Record-breaking predictions

What are the absolute limits to human performance levels?

Everything’s up to date in Kansas City – they’ve gone about as far as they can go!

The song in question, in a musical set in the 1890s, was actually written a hundred years later; the irony lay in the hindsight. In each generation over the last century we have seen new records set, with performances once regarded as superhuman becoming routine and unremarkable, writes Bruce Tulloh.

In 1912, for example, there was a titanic track race over 5,000m between Finland’s Hannes Kohlemainen and Jean Bouin of France, which took the world record from 15:01 to what was regarded as an ‘ultimate’ 14:36. In 1954 I can remember seeing Vladimir Kuts push Chris Chataway to a world record of 13:51.6 for 5,000m, when Emil Zatopek and Gunder Haegg were the only other men to have gone under 14 minutes. Three years later, Kuts ran an amazing 13:35, and this was regarded as an ultimate until the 1960s when Kip Keino and Ron Clarke came along... and so on.

The pattern of record-breaking

In cycling, the criterion of individual performance has always been the one-hour record. Although the sport has been largely confined to Europeans, the immense rewards available to the world number one have created a continuous pressure to improve, broken only by the two world wars. As you can see from the panel on page 4, the record improves very steeply at first – by 3k in the first year and then a further 3k over the next 10 years.

Another 10 years brings a further 3k improvement, so the progression seems to be linear, but after WWI it takes nearly 30 years for the record to improve by just 1.5k. In spite of much better bikes, better tracks and the use of altitude training, the record set in 1972 by Eddy Merckx, the Lance Armstrong of his day, was only 3.5k better than that of Fausto Coppi 30 years before.

From then on, improved technology applied to wheels, helmets, tracks and riding position have taken the record much further on; nevertheless, Chris Boardman in 2000 and Ondrej Sosenka in 2005 managed to set ‘official’ records using conventional helmets and wheels (see panel lower down). In the 33 years since Eddy Merckx’s record, they have moved it up by all of 269m!

The record progression is now following the classic asymptotic curve – flattening out more and more over the years.

In cycling, records are frequently influenced by the equipment involved. The two sports where performance depends solely on unaided physical effort, with little or no help from equipment, are running and swimming, and in both of these we have seen continuous improvement since records began.

Of course, the key difference between the two sports is that swimming records are restricted to countries with plenty of swimming pools, whereas anyone anywhere can run.

When writing a book in 1967 (1), I tried to predict where the records would be in 1980. Plotting the men’s world records for the middle- and long- distance events from 1900 to 1966, the progression seemed to be linear and I was able to make projections simply by extending the straight line. The comparison between those predictions and the actual progression that occurred is shown in table 1, below.

Table 1: Tulloh’s record predictions, 1967-1980
Event 1967 Actual 1980 Predicted 1980 Actual 2000 Actual
800m 1:44.3 1:42.7 1:42.3 1:41.1
1,500m 3:33.1 3:28.0 3:32.0 3:26.0
5,000m 13:16.6 12:57.0 13:08.4 12:39.3
10,000m 27:39.0 26:48.0 27:22.5 26:22.7

From this table, we can see that my linear projection of the records was misleading. The figures I predicted for 1980 were actually achieved only in the 800m, although by 2000 all my predictions had been exceeded.

Why do records go on improving? Simply put, each generation of young athletes looks at the current records and works harder and harder to reach them. Human nature being what it is, people train just hard enough and long enough to achieve the goal – but no more.

It is partly a question of the intensity of training and partly a question of numbers. In the nineteenth century the amateur athletes of the day did not train all year round. It only gradually became socially acceptable to spend a lot of time on sport. The ‘Flying Finn’ Paavo Nurmi was regarded as a boringly serious person in the 1920s, but he revolutionised the records by his attitude to training, just as Emil Zatopek did 30 years later.

Economic factors are also important. Over the last 50 years, more and more countries have reached an economic level that allows them to take part in sport at a global level. Before World War II there were never more than 50 nations sending teams to the Olympics; now there are more than 200 competing nations, and in most of these, the levels of nutrition and education are increasing year on year.

