Fair Decathlon Model. Part 9: What Should a Decathlon Scoring Table Actually Measure? (2)

FAIR DECATHLON MODEL: PART IX

Ten Events. One Decathlon.

What Should a Decathlon Scoring Table Actually Measure?

A conceptual continuation of the Fair Decathlon Model series

Rafal Snoch   |   for Decathlon2000.com   |   August 2026

After the zero-point audit

Part VIII completed a task that had remained unresolved. The zero-point parameter B was audited event by event against the lower end of the decathlon population, rather than being left in place simply because an earlier table had put it there. Some boundaries moved. Others survived the audit. The classical FDM framework was finally complete on its own terms.

One practical piece, however, was missing. Part VIII published the new FDM-B formulas, but did not show their most intuitive consequence: what actual performances now correspond to familiar point levels, especially near the lower end of the scale where the changes in B matter most. I should have included that table there. To correct the omission, the annex to this part gives the first valid performance reaching every 100-point threshold between 100 and 1200 points in each event.

That technical exercise also exposed a more basic question. B tells a formula where a performance reaches zero points. It does not tell us what that zero is supposed to mean, or what a decathlon scoring table is ultimately trying to measure.

To ask that question properly, it helps to leave the coefficients alone for a moment and return to the stadium.

Ten events, one decathlon

Imagine gathering a large group of healthy adult men who have never trained for athletics.

Give them basic instruction, ensure that every event can be performed safely, and bring them to a stadium for two days. Their task is not to produce ten isolated personal bests. Their task is to complete a decathlon.

That distinction changes everything. They must not run the 400 metres so hard that they cannot return the next morning. They cannot treat the 1500 metres as a fresh endurance test. They must manage risk in the jumps, learn enough technique to clear a height in the pole vault, and preserve enough physical and mental energy to finish all ten events.

Each participant should be evaluated only after completing the full sequence. The experiment would not merely tell us how far an untrained man can throw a shot or how quickly he can run 400 metres. It would show us what those performances look like inside a decathlon.

A shot put or a “shot drop”?

At the lowest end of a scoring table, the distinction becomes obvious.

A very short shot-put result may technically be measurable. The implement leaves the athlete’s hand, lands inside the sector and produces a number. But at some point the performance stops resembling a genuine athletic throw and begins to resemble what Mikko Malmivuo aptly called a “shot drop.”

Part VIII asked where the scoring boundary should sit by using real decathlon data. The thought experiment here asks something different: what sporting distinction is that boundary trying to represent? A healthy but untrained participant who genuinely attempts the full decathlon may help us recognise the difference between a weak performance, a minimal technical completion and an action that merely satisfies the formal requirement of recording a mark. But his average result would not automatically become a new value of B.

The same distinction appears in a less visual form in the 1500 metres. FDM retains 480 seconds as the zero-point boundary. A completed 8:01 lies on the zero side of the scoring function. A non-finish is also not a positive scoring performance. Numerically, the formula has reached the same floor; sportingly, the two situations are not the same. One athlete completed the tenth event. The other did not.

That is an important limit of B. A zero-point boundary is a property of a scoring function. It is not, by itself, a complete definition of what counts as completing the decathlon.

Ten separate tests would answer the wrong question

Suppose the same men were tested differently. On Monday, they run 100 metres while fully rested. On Tuesday, they perform the long jump. On Wednesday, they throw the shot. Each event is separated by enough recovery for everyone to produce the best possible isolated result.

Those results might be useful for general fitness research. But they would not describe the lower end of the decathlon.

A decathlon is not a laboratory in which every event begins from the same physical condition. The athlete entering the long jump has already run 100 metres. The athlete entering the shot put has completed a sprint and a jumping competition. By the time he reaches the 400 metres, he has performed four events.

The second day begins after the accumulated fatigue of the first. The discus follows the hurdles. The pole vault is the eighth event and follows seven completed events. The javelin comes after a potentially long and exhausting pole-vault competition. The 1500 metres is performed after everything else.

Each performance carries the weight of those that came before it.

What does “effort” mean?

A common claim is that a scoring table should represent the effort required to achieve a result. That sounds reasonable. But what exactly is meant by effort?

If we consider only the visible effort made during the event itself, we miss a fundamental part of the decathlon.

