Stroke Rate for Distance

[ukgm]

Personally, I advocated use of the ‘Stroke Index’ (SI). To calculate your SI, you merely multiply your average speed in metres per second by its average distance per stroke in metres. This goes a little way to normalising the changes you might make when paddling by producing a score that really is a near surrogate for your paddling efficiency (using the metrics we can realistically access at the moment).

Hello Bryce,
I probably need to get more detail from you on the topic.

It seems to me that the SI works only when:
1) The level of effort is constant (which I believe you included in your assumptions in your various articles)
2) Short distances where the condition do not affect the results
3) Over the last months I have steadily increased my SPM from 36 to 40 which resulted in shorter DPS. I have also reduced the surface of my paddle blades.

Looking at the SI now is bringing more confusion in my mind as it seems to revoke the idea of smaller blade, higher rate and fastest speed that I have been experiencing.

1) Yes. If the effort isn't constant, you'll get too much noise in the data and it probably won't be too much good. In other words, it's a good system for flatwater distance races but if messy and choppy stuff is your thing, your perceived and emotional experiences may well be as relevant. If you tested in messy conditions, the data is likely too poor in quality to tell you what you need. (how many runs of each paddle did you do ?)

2) Any distance is fine (I used one that had been validated in the journals that would allow as many good repeated runs as possible) but yes, you need still calm conditions to get the accuracy and precision in the results. However, this just means on a windy or more chaotic day, the benefits (or penalties) are being hidden in the noise of the tests (rather than meaning the test results don't apply at all).

3) That's pretty normal. Speeding up will shorten the stroke. However, if the SI has improved, I personally wouldn't worry about it (although i would allow the knowledge of that to shape some of the training I'd do).

[ukgm]

But seriously, I don't see how SI can work for open water, too many variables to be reliable or useful.

This is absolutely true. Don't test SI in messy conditions. You can check this by including the standard deviation as error bars on your graph of all the runs you did or calculate the coefficient of variation (CV= SD/Mean X 100). If your Cv of your tests runs (at least 4-5 I'd suggest) is more than ~6%, the data is junk.

A lot of the gains you find in the controlled conditions will translate across to any water in principle - it's just that if you try and actually test in crappy conditions, there is a good chance the data won't be any good.

The point here is:
1) Test in controlled conditions.
2) Establish the margin of error of your testing process by using several test runs.
3) Do your comparative tests of different equipment.
4) Take the gains to more chaotic conditions (but accept that not all gains will translate across - fins being one).

My test was developed to assess equipment for flatwater races.

[ukgm]

Yeah I wouldn’t worry too much about Bryce’s formulae: He would probably like SUP races to be held in an indoor tank where everything can be measured and every session is the same like this:

The SI actually isn't not mine. Its actually adapted from peer reviewed journal papers in swimming and kayaking (Larry Cain also uses it due to adaptation from his time coaching with the C1 national squad). If you test in controlled conditions, you'll get good data. The question is how much of the gain found in still conditions translates across to more chaotic ones. In some forms of equipment, its all of it. However, my articles (and journal paper on the subject) openly concede that the skills and parameters for good performance in open water racing are going likely going to be in different (or in different proportions) to those found in flatwater.

To be honest though, I've seen this level of skepticism in sport before. I saw it in cycling when wind tunnels were first used and in flume tanks with watercraft or swimmers. People didn't feel the results were transferable to the 'real world'. In all occasions, (and provided you knew the limitations of the tests and the results), they were wrong and those that didn't take measured performance seriously, fell behind those that did. If Onegiantleap had released their SUP power meter, this would all be considerably easier !

[ukgm]

The SI actually isn't not mine.

(Sorry - poor early morning grammar.  The SI is not my concept !  )

[yugi]

I’m sorry but I’m really not following the base premise and thinking on your use of SI. It is just not clicking as a base on which to do your data analysis. Maybe you could point me to some background material on SI and how it is used. On how Larry is using it for instance. I understand his thinking.

I do not believe that to go faster a paddler just spins a higher cadence like a fixie cyclist. Have you ever seen Titou accelerate? He digs deeper not faster, like a good swimmer.
 
From what I know about paddling:
 - for a given speed: a better swimmer and paddler does more distance per stroke  than a spinning thrasher.
 - cadence and speed are not correlated. (same refection as above really)

Are you checking all this stuff against yourself only? If that’s the case, seriously, rethink that. Especially since we don’t know how you paddle. Double check it on an elite paddler.
 
