Why a negative split is your best bet for running a PB

Yes, I am biased. I love the negative split. So much so that I wanted to name my coaching company after it. But what is it, and why do I like it so much? Here's a definition that neatly summarises it:

A strategic approach to racing, involving running the second half of a race faster than the first. It hinges on a deliberate decision to start at a moderate pace, followed by a gradual or sudden acceleration towards the finish. ​ "I ran a negative split for my last 5k race"

When you first start running, it's much more likely that you will be drawn to the opposite of this strategy, namely, the positive split. In fact it probably wasn't the result of 'a strategy', but rather the natural way that most people approach racing. Go out hard, suffer, and try to hold on. For shorter distances, this is even the common approach used by elites. The 400 and 800m in particular are often referred to as races won by the athlete who 'dies the best'.

The problem is once you start to increase the race distance, going out hard isn't just a case of then slowing down relative to this, and achieving a neutral average pace. For example, if you believe your fitness to be ready for a 3-hour marathon, running a 1.25 first half doesn't mean that you will then be able to run a 1.35 second half, and come out with your target time. This is because going out hard elevates your heart rate and in turn starts expending more energy than if you go out slower. If you've heard of the phrase, "hitting the wall", this is what is happening. You are out of energy.

If you flip those splits around, then you are less likely to be working your energy system to a degree that depletes it in the same way, and theoretically you should be able to increase your speed in the second half of the race to come out with the time you are after.

Of course, it's a skill and if you get it wrong, you may end up selling yourself short. If you end up powering over the finish line and feel like you could run the race again, then you've probably mistimed things, gone out more cautiously than you should have, or not sped up early enough. But hey, you didn't blow up, and scalping lots of runners in the latter stages of a race never fails to be fun and provide ongoing motivation even when you're getting tired.

When I first ran the London marathon, I was aiming to scrape a sub-3. Here's how I ran it:

Data showing the mile splits from my first London Marathon, where I ran a positive split

You can see that I was running 6:40/mile (2:55 marathon pace) early on, and this is what did for me in the latter stage of the race, where I crashed and burned. And believe me, it was miserable. I trundled over the line well over my target time. The truth is that I probably wasn't in sub-3 shape, but maybe I was in 3:05 shape, or 3:03. The point is that I'll never know. The time run faster than my average pace was not good time 'banked', as by the end I was shuffling along slower than 8 minutes/mile.

I ran London earlier this year, and thankfully I stuck to my plan and roughly speaking, executed it well (I can confirm that I did not drop a 5.33 at mile 19, I believe that was due to a GPS error at Canary Wharf).

Data showing the mile splits from my most recent London Marathon, where I ran a negative split and achieved a PB

You can see here that my fastest 5k segments where the last two, and I had enough energy left at the end to accelerate for the finish. Now, I'm definitely fitter than I was back in 2019, however by adopting this strategy I am confident that I got closer to achieving my fitness potential than had I gone with a more aggressive approach. Maybe I was in 2:45 shape. But at least I didn't end up with 2:55 or slower.

And unlike with London the first time, I now have real confidence based on the successful completion of the race that I can dial it up and go faster next time. Obviously with marathons they are normally quite far apart and a lot can happen in between. But with 5k/10k I find that this approach really comes into its own, as provided that you run a negative split, you can race again shortly after and know that you can be that touch more aggressive and have confidence you'll run a faster time.

That's not to say you should approach every race this way. For shorter races, I definitely think it's worth testing yourself - you never know, you might be stronger than you think and hold on long enough to make the positive split worth it. The worst case scenario in a 5k for example is likely to be no more than 5 minutes of pain. Over the marathon distance, personally I don't think it's worth the risk. Blowing up somewhere between mile 16-20 as is common might mean over an hour of struggle, and having been there I can vouch that I never want to go back!