The goal is a number I have never hit. A marathon age grade above 80.15%, at age 48, on a repaired knee.
Two years ago I would have called that backwards. You do not chase personal bests at 48, a year out from meniscus surgery, off the back of the slowest marathon block of your life. You manage decline. You make peace.
But age grading does not measure the clock. It measures the clock against what is possible at your age. And that second number moves in your favour as you get older, quietly, every single year. So the goal is not backwards at all. It is exactly the kind of goal that gets more reachable while your raw speed gets less.
Aging brought an injury I could not name
The knee did not stop me running entirely. That was the trap. I could still run short — about 45 minutes, and then the knee would start to hurt and I would have to stop. Anything longer was off the table. Run short, ache, recover, run short again. For months.
The volume tells the story better than the pain does. Healthy, I run 80 to 100 kilometres a week, five or six days out of seven. Through the injury I was down to two or three runs a week, 20 to 30 kilometres at the outside, and some weeks none at all. A third of my normal load, on a good week.
So most of my training stopped being running. I went to the gym. I got on the bike. Cross-training — the stuff runners quietly resent, the hours that feel like a tax on not being able to do the thing you actually want to do. I did them because a short run was all the knee would give me.
For a long time nobody could tell me why. I had an MRI scan in October 2024. It came back clean — no tear, nothing structural, just a knee that hurt. So I kept running on it, short and careful, because short was possible, and possible is a dangerous word.
The pain did not agree with the scan. A second MRI in March 2025 finally found it: a meniscus tear the first scan had missed. Surgery, a repair, came in late April 2025. I was running again by early June.
The peak I did not know was a peak
Rewind to the winter before all that. The block from October 2023 into spring 2024 was the best training of my life, and I did not know I was standing on top of it. It ended at London on 21 April 2024, where I ran my marathon PR — 2:41:56 at age 46.
On the Alan Jones 2025 tables that scores 80.15%. For a 46-year-old man the marathon age factor is 0.9291, and the open-class standard is Kiptum's 2:00:35. So 80.15% is four-fifths of the way to the current world best, adjusted for my age — my best marathon grade, and the number Amsterdam is measured against.
You never feel a ceiling while you are under it. You feel it later, when your knee starts aching after every run and you realise the top was months ago.
What you keep, and what you let go
Here is the reframe that got me off the couch. I may never run 2:41 again. That clock PR belongs to a 46-year-old with a healthy knee, and matching it at 48 off a repaired one is a long shot. If I measured myself by the stopwatch alone, the honest verdict would be: probably past it, manage the decline.
Age grading refuses that verdict. It asks a different question — how close are you to the best a runner your age could do? — and the answer to that has not walked away. My raw times may never match the old ones. My age grade can, and it travels across every distance I race, not just the marathon.
The climb, in real numbers
Here is the marathon arc since the injury, run through the same engine — Alan Jones 2025, my exact age on each race date, no rounding of the factor.
| Race | Date | Time | Age | AG% |
|---|---|---|---|---|
| London | 21 Apr 2024 | 2:41:56 | 46 | 80.15% |
| NYC | 2 Nov 2025 | 2:56:19 | 48 | 74.93% |
| Two Rivers | 15 Mar 2026 | 2:50:05 | 48 | 77.68% |
NYC was my first marathon after surgery, five months into running again. 2:56:19 on a hilly course — 294 metres of climb — and my quads were gone by kilometre 28. A flat-course equivalent might be a minute or two quicker, but that is an estimate, not a computed grade, so I will not dress it up as one.
Two Rivers, four months later on a flat course with 87 metres of gain, is the cleaner marker. 2:50:05, 77.68%. That is my honest post-op fitness, and it is the number I trust. Two and a half points under London, and climbing.
The Amsterdam question: a marathon age-grade PR
My next marathon is Amsterdam, on 18 October 2026. I will still be 48 on race day — just, with my birthday days away. The goal is a marathon age-grade PR — beat 80.15%.
For a while I framed it as a clock time, 2:45:00. Round, ambitious, the kind of number you write on a fridge. Then I did the arithmetic. At 48, with a marathon factor of 0.9127, 2:45:00 grades 80.07% — 0.08 under London. Close enough to touch, and still a miss.
Run the inverse — what time at 48 equals 80.15% — and it comes out at 2:44:51. That is break-even. To actually beat the number, clearly rather than by a rounding error, I want about three tenths of a point of daylight: sub-2:44:14, which grades 80.44%. From Two Rivers' 2:50:05 that is roughly five minutes and fifty seconds to find. A training block, not a different life.
Getting older is on my side
Here is the part that still surprises me. Some of the ground between London and Amsterdam is not mine to close. The tables close it for me, just by the calendar turning.
Take my actual London clock, 2:41:56, and grade it as if I ran it at 48 instead of 46. It scores 81.59% — a point and a half higher than the 80.15% it earned at 46. Same time, older man, better grade. That 1.44 points is pure age-factor easing.
You can see it in the break-even too. To match London's grade I am now allowed to run 2:44:51 — nearly three minutes slower than I actually ran at 46 — and the tables call it the same performance. At a clock PR that easing means nothing. At an age-grade PR it is the wind at my back.
So the honest arithmetic is a hopeful one. I need real fitness — roughly six minutes off Two Rivers, earned in training, on a knee that now holds up. But I do not need to be the runner I was at 46. I need to be a slightly slower runner whom the tables, two years kinder, grade a fraction higher. Persistence supplies the fitness. Age supplies the rest.
What my story actually shows
Three things, if you are a runner watching your own clock times drift the wrong way.
First, the injuries change as you age. Mine let me run for the better part of a year before anyone named it, and it cost me a season and a surgery. Take the grumble seriously earlier than I did.
Second, the absolute records may be gone, and that is survivable. A stopwatch PR is one number on one day at one age. An age grade travels with you. You can lose the first and keep chasing the second for decades.
Third, persistence works, and the tables meet you halfway. My post-op marathon grade has climbed from a hard restart to 77.68% and is still rising, and the gap left to my best is a training block, not a life sentence. Getting older hands me a new marathon age-grade PR before it hands me back 2:41 — the age grade is the nearer target, and the one I am chasing first.
And if the knee holds? I have not fully closed the door on 2:41:56 either. Stay healthy for a couple of clean blocks and the clock PR is not obviously beyond me. That is the long-odds dream. The age-grade PR is the one I actually plan to catch — and it is a record worth wanting on its own.
Sources
- Alan Jones, Age-Grade Tables (road, 2025 edition — approved 2025-01-10 by the USATF Masters Long Distance Running Council) — retrieved 2026-07-27, https://github.com/AlanLyttonJones/Age-Grade-Tables
- World Athletics, "Ratified: world marathon record for Kelvin Kiptum" (2:00:35, Chicago 2023, ratified 2024-02-06 — the open-class marathon standard in the 2025 tables) — retrieved 2026-07-28, https://worldathletics.org/news/press-releases/ratified-world-marathon-record-kelvin-kiptum
- USATF Masters, Age Grading (method reference) — retrieved 2026-07-27, https://usatfmasters.org/age-grading/
- RaceRecords, Methodology — retrieved 2026-07-27, https://racerecords.run/methodology