I don’t see how your data works to prove a better paddle or fin. I  don’t get it.

and I disagree on this:

…
 Speeding up will shorten the stroke.
 …
 

[burchas]

But seriously, I don't see how SI can work for open water, too many variables to be reliable or useful.

This is absolutely true. Don't test SI in messy conditions. You can check this by including the standard deviation as error bars on your graph of all the runs you did or calculate the coefficient of variation (CV= SD/Mean X 100). If your Cv of your tests runs (at least 4-5 I'd suggest) is more than ~6%, the data is junk.

A lot of the gains you find in the controlled conditions will translate across to any water in principle - it's just that if you try and actually test in crappy conditions, there is a good chance the data won't be any good.

The point here is:
1) Test in controlled conditions.
2) Establish the margin of error of your testing process by using several test runs.
3) Do your comparative tests of different equipment.
4) Take the gains to more chaotic conditions (but accept that not all gains will translate across - fins being one).

My test was developed to assess equipment for flatwater races.

Not at all a skeptic. I totally see and agree with the point you're making.
If I was a serious racer (or racer at all) it would be my M.O. as well no doubt.

I would even push for the water tank conditions if I could Love sport science.

[burchas]

Thanks very much!

Think I'm going to take Burchas recommendation and go down in size a bit to a mana 82, see what happens.

Thought the list of things that increase distance per stroke but were suboptimal were interesting - gliding too long, etc.  I currently use werner adjustable paddles and am really struggling to land on one paddle length - move it around depending on how I'm feeling.

I can't land on one paddle length either. Paddling in mixed conditions often presents this problem. Flat water I like my paddles 74inches. Open water 73. downwind/upwind 72.
I feel that my larger blades (85-90sqi) exacerbate the issue even further. With my ZRE 75sqi and the Mana 82 I can wing it with one size no issues.

[Area 10]

“Paddling efficiency” in the real world has to take account of the physiological impact of the stroke/cadence upon the paddler. The SI will not do that.

Then, given that the most effective stroke (and equipment) for ultra long distance will be different from that for a short sprint, it’s easy to see that while there may be some principles discovered from eg. SI when they are applied to understanding one particular format of race, once you depart from that the applicability of the what you have learned is likely to become rather sketchy.

This is not to say that the application of scientific principles to understanding the dynamics of stroke techniques is fruitless, it’s just to say that you have to be cautious about claiming how much you understand. Sports scientists are just as prone to “over-selling” as any other profession. But this certainly does not mean that it is all nonsense - and it is better than just giving up and doing whatever any fruitcake with a marketable idea wants to tell you.

[yugi]

I can safely say I don't get it.

As a data scientist I prefer to measure grin factor.

[ukgm]

1) Maybe you could point me to some background material on SI and how it is used. On how Larry is using it for instance. I understand his thinking.

2) I do not believe that to go faster a paddler just spins a higher cadence like a fixie cyclist. Have you ever seen Titou accelerate? He digs deeper not faster, like a good swimmer.
 
3) From what I know about paddling:
- for a given speed: a better swimmer and paddler does more distance per stroke  than a spinning thrasher.
- cadence and speed are not correlated. (same refection as above really)

4) Are you checking all this stuff against yourself only? If that’s the case, seriously, rethink that. Especially since we don’t know how you paddle. Double check it on an elite paddler.
 
5) I don’t see how your data works to prove a better paddle or fin. I  don’t get it.

6) and I disagree on this:

…
 Speeding up will shorten the stroke.
 …
 

Ok, bear with me and I'll work through your various points.

1) Here's one you should be able to access: https://ojs.ub.uni-konstanz.de/cpa/article/viewFile/620/545. Larry uses it exactly the same way I do as its the same formula. If you look on his blog for some of the allstar testing he did, he uses it the same way. The only difference is I learnt it via academic sources and he got it via the C1 coaching system.

2) I never said they did. SI is calculated by multiplying your average speed in metres per second for a test run by its average distance per stroke in metres. Cadence doesn't form part of the equation. However, I would add that 'in well trained' paddling athletes, in rowing and K1 in particular, cadence has been shown to correlate with speed (although companies like Vaaka overhype this). I would say we're limited as the real metrics we need, can't be measured commercially yet. Note though, if you're not a paddler with decent technique, focusing too much on cadence is an extremely bad idea.

