Sometimes running is nice, sometimes it is not going perfect. At some distances you often try to improve your best time, but others you hardly ever run. So some of your records are strong, but others aren't.
With world records the same. But which ones are strong? How much faster the weak records could be at least? And how do women and men differ? Are women relatively better at long distances than men, as sometimes is said?



In the graph to the right horizontally distances and vertically the number of seconds of world records (women and men) are shown. The upper line shows women's records, they need more seconds. (In all graphs the line of the women is brownish red, of men dark blue.) This graph is useless because all distances up to the marathon clump together in the lower left cell! (Almost all records used come from the IAAF-site with statistics.)

[graph] A way to circumvent this is a logarithmic graph, see the graph to the left. Big numbers are compressed, therefore all distances now are rather equally spaced on the horizontal axis, as are the times on the vertical one.
We are ourselves logarithmic in a sense. We like having records at 100 and 200 m, differing 100 m, but we see no use in having records at 5000 and 5100 m, also differing 100m. That would feel as nearly the same distance, so we like to have only records at doubling distances, or something like that. (Every doubling on a logarithmic axis takes the same amount of millimeters.) This has led to the official running distances. Added are very short distances which fit into indoor accomodations, distances that are dictated by our measuring units (1 km, 1 mile, 2 km, 2 mile) and records over a certain amount of time (1, 12, 24 hour). The marathon (42195 m) and half of it also are added.
The whole series of running distances than becomes (M = mile = 1609.344 m): 50, 60, 100, 200, 400, 800, 1000, 1500, 1 M, 2000, 3000, 2 M, 5000, 10 000, 15 000, 10 M, 20 000, semi-marathon, 25 000, 30 000, marathon, 30 M, 50 000, 40 M, 50 M, 100 000, 150 000, 100 M, 200 000 plus races of 1 hour, 12 and 24 hours. Other distances exist (a quarter of a marathon, 6 hours) but they are less official and they are not shown in my graphs.
The second graph still does not lead to much insight because all points nearly form a straight line.

[graph] A bit more can be concluded when vertically the mean velocity (km per hour; not logarithmic) is plotted (graph to the right; the upper line are men's records), but still the rectangle of the graph is mainly empty. (You can open an enlargement in its own screen (24K).)

[graph] I have tried to find a graph with all points more spreaded, which resulted in the graph to the left. This shows how much records deviate from the straight line in the upper left graph. Unfortunately the numbers along the vertiacl axis have become senseless now. (They indicate velocities in an unclear way; the upper line are men's records.) This picture (enlargement: 24 K) easily illustrates how different the strength of records can be!
In general distances ran more often show stronger world records.



Leaving out weak records makes the line less chaotic, resulting into the graph to the left. Some chaos remains, at distances longer than the marathon. More about that below ('the ultra-problem').
What does the graph show? (Enlargement: 24 K.)The upper line are men's records. The very small waves of the upper left graph now are enlarged. It is a curve with two summits, having four parts. At first the line inclines (sprint), than it declines (mid distances), but soon it raises again (long distances) almost as a straight line. Than in an unclear way it goes down: ultra running. (At distances longer than 24 hours decline is even steeper. I did not take them into account.)

Four systems
The waving line exists of four parts and we have four energy systems. That will not be accidentally. The graph does not show at which distances a certain energy system is dominant, but the four sytems each lead to a characteristic decline, I think. 'Decline' means the way we go slower at longer distances.
The first part shows the sprints. Here the alactic anaerobic energy system is dominant. This is working up to about 20 seconds. The line in the graph goes up, which means that our running is rather efficient on short distances. Decline is low. It always will have been favourable to sprint to trees or shelter when tigers or lions were approaching!
The second part shows the mid distances. According to the this graph distances from say 250 to 1250 meters are concerned. The energy system is called lactic anaerobic, souring legs that means. The line goes down, decline is strong, we are not good at it.
Third part: long distance running. For men it are distances up to about 40 kilometers, for women up to around 45 kilometers. Below more about this interesting difference. The enery system determining the decline pattern is the aerobic system, in particular the burning of carbohydrates (sugars). The line goes up, we are good at it. It is said that humans can catch up with all animals because we can go on and on. Faster animals give up but that obstinate two-legged ape doesn't. As long as souring of the legs is standable.
The fourth parth of the graph exists of ultra running. Decline is strong, the body works on the burning of fat. We have plenty of that, but making energy free from fat goes slow.
In fact a fifth part of the graph exists: distances longer than 24 hours. Here decline is much stronger, taking some rest underway is unavoidable. Obstinacy reaches its limits. I myself have once done a walk of 24 hours and although walking costs much less energy than ultra running I nevertheless fell asleep when walking. Happily I woke immediately so I did not fall on the pavement.

Predicting cautiously
Sometimes someone predicts the development of world records in the future. Honestly, I do not believe those predictions. What I have done is a kind of making predictions too, but only about possibilities of weak records. Cautiously the weak records were screwed up until a smooth graph resulted. This we do know: in the future improvements of records will come to an end, then all records will be just as strong and decline will be systematic. A graph will not be chaotic any more.
About todays strong records few if any hypotheses can be made. I have no idea which one is the strongest of all records. Statistically extrapolations always can be made, but it remains believing in clairvoyance. That's why I only say things about weak records.
Because I only made cautiuos predictions every day the whole system can be overthrown, when one of the strong records is improved. But this is a rare event.



