I hypothesise that doing a parkrun the morning after drinking an undisclosed number of alcoholic beverages at the work Christmas party, followed by dancing into the small hours, is not conducive to running well.
I have conducted an experiment to establish
a) if this is true
b) if it is true – to what extent
Legs, rest of body, trail shoes, rest of clothes, a parkrun (including: venue, timing device, friendly volunteers, barcode, access to previous run times courtesy of parkrun statistics), calculator/brain.
1. Go out for a nice evening of good food, laughing, a few drinks, plenty of dancing – that kind of thing.
2. Go to bed at approximately 01:45
3. Wake up at approximately 07:30 to tidy the kitchen, mop the floor, feed the rabbits, etc. thus buying a few brownie points and facilitate the imminent departure for the required parkrun, given that no time has been spent with family since before dinner the previous evening.
4. Run a parkrun. For purposes of statistical analysis and scientific rigour, it is important that this parkrun is one that has been run on numerous previous occasions although the precise number of previous occasions is an area of considerable dispute within the scientific community (see ‘conclusion’ for further discussion)
5. Don’t forget your barcode*.
*For the uninitiated (and there may still be some), a parkrun is a free, timed 5 kilometre run facilitated by volunteers that takes place at various places around the country every Saturday morning at 09:00, but to receive your results you first need to register, and then take along your printed barcode so that can match your time with your details shortly after the run has finished.
6. Await results to be texted out.
7. Compare results with those achieved previously.
7.5 Analyse results.
8. Write a blog post about it.
9. Patiently await Nobel prize for contribution to science.
Time taken to run the parkrun following work Christmas do = 23:50
Previous 6 parkruns at the same venue = 23:13, 23:54, 23:48, 23:13, 23:46, 23:55
Average of those previous 6 parkruns = 23:38 (.166666667)
Converting these times to seconds (to facilitate calculation of percentage decease in performance)…
Average of previous 6 parkruns = 1418 (.166666667) seconds ; post-party parkrun = 1430 seconds
_____________ = 0.99172494172
The above results appear to demonstrate that in percentage terms there was only a 0.8% (to one decimal place) decrease in performance.
However, it is not possible to conclude with any scientific certainty that this decrease is due to the previous evening’s activity and not just the natural variation in performance already evidenced in the times recorded on the six previous occasions (two of which, if taken in isolation, would actually indicate an improvement).
It could also be argued that instead of taking an average of previous runs, only the most recent run should be considered.
This would therefore suggest a percentage performance decrease now of 2.6% (to one decimal place)
It must also be noted that to be statistically significant, it would perhaps be better to do this more than once…
…so see you next Christmas.