About race time predictions

How many times have you typed into Google “how do I know my marathon pace?” or “workouts to predict my marathon time?”. I know I have. More than once.

It truly is fascinating how, during training, we spend weeks doing long runs with marathon pace in them, specific marathon pace runs, threshold runs, steady runs and even Yasso 800s, all of them with a race day pace in mind but then taper arrives and the doubts start creeping in.

Suddenly we feel like we haven’t had a good read on marathon pace and start finding reasons for that inability to make a decision, “that session I remember was too windy”, “too hot”, “my heart rate strap wasn’t working properly”, “I was too fatigued”, … Whatever it takes not to be convinced about what we can do on the day.

Even when training has gone well, with several excellent sessions, and no need for any last minute additional work, we start hoping we could have “just a couple more long ones”. We most definitely feel unprepared and full of doubts.

And that’s normal! For what it’s worth, I don’t necessarily believe in one session to determine our readiness. For shorter distances it’s fairly simple to test the pace during the same time we will be running but introducing some rests, and get a good understanding of what we can do. 5 or 6 x 1km with 60" rests would definitely give you a pretty good indicator of a 5km time or 4x3000 at 10k pace with 90" recoveries might also give you a fairly good indicator of your 10k possibilities.

But, unfortunately, there’s no such luck with the marathon. No single workout other than the race itself will tell you if you’re ready or not. Mainly because of its unique nature, the distance.

We can’t accurately or fully recreate the conditions of race day, the way the body fatigues, the energy is consumed, etc. so we have to find other ways of predicting what we might be capable of running.

Needless to say that if we’re just going to run within ourselves and the aim is to reach a very achievable target we don’t have much to ponder and can just enjoy the day. I guess I’m more referring to trying to run to the best of our potential on the day. What’s the fastest I can cover 42.195km.

I make that distinction because there’s a fundamental difference in how our body reacts when pushing it to the limit of what it’s capable of versus just handling the distance at a pace that will not test us too much. Of course you still have to complete the distance but the overall proposition changes when going out perhaps 10 to 15 seconds slower per km than what you could potentially do. It minimises the risk around energy and oxygen consumption and it places the main challenge on mental and physical endurance. Still a great challenge, but a safer one.

So are there any predictors out there that could help? I mentioned Yasso 800s. Is that a good predictor? Well, if we’re strictly talking about predicting… then no. There is a glaring issue with using an 800 meter interval session to predict a 42.195km race outcome. Specificity. A great middle distance runner would have no problem acing that workout but would struggle to complete the marathon distance in the predicted time. There are different energy systems and muscular requirements at play.

I am not going to describe the session in detail here but we’re talking about 10x800 meters with the same time of active recovery between sets that it takes the runner to cover the 800 distance. The coincidence is that a large percentage of runners are able to complete the average 800 distance in the same numbers (minutes and seconds) as they go on to complete the marathon (hours and minutes). So a 3 hour marathon hopeful would have to run the average 800 in 3 minutes dead with also 3 minute jogs in between.

The fact that minutes turn into hours and seconds into minutes is a very clear sign of no actual correlation. It’s just a coincidental data outcome.

It is true, however, that if the runner is not able to complete the session averaging the time they went out for, it is very unlikely they will have the speed to maintain the marathon pace for the entre distance as they had hoped.

So… what about times achieved in shorter distances? Are they a good marathon time predictor? They can be. Absolutely. But obviously common sense should be applied.

Following the same principle of specificity we can deduct that the longer the race we’re using, the better. A half marathon is better than a 5K, and a 10K is better than a mile race.

Recently I looked into what the internet has to offer on that front and I saw predictors taking your HM time, doubling it and adding 4 minutes and 30 seconds, one was adding 6 minutes and a lot use 8 minutes.

I even found this… “T2 = T1 x (D2/D1) 1.06 where T1 is the given time, D1 is the given distance, D2 is the distance to predict a time for, and T2 is the calculated time for D2”

When I first started running marathons my conversions from a half were poor. The first handful of attempts I was closer to HMx2 + 15 minutes. The last few good marathons I’ve completed this has come down to HMx2 + 7 minutes.

So again, it would appear that the outcome will actually depend on the training done. The number of years we’ve been running, the miles accumulated and the experience gained.

I believe that in our first few marathons, we’re not very marathon strong or efficient so we can’t really get that close to what we could achieve with years of work behind us. At that point looking at HMx2 + 12 to 15 minutes is a lot more realistic if we have done the work.

As we get more experienced and get more precise and specific with our training we can bring that down to HMx2 + 8 to 10 minutes.

And it’s only when we truly specialise in the distance and maximise every aspect of our training after years of work that we can get to HMx2 + 5 to 7 minutes.

My last competitive marathon I did 01:13:56 and a few months later 02:34:52. That’s HMx2 +7 minutes.

Unfortunately this time around (less than 3 weeks to race day) I don’t have a HM time to use. I do have about 6 very specific workouts in the last 20 days totalling 125km at marathon pace or under including a 25km run at 3'27"/km (HM in 01:12:45).

The numbers of the last 20 days before the final 3 weeks are:

• 402km (20.1km per day average)
• 6 quality workouts (2 per week)
• 125km at marathon pace or quicker in those 6 workouts (20.8km per session average)
• 31% of quality work vs 69% recovery and easy
• Threshold pace at around 3'18"/km

So the question is… what is my marathon pace? And, am I ready for it? The answer is I do not know.

I suspect, based on experience, that I could be between 2:28 and 2:34. The reason I can’t reduce that window of 6 minutes is the very little specific work outside those 20 days. It’s not so much the overall mileage (I’m still averaging over 19km per day this year; 1,806km in 95 days) but the quality work required for longer than 20 days.

Ideally I would have a good 10 weeks of solid work, not 5 (which is what I feel I’ve had in terms of quality work). I have also ONLY completed one traditional long run. Only one hilly 2h30m run.

Obviously the 6 workouts averaging 20km tend to add up to longer when including warm ups, cool downs or recoveries/easy running in between. That means some of them have even added up to 34km and 37km but there’s so much to be said for completing a handful of good old traditional long ones (at a steady-ish pace).

I still have just over 2 and a half weeks to go and, especially this week, I am hoping to throw in a couple more decent workouts before tapering properly, and maybe during this period I will be able to be more precise on what marathon race day strategy can look like, but at this very point I have a few doubts on what I could realistically achieve.

One little positive aspect in my planning is that this is not even my target race for the year. I have been lucky enough to have been included in this elite marathon (at a late stage) and I have been able to make a qualitative jump in my training lately giving me a surprising chance to potentially take a good chunk off my PB, so in a way it feels like a win win situation.

I can just relax and be ambitious on the day without any pressures of having to run a particular time, only focusing on my own performance. I am not going to lie though, a sub 2:30 this early on in the year (before London or Valencia) would be awesome. Not only because it’s a bit of a milestone to go under 2:30 for the marathon but because it would give me renewed energy and motivation to even go faster later in the year after 12 weeks of solid specific work. What did that Kenyan guy, who also runs, say once…? Impossible is nothing? ;-)