In Nigeria, for example, with a population of 90 million, the percentage of people able to take part in sport will have increased from perhaps 10% when the country first started sending teams to the Olympics to something over 60% now, bringing 50 million more people into contention in one country alone.

These factors account for the linear progression of the records over most of the modern era, but we are now seeing a reduction in the rate of improvement. It took 70 years for the 100m record to come down from 11.0 to 10.0, but since Jim Hines ran 9.95 in 1968, the record has come down by only 0.18 seconds over 37 years. Similarly, the 200m record has improved by only four-tenths of a second since 1968. Clearly we are approaching human physical limits in the sprint events, but is the same true of longer distances?

In his book Lore of Running, Tim Noakes examines a number of mathematical treatments of world records (2). Those which assume a linear progression have already been shown to be ridiculous by predicting, for example, a two-hour marathon by 2004.

A much more convincing model, quoted by Noakes, was devised by New Zealand statistician Hugh Morton (3). Using figures for the 1,500m record, he produced good evidence that the progression is asymptotic – a curve that flattens out as it gets nearer to the ultimate. Some of his predictions, made in 1983, are shown in table 2, opposite. Again, all these are for men’s records.

We can see that the predictions made by Morton in table 2 are more accurate than my own in table 1. But they are still flawed because other factors have come in and changed the pattern of progression.

Table 2: Morton’s record predictions, 1980-2010
  1980 2000 2000 2010 2005
Event Actual Predicted Actual Predicted Actual
800m 01:41.7 01:40.9 01:41.1 01:40.1 01:41.1
1,500m 03:30.8 03:27.5 03:26.0 03:25.4 03:26.0
5,000m 13:08.4 12:51.0 12:39.3 12:42.6 12:37.3
10,000m 27:22.5 27:03.0 26:22.7 26:46.0 26:17.8

Whatever the mathematical predictions, the actual improvements in records depend on physical improvements in the athletes involved. Muscle strength is obviously a major factor, and this depends on the number and size of the muscle fibres, which can be improved by training. Reaction time is not capable of significant improvement, because it is limited by the speed of conduction of nerve impulses along the fibres.

However, technique can be improved – by better recruitment of fibres, better coordination and better relaxation. Maximum power output can be maintained for longer by having larger stores of creatine phosphate. These improvements will be measured only in hundredths of a second.

We have seen how the use of anabolic steroids has produced some remarkable improvements in sprinting performances. Ben Johnson produced a world 100m record, subsequently disallowed, in the 1988 Olympic final, and it is likely that his performance in the 1987 World Championships was similarly steroid-enhanced.

In the women’s events, the existing world records, some dating back more than 20 years, are far in advance of the performances of the last few years. Compare the 400m record of Marita Koch, 47.6 in 1985, with the battle in this year’s World Championships, which was won in 49.55. In the case of the East German records, we know that they were drug-aided.

Only an idealist or a fool would claim that the other sprint records are genuine, but I think it is reasonable to expect that any record set with the aid of drugs can eventually be achieved by legal means.

In the men’s sprints we can expect only marginal improvements – less than a tenth of a second off the 100m record over the next 20 years, while the women’s events will slowly edge towards the existing world records, as long as only legal methods are permitted.

In the field events, only Jan Zelezny’s javelin record and Jonathan Edwards’ triple jump records were set within the last 10 years. Plotting record progression is complicated by two factors: the use of steroids in many countries in the 1970s and 80s, and improvements in equipment.

Looking at the results of the last three global championships, it appears that the progression is now asymptotic rather than linear, but there could be surprises. The more sophisticated events are dominated by Americans and Europeans, but we may see changes when more countries move into these areas of sport. We are seeing the beginnings of an explosion of interest in China, while the huge populations of India and South America are as yet largely uninvolved.

As far as endurance events are concerned, the most important physiological factor is transport of oxygen, which involves heart, lungs, blood vessels, numbers of red blood cells and haemoglobin levels. Other relevant factors are power-to-weight ratio, body fat percentages, heat dissipation and the ability to absorb a large volume of training.

