Thursday, November 7, 2013

Kona 2013 Run Data - Pro Men

Two years ago, I wrote a couple posts on the pro men and women, and how they ran the marathon in Kona. One of the tendencies of the race is that the run starts off similar to the bike, with a big surge. Much like the bike has the early town loop with the 2 climbs and everyone fighting to make the lead pack onto the Queen K, there seems to be a similar surge out of T2, pressing to open a gap and break the strings connecting them to the other competitors.

This happens in both races, men and women. In the past, I have stated in multiple interviews and writings, that Mirinda Carfrae likely lost to Chrissie Wellington in 2011 because she paced the run too poorly early on. She played right into Chrissie's hands, possibly missing out the victory which would have validated her entire career from that point on. (Note: She seems to have validated now with Kona course record, but a win over Chrissie is certainly something which will always elude her). Chrissie was running scared in 2011, fighting to catch the girls in front of her for the first time off the bike, and made a big error in early pacing, and Mirinda wasn't able to capitalize, since she made the same mistake early. I wasn't in Kona for the 2012 race to get the data, but would have been interesting to see if she made a similar mistake. More on the women's 2013 run data next week.

It was also close in the men's race in 2011, with Pete Jacobs possibly costing him the title, as he poorly paced himself early on. It'd be a stretch to say he would have beaten Crowie on that course-record setting day, but he would have certainly been closer, and who knows what happens as he gets closer, with Crowie cramping and stopping late in the run.

A lot of people believe that the race is all about the run split in Kona. I disagree. The number one metric for predicting performance in Kona is Total Race Time to T2. How quickly and well positioned in the top places into T2 an athlete is, is their best predictor of how they will do in Kona. The run split is the second best predictor.

With the run split still being so important, and coming into T2 greatly fatigued from a hard swim and ride, there is little room for error in pacing, but a majority of the pro's still make large pacing errors. Whether that is because of a mental boost to drop athletes near them, meaning it isn't an error in pacing, is another debate in itself.

My biggest question is, with the quality of fields getting deeper, are the top performers really fitter, or do they just execute better pacing?

So what happened in 2013? I wanted to look a bit more at how the athletes paced themselves the first mile, but also how the pacing compared as they went thru the first 5K, and their final marathon split. (I've listed a projected marathon time, which is based on their first mile split).

Disclaimer: This requires an assumption that all markers are accurately placed, and the course is accurately measured as a whole. There was a timing mat at the 5K mark, but it might have been off, considering how many athletes ran faster for the 5K pace than the first mile. However, I am confident in the measurement of the 1st mile split, which is what most of the analysis is based off. 

More specifically, this time I wanted to see what the differential off the actual pace they ran for the entire marathon was, compared with what they ran the first mile in. Here's how it shook out, by finish order of the top 20 Pro Men... (Click on images to enlarge)



A few highlights of the top 10 men...
Average 1st mile split: 5:59
Average 1st mile to actual pace differential: 39 secs
Range of 1st mile to actual pace differential: 7 secs to 1:28


Question: Did Fredrik Van Lierde, (FVL), secure the win with better pacing than his closest competitors, like it was shown he did on the bike? 

Only two athletes in the top 20 ran a smaller or equal pace differential for the first mile compared to actual split, than FVL, and of those two, only one actually ran faster than him. Only 3 guys in the race ran faster than FVL, and one of those was by only 11 seconds.

Answer: FVL certainly executed a smart pacing strategy. It wasn't perfect, but it was much better than his closest competitors.



Question: Did the top 10 men just pace themselves better than the 11-20th place men?

Answer: No. First, remember that the top predictor of performance is total race time to T2. You can see in the chart that these guys had slower times to T2 on average. Other highlights of places 11-20th...
Average 1st mile split: 6:19
Average 1st mile to actual pace differential: 38 secs
Range of 1st mile to actual pace differential: 21 secs to 1:10

The 11th-20th places did have the slower first mile on average, but almost identical differential in 1st mile to actual pace, 38 vs 39 secs. Their range of differential shows they paced about the same as the leaders.


The fact the pacing of FVL was so solid, means athletes have to execute proper pacing on the run to win now. I think this is the next trend we will see in Kona, with athletes getting smarter about their pacing on the run, maximizing their potential. We would never encourage an athlete to start off 40 seconds or faster for the first mile of a marathon fresh, so why would we think it is the correct strategy coming off the bike in Kona, in an extremely fatigued and fragile state?

Two athletes heading onto the run in Kona, of equal run ability, and everything else being equal, the one who executes the better pacing will win. If you're not the best runner in the field, you have to be smarter to make up for the difference in ability, preparation, etc. This is the next area of the sport, where we will see an athlete break through, winning or performing better than expected, due to better pacing. (Luke McKenzie wasn't awful in his pacing, either).

Coach Vance


No comments: