The conditions are part of the challenge

Researchers in Paris gathered ten years of results from six of the biggest marathons in the world. The races were Paris, London and Berlin, and Boston, Chicago and New York. Across all the races, that came to about 1.8 million finishes. Of everything they measured, air temperature mattered most.

The best times came in cool air, between about four and ten degrees. Above that, runners at every level slowed down. More of them dropped out too. London in 2007 was run at 19 degrees. In the other nine years it averaged about 12. That year the average finishing time was 17 minutes slower than usual.

Another study followed seven North American marathons, some for more than 30 years. As the weather warmed, the winners slowed. Even among the first 300, those further back slowed more. Part of the reason is that a slower runner is out in the heat for longer.

Heat costs you because your blood has two jobs. It carries oxygen to your working muscles. It also carries heat out to your skin, where sweat takes it away. In cool air your skin needs little blood, so most of it goes to your muscles. In warm air more of it goes to your skin. Your heart has to beat faster to serve both, so the same speed asks more of it.

For a fast time, choose a flat course in a cool season. A spring or autumn race is the usual choice. If you want hills or a swim in the sea, choose them for that. Then set a goal that fits them, not the time you would do on the flat.

Your goal might be a time, a place in the field or a strong finish. You set it months ago, for a cool, still morning. On race morning you can see the day you got. Look at the temperature, the wind, the hills and the water. Set your goal from them.

On a cool, still morning, your plan can stand. On a hot one, aim slower than you planned. In a heatwave or a rough sea, drop the time and race for a good finish.

The first kilometres of any race feel cheap. You are fresh, and everyone around you is going fast. So start by effort. That means starting at an effort you could hold all the way to the finish, whatever speed it gives you.

A study of 1.7 million recreational marathon runners showed what the other way costs. Runners who started too fast usually started far too fast. They slowed badly later in the race. They finished later for it.

In open water, the start costs more again. Cold water on your skin makes you breathe fast and hard for the first minutes. Your heart races too. This is called cold shock, and nerves make it bigger.

Researchers lowered 11 people into water at 15 degrees for seven minutes, twice. Before one of the two times, they told them the water would be five degrees colder. It was not. That time they were more anxious. Their hearts beat faster, and they breathed harder for longer.

A crowded start, with arms and feet all around you, brings the same nerves. So start calmer than you feel. Build once your breathing settles.

On hills and in wind, hold the effort and let the pace go. On race day you can sharpen that. Put in a little more where you are slow, and a little less where you are fast. Minutes spent going slowly are long minutes. So effort spent there saves the most time.

Researchers tested this on seven male cyclists. Each rode a 16-kilometre time trial on an indoor trainer. The trainer gave them a headwind for the first half and a tailwind for the second. First they rode it at their own pace. Then they rode it twice more, at the same average power.

One time they held that power steady. The other time they rode 5 percent harder into the wind and eased off with it. Riding steadily took ten seconds off their own-paced time. Riding harder into the wind took off 12. It also felt the easiest of the three.

Hills work the same way. A climb costs far more than the descent gives back. Researchers put ten runners on a treadmill and tilted it up and down. On a slope of 45 percent, steeper than any road, each metre up cost more than five times as much as the flat. Running down cost at best about half.

So give a little more on the way up, where you are slow. Then flow down the other side with quick light steps. Let your legs roll under you. Let gravity do the work, and the speed comes cheap.

The first time, the distance did not beat you. A pace meant for a cooler day did. So run that warm April morning again. You set a goal some minutes slower than last year. You start at an effort you could hold to the end, and you let the crowd go.

At 28 kilometres you are still moving well. You finish slower than in the cool year, but you finish strong. In the last kilometres you pass people who started at last year's speed.