Planning and Tracking Features

[Robert BB]

Hi All,

I've seen reference to v4.0 in a few places, which were said to include planning and tracking tools. Doesn't seem to be much since 2016 though - just wondering if this is still in the works and if so if there is an ETA?

Cheers,

Robert.

[Mark Liversedge]

On Wednesday, 16 May 2018 08:36:10 UTC+1, Robert BB wrote:

I've seen reference to v4.0 in a few places, which were said to include planning and tracking tools. Doesn't seem to be much since 2016 though - just wondering if this is still in the works and if so if there is an ETA?

Yes.

The version 4.0 monicker was dropped, we will keep with the current numbering, so the next release in July is 3.5, its pretty much done apart from the Overview page I need to complete.

Version 3.6 development will start once thats done. On a personal note I will ONLY be working on Planning, I keep getting sidetracked onto things that interest me, but don't add a lot of value and Planning drifts further and further away.

I want to get away from developing a glorified calendar function (ok, so sometimes they have load planning and periodization). I want to use models to help, that means something like banister, with some machine learning and making it adapt to reality (or at least feedback to the user) in a closed-loop manner. I want to incorporate recovery planning, injury prevention, athlete feedback (RPE/ROF) as well as goal based optimization. Goodness knows how long it will take. It will need collaboration from ML and Sports Phys types and a lot of testing, fortunately we have some friends with the right skills, but always open to expert advice. What we build will likely look completely different when we get there !

One of the motivations behind the recent OpenData initiative was to collect enough data to help develop and validate the models (we already have 260 athletes training history, 180,000 workouts -- I need to justify that generosity and develop something !). We might have to mandate regular testing for the modeling to work. Its best practice, but not neccessarily common practice with self-coached. We will also need to do some educational videos and articles to explain how it works etc.

Mark

[Arnaud LECLÈRE]

Dear Mark,

Does the work progress as you have planned?

Does the amount of data (Athletes and Workouts) of the OpenData initiative validate your model as expected?

Do you have better visibility about the availability of these new tools? (ETA)

Thank you for the wonderful job.

BRegards,

Arnaud

[Mark Liversedge]

On Friday, 14 September 2018 08:48:15 UTC+1, Arnaud LECLÈRE wrote:

Does the work progress as you have planned?

3.5 burn down for release has been slow, this is mostly due to rest of life issues.

Does the amount of data (Athletes and Workouts) of the OpenData initiative validate your model as expected?

We have lots of data now, 500k workouts, 900+ athletes but the analysis is only just starting.

WRT any model, it is likely to start with banister using classic model fitting techniques and mature into NN.

Do you have better visibility about the availability of these new tools? (ETA)

Likely next year, possibly start to see functionality over Christmas.

Thank you for the wonderful job.

You're welcome.

[Arnaud LECLÈRE]

OK perfect

Good continuation

Arnaud