Speed is one of two fundamental forces in our sport (alongside aerobic strength) and a bit like the annoying younger sibling. The conventional wisdom is that it’s mostly talent-based. It burns bright and fizzles fast as we age, while strength is more the cumulative resource. While there’s certainly truth to this — maxing out every day is fine when you’re a squirmy kid, but hurts once you’re paying your own health insurance — performance potential doesn’t play by the same rules.
You can continue to improve as you get older, you just have to touch 👉 speed consistently and rely on your mental aptitude to surf the conditions you’re presented. Be gracious with yourself when you aren’t feeling 100%, and extra deadly when you are. The adventure of solving this little puzzle is the great joy of a distance running career.
Solve it successfully and you get a great reward 🍯 — your lifetime bests!
Table of Contents
THE 👏
The small group of us training in San Luis Obispo, under capitan-in-chief Joe Rubio1, have dubbed our training philosophy “the clap”. It’s an approach to periodizaion wherein oppositely flavored blocks of training converge systematically onto the target event.
Say, you want to peak for a 5K in June, you might begin by building an aerobic base in the fall, race cross-country or roads in the winter, then shift down to speed training (800-1500m) in eary spring before finally emphasizing VO2max work in the weeks leading up to the target event. In those final few workouts BOOM! the hands collide and you reach the height of your fitness.
A Rubio-ism for the occassion: “Hey man, it’s not that complicated. They’ve been doing this stuff since the seventies!”
THE GOAL
So my initial plan for January was to be aerobically on-it with long workouts, etc., and then travel to Houston to try and improve on my personal best in the half-marathon. When a hamstring injury (during a high-volume week in November) kind of interrupted this initial plan, I started to consider taking a stab at an indoor track season instead. This isn’t too much different in the grand scheme, I’d just postpone focusing on “strength” and start focusing on “speed” a bit earlier. The four-minute mile isn’t something I ever really believed to be in my wheelhouse, until maybe two years ago. Partly, this is because the empirical evidence shows I’m already too old – peak performance age is well-studied to be ~25.5 +/- not enough years to include me . But moreso, growing up, my top-end speed was never that exceptional. I was always the kid who could keep running the longest, not necessarily the fastest. I don’t think I even broke 60 seconds in the 400 until I was, like, 17 – but could run a 34-minute 10K at 14.
As a result, I never emphasized speedwork or power development as extensively as a mid-distance specialist. So, I’ve been curious what effect a dedicated mid-distance mini-season might have, and with the clock always ticking against my performance potential, I decided this was a gift from the gods to go for the sub-4.
I ultimately failed, as mentioned, but significantly improved both my mile and 3K personal bests from a decade ago, and even ran my fastest-ever 200m, all within the same week. All these performances align closely with my best at any distance—and have happened earlier in the year than usual. So, especially if PBs continue to fall, the mini-season also feels like a huge success.
THE CURVE
To have confidence in your training plan and goals, it helps to have an objective picture of your abilities. For a younger athlete, say below 20 years, there won’t be enough useful data points to construct a stable picture. You’ll have outlier performances, and the most extreme of those will be the best approximations of eventual peak ability. Anyway, the focus at that stage should be on building durability and intuition, not on performance optimization.
Experienced athletes in their 20’s and 30’s will have more stable data points to draw from. For example, below are my personal bests, as of Feb 2025:
discipline date age time score half-marathon (road) 15 Jan 2023 29.5 1:03:16 1058 10-mile (road) 07 Apr 2024 30.7 47:44 1054 10,000m (track, outdoor) 08 Jun 2024 30.9 29:02.88 1037 5,000m (track, outdoor) 28 Jun 2024 31.0 13:45.88 1047 3,000m (track, indoor) 14 Feb 2025 31.6 8:00.34 1076 1-mile (track, indoor) 14 Feb 2025 31.6 4:02.45 1098 1,500m (track, outdoor) 06 May 2023 29.8 3:44.00 1053 800m (track, outdoor) 21 Apr 2023 29.7 1:53.81 920
The same set of points can be plotted {as m/s … on a curve}…
THE PLAN
The World Athletics score derives from the statistical relevance of each performance. It’s always been obvious to me that my “range” ended well before the sprints; nonetheless thrilling to see this confirmed . Otherwise, the scores are relatively balanced from 1500m to half-marathon, which I think is accurate, though I tend to target the 5000m as my main event. Conventional wisdom tells us that “the clock is ticking” on the window of opportunity for setting personal bests, particularly at shorter distances. There are lots of mundane and macabre physiological explanations for this – reduced fast-twitch muscle mass, decreased maximum heart rate, reduced tendon elasticity. All boil down to some pathway of cellular degradation. All can be moderated by taking good care of yourself (sleep, diet, exercise), but are, rather tragically, perfectly “natural” anyway, and in any case, the hand of time can’t ever be stopped. This window of opportunity is also believed to shift up in distance with time, with peak performance age expected later in longer events, like the marathon. The explanation – you’re better able to take advantage of compounded, multi-year muscular and aerobic adaptations, but gradually lose top-end speed due to the decline of your max heart rate, fast-twitch fibers, elasticity, etc. To accommodate this, a pattern many elite runners follow is that of “moving up” gradually to the next successive event group as they age. Typically, this involves following tailoring training programs ever more for optimizing “strength” vs “speed”. I’ve always been drawn to the opposite approach, of tailoring training gradually more to address where a physiological decline should be taking place, even if at the expense of
An exponential decay
f(x)=a_0 ⅇ^(-b_0 x)+a_1 ⅇ^(-b_1 x)+c a_0= 1.5043680963 b_0= 0.0005418255 a_1= 0.9818982345 b_1= 0.0000382296 c= 5.0962947196
a_0= 1.504368 b_0= 0.000542 a_1= 0.981898 b_1= 0.000038 c= 5.096294
As such, the line-of-best-fit is really an exponential decay function (excel offered a “power” trendline I included, which is passable). And the