When will the first two-hour marathon be run?

I found this piece by Alex Hutchinson very interesting, here is one excerpt on the issue of incentives:

One reason marathoners are running faster is that road racing is more lucrative. When the Sheikh of Dubai put up $1 million in prize money plus a $1 million world-record bonus in 2008, the Dubai Marathon instantly became one of the world’s fastest, despite its desert temps (average high in January, when the race is held, is 75°F). In fact, prize money for road races more than doubled since 1998, while track racing purses have gotten smaller (see below). As a result, runners are increasingly heading straight to the marathon. But big money can also draw the fastest runners away from the fastest courses, and the standard winner-takes-most prize structure favors cat-and-mouse tactics as runners race each other instead of the clock. When the Amsterdam Marathon switched to time-based prizing in 1999, four different runners immediately smashed the course record by 90 seconds. The sub-two-hour solution? A big pot of money that runners can win no matter where they race, and that is shared equally among all who break 2:00 in that event.

It is not obvious to me why a big first prize is not a good incentive, for instance why does it militate against speed to “race each other instead of the clock”?  Is it that the runners stay on too few courses, thus lowering the variance of performance outcomes?

In any case the hat tip goes to Vic Sarjoo.


There's a certain amount of wind resistance to overcome in distance running, so professional "rabbits" who run a predetermined pace at the front of the pack and then step aside are a fixture in big track meets.

There's the wind resistance, and also the fact that if you are running a bit quicker than your comfort zone, it's easier to push yourself and to keep a perfectly even pace if you are running behind someone else with a perfectly even pace. This is true also for much slower runners, where the wind resistance is even less of a factor.

This is ego depletion, mental exhaustion from being the pace setter. Following is less taxing physically and mentally.

I've heard of fast runners expressing their admiration for the mental endurance of slow runners who can keep plugging away, often in isolation, for four or more hours.

Marathons are so debilitating physically that contenders put a lot of effort during the race into trying to psychologically undermine rivals into throwing in the towel, which can be economically rational for likely losers. If your archrival pulls away at, say, the 18 mile point, do you take off after him and risk exhausting yourself so badly you won't be able to compete in another marathon until next year? Or do you say, "I guess it's not my day," jog in, and go register for the big money marathon in two months?

To break 2:00:00, they could make it a team process like climbing Mt. Everest in 1953 where members of the team who wouldn't go to the top did most of the heavy lifting, with the man in charge eventually picking the two strongest-looking and best-rested climbers to go for the summit.

A pack of paid, professional "rabbits"?


That's exactly how Roger Bannister broke the 4 minute mile. He had two "rabbits", Brasher and Chataway, who paced him through the first three laps. After the third lap, Brasher and Chataway dropped back and Bannister went on to finish in 3 min 59.4 sec.

The idea that Bannister was able to break the 4 minute mark because he was the only one who believed it could be done is a myth. Bannister's run was a team effort.

I met a man who was there as a 12 years old and could recall Norrish McWhirter (future co-editor of the Guinness Book of World Records) public address announcement:

"As a result of Event Four, the one mile, the winner was R.G. Bannister of Exeter and Merton colleges, in a time which, subject to ratification, is a track record, an English native record, a United Kingdom record, a European record, in a time of three minutes..."

Indeed. Pick a flat course, get some rabbits (relatively inexpensive) and some top runners looking for a payday and fame. You don't even need a big prize, just some money somewhere -- why not a 3 hour special on ESPN8 ("the Ocho")? Has to be more interesting than the first hours of a Tour de France stage which is heading to a sprint finish.

Also, in racing each other, there is a lot of strategy around running just fast enough to win, or rather, to not lose versus hitting a certain pace each mile.

The flurry of fast times in distance running in the later 1990s had much to do with the anti-anemia drug epo spreading everywhere.

I wonder what the slowest marathon is for the average participant? Perhaps Los Angeles, where the authorities shut down major streets like Sunset and Santa Monica for most of the day for the convenience of the runners and walkers, who take up to 9 hours or more to finish.

The Atlanta Marathon is run early Sunday a.m. and I get the impression they start rolling up the course around the 4 hr. mark, which still isn't enough to discourage obese people training to be slow and weak (and ultimately even fatter) who want to say they "ran a marathon."

If these were actual spectator sports, as opposed to pseudo street-festivals with no real fare, they'd just have some number of top qualifiers, say 25, in each age class.

Marathons are a pretty terrible spectator sport. Without Herodotus as its press agent, it's hard to imagine modern people rationally coming up with the marathon. The marathon is a product of 19th Century romantic hellenophilia.