3) You might well be right. As I said above, SI is not calculated using cadence however, its is pretty well cited that stroke length has a tendency to shorten as paddling speed is increased.

4) Yes. My article was to prove the test procedure itself was stable..... which it was..... to near lab levels of variation. However, if you are trying to see if an innovation generally is faster overall, you'll need a test pool of willing subjects.

5) Have you read my two web articles on this ? I would post the links on here but last time I did admin deleted it. I will post the link to the journal paper when it comes out (its in press at the moment but approved for print).

6) See 3). Its hard for a paddler to isolate stroke rate from stroke length. This is easy to measure. Put tape markers on your board to act as targets and then do a series of runs at cadences from 30spm upto 70spm in 10spm increments and let us know what a speedcoach (or similar) says about your stroke length. It'll tell you your stroke generally shortens as your cadence rises (provided the goal is you increasing your overall speed).

[ukgm]

1) “Paddling efficiency” in the real world has to take account of the physiological impact of the stroke/cadence upon the paddler. The SI will not do that.

2) Then, given that the most effective stroke (and equipment) for ultra long distance will be different from that for a short sprint, it’s easy to see that while there may be some principles discovered from eg. SI when they are applied to understanding one particular format of race, once you depart from that the applicability of the what you have learned is likely to become rather sketchy.

3) This is not to say that the application of scientific principles to understanding the dynamics of stroke techniques is fruitless, it’s just to say that you have to be cautious about claiming how much you understand. Sports scientists are just as prone to “over-selling” as any other profession. But this certainly does not mean that it is all nonsense - and it is better than just giving up and doing whatever any fruitcake with a marketable idea wants to tell you.

1) Agreed, no it won't (or at least, only partially). But until we get to the point when such metrics can be measured, we have to use surrogates that will only go partway there. Even traditional faves like V02 max are now being reined in that there it is accepted that there is a lot more to elite level performance and selection than its use (particularly as more than a few elite endurance athletes don't always have spectacular levels of Vo2 max). We both know that as long as you concede/state the limitations, it has a purpose if used correctly and within its intended context. I would also add that tailoring and designing  paddlers training to such things also helps adaptation.

2) Agreed and I concede as much in my articles. I would say that SI is fine for determining the benefits of equipment in long or short events but you then must make sure you tailor the test intervals accordingly. The paddling speed is key for such tests. That's why I tailor my tests to the speed i'm going to be racing at (9.5kph). This does make the assumption that I'm substituting flatwater event distance for speed generally (as the drag and technique alter with paddling effort or board speed).

3) I know that you know that we both know that. 

[PonoBill]

I think I know why the SUP paddle power meter was never released (beyond limited market) and how to fix the issue. The easy way to add a power meter to a paddle would be a strain gauge on the shaft. I did that with my paddle pod and never got values that seemed usable. Over long-ish distances attempting to measure the force while paddling with consistent effort applied to the shaft the strain gauge values fluctuated wildly. The relatively consistent accelerometer values and vibration curves didn't show any useful correlation to strain. I only measured the acceleration of the paddle blade, not the board, since what I was trying to measure was relative performance of the blades.

I tried various mounting schemes, including increasing the distance of the flex measurement by fixing one end of the strain gauge to the shaft and the other to a six-inch strap of carbon fixed further down the shaft. Measuring strain over a longer distance didn't help, the output still looks like noise. I don't have a theory for why, and I gave up quickly because I had a lot of tests to run.

While it would certainly be nice to precisely measure the input force, the resultant power is pretty easy to get at, especially if all you want is a relative value. Board acceleration is directly related to force. Of course the higher the speed is, the less efficiently the force applied is converted to acceleration since drag becomes greater, rising exponentially when wave drag becomes dominant, so you need an accurate speed measurement, some calibration runs and a bit of math, but it's feasible, and not even particularly expensive to build. There's no useful market since it can't be universally calibrated--rider and board weight, stroke effects, current, wind, etc. will screw up any direct reading. I used a dragged free-spinning prop to measure board speed. I doubt that would be popular since it was a huge pain in the ass, but actually, a Doppler-based speed reading would be accurate enough if the sample rate was high enough.

