Graham O’Grady and Candice Hammond dominated the 2013 edition of the Taupo Half with times close to the best ever achieved on the course. From a statistical viewpoint it’s what happened behind them that is most interesting – allowing us to draw conclusions about the relative difficulty compared to previous years.
Both the finish times and conditions initially point at a fast day and further analysis shows that it was nearly the fastest day we’ve ever had for the current course. Which is going to be bad news for those who didn’t achieve a PB or thought that their improvement was a result of training…
This is the dashboard of Taupo results through the years:
A couple of facts are instantly clear:
- the second highest number of finishers ever is a credit to the organisers
- fastest average times for bike, run and overall on current course show it was a PB day
- swim was faster than 2012 as water was smooth
Not shown are the other summary statistics – suffice to say (at the risk of losing the audience) the Medians & Quartiles show the same trends so the skew of the data is consistent.
Additionally Candice set a new womens run record and Amelia Rose Watkinson a new womens T1 record.
Next up is the ‘same athlete analysis’ – or as I refer to them ‘repeat offenders’. I track the average performance for each athlete and compare their performance in any given year to their overall average (obviously only athletes that have completed at least two events). Yielding the following:
Candice setting a new run record notwithstanding – the only thing that stopped 2013 from being a overall record breaking year was a slow run.
And it’s clear that the smooth water aided swim times quite a lot, although the lead pack swimmers were slower than expected (I’m told that there were navigational issues).
For athletes that competed in both 2013 and 2012:
So both overall averages and ‘same athlete’ averages show that 2013 was 10mins faster than 2012. Unless you are me and felt terrible from the beginning, eventually stopping after the ride.
I believe I may have mentioned once or twice (or every chance I get) that I like to construct course models of events so that I can understand the effects of weather, equipment & pacing. So I fired up the Taupo Half MKIII model and went to wunderground for some weather data.
I set up an athlete who completed the ride in 2.52:41 in 2012 conditions then switched the model to 2013. The predicted time for that same athlete was 2.40min faster. Higher temperatures reduce air density and decrease the rolling resistance of your tyres, plus the wind direction was ideal this year (tailwind home).
This leaves nearly 5mins of improvement that can’t be explained just by the model. I suspect part of it is the mental factor – the homeward leg is so much easier with an assist. And athletes improving year on year (training/equipment) is certainly a possibility (though I typically don’t really see this as a trend overall). The new 1m drafting rule that many riders seemed to be adhering to may have helped also.
Unfortunately higher temperatures come back to bite you on the run where you are moving slower so don’t get the same cooling effect that you do on the bike. So 2013 was one of the slowest years on record (for repeat competitors).
Ultimately the bike is the biggest determinant of overall performance (as it is the longest discipline and has a big follow on effect for the run) so the overall times were still at PB levels despite the unfavourable Run conditions.
2013 was a fast year for the Taupo Half because of near ideal weather conditions on the bike. It’s possible that the only thing that stopped it from being a record breaking year was the heat on the run (though that had led to faster ride times)
And clearly, while Graham O’Gradys lead from the front performance was impressive, he was not pushing as hard as last year given his slower time despite the better conditions.