Only if you close your eyes. Just survey the crowds at any big marathon. Apparently, the spectators disagree with Sailer.

The Nielsen ratings for marathons are sky-high.

What's the meaning of an "athletic" event that gets 30,000 entrants, the Boston Marathon, for instance? An ordinary distance runner could trot 26+ miles in a circuit of his own town without stopping traffic for hours. Obviously, the attraction is being a single sardine in a gigantic school, a member of a like-minded group, the antithesis of the individuality displayed by the solitary runner training for it alone. The organized marathon is validation of the insanity of running that culminates in hip and knee replacement.

In Los Angeles, they let you walk the marathon for nine hours down major thoroughfares. It's hell on traffic, but it gets a diverse turnout of walkers.

Back in the 1980s, my wife and I hired 3 guys and a truck to move us one mile, but it turned out that Moving Day that fall was the same day as the Chicago Marathon. Tt took the truck 3 hours to go the one mile across the marathon route, at $90 per hour or whatever we were paying.

I've been jaundiced about marathons ever since.

Shallow analysis, as usual, Steve.

The participants are the spectators. Same as rock music in dive bars.

Does this explain other iron man type of competitions? It's not just marathons, though marathons are the most popular.

>It is not obvious to me why a big first prize is not a good incentive, for instance why does it militate against speed to “race each other instead of the clock”? - See more at: http://marginalrevolution.com/marginalrevolution/2014/10/when-will-the-first-two-hour-marathon-be-run.html#comments

If two runners are far ahead of the rest, they won't want to push each other too hard or they'll risk both getting exhausted and eventually passed by the runner in 3rd place. They can match each other's speed then compete for the win near with a short burst near the end.

(Not that I know much about competitive running).

Right, to lower the current best mark considerably you'd want to pace yourself to use up all your energy with your final stride, then collapse in a heap. But maybe you overestimate ever so slightly your capabilities today and instead of collapsing after you cross the finish line, you'd run out of gas within sight of the finish line and be passed by less ambitious runners. Olympics marathons have a long history of runners crashing and burning on the last lap inside the stadium.

I know quite a bit about competitive running, but I will try to be brief.

1. The Science of Sport blog has had a number of posts on why a 2-hour marathon is a daunting achievement, particularly in the light of 10k times. Here is a recent example: http://sportsscientists.com/2013/10/is-the-sub-2-hour-marathon-imminent-dont-hold-your-breath-and-heres-why/

2. Elite runners can only truly compete in two or at most three marathons in a year. Therefore they have to choose their races carefully and indeed have an incentive to drop out early if they are certain they are out of the money. There is a sportsmanship issue here which is also a regulatory issue since technically you are not allowed to stop competing in the middle (this is a rule found in many professional sports, probably most - e.g. a soccer team can't just send in all its weakest players just because they are behind 6-1). But runners do seek ways to get that DNF (did not finish).

3. The doping issue is crucial. On the one hand there is reason to believe that today's top marathoners don't dope. The biggest reason is that their training camps are often in remote areas in East Africa without adequate refrigeration. Anyone who followed the Lance Armstrong saga can figure out the implications. Another reason is that to the best of my knowledge no top runner has been caught doping. It is pretty easy for any given competitor to avoid getting caught but it is pretty unusual for no one to get caught if all the top competitors are doing it. Since EPO and steroids are extremely effective (again, compare cycling) if top competitors started using them the times would go way down. But two hours is still pretty daunting.

4. The question of why tactical races are ever run remains a huge mystery to me. The research is completely unambiguous - you will run fastest if your effort over the entire race is even. Spurts only hurt your time. So keeping up with someone who throws in a burst in the middle of the race is always a mistake. It is ALWAYS wiser to speed up only a little bit to make up the time gained in the sprint evenly over the remaining kilometers. Yet virtually no runners do this, they always try to keep up. I have tried various models, based on wind resistance, asymmetric information/signaling (I know how well I feel but you don't), etc. but I have never found a way to make keeping up with a sprint faster than the fastest plausible maintanable speed not dominated by spreading out the lost time evenly. Maybe some reader can help me with an idea or perhaps someone has published something.

' The research is completely unambiguous – you will run fastest if your effort over the entire race is even.'

And oddly enough, that person who keeps their effort even, except a burst at the very end, may not run their theoretical best time - but they often win the race.

At least in the longer distances, with two runners seemingly evenly matched. Racing is not just about theoretical maximums - it is about racing.

(In a similar vein, a car racer willing to destroy their motor has an advantage over one who doesn't - which was considered, anecdotally, the reason why Enrico Ferrari was never a truly great driver, as his engineer's perspective kept him from ruining his motor in the pursuit of victory.)

