Could I beat my 5 km personal record set only 6 weeks ago?
Where has the week gone!
It’s been 7 days since my last race and I still have not written this race report. So here goes…
Last Sunday morning (20th August) I ran the 13th race of the 2023 Sportlink Grand Prix, the Dereham 5k. As per my pre-race article this is a two lap 5 kilometre road race around the mid-Norfolk market town of Dereham. A flat course which, apart from quite a few corners and a couple of 180° hairpin turns, is fast and known for its great potential for getting personal records.
Would the forecast 80% humidity stop me achieving my target time of 19:10?
And talking of personal records. This was very much my focus for the day. To beat my current 5k personal record I had set only six weeks previously at the Wroxham 5k. Consequently I was hoping for friendly weather, and that the forecast 80% humidity was not going to get in the way of me achieving my goal.
Dereham is my nearest town and so it was a later start than normal to get to the race HQ, and whilst the temperature was the forecast 18/19° C, there was a moderate breeze taking the edge off of the humidity. Good conditions for a personal record!
My Race
The start was the usual manic affair of a short race, as people rush out fast to get into their running stride. Being highly focused on my target 3:50 min/km pace, I did my best not to get carried away with the crowd. A flat 5 kilometre race such as this, needs to be run hard from the start, but it must be consistently hard throughout the race. Like the longer races, you can still set off too fast and pay for it later.
At the 1 kilometer marker, I was rewarded for my diligence, 3:47. 3 seconds ahead of schedule, and all things considered I was feeling okay. I pressed on.
I knew that I would probably not be able to maintain this speed…
The second kilometre marker was reached in a repeat 3:47 time. That would make me 6 seconds up on my target. I was certainly feeling the effects of the pace now, and knew that I would probably not be able to maintain this speed for another 3 kilometres. However, I knew I could maintain the effort!
As I crossed the start line and set out for my second lap I resisted the usual psychological lull that can happen, instead focusing purely on the fact that I had only another 2.5 kilometres to go. The 3rd kilometre marker was reached in a split time of 3:52. I was slowing. The cooling breeze had dropped, and it was getting hot.
Getting my personal record was still very much on the cards.
I was now 4 seconds ahead of my target. Could I maintain my pace and hit my target, or would I slow? Regardless, I knew I was going to give it a go, and I was pretty sure that I would not slow too much. Getting my personal record was still very much on the cards.
The 4th kilometre was tough. I knew I was slowing, but by how much? As I crossed the marker, my watch flashed up 3:54. I was now bang on target. I needed a 3:50 min/km final kilometre to come in at 19m10s, and at least faster than a 4:05 min/km pace to get my personal record. The race was on!
It was no sprint finish for me…
For that final k’ I did all I could to keep my legs moving, and it was a welcome relief to turn the last corner for the final 100 metres to the finish line. Whilst I did manage to up the pace to cross the line, it was no sprint finish for me – my legs were spent, and I fell to the ground to recover.
Was it worth it? Of course it was! A final kilometer in a pace of 3:49 min/km. I was 1 second ahead of my target! My new 5km personal record stands at 19:09. Job done!
Analysis
Below is a table summarising the predictions from various algorithms / sources, along with their error from my actual performance.
| Algorithm | Predicted Pace (min/km) | Predicted Time (mm:ss) | Error (seconds) | Error (percent) |
| Riegel (6 min assessment) | 4:19 | 21:35 | + 146 | + 12.7 % |
| Riegel (3.2 km assessment) | 4:10 | 20:50 | + 101 | + 7.9 % |
| Athlete Data Analysis Platform | 4:05 | 20:25 | + 76 | + 6.6 % |
| Riegel (Half Marathon) | 3:59 | 19:55 | + 46 | + 4.1 % |
| Running Watch | 3:58 | 19:52 | + 43 | + 3.8 % |
| TrainAsONE | 3:51 | 19:16 | + 7 | + 0.5 % |
So there you have it. I beat all the predictions, though TrainAsONE came extremely close, suggesting I would run 0.5 % (or 7 seconds) slower. I think anything below 5 % is good, so 0.5% is amazing. (And as we’ve seen over the season it has consistently been accurate.)
Finishing Up
15 seconds off my personal record, and my 7th of the season – so far…
My next race is a half marathon in only 7 days time. Can I bag another personal record? I’ll keep you posted.
As ever, a big thank you to the race organisers, volunteers and all those involved. A thoroughly enjoyable and great race. Thank you.
Till next time.
Well done, another PR! But I keep thinking, isn’t there a confirmation bias in the results? It would be fairer if you blind yourself to the predicted times and then run as fast as you can. Maybe an idea for the next blog?
Hi Steven, Thank you (re. my 5k PR). Yes, the results of the predicted vs performed is in no way statistically significant and biased in a number of aspects. In some of my earlier articles I provided error ranges for the TrainAsONE and Riegel predictions (we do not have access to these for my Running Watch or the Athlete Data Analysis Platform) to help illustrate the overall picture. My intention is to write a summary article covering the whole season, and these should definitely be discussed in that. I have contemplated running blind, but given that it is still a ‘sample of one’ did not see too much value in it. What I do find most intriguing in the results is that TrainAsONE has invariable predicted the fastest time, and I have either exceeded or come very close to this. Often this has been 5% or so faster than the other platforms. So even if there is confirmation bias present, the fact that I have raced faster than other predictions say is possible is very interesting.