So those are screenshots of my little ongoing program. My program works with splits. So first screenshot called Query says to pull all activities that have splits of 2000 meters, plus or minus 2m, and color the splits in yellow when I select the activity.

I then show 4 results that were all done on the indoor track in the last 4 weeks. The result of 2020-02-25, there's actually a third split that is not in yellow because I got side stitches, and quit before the end of the 2000, therefore it doesn't fit the 2000 +/- 2 criteria.

First thing I noticed, is how the air power was not consistent (orange box). Whereas all those splits are pretty much at the same pace (sorry for the pace in km, for reference 4:00 / km is equal to 6:26 per mile), the air power goes from 4.8 to as high as 11.5 watts on different days. I guess the sensor in not all that precise, because it is an indoor track. No wind.

So knowing in how in the new Stryd, they just added the calculated air power to the power and made that the new power, I calculated a running effectiveness with full power (called RE) and also one in which I substracted air power from total power, and called that RE no Air (dark blue box), and I'm using that for analysis. You may agree with this or not, but that's what I'm doing. Actually, since both values are there, I look at them both

On 2020-02-18, I was not thinking about cadence. The activity is there for reference. You can see that I increased slightly the pace from splits 1 to 3, cadence (black box) increased with pace, ground time (green box) decreased with increased pace, and vertical oscillation (light blue box) decreased with increased pace. Running Efficiency No Air (dark blue) increased with pace. You can also note that RE (with air) actually decreased when going faster, which is consistent with Air Power (ornage box) increasing.

On 2020-02-25, I was also not thinking about cadence. Remember that you can look at a third split in this activity that is not in yellow. This one is also there for reference. Air Power values are not consistent with the activity from the week before, and many of the hasty conclusions we could have made by just looking at the other activity become kind of muddled here.

On 2020-03-03, I decreased my cadence on the third split. With Air Power values that are not consistent with either of the two previous activities, this is where I decided to add a RE with no air value and work with that.

I'll note that RE is actually your speed in meters/second divided by your power to mass ratio. In my case my weight has not changed, so you can look at it that the faster (increased speed) you go for an amount of power, the higher your RE will be. In other words, the higher your RE, the more efficient you are.

It's actually hard to lower or increase cadence over 2 kilometers during a workout that is hard, so maybe I should do these tests at maybe MP, where it would less of a struggle to just keep up with pace.

Anyhow, in the third split, you can see my cadence at 180, which is the lowest in all 4 activities, and my RE dropped a little bit. It is a very small drop. Vertical oscillation has increased in that third split, and Ground Time has not changed enough to conclude.

On 2020-03-10, I increased my cadence on the third split. Vertical oscillation became lower, Ground Time is not very conclusive, and RE increases a bit.

Overall, I still find it hard to draw hard conclusions. The only one that seems obvious to me is that cadence is inversely proportional to vertical oscillation. And that seems to have an effect on RE.

But Stryd doesn't measure Power, it calculates it from other values. So by changing variables, it's necessarily going to affect power, so I'm not sure I can conclude to anything on RE by changing how I run using Stryd, given how power is kinda calculated from measurements of how I run.

So, for now, I put in way more work into this that what I'm getting back. Oh well. I'm a geek. Gotta own it.