Friday, August 24, 2012

Number Crunching

Robert Osfield left a comment: (I hope you don't mind, Robert)
I'm curious, how does your pace/HR/Calories per mile match to what it was like before Connemara?
I do wonder if you might have enough data to do a similar analysis, do you log your HR, Calories consume, mileage, elevation/descents for each run? It might be interesting to plot the trends and any correlations to race performance that might spring up.

I do have my data of the last few months available on Garmin Connect, so I guess the data would be there, at least for a few months. I have data going back years in a different format, but that's harder to access. To be honest, I wouldn't know where to start with a detailed analysis. I do keep a very simple spreadsheet where I track my pace/HR correlation. I deliberately kept it very simple, ignoring even things like elevation or distance. What I can see are some general trends and the fact that my numbers have never been as good as right now except for February/March this year when I was nearing peak fitness for Connemara. I have never seen figures like that in base training and that's why I think my numbers are very good at the moment, assuming my HRM was working ok (which may or may not be the case).

Talking about the HRM, I do believe most numbers are ok but the numbers of the last 3 days were definitely wrong. It might be just another dead battery, but since the thing is definitely reaching the end of its useful life (it has a crack in it), I have ordered a replacement. I'll see what the figures will be like when that turns up.

I went back on the Kerry Way on Wednesday. One thing I noticed is that I automatically run faster on the rough terrain when I am wearing my glasses, especially on the steep downhills. That's another lesson learned (maybe I'll eventually turn into a real trail runner [no, I don't think so either]). I can feel definitely feel the leg strength coming back. It stood me well in Bangor, hopefully it will go as well in Dingle.

I did my last real workout before Dingle this morning, a repeat of Sunday's 10 miles with 7 at tempo effort. During the week I thought I had gotten the better of the Achilles problem in my left leg by doing eccentric calf raises, but the discomfort returned when running in my Lunaracers. That obviously puts the use of those shoes for Dingle in severe doubt. I will give them one more go and if the result is the same then I'll wear the same pair of Green Silence that got me through Bangor instead. They are nice, comfortable and light as well and would do just fine.

The tempo run itself went a lot better than on Sunday. I was already doing a decent enough pace when I spotted two girls ahead of me in the distance and spent the next mile catching up, which gave me a significantly faster average pace. Funny that. Just don't tell Niamh.

22 Aug
12.25 miles, 1:51:30, 9:06 pace
   Kerry Way, Windy Gap x 2
23 Aug
5 miles, ~40:00, ~8:00 pace
   no Garmin (!)
24 Aug
10 miles, 1:08:53, 6:53 pace
   incl. 7 miles @ 6:38 pace


  1. Hi Thomas,

    I have software I wrote that keeps track of YPB (yards-per-beat of the heart)
    which you track, even if you don't call it that 8) (measure of cardiac efficiency)

    If there was a way I could get access to your tcx files I could look at your data any way you like.

    If you are a tech-y type I could send you my stuff (python) and you could do this crunching yourself....??


  2. I have that problem too - but I never catch them!

    Great sign if you're near your best at HR/pace. The 100k training & race has given you a boost. One thing I've found (that might not affect you for another 10 years) is that as max HR declines with age, HR/pace numbers need to be better in order to run the same times in races you might have done 5+ years ago. Pretty obvious really.

  3. It would be interesting to check your Hr data around the time you spotted the two girls ;-) All the best in Dingle. Your prep has gone very well and i'm expecting great things.

  4. Thanks for posting the spreadsheet Thomas. How do you compute VDOT?

    I had a bash at adding a graph for each month of the HR to speed and this worked great on my local copy of the spreedsheet with LibreOffice but when I uploaded it to googledocs I lost the graph, and as yet still haven't leant to use googledocs so haven't been able to add them back..

    I find the the scatter and slope of the HR to speed graph useful in gauging how my aerobic and anaerobic fitness is doing, in particular I use the speed at a HR of 140 and HR of 160 to give me a means of comparing how my lower to upper aerobic range is doing.

    Looking at these figures it looks like February and March looked best, and this month it does look like things are improving back to where you were. It would be interesting to do a quicker flat run to see how your speed is at your upper aerobic levels, but given your are supposed to be tapering I guess you'll need to pass on this.

    I need to publish my own results - currently busy doing DIY and getting on with work so tinkering on my blog is lower priority. However, I did acquire another useful datapoint at the weekend - I did a PB in Killin 10k, I'm now a sub 40 min 10k runner :-)

    My interest in all the analysis is work out what types of training work for me, and what paces are reasonable to aim for when running events based on what the training suggests. This worked out for me this weekend, next weekend I'm thinking about running the Aberfeldy marathon so could have another datapoint to calibrate things by soon.

    1. Robert, I have a spreadsheet that translates HR and pace into a "VDOT". It has to be noted, though, that this does not match the VDOT race predictor numbers. My VDOT for races is always a good few points slower.

    2. What is the equation you use? I did a search on the web a got references to a VDOT that relates to velocity of VO2Max. For instance a VDOT calculator:

      Does this have any baring on the VDOT that you are calculating?

      As for races diverging from training efficiencies - in my 10k at the weekend I averaged a HR of 178 which is more than I'd usually expect when doing a tempo run or even a 10k race, and looking at my graph of HR to pace their is dip to the 10k race results. This suggests that my HR was elevated and theoretical efficiency reduced due just being in race, I'd guess it's down to adrenalin. This result is consistent with your own VDOT results.

    3. One possibel spreadsheet is here:

      "Race VDOT" is lower because the adrenaline causes a higher heart rate and because the Garmin tends to over-estimate distances in training. However, as long as the VDOT figures are consistent, that's good enough to spot trends and developments.

  5. I really like the Green Silence shoes too :0]
    Be careful with your tendon injury, you don't want to be sidelined for two months like I was at the start of the year!