A precise speed reading with a high sample rate is all we need for measuring acceleration, and that by itself is a revelation. Any notion you might have of the glide of a SUP goes away quickly. So does any notion of adding power by pulling near your feet. Glide is only significant when the board has slowed enough for the drag curve to be flat. In other words, much slower than anyone wants to go. You don't need equipment to verify this. Get up to speed and stop paddling. The board glides along, but pay close attention to the speed. You'll have to do this visually, I haven't seen any GPS-based speed device that updates often enough to be useful. Initial slowing is extremely rapid (remember that exponential drag), the overwhelming majority of the glide is at roughly 1/4 top speed.

The reason why higher cadence results in higher speed isn't exactly what first comes to mind--a simple blade speed multiplied by rate idea. In part it's because the deceleration curve is very steep. With no practical glide all the deceleration needs to be made up with acceleration against high drag. With a slower cadence you're catching the curve at a lower speed and there's a lot more area under that curve (power required) than there is if you catch it sooner.

Quick recovery in the air and pulling the blade out of the water as soon as you can no longer effectively add force just means you don't have to add as much power to maintain the speed. Apparently the penalty of not doing that can be overcome with conditioning and other stroke efficiencies because there are some very fast paddlers that have relatively slow strokes, but it DOES need to be overcome.

I think some of the six-man outrigger canoe notions that get applied to SUP paddling don't really work. "Long in the water" is probably fine when your craft weighs 300 pounds and has 800 pounds of paddlers in it. A single paddler isn't going to accelerate that mass, so it's much more important to match blade speed to boat speed. But a 30 pound board and 165 pound rider is a substantially different business. Deceleration and acceleration is rapid, significant with every stroke, and directly related to what you do with the paddle.

[yugi]

…
3) From what I know about paddling:
- for a given speed: a better swimmer and paddler does more distance per stroke  than a spinning thrasher.
- cadence and speed are not correlated. (same refection as above really)
 …
 

Ok, bear with me and I'll work through your various points.
…
3) You might well be right. As I said above, SI is not calculated using cadence however, its is pretty well cited that stroke length has a tendency to shorten as paddling speed is increased.
…

^ for context

3) You might well be right. As I said above, SI is not calculated using cadence however, its is pretty well cited that stroke length has a tendency to shorten as paddling speed is increased.
…

 

“Stroke Length” being "distance covered per stroke" I presume. [referring to the article you posted]

Let me just check I’m getting this right. You are pretending that as speed of the SUP’er increases the "stroke length" is decreasing?

 

I beg to differ. We seem to not be on the same page. I do not understand how that is remotely possible. Something is amiss.

[Luc Benac]

The reason why higher cadence results in higher speed .... in part it's because the deceleration curve is very steep. With no practical glide all the deceleration needs to be made up with acceleration against high drag. With a slower cadence you're catching the curve at a lower speed and there's a lot more area under that curve (power required) than there is if you catch it sooner.

I can relate to that. While a slower cadence feels more rewarding as you feel the board accelerating again, GPS seems to say that a shorter stroke, faster rate keeps the board at speed more consistently even if you do not have the same feeling of achievement. At least within reason.

[yugi]

I think I know why the SUP paddle power meter was never released (beyond limited market) and how to fix the issue. The easy way to add a power meter to a paddle would be a strain gauge on the shaft. I did that with my paddle pod and never got values that seemed usable. Over long-ish distances attempting to measure the force while paddling with consistent effort applied to the shaft the strain gauge values fluctuated wildly. The relatively consistent accelerometer values and vibration curves didn't show any useful correlation to strain. I only measured the acceleration of the paddle blade, not the board, since what I was trying to measure was relative performance of the blades.

...

I bet that a good measure of paddler efficiency would be speed divided by the force exerted on the shaft.

There is hope for your "power meter" . It is just better applied at teaching efficient paddle stroke rather than testing paddle blades. You should market it!

Said in another way: I bet Travis or Tituan cover a lot more ground than other paddlers for a given amount of force applied to the paddle shaft.

It’s all about how efficiently you can get the board to scoot across the water. Any sailor or swimmer knows this.

I stopped by a motocross enduro event last weekend. Just like a swim or SUP event the first motocross racer was so much faster and so much smoother. I bet he, the winner, spent a lot less fuel than those behind him [given their motorcycles had about fuel efficiency bikes]. He  flew over the course with the greatest of ease. He never sounded like he was throttling it like other racers were and yet he flew so much faster out of corners and over obstacles.

Which is pretty much the point I object to in all the above mentioned [so called] “scientific data analysis”. It is missing the basic point.

[disclaimer: I’m an engineer]