I mentioned many time the best rolling plots are my favorite feature of ConnectStats. After adding a new ability to visually see when you reached specific best, I have now a first version of a feature I have been planning to do for a while now: time analysis of specific slice of your best achievement.
So today, I submitted to Apple for release a first version of this feature in ConnectStats 6.1. The best rolling curve shows you what best power, heart rate or pace you reached for a given time or distance. You could previously see what that curve looked like for a month or year. Now you can see how specific point evolved over time.
This screen is displayed at the bottom of the summary page. You can configure it by doing a long press on the graph.
You can display the max over a rolling period: What was the max 1 minute power reached over a rolling 2w period as above, or the trend over the last few weeks.
The intention is to see how you pushed yourself over time or improved. In my case above you can see that since May the max power of 30min has increased slightly but around late feb/early march I had a peak power. You can see in the following demonstration video how this and other related features can help you understand where the peak came from.
It can be interesting to try to understand what each of the graph of the summary tells you. This is my graphs for running over the last few months.
The fatigue/fitness graph tells me that I have been increasing my training intensity quite a bit since march. It makes sense as since the lock down I have switch to working from home, which means less lower intensity bike commutes and more runs in the morning. Around may I reduced slightly the intensity to rest, but maintained the fitness. The best Average power tells me that over the last few months I have been quite constant in how I pushed myself, but I am quite lower than what I achieved last year. And the new graph, shows that even though last few months were quite consistent there is some increase on the 30min power, I definitely pushed myself more recently…
I have a few more improvements to do on the new graphs, but I wanted to release a first version to get early feedback.
I have struggled to decide if looking at trend was more interesting or looking at the max/best over a period. Feel free to comment, you can change between the two in the details screen.
I will add the ability to change the points you look at in the history (currently only 1min and 30min). And I’ll probably add the ability to remember the preferred configuration for each user.
Not sure I understand what this means! My 30 minute line increases, I think that’s because I did a few 10ks and cut down on the easy runs. But I’m guessing if I stick to a 80/20 strategy the line should be flat?
Your power line increases? Currently by default it shows the max, so yes, if it increases it means you have higher max. Though having more easy run should not affect it. If you switch to trend (currently you have to do it manually), the line then would be more affected by a change in 80/20 or other strategy. I am quite debating what is more useful to look at, the max or the trend. let me know what you think?
Hi Brice, I really like the power-duration curves you’ve built into the latest versions of the app. I think these, with the month indications, give a good indication of trend – they tell me if I’m improving in short or longer distances for example. The new chart showing 1 and 30 min max power over time seems harder to interpret – I’d be more interested in a „known“ metric like CP/FTP or RSS plotted in time. There’s a long discussion on FB about stryds new analysis of fatigue resistance with the same problem – too hard to understand what influences it…
Thanks a lot for the feedback.
I just went to look at FTP/CP. It is typically calculated by the the maximum power you can maintain over 45 min or 60min. So I think that last plot is basically your “empirical/realised” FTP/CP plotted over time but for 30min and 1min… I will provide an ability to change the time frame it’s calculated instead of hard coded 1min/30min, so people can switch to 45min or other standard/arbitrary period.
I think I will rename the plot to CP 30min. That is a more accurate description of the calculation actually. I just need to make sure there is no copyright/trademark on the name… But looks like it’s just a generic term.