medians show it:
| Pro men, median | Cozumel 2021 | Florida 2026 | Course cost |
|---|---|---|---|
| T1 (swim → bike) | 2:09 | 3:45 | +1:36 |
| T2 (bike → run) | 2:10 | 2:31 | +0:21 |
Ditlev gave away 3:08 in transitions. About 2 minutes of that is the venue, so he lost roughly 1 minute through his own handling.
Bike: power, air and wind
Neither athlete published power data for these races, so we estimated it. On flat roads, almost all of a rider's power goes into pushing air out of the way. The rest goes into the tires:
We used the same assumptions for both: a typical elite time-trial position (CdA 0.21 m²), fast race tires (Crr 0.0035), 90 kg for rider plus bike, and 97.5% drivetrain efficiency. The only inputs that change are each rider's speed and the density of the air (ρ) on his race day.
| Speed | Air density | Estimated power | |
|---|---|---|---|
| Ditlev · Florida | 46.7 km/h | 1.204 kg/m³ | ≈ 325 W |
| Blummenfelt · Cozumel | 44.5 km/h | 1.170 kg/m³ | ≈ 275 W |
At these speeds about 85% of the power goes into air drag. Air drag grows with the cube of speed, so 2.2 km/h more costs about 50 W more. These are estimates, not measurements. A smaller CdA lowers both numbers by the same proportion, but the gap between the riders stays.
Air density: cold, dry air is thicker
Florida raced the morning after a cold front. ERA5 reanalysis for the bike hours gives about 19 °C and 60% humidity. Cozumel gives about 26 °C and 84% humidity. Warm, humid air is less dense: water vapour is lighter than the nitrogen and oxygen it replaces. Ditlev rode through air that was about 3% denser. With the same power in Cozumel's air, our model makes him about 2 minutes faster. Blummenfelt in Florida's air would have been about 2:15 slower.
Wind: you can see it in the timing mats
Florida's race day had a north wind of 13-16 km/h with gusts near 30 km/h (ERA5). The bike course is one big loop with several out-and-back sections. On the stretch heading north toward the turnaround (km 93-113), Ditlev averaged 41.4 km/h. Heading back south (km 113-144) he averaged 54.9 km/h. Out-and-back sections only partly cancel the wind, because a rider spends longer in the headwind than in the tailwind. For Cozumel 2021, the reanalysis shows light winds of 5-10 km/h, which is calm for an island whose east coast is usually windy. The reanalysis grid is coarse, though, so treat that as a rough figure.
Drafting, roads and distance
- Draft zone: 12 m in 2021, 20 m since March 2026. Both men rode most of the day alone at the front, so drafting doesn't explain either time.
- Road surface: neither course publishes rolling-resistance data. As a sense of scale, a rougher road (Crr 0.0045 instead of 0.0035) costs about 3 minutes over 180 km at Ditlev's power.
- Elevation: both courses are near sea level with only small rollers. At 45+ km/h they don't decide anything.
- Distance: Slowtwitch listed Cozumel 2021's bike as 110 miles (177 km). If that's right, Blummenfelt averaged 43.8 km/h at about 265 W. IRONMAN said the course was accurate.
Both riders beat their nearest rival by a similar margin: Ditlev by 8:43, Blummenfelt by 8:07. Both were the strongest cyclists in the race by a long way. Ditlev did it faster and in thicker air.
Run: heat is the hidden handicap
Blummenfelt ran 2:35:24 in about 26-27 °C and 80-85% humidity. Ditlev ran 2:33:31 in 17-23 °C and dry air. Humidity is what hurts: sweat only cools you when it evaporates, and saturated air slows evaporation. A study of seven big-city marathons over decades found that times slow steadily as the WBGT heat index rises, and the fastest runners slow too (Ely et al., 2007). Cozumel's conditions sit roughly 8-10 °C higher on that index than Florida's. For an elite runner that usually means a 2-4% slower marathon, which is 3-6 minutes here. It is likely worse in an IRONMAN, where the body is already hot after a 4-hour ride.
Florida's run was fast that day: Matt Hanson ran 2:28:07, the fastest in the race. Blummenfelt was the fastest runner in Cozumel by more than 6 minutes. Adjusted for heat, his marathon was probably the better of the two.
Fueling: the science behind 7 hours at the limit
Neither athlete published his race-day nutrition, so here are the physiological limits they work within:
- Energy: a cyclist turns about 23% of the energy burned into power at the pedals. At ~325 W that is roughly 1,200 kcal per hour, or about 4,700 kcal on the bike alone. No athlete can eat that much during a race. The gap comes from stored glycogen and fat.
- Carbohydrate: the gut absorbs about 60 g of glucose per hour. Mixing glucose with fructose, which uses a different transporter, raises that to around 90 g/h and more (Jeukendrup, 2010). That is why pros now aim for 90-120 g/h and train the gut to handle it.
- Fluid and salt: in Cozumel's humidity, sweat rates of 1.5-2 L per hour or more are common. Sodium and fluid matter more there than in Florida's cool, dry air, especially on the run.
The verdict: one course, same conditions
What if Blummenfelt had raced Florida 2026? Adding up the course effects above:
| Florida vs Cozumel, for Blummenfelt | Effect |
|---|---|
| Swim with the leaders and no current (Ditlev's swim) | +8:11 |
| Longer transitions (field medians) | +1:57 |
| Same power in denser air | ≈ +2:15 |
| Cooler, drier marathon | ≈ −3 to −6 min |
| Blummenfelt's 7:21:12 moved to Florida | ≈ 7:28 to 7:31 |
On equal terms, Ditlev's day was about 9-12 minutes faster. It rests on the best bike split ever recorded at an IRONMAN, ridden mostly alone in thick air. Blummenfelt's Cozumel run in the heat stays one of the great IRONMAN marathons. What made his total so fast was the water.
What this estimate leaves out. We don't know the wind Blummenfelt would have met in Florida, the actual road surfaces, either athlete's real CdA or power, or how much the 2021 course distances really differed. Each of these could move the result by a minute or two. None of them can close a 9-minute gap.
Method and sources
- Finisher results database: every finisher of IRONMAN Cozumel 2021 (2,336) and IRONMAN Florida 2026 (2,152), with timing-mat splits. Field medians count finishers only.
- Weather: ERA5 reanalysis (ECMWF / Copernicus) through the Open-Meteo historical API: hourly temperature, humidity, pressure and wind at each venue (sea-surface temperature from the Open-Meteo marine model). Reanalysis is a model on a ~25 km grid, not a weather-station reading. Air density is computed from temperature, pressure and humidity.
- Athié, G., Candela, J., Sheinbaum, J., Badan, A., Ochoa, J. (2011). Yucatan Current variability through the Cozumel and Yucatan channels. Ciencias Marinas 37(4A): 471-492.
- Ely, M. R., Cheuvront, S. N., Roberts, W. O., Montain, S. J. (2007). Impact of weather on marathon-running performance. Medicine & Science in Sports & Exercise 39(3): 487-493.
- Jeukendrup, A. E. (2010). Carbohydrate and exercise performance: the role of multiple transportable carbohydrates. Current Opinion in Clinical Nutrition and Metabolic Care 13(4): 452-457.
- Race results and context: Slowtwitch, IRONMAN Cozumel 2021; Slowtwitch, IRONMAN Florida 2026; Triathlete, IRONMAN recognises the Cozumel times.
- Power, energy and heat figures on this page are model estimates under the stated assumptions, not measured data.
Race anyone. Even your past self.
Finisher has 2.5M+ IRONMAN and 70.3 results with every split. Compare any two athletes, or two of your own races, leg by leg.
