“Start slow, speed up”
The data says no. The first run looks fast because the first lap is shorter at most venues. At the same fitness, a fast start costs you nothing later. Run your own pace.
n = 9,134 races with full splits
Free · no email · updated August 2026
Answer 3 questions. Get a week-by-week plan built for your race date. Every session is backed by real race results. Free, as a PDF.
34,076
race results
14
venues measured
86
claims tested
The plan counts back from race day. No race booked yet works too.
Updated August 2026 · calibrated against seasons 8 and 9 official results
The plan counts back from race day. It splits your weeks into base, build, peak and taper.
Your level and your days per week decide what fills each week. Runs, stations, strength, simulations.
Every session shows its source. Download it as a PDF. Free, no email.
The numbers come from real data. We read 2,061 pieces of Hyrox advice, kept the 86 claims most people agree on, and checked them against 34,076 official race results.
Most training advice was never checked against results. We checked. 3 popular claims that fail:
The data says no. The first run looks fast because the first lap is shorter at most venues. At the same fitness, a fast start costs you nothing later. Run your own pace.
n = 9,134 races with full splits
528 upvotes say fast wall balls prove you went too easy. The results say the opposite. Athletes with fast wall balls finish about 19 minutes ahead and slow down less. True at all 28 venues we checked.
28 of 28 venue checks agree
Depends on the building. Glasgow costs 3:49. London Excel costs 10:00. Same athlete, same fitness. The plan uses your venue's number instead of a guess.
2.6 x between venues
When the community agrees, the plan says so. When the data disagrees, the plan says that too.
Transition time depends on the building. Sleds feel different at every venue. Measured August 2026 from official results, comparing athletes of the same level.
| Venue | Roxzone median | Course runs | Sleds | Measured |
|---|---|---|---|---|
| Glasgow | 3:49 | ~9.20 km | fast | 2026 |
| Manchester | 4:09 | ~9.17 km | average | 2026 |
| New York | 4:46 | ~9.28 km | average | 2026 |
| London Olympia | 4:47 | ~9.16 km | average | 2026 |
| Paris | 4:57 | ~9.15 km | slow | 2025 |
| Chengdu | 5:00 | ~9.41 km | fast | 2026 |
| Madrid | 5:22 | ~9.15 km | average | 2025 |
| Jakarta | 6:09 | ~9.46 km | average | 2026 |
| Paris GP | 6:13 | ~9.20 km | mixed | 2026 |
| Sydney | 7:02 | ~9.45 km | fast | 2026 |
| Istanbul | 7:30 | ~9.38 km | average | 2026 |
| Hangzhou | 6:10 | ~9.42 km | slow | 2026 |
| Delhi | 8:42 | ~9.46 km | mixed | 2026 |
| London Excel | 10:00 | ~8.95 km | mixed | 2025 |
Course runs = how far you actually cover, transitions included. Sled verdicts compare each venue's sled times to its erg times. The ergs are the same machines everywhere, so they make a fair baseline.