It is true that since everyone is running tactical races it is often possible to win this way. But I can't figure out why this is not a dominated strategy. You are correct that running is about racing, but the question is why that should be a stable strategy.

Obviously the willingness to endure pain and injury is a big piece of the success picture. Money plays a big role in this as poor East Africans who don't find running fun at all may still be willing to endure a lot of pain and discomfort for the huge paydays the sport produces by East African standards.

One thing I did not point out is that the same dynamic is at work in the ultramarathons. In the past these did not attract any gifted competitors because a. there was little money in them and b. they take you out of commission for even longer than a marathon. Thus, a way, way past his prime Alberto Salazar was easily able to take the Comrades marathon. But lately there are bigger pots. Depending on the supply elasticity, this could increase or decrease the competitiveness of "regular" marathon fields.


Thanks - and even using the accurate name, I still couldn't find the source for that anecdote.

How do you model motivation? It must be far easier to motivate yourself to immediately run faster to keep up with a competitor who has sped up a bit than to let him spurt & then mentally push yourself to give a tiny push spread out over kilometers. Getting the mind to coax your body into pushing itself to the limits might need these seemingly non-rational tactics.


Paavo Nurmi ran with a stopwatch in his hand almost a century ago, winning, what?, nine gold medals pacing himself against the time, not his rivals. That seemed pretty convincing, but, I guess, as Keynes would say, "animals spirits" matter.

On your 4th point, how true is this at the most elite level? They run quite steadily at a pace that few can sustain over even a 5k. There isn't much room for spurts anyway. But record breakers consistently run slightly faster in the second half of their marathons, unlike most of the rest of us (I think ~85% overall are slower H2). So I'd have thought tactics would matter at the elite level, and potential aerodynamic advantage of following closely is likely to matter more too. Genuine questions - I don't know the answer and would be interested if you have insight into why that might not be the case.

"On the one hand there is reason to believe that today’s top marathoners don’t dope." Really? I don't think so... Let me give an example: in the norwegian cross-country skiing team, there was never anyone caught for doping, but it is obvious there are some skiers in their team who must be doping right now (Marit Bjoergen, specially...). I really don't believe there is a sport who can claim to be doping-free. Specially marathon running, since doping is widespread in sprint distances (Some are too big too fail like Usain Bolt... for now), there is no reason to think otherwise, with a longer distance.


Good comment. I find it impossible to believe top marathoners aren't doping - the gains to doping are so high in endurance sports, and the tests are so awful... It wasn't very long ago when Lance Armstrong, who was about 30% PED by mass, was one of the world's most popular and scrutinized athletes. He didn't seem to be failing drug tests.

http://sportsscientists.com is great.

"4. The question of why tactical races are ever run remains a huge mystery to me. The research is completely unambiguous – you will run fastest if your effort over the entire race is even. Spurts only hurt your time. So keeping up with someone who throws in a burst in the middle of the race is always a mistake. It is ALWAYS wiser to speed up only a little bit to make up the time gained in the sprint evenly over the remaining kilometers. Yet virtually no runners do this, they always try to keep up. I have tried various models, based on wind resistance, asymmetric information/signaling (I know how well I feel but you don’t), etc. but I have never found a way to make keeping up with a sprint faster than the fastest plausible maintanable speed not dominated by spreading out the lost time evenly. Maybe some reader can help me with an idea or perhaps someone has published something."

I have two hypothesis.

First, it's a valid race tactic.

During the race runner's bodies will go through a lot of stages and won't always be in an optimal state to perform top speed. If you are in a physiological state to cheaply surge it may be worthwhile to do so on the assumption that you'll drag your opponent with you and waste their energy on an inefficient surge, or you may form enough of a gap that they can no longer draft off of you.

As the follower it's still valuable to draft, even if the person is moving at too high a speed during a surge. First the draft effect itself will offset the excess physiological cost. But second, even at elite levels, the athlete's metal state plays a big factor in determining their physiological reserves. Drafting allows the following athlete to offset a lot of the motivation and preserve mental reserves for later in the race.

Second, even if race tactics can help, they're used far too extensively and are mostly a tactical error.

I think that quite often the runner who surges is simply the stronger runner, but people attribute the win to their surguing tactics and reinforce the idea that surging is highly effective. Also the factor that elite runners are going to be ultra-competitive and not inclined to let someone pull away when they have the ability to make a slight increase and keep pace.

Re: your point #4, at marathon speeds it is extremely difficult to differentiate between a general pace increase and a genuine surge. It's not like a 10K, where a surge feels like a surge. Letting someone get out of your sights is costly if it's not a true surge.

