Goodbye long intervals – 4x2km OTW, plus some info on

The training

Last time I did this workout, 4x2km with 5 minutes rest, was on April 29th. Then I managed a 2:12.2 average, so today I was shooting for “faster than 2:12”.

In line with my plan, this would be the final “long interval” session of this OTW season. After today, it’s “so long, long intervals” until some time in September. 🙂

As I have a day off, I also wanted to take the time and experiment with the footstretcher location. After watching my video, several people suggested to move the footstretcher slightly sternward. So I did, but not immediately. The plan was to row the first two 2kms with the old rigging, and then change. I would take a bit more time between intervals #2 and #3 to move the stretcher and to get a bit more rest, so that hopefully fatigue would not influence the experimental results too much.

Wind was between 1 and 2 m/s varying between North and NorthWest, according to SportTracks, and it was between 2.5 m/s and 3.5 m/s NNW according to WindFinder’s prediction. On the water, I definitely noticed that the wind was growing. Between the tailwind intervals #1 and #3 there was definitely a stronger tailwind in #3, but also the water changed from “ripples” to “small chop”, and I wasn’t faster in #3 (with the new footstretcher setting). For the headwind intervals #2 and #4, pace went down by 3 seconds (from 2:09 to 2:12) with #4 being the slowest interval. Definitely a interval #4 was the hardest battle seeing the slowest splits at some points. I also noticed that interval #4 got quite sloppy in one point. When I realized it, with about 500m to go, I started to focus more on technique, which gave me a few seconds of split in the right direction (it could also have been the effect of a slight turn in the course, which made the wind less head-on and more cross-wind).

More about wind in the second half of this blog. First, the plots:

Stroke length info
Stroke length info. Did my strokes get shorter with the new rig? Or is it the effect of fatigue?

And the stats:

Workout Summary - media/20160628-105331-2016-06-28-0805.CSV
Workout Details
01|02980| 15:08 |02:31.5|19.6|131.0|161.0|10.1
02|02000| 08:06 |02:01.5|27.0|165.0|173.0|09.1 | tailwind standing start
03|02000| 08:35 |02:08.9|27.2|172.0|177.0|08.5 | headwind rolling start
04|00781| 04:30 |02:36.2|19.9|134.0|144.0|09.6 | tailwind standing start, footstretcher moved
05|02000| 08:06 |02:01.7|28.2|169.0|176.0|08.7 | headwind rolling start
06|02000| 08:49 |02:12.5|27.4|171.0|177.0|08.3
07|01767| 09:25 |02:40.1|18.1|132.0|139.0|10.3

You can see more on the workout page or the interactive plot on

After the session I did some rigging measurements, just to make sure I had all the data, and also to compare with on-line rigging charts for single sculls. The problem of course was that I never measured “work-through” myself, so I measured this differently than the tables. What I measured:

  • Measured distance between stern (“front”) side of front stops and the line connecting the oarlock pins: 18.0cm, with front stops being more sternward.
  • Measured distance between “ball of foot” location on footstretcher and the line connecting the oarlock pins: 57.8cm

What people measure:

  • Distance between knee and the line connecting the oarlock pins at “half stroke” (between -2cm and +2cm recommended)
  • Measure hip joint position vs that same line at catch position

I am actually not sure how people measure this without help, and I am not sure how accurate these measurements are … OK, will do more measurements tomorrow. Also, I forgot to measure footstretcher height and angle.

I will keep the footstretcher at the new position for a while. Boat didn’t feel significantly heavier at the catch.

Uploading trainings to Concept2 and comparing trainings

Regular readers know that I have been building a website to support creating those colorful plots that I show in my blogs. The site is at There has been quite an evolution over the past few weeks, probably because I put a lot of my free time into it. So here is an overview of the functionality:

  • Upload data files for OTE and OTW rows (Painsled, ErgData, ErgStick, CrewNerd, RiM, SpeedCoach) and make plot image files or use interactive plots
  • Synchronize your data (including heart rate, stroke rate, pace vs distance and time) to the Concept2 logbook
  • Import your data from the Concept2 logbook for advanced plotting and sharing. Concept2 doesn’t allow to see the rows of other people. allows you to overcome that.
  • Synchronize (import and export) your data with Strava and SportTracks
  • Comparing workouts

Actually, everything I do on this blog is now going through that site. Bugs are starting to get relatively minor and seldom, so I encourage you to take a look, register, and try it out. Comments here or on

That last functionality, comparing workouts, is relatively new, but quite interesting. Here are a few example plots for today’s workout vs the same workout done on April 29th:


The blue line is today’s workout. The red line is the same workout on April 29th.
The blue line is today's workout. The red line is the same workout on April 29th.
The blue line is today’s workout. The red line is the same workout on April 29th.
The blue line is today’s workout. The red line is the same workout on April 29th.

The distance axis is not completely similar between the two workouts, because I took more time and distance for the warming up. Still it is clear that I managed significantly higher stroke rates and faster paces today. I call that progress (and perhaps the effect of the low i sculls). It cannot be only due to the warming up of the water …

Geeky stuff

Another thing that I am working on is correction for wind. The trouble is of course to get good wind data. Must develop a portable data logging wind meter for rowers …

Using the GPS track (and some data smoothing) you can calculate the boat’s bearing, and thus its angle with respect to the wind. Decomposing that wind into head/tailwind and crosswind components is then child’s play. Here’s my track of today:

4x2km OTW 6-28-2016
Today’s track. The start of the tailwind 2ks are at the north-northwest end of the straight legs. The headwind 2ks start at the bottom right (near “Rakovecka” street) and take a slight turn towards NW.

Here is the wind according to SportTracks, taking data from nearby amateur weather stations:

Wind speed and direction vs distance. North is 0/360. South is 90. East is 45 degrees. The black line is the head/tailwind for my scull, using the bearing calculated from the GPS course

Taking the prediction from WindFinder, the same plot looks like this:


Both plots were generated directly on using the wind editor. I haven’t implemented wind correction on the site yet, but I have it working on the desktop. Please don’t look at the “Erg Power” and “Erg Pace” lines. They are off and I know it. That’s the next step. As the whole calculation takes about 10 minutes, I need to program this as a “background task”, enabling the users to come back to the calculations once they are done (and sending them a message when that is the case).

Using SportTracks data
Same row but using WindFinder data for the correction. Calculation failed on the headwind sections
Same row but using WindFinder data for the correction. Calculation failed on the headwind sections

As you can see from the second plot, the calculation sometimes fails. Need to look into this a bit more.

Comparing the tailwind section’s comparisons, I think the truth is somewhere between the two plots.