The curve here is Kurt Jensen’s, built from the power profiles of elite rowers, whose power over 10 seconds, a minute, 6k and the hour holds near-constant ratios to their 2k. Those ratios put every max-effort piece you row on one scale: the 2k it would be worth to an athlete with that curve. They are an average, and your own curve has its own shape, so read the number as a yardstick rather than a prediction of your 2k. Connect your data and the band with the lowest 2k-equivalent gives insight into the energy system to train in order to improve your 2k.
| Piece | Split | Watts | Band | 2k-equivalent | |
|---|---|---|---|---|---|
| 20s inside a 6,000m rowdiscovered | 1:28.0 | 514 | 10 sec | 6:56.6 | |
| 500m | 1:31.4 | 458 | 60 sec | 6:52.0 | |
| 1k | 1:35.3 | 404 | 2k | 6:44.0 | |
| 2k test | 1:43.5 | 316 | 2k | 6:54.0 | |
| 5k | 1:45.5 | 298 | 6k | 6:42.9 | |
| 6k test | 1:47.5 | 281 | 6k | 6:47.5 | |
| 60 min | 1:56.1 | 223 | 60 min | 7:03.9 |
Kurt Jensen found that elite rowers hold a near-constant ratio between their 2k power and their power over 10s, 60s, 6k and the hour. Running that backwards gives any piece a 2k-equivalent: the 2k it would belong to on that curve. It is a unit of comparison rather than a prediction. The ratios are an average of elite profiles rather than anyone’s actual curve, so a 10s and an hour from the same week will rarely agree, and what you would actually pull for a 2k is its own test. The disagreement is what the page is for: the band with the lowest 2k-equivalent sits furthest off the elite balance, which makes it the strongest indicator here of what might train a faster 2k. That is the band the page calls your limiter.
The limiter is the band giving away the most watts against your own ceiling, so it is where the most is available — not where you are worst in absolute terms. A short band lagging points at the anaerobic end: starts, sprints, strength work. The hour lagging points at aerobic base, which is volume at low rate rather than harder pieces. The middle lagging while both ends hold up points at threshold and race-specific work.
Two things it cannot tell you. An untested band looks exactly like a weak one — if you have never rowed an hour flat out, that band is empty rather than slow, and what it is asking for is a test rather than a training block. And nothing here weighs recency: a band propped up by one piece from last winter is a claim about last winter, so check the dates in the table before rebuilding a season around it.
Jensen makes the case himself in World Rowing’s Using the indoor rower for testing: why do we test different distances? — the reason to test at several distances is to find out whether a 2k rests on aerobic power or anaerobic power, because that is what decides which gap is worth training.
Jensen published five ratios and the longest of them is an hour, so everything to the right of the 60 min mark is the curve guessing. It guesses with a power law — watts falling as a fixed power of duration — because that is what Jensen’s model already is. Drake, Finke and Ferguson showed that his Golden Standard, along with cycling’s FTP, Daniels’ VDOT and Riegel’s race calculator, are all the same power law wearing different clothes, each with its rate of decay fixed in advance rather than measured. Their companion paper, which is open access, argues the same law beats the textbook critical-power model across a far wider span of durations.
Measured rather than assumed, on the same rowing data Drake, Finke and Ferguson used: 45,245 bests self-reported to nonathlon.com, in the exact copy those authors published alongside their code. The fit reads the 4,630 of them that run longer than an hour and sit alongside the same athlete’s own 60 min, and asks how much further the power has fallen. It falls as duration−0.18 — a Riegel fatigue factor of 1.064 on pace. Note that it steepens past the hour rather than levelling off, which is the opposite of what a curve drawn through Jensen’s last two anchors would suggest.
Treat the dashed marks as the weakest numbers on the chart. Jensen tested elite rowers, so the curve runs optimistic for the rest of us at every duration; the exponent is set so that extrapolating does not make that any worse than it already is at the hour.