"On the one hand there is reason to believe that today’s top marathoners don’t dope. The biggest reason is that their training camps are often in remote areas in East Africa without adequate refrigeration."

Maybe they, like, run to some place that has a refrigerator?

Records have recently been broken where there are two contenders. It helps to have a near competitor as pacer, and also to reduce wind resistance (taking time at the front - less important than cycling, but still a factor at elite level). If you get more than two (as often happens in London) then people hold back to avoid giving advantage to others. More here: http://www.bbc.co.uk/news/magazine-29480460 I think that explains reason for time prize not winner prize.

Thanks for this interesting link. In terms of percentage improvement ( approx 1.5% in the 11 years from 2003 to 2014) wonder how the Marathon record improvement compares to other Track events.

A lot of current women's world records in track are still from the 1980s, because women got more bang from the buck from synthetic male hormones because they have so much less of the real stuff:


Didn't anyone try doping men with gazelle hormones or something?

Are you the real Michael Savage? Did you know Steve Sailer hates Mexicans more than you?

Actual I am a real Michael Savage, as opposed to the well-known shock jock who took it as a pseudonym. We have nothing whatsoever else in common.

My wife and I are about the same speed in endurance events, running and tri. I am not fast enough to win anything, but she is, and I sometimes pace her.

We have run races before where we spend the whole race about 10 seconds ahead of the second-place woman. I look back occasionally, and if it seems like the contender is closing the gap, we speed up. If she is dropping off, we back off a bit. The goal is to win by a little bit.

Probably if we red-lined and ran really really hard the entire race, my wife could win by a lot rather than a little and also PR. But that would involve things like: puking at the end, maybe a trip to the hospital, spending the rest of the day in bed, not being able to train for days or longer, not being able to race for several weeks, not being able to drink beer that afternoon on the boat, etc.

I think one thing to remember is that running hard for a long time is incredibly painful, and your brain doesn't work normally when you are out there. You aren't totally rational, and often you end up telling yourself something about why you should slow down or quit.

>for instance why does it militate against speed to “race each other instead of the clock”?


Because it's a marathon, not a sprint. They are not running at max speed at all times.

This article is so confusing. I thought people ran marathons just so they could put those "26.2" stickers on the back of their Suburu Outbacks.

Nobody would run marathons if you needed to sign a confidentiality agreement first.

But as people above have pointed out, signaling explains 29,980 people in the 30,000 person field. There's a sport going on for the few contenders. A dangerous and mind-numbingly boring sport, but a sport nonetheless.

That's why the participants receive commemorative T-shirts that they can wear when they go shopping at Trader Joe's.

When they drive their Prius to TJ.

Jeez, so much hate. Perhaps people participate in races because they enjoy running? I like the feeling of running and I like the competitive nature, picking people ahead of me to pass or try to catch up to, to push myself. If you don't like what other people are doing, and it doesn't affect you, then don't pay attention to it.

I think the time vs. place issue applies to just about anything.

Take mathematics. I would expect the Millenium Prizes to generate far more impressive results than simply the "prize" of getting tenure at a top department.

Put another way, if the goal is to optimize for fastest marathon time, then it makes sense to reward that goal. Rewarding a different goal will almost always fail to optimize for the desired goal, since the two goals are unlikely to be perfectly aligned.

Take a look at Bicycle "Sprinting." By no means are they trying to get the fastest time. Rather, they try to outmaneuver one another with the use of position, altitude, and visibility. It is more like a WWI dogfight than a race.

A thought about the negative aspects of a big first prize.

If my goal is "simply" to run a sub 2:00 marathon, then I don't really care about someone "taking off" at any point. I know my pace and that which is needed to achieve the sub 2:00. I can ignore that person who blasts out at a pace to "win" if they are at a pace I don't need to win the big 2:00 prize.

What I need is excellent other runners all focused on achieving the sub 2:00. Run as a pack at the exact pace necessary with no worries about losing to someone faster. Just a worry about keeping to the sub 2:00 big money goal.

'why does it militate against speed to “race each other instead of the clock”?'

There are at least two reasons. First, elite runners are just fast enough for drafting to make a measurable difference in energy economy. It's a far cry from cycling, but in a competitive situation, a few seconds over a two hour interval matter. Second, while it may be true that an even pace is optimal, it is impossible to know in advance what that pace must be in order to win. One can target a pace that would imply a new record if successful, but one cannot guarantee success. By definition, in order to set a record it is necessary to choose a faster pace than has ever been achieve before. And there is a risk to setting out too fast, which is that one may blow up and run a slow time.