The key question is – why has the rate of improvement in long distance events been so much greater than in the explosive events?

There are two reasons: firstly, oxygen uptake responds to large volumes of training; secondly, distance running is a cheap and simple sport, whose rewards have attracted thousands of talented young runners from East Africa. Although fewer and fewer Europeans are willing to make the sacrifices involved in training for distance events, thousands of third world runners are more than willing. And as more women get involved, their performances are likely to improve at a faster rate than the men’s.

Drugs feature in distance events too. In the women’s 3,000m, the world record stood in the mid-8:20s from the mid-seventies, but on one day in 1993 five young Chinese runners, most of them teenagers, broke the world record, with the best of them, Wang Junxia, running 8:06.11. Although these runners and their coach were subsequently banned by the Chinese authorities, none of them had tested positive for a banned substance and, to this day, all the women’s records from 1,500m to 10,000m are held by China.

The human factor

Leaving aside mathematical predictions, coaching schools and economic factors, there is one more major and unpredictable factor – the exceptional individual. He or she may appear anywhere in the world, combining a massive amount of natural ability with huge ambition. If that person also has the intelligence and the application to work hard for a long time, and happens to find the right coach or the right opportunity, anything is possible.

In the modern era, we have witnessed great field athletes like Javier Sotomayor, the Cuban high jumper, or Jan Zelezny, Hungary’s multi- medallist in the javelin; Michael Johnson of the US in the sprints; and distance runners Haile Gebreselassie (Ethiopia) and the UK’s Paula Radcliffe. These are the kind of people who make a nonsense of mathematical analyses and change the world’s perception of their sport.

When US long jumper Bob Beamon jumped 8.90m in 1968, this represented an improvement of nearly 22ins in imperial terms – a feat that statisticians said was 50 years ahead of its time. However, once he had done it, others set that as their target, though it was 23 years before Mike Powell jumped 8.95m.

Will women’s records ever catch up with men’s? This questions is dealt with in more detail in the article on page 11, but briefly the answer is no. Across the whole range of athletic performances, women’s records remain consistently at 90% of men’s. Although a top-class woman will beat most men, she will not beat the very best.

If we compare current records, we can see that, whereas Radcliffe’s marathon record stands at 92% of the men’s, it would take an 8m long jump or a 16m triple jump to reach the 90% mark in those events. It is interesting to note that if we add 10% to Daniel Komen’s current 3,000m record we get 8:08, which is within reach of Wang Junxia’s ‘impossible’ 8:06.

As more countries allow women an equal chance, the women’s records will continue to improve more rapidly. In events new to women, such as the hammer, the steeplechase and the 10,000m, improvement will probably continue to be linear for at least the next 10 years, but then the curve will flatten out, as it has done for men.

Where will it all end? Looking at film of the Paralympics, it did occur to me that artificial limbs may one day come to be more efficient than homegrown ones, but we had better not go down that route!

Despite the inevitable flattening of improvement curves, records will continue to be set – the financial rewards involved will make sure of that! I can predict that in 20 years’ time the women’s standards, unaided by drugs, will have moved up to the level of the drug-assisted records – which means a 47-something 400m and a sub-3:50 1,500m. There is a tremendous crop of young African women, who can be expected to take the distance records down to around 14 minutes for 5k and 29 minutes for 10k.

Progress will be slower in men’s events, but my predictions for 2025 are set out in table 3 below.

Table 3: Tulloh’s record predictions for 2025
Event 2005 Actual 2025 Predicted
100m 9.77 9.69
200m 19.32 19.20
400m 43.32 42.54
800m 1:41.11 1:39.40
1,500m 3:26.00 3:23
5,000m 12:37.35 12:22
1,0000m 26:17.5 25:44
Marathon 2:04:55 1:59:50

The danger is that, as the records move further and further away from the abilities of the ordinary person, pressure to achieve these standards will demand greater and greater sacrifices from the aspiring young athlete, whatever his or her sport. Training becomes harder and starts younger, becoming more and more damaging. Should we encourage young people to go to these levels?

