Watts versus split: why you take a percentage of the watts

Concept2 publishes the conversion its monitors use: watts = 2.80 divided by pace cubed, where pace is the time in seconds over the distance in metres [1]. On the rower and SkiErg the monitor shows pace per 500 m, so a split of 2:05 per 500 m works out at 179.2 watts, which is Concept2's own worked example [1].

The cube is the whole story. Speed and power are linked by the cube, so the same step in speed costs more power the faster you already are. Everything below comes from that one line, and every split used here is an example, not a target.

Take the same ten seconds off your split at two different speeds. From 2:10 to 2:00 asks for 43.2 more watts. From 2:00 to 1:50 asks for 60.4 more. The second ten seconds is a bigger ask in power, even though it is the same on the clock.

The same is true one second at a time. Going from 2:30 to 2:29 adds about 2.0% to your power. From 2:00 to 1:59 it is about 2.5%, and from 1:45 to 1:44 about 2.9%. One second gets more expensive as you get quicker.

Concept2 makes the same point with its own example. Improving from a 2:30 split to a 2:00 split takes roughly 100 more watts, about 104 W up to 203 W. Going from 2:00 to 1:45, half the time saved, takes a nearly identical extra 100 W, from 203 W to 302 W [2].

Run it the other way round. A split of 2:00.5 is 200 watts. Double the power to 400 watts and the split is 1:35.6. That is about 21% off the time (the split is 2^(-1/3), or 0.794, of what it was), not half.

Ten percent faster is the same idea. Going from 2:00.0 (202.5 W) to 1:49.1 (269.6 W) is 10% faster, and it costs 33% more power, because 1.1 cubed is 1.331.

When a session is written as a share of your effort, such as 80%, the share belongs to the power, not to the split. Say your base is a 2:00 split, which is 202.5 watts. Eighty percent of that is 162.0 watts, which is a split of 2:09.3.

The tempting shortcut is to add 20% to the split instead. Twenty percent on a 2:00 split is 2:24.0, which is only 117.2 watts. That is 57.9% of your power, far lighter than the 80% that was asked for.

A study of eight well-trained rowers on a Concept2 found the same relationship from the other side. Changes in time to complete a test were approximately one-third those of mean power, apparently because simulated velocity is proportional to the cube root of power on this ergometer [3].

In the post's example, a 3% rise in power is a speed change of about 1%. Read a small change in your split as a bigger change in the work behind it.

The maths here is for the RowErg and SkiErg, which show pace per 500 m. The BikeErg shows pace per 1000 m. Concept2 sets it so the wattage needed for a pace per 500 m on the rower is the wattage needed for that pace per 1000 m on the BikeErg [4]. We do not claim anything about bike splits beyond that.

Watts also make a shared language across machines, but for your own progress compare ski with ski and row with row. That last line is our coaching line, not Concept2's.