Paying people to set records incentivizes them to set records. Paying people to win incentivizes them to win. A record will usually win, but a win is seldom a record.

Here is an explanation of pacing, strategy, and why runners in elite middle distance races often hold back, "racing each other instead of the clock": http://thesportseconomist.com/2012/08/10/olympic-distance-running-dominated-strategies/

The third and final explanation offered is the most compelling, and is even more pronounced in longer-distance racing (and cycling, though wind resistance is a much greater factor in that sport, and fosters its own distinct strategies). Distance runners are risk averse: a runner's absolute maximal capabilities may vary from day to day and week to week, and even the runner may not know how sustainable a perceived maximal effort will prove to be on a given day. So when racing for big money in a distance race, a runner often decides to run at a more controlled pace, and bet on his ability to win a fast scramble (known as a "kick") over the last stage of a race rather than risk it all by gunning the pace to what he thinks would be his maximal even pace, risking that he overestimates and flames out with enough distance left that his competitors could overtake him before the finish line.

I don't know if it's ever been formally against the rules, but rabbits used to be frowned upon. What non-runners generally don't appreciate is that running is a sport, not just a question of covering distance at a high speed. That's why runners generally compete against each other instead of the clock. That's why many of us feel that using a rabbit or a pacer is cheating. That's why people who draft off the front-runners are cowards.

I would love to see a sub-2 marathon, but not at the cost of what the sport should be.

A prize shared by all runners under X minutes seems like a good way to get a lot of clumping around 99% of X and less incentive to be much below that (to the extent it risks over-exertion or other problems).

What if the Sultan of Dubai offered $10,000,000 to anybody who runs under 2 hours, the leader enters the stadium heading for a 1:59, he gets 3/4ths of the way around the final lap, the eyes of the world are upon him, and he drops dead?

1. Marathon running would become more popular
2. Lots of lectures about post-colonialism

In sailboat racing this is called covering your opponent, meaning once you are ahead, you don't risk changing the strategy relative to your opponent's strategy. If he tacks behind you, you tack to cover. You only need to beat him, not break any time based record.

The 1000 lbs raw deadlift record was just broken a couple of months ago. I seriously doubt anyone in their right mind would care about some emaciated yuppies jogging slightly faster, when a man has just lifted the equivalent of a full grown crocodile.

A bit ago I predicted a 2038 sub-2 (link below, marathon records are very logarithmic). Unless the talent-sorting mechanism is poor (i.e. there are gene variants that haven't been found yet) I think this is safe. Training improvements aren't likely to yield much more.


That's a genuinely great blog post. Thanks for sharing.

I never understood why people run the whole thing. I only run half marathons and then simply double my time if I want to know how long a marathon would take me.

Aren't the incentives obvious?

Winning a race is a large benefit. There is substantial cost to professional marathoners (in effort and in recovery) to putting their all into the race. There are only a small number of racers in every race that have a serious chance of winning.

Now what is required to win? All you need to do is provide evidence to the other serious competitors that you will beat them. One way this could happen is simply the first person to pull away with a significant lead succeeds in establishing a strong advantage. In particular any serious competitor would have been better off staying up right behind them and drafting so failure to do so essentially creates significant evidence in everyone's eyes that you will win. Perhaps each racer has an internal sense of how good a day it is so a racer tries to pull away when they guess their skill + internal sense of how good a day it is enough larger than anyone else's they can burn enough energy to pull away and still win. If you guess wrong and another racer stays behind you then you just burned a bunch of excess energy while they continued to draft you so your incentive is to either pull away on first attempt or give up when it's clear you've burned so much more energy than your pursuer they will win. Once a serious competitor has pulled away they are unlikely to be challenged so they have a incentive to do so as soon as they think they will succeed since it reduces their overall expenditure. On this model essentially each racer plays a mixed strategy based on their estimate of other racer's energy distribution about when to make their break. This is probably disproved by empirical evidence.

Another possibility is that these players are essentially collaborating in an iterated prisoner's dilemma. Based on some kind of surreptitious signalling they conspire to decide who gets the race (they could play rock paper scissors ;-) ). More likely is a shame/reputation bidding game. The racers talk shit in ways that, if they fail to live up to their precommitments, harms them as they suffer a serious loss in ability to threaten away competitors and win without huge effort expenditure. I think with the right assumptions this could yield the observed result. Maybe talking trash in front of loved one's etc.. can up the commitment.

Another possibility is that weakness or strength on a given day is fairly easily evaluated either through back channel information or observation. All racers end up knowing quickly their overall place in the race and the competitive racers accept it without pushing for more.

Comments for this post are closed