We hear of the successes – Tiger Woods, Carl Lewis, the Williams sisters, Olga Korbut – but we seldom hear of those who made great sacrifices but didn’t quite make it and finish up crippled in body or mind.

Leaving aside the dangers of drugs and the prosthetics we see in the Paralympics, the biggest threat to sport lies in the realm of genetics. There is a distinct possibility of the emergence of genetically- modified athletes within the next 20 years, as medical research discovers the genes for stronger muscles, larger size, more red blood cells and so on.

Unscrupulous scientists, eager to have stronger and faster football players on US college teams, could use this knowledge either to modify young athletes or to select embryos with the right genes in them. Just as natural selection produced Tyrannosaurus Rex, so artificial selection may produce 3-metre giants who will transform all existing ideas of what is possible.

The UCI (Union de Cyclisme Internationale) one-hour record list demonstrates the effect of economic development and the influence of technology on modern sport.

Note first the northern European bias, corroborating my point about the erstwhile stranglehold on peak sporting performance enjoyed by developed nations. Willie Hamilton stands out as the only non-European – and non-white. New record holder Ondrej Sosenka perhaps represents a new wave of riders from other regions.

Then there is the remarkable record that no man will ever beat. In fact, Chris Boardman’s 1996 performance is now listed by the governing body not as a record, but as the ‘Best Hour Performance’ as opposed to the ‘UCI Hour Record’, held by the great Eddy Merckx for 12 years from 1972 and again for five years from 2000 to earlier this year. Confused that a record enjoys two separate stretches as the figure to beat? The answer lies in arguments over new aerodynamic bike and helmet design – and unconventional riding positions – as exemplified by Graeme Obree’s controversial ‘superman’ profile. The UCI announced in September 2000 that ‘in view of the fact that new regulations would make the current hour record virtually impossible to beat’ they were winding the clock back to Merckx’s 1972 ride.

Casting further doubt on what truly is a record is the fact that Mexico is at high altitude, while Sosenka’s 2005 performance was in lowly Moscow.

Progression of world one-hour cycling record

Progression of world one-hour cycling record

Evolution of one-hour cycling record
Year Rider Country Distance (km)
1893 H Desgranges FRA 35.325
1894 J Dubois FRA 38.22
1897 O v d Eynde BEL 39.24
1898 W Hamilton USA 40.781
1905 L Petit-Breton FRA 41.11
1907 M Berthet FRA 41.52
1912 O Egg SUI 42.36
1913 M Berthet FRA 42.741
1913 O Egg SUI 43.525
1913 M Berthet FRA 43.775
1914 O Egg SUI 44.247
1933 M Richard FRA 44.777
1935 G Olmo ITA 45.09
1936 M Richard FRA 45.398
1937 F Slaats NED 45.558
1937 M Archambaud FRA 45.84
1942 F Coppi ITA 45.871
1956 J Anquetil FRA 46.159
1956 E Baldini ITA 46.393
1957 R Rivere FRA 46.923
1958 R Rivere FRA 47.346
1967 F Bracke BEL 48.093
1968 O Ritter DEN 48.653
1972 E Merckx BEL 49.431
1984 F Moser ITA 50.808
1984 F Moser ITA 51.151
1993 G Obree GBR 51.596
1993 C Boardman GBR 52.27
1994 G Obree GBR 52.719
1994 M Indurain ESP 53.04
1994 T Rominger SUI 53.832
1994 T Rominger SUI 55.291
1996 C Boardman GBR 56.375
2000 C Boardman GBR 49.441
2005 O Sosenka CZR 49.7
My thanks to Joe Beer for cycling statistics

Bruce Tulloh was European 5,000m champion in 1962 in a time of 14:00.6. The championship record is now 13:10, but the 2002 title was claimed in 13:38


  1. Tulloh on Running, Bruce Tulloh, Heinemann 1968
  2. Lore of Running, Tim Noakes, Oxford University Press, 1985
  3. Aus Jour Sports Science 3,7-10

Get on the road to gold-medal form and smash your competition.
Try Peak Performance today for just $1.97.

Tagged in winning
Privacy Policy [opens in new window]