Four 5ks, four results

I've run a 5k race each of the past four weeks. Each week, I've had the same goal: Break 18:40. Each week the strategy was slightly tweaked, but the basic plan was pretty much the same -- run a slight positive split and give myself enough cushion to beat the time goal even if something went wrong. But I was only able to do it on the final race. Lets take a detailed look at the numbers and see if we can uncover any patterns.

Here's the basic data for the four races (in the order I ran them):

GAP = Grade-Adjusted Pace

As you can see, my pace improved for each race except the 5Kahne, which was the hilliest race and run on a warm day.

The dramatic difference in time for Balloonfest is primarily due to the fact that my Garmin distance matched the course distance almost exactly, where it was between .04 and .07 of a mile off for the other three races. That may not seem like much, but at a 5:54 pace, .04 miles takes 14 seconds to run!

That said, even if Balloonfest had come out to 3.17 miles, equal to the longest run of the group, I still would very likely have broken 18:40, assuming I could have maintained the 5:54 pace for the entire distances.

So did my overall fitness improve over those four weeks, or did I just have a better strategy? Let's take a look at my splits for each of the races:

Garmin splits for each race

I had the planned positive mile splits for the first three races, but at Balloonfest I negative split. I started Balloonfest slower than any of the races and finished faster. My Mile 2 split was nearly identical for each race, but Mile 3 was much faster at Balloonfest. But a slower start isn't necessarily best -- My second slowest start was at the 5Kahne, which was my third best race.

Grade-adjusted splits might be a better way to view this data. Strava gives a grade-adjusted pace for each mile, and we can graph those as well.

Strava's grade-adjusted splits for each race

Now the races are starting to look more similar. When adjusted for grade, I actually positive-split the first three miles of each race, then picked it up at the end of all except Balloonfest. Once again Mile 2 was similar for each race, but I started USA Masters quite a bit slower. Interestingly, my Mile 3 was fastest at Balloonfest, even adjusting for the fact that it was mostly downhill.

The reason Balloonfest has an off-the-chart-slow final 0.1 mile is because it was a fairly steep downhill, but the slope was rough, so I did not want to risk going all out. Even so, on an absolute scale, it was the fastest finish of any of the races. 

All this suggests to me that my fitness actually improved over the course of the races. My worst grade-adjusted Mile 3 was in the first race. Yes, I was slower in Mile 3 in week 3 than week 2, but that's probably due more to the relatively warm, humid weather that day and a conscious (if misguided) decision to slow down than a decline in fitness. The combination of an overly-ambitious start at Lungstrong and my poorest fitness of the bunch led to my poor showing there.

Here's hoping I will improve even more over the coming weeks. I won't be racing as much, but when I do, I'll be looking to build on the gains of the past month.