TRIMPS to estimate TSS

You can use Training Load plugin in SportTracks to do it.

Indeed for me, I find an HR based mechanism of tracking Training Stress and identifying when I will perform well, and when I'm too fatigued / under trained to do well more accurate than a power based one.

see
http://www.tritalk.co.uk/forums/viewtopic.php?t=67118

doyler78 wrote:

So what I'm looking for is a TRIMPs calculator that will analyse my polar hr data and give me an estimated TSS. I thought that WKO+ done that automatically if power data wasn't available but if it does I can't find it.

I would suggest that rather than go by HR to get a TSS, one is simply better off assigning an estimated IF for the ride and calculating TSS from that. But once you get past about 10% of rides being manually estimated, then you potentially introduce some error into the Performance Manager that may provide some misleading indicators.

http://cyclecoach.com/forum/viewtopic.php?p=125#p125

jibberjim wrote:

You can use Training Load plugin in SportTracks to do it.

Indeed for me, I find an HR based mechanism of tracking Training Stress and identifying when I will perform well, and when I'm too fatigued / under trained to do well more accurate than a power based one.

see
http://www.tritalk.co.uk/forums/viewtopic.php?t=67118

Perhaps you are not using the power based methodology well? Although you are trying to apply the impulse response model across more than one sport (and only cycling enables the use of power), and that in itself I think is problematic and needs to be thought through some more. Certainly Phil Skiba has given it plenty of thought.

A HR based metric significantly undervalues hard efforts, since by definition HR can't extend past a maximum value. HR can also significantly overvalue low efforts (since when we stop riding hard, HR does not drop away in the way power does). I'm not sure these two wrongs make a right.

I don't agree that there is a "strong correlation" between HR and intensity. I would say there is a good correlation but not necessarily a strong one. The correlation is higher indoors and/or in a controlled setting but once the efforts become quite variable (like they do when training outdoors) and/or involve hard efforts, the correlation falls away significantly.

Your example of a 30-min ride break overly inflating TSS would suggest to me that the ride was in fact two rides and not one (a break long enough to have a beer and eat a burrito). What happens when you create two rides with your data?

I also note the comment about not being concerned about the absolute accuracy of the TSS (well FTP really) and it's more about the overall trends/shapes. While that's certainly true to some extent, I think once you start to examine multiple seasons of data for an athlete, you'll realise the accuracy of having a pretty good handle on FTP makes a difference in what you can learn about training impulse/response (especially when an athlete is training significantly).

So rather than fudge it, just get good data to start with. It's not hard to do.

All up though, these are not major issues (only interesting discussion points) , as for the majority of first time users the charts simply reveal the bleeding obvious - i.e. they don't train enough, or with sufficient consistency, or try to do too much in too short a time span and suffer the consequences.