To the right you see a graph with horizontally distances up to 100 km (logarithmic) and vertical the proportion between men's and women's records. (Here level '1.10' for a certain record means that women ran 10% slower than men at that distance, so multiplicating the record of the men by 1.10 gives the women's time as result.)
But how chaotic that picture is, what about that? The green line shows proportions as following from existing world records. The orange line shows a first attempt to make things smoother, but still no one will believe that the difference between men and women is so chaotic. With some records something will be wrong, but what?

[graph] To get smooth curves I have thrown away several records! The graph to the left shows the result (enlargement: 20K).
I am not the only one who distrusts some of women's world records. It is known that Eastern Europeans have done tricky things before the downfall of the Berlin Wall. It remains strange that they wanted to prove the superiority of their socialism by way of results in sports, but that they used extremely inferior means for that goal, often without informing the athletes. After the downfall of the Wall taking drugs could have been a way to overcome deplorable financial situations. I do not say that things like that have happened, but I am not completey free from distrust.
It is not known which records have been tricked, but when I threw away some records suddenly the graphs became smooth! This does not prove anything of course. It also does not prove that the remaining records are pure and not tricked.
By the way: I did not throw away records until the curves were smooth enough, I threw away those records that I distrusted most and than suddenly the graphs were as shown.

[graph] As 'world records' for these calculations I have adopted the following: at 100 and 200 m Marion Jones (number 2 of the all-time list), at 400 m Marie-José Pérec (nr 3), at 800 m Ana Quirot (nr 3), at 1500 m Hassina Boulmerka (nr 12), at 3000 m Gabriela Szabo (nr 5), at 10 000 Paula Radcliffe (nr 2). The difference between real and adopted records can be judged on the graph to the right. The yellow line (real records) resembles more the form of the men's line, and that is exactly what can be aspected when drugs are used.
By throwing away some records my predictions became more cautious, and that is what I wanted. After that action all curves became smooth. Except...

The 200-problem
Michael Johnson's world record at 200 m seems unstatistically strong. At least it leads to a big bump on an otherwise smooth graph of female/male proportions (green line of the graph to the left). The bump disappeared by taking the second best time ever (orange line). This is Michael Johnson too, also at the Atlanta Olympics, but the semi-final.
Is Johnson's record really that special or is something else at stake? You can benefit much from a strong wind during the curve when running 200 m whilst having legal wind assistance during the last 100 m. There wind speed is measured, so sometimes the measurements does not lead to correct judgement. Who knows the Atalanta story?
A small orange bump can be seen at 400 m. I decided to keep it like that, because all solutions looked artificial.

Removing the big bump or not, the conclusion must be that women are relatively best at marathon to 100 km, and at 100 m they are almost as good. The difference between women and men peaks at about 2400 meter, six laps on a track.



To the right you see a part (100 - 100 000 m, situation dec 2002) of the graph already shown above ('strong and weak records'). The graph is unfinished for distances longer than a marathon. The number of people running ultra distances is (still) low and so little can be said about the strength of ultra records. At most two of the men's records can be called 'strong' (100 km and 24 hour) and only one women's record (100 km).
At which distance long-distance running becomes ultra running? A few years ago I started to make these graphs. At that moment it seemed that records of men and women on long distances (1500 - 30 000) showed the same pattern, it seemed that women were a bit better at the marathon, but their ultra running was much weaker.
But after throwing away several records (see above) suddenly the patterns of 1500 - 30 000 were different for women and men. Records on the marathon were improved also and now it seems clear that for women the marathon is the longest long-distance event while it is the shortest ultra running distance for men. For the rest everything is unclear so I have not made the graph smoother.
It is interesting that women can go on for a longer time before they enter the ultra running domain. The lower yellow line in the graph to the right shows personal bests of Tegla Loroupe (former marathon record holder). It is almost a straight line, there is no turning downwards before the marathon is reached. The upper yellow line shows personal bests of Haile Gebrselassi. When he should be able to run such a straight line as Tegla Loroupe he would run a 2:02 marathon. But his line bends downwards. As an aside: his world record on 5000 m turned out to be not as strong as his 10 000 m, but as you can see it is fully in line with his other bests. His personal peak maybe is at about 10 000 m, but some day there will be a runner who has his peak at 5000 m.

Because sometimes an ultra-run has been won by a women it is said that women are relatively much better than men at those distances. But record lists did not support that thesis, on the contrary. Women's records for ultra distances were all much weaker than men's records, until a rather good Japanese marathon runner (2:26:09) tried a 100 km. Tomoe Abe smashed the women's record and at the same from that moment all men's records did look pale. Two conclusions are possible: all men's record are weak or very weak (we have to wait until good marathon runners become ultra runners), or indeed women are relatively much better in ultra running, or both. That's all that can be said at this moment.
Because I wanted to make some predictions I took for men a straight line from marathon to 24 hours and two lines for women: from marathon to 100 km and from 100 km to the (weak) 24 hours record (ran on an indoor track!).


[table as picture]

Now a table with my predictions. Women left, men right. First the existing record is given (or blue an adopted record), than my prediction, and how many percent that improvement would be (Dutch 'eraf' means 'lower'). For distances above 100 km preditions for women's records are much too cautious. Predictions for men's ultra records likely are also a bit cautious.
Predictions for records of 1 and 12 hours are given below. Percentages mentioned here are not comparable to percentages of the bigger table (Dutch 'erbij' means 'more').
When you like to become famous specialize on distances with a very weak record!

[table as picture]

Weia Reinboud (weiatletiek (@) (More on athletics here.)