The effort required to run 400 metres in a decathlon is not only the effort produced during those 400 metres. The athlete reaches the start carrying four earlier events and a strategic constraint: he still has a second day to complete.

Later events intensify the same problem. By then, fatigue, technical disruption, risk management and the length of earlier competitions have accumulated into the conditions under which the next performance is produced.

It does not begin when the gun is fired or when the athlete enters the circle. It includes the burden carried into the event from everything that has already happened, as well as the need to preserve enough resources for everything still to come.

Reverse the order

A simple thought experiment reveals how important the sequence is. Imagine reversing the decathlon:

1. 1500 metres
2. Javelin
3. Pole vault
4. Discus
5. 110-metre hurdles
6. 400 metres
7. High jump
8. Shot put
9. Long jump
10. 100 metres

Damian Warner has run 10.12 seconds in the 100 metres during a decathlon. Would he run 10.12 if the 100 metres were the final event, after two days and nine previous competitions? Almost certainly not.

The obvious reason is fatigue. The mark of 10.12 was achieved in the opening event, when Warner was physically fresh and before the accumulated cost of the decathlon had begun to build.

But reversing the order changes something else as well: the competitive purpose of the race.

Imagine a major championship in which Warner reaches the final event twenty points ahead of Niklas Kaul, while the next athlete is already two or three hundred points behind. Kaul typically runs the 100 metres in roughly the 11.20–11.40 range. Warner would not need to chase 10.12. His task would be to run only as fast as necessary to protect the lead. For a sprinter of his quality, a controlled race around eleven seconds might be entirely sufficient.

The 100 metres would suddenly begin to resemble the role that the 1500 metres often plays today. The athlete would know the standings, know roughly what his rival can do, and run against the required margin rather than against his absolute personal limit. “How fast can I run?” would become “How fast do I need to run?”

The reverse would also apply to the 1500 metres. Placed first, it would be run on fresh legs, without nine-event standings already dictating the target. It would no longer be the final tactical equation of the competition.

So event order changes not only fatigue. It changes information, incentives and strategy. The same athlete, possessing the same underlying abilities, would produce a different profile because the meaning of each event inside the competition had changed.

Invite the world-record holders

Part VIII ended with one methodological fact left deliberately unexplained: throughout this project, FDM has never used specialist world records to determine the relative value of the ten events. The reason begins with the same distinction between an isolated performance and a performance produced inside the whole decathlon.

Now repeat the thought experiment at the opposite end of the performance spectrum. Invite the world-record holders in the ten individual events to enter one decathlon.

Every athlete is given the same rule: his result in his specialist event will enter the event ranking only if he records valid marks in all ten disciplines and completes the 1500 metres. If Usain Bolt runs 9.58 in the opening 100 metres and then goes home, the 9.58 does not count for this experiment.

This condition matters even for the first event. Bolt could reach his speciality before fatigue has accumulated, but he would still have to accept the obligation attached to the performance: nine more events and a finish on the second day. The experiment is not asking for ten exhibition marks. It is asking what the best specialist in the world can produce while also carrying the obligation to complete an entire decathlon.

The later specialists face the sequence more directly. Would Javier Sotomayor really clear 2.45 metres in the fourth event, after the 100 metres, long jump and shot put? Would Armand Duplantis clear 6.30 metres or more as the eighth event, after a full first day, the hurdles and the discus? Would Hicham El Guerrouj run 3:26.00 after nine completed events?

We cannot know the answers without staging the competition. That uncertainty is precisely the point. A specialist world record tells us what an athlete achieved when his preparation, warm-up, risk, energy and competition day were organised around one event. It does not tell us what the same athlete would achieve in that event while carrying the obligations of the other nine.

A real example provides a rare glimpse of this difference. Renaud Lavillenie did not enter the 2013 Decastar in Talence merely for amusement. He arrived with a double ambition: to complete his first decathlon and to challenge Tim Lobinger’s 5.76-metre best pole-vault performance achieved within a decathlon.

The second goal could not be separated from the first. Under the criterion then applied to recognised decathlon bests, Lavillenie also needed to finish all ten events with more than 7000 points. A great vault followed by withdrawal would not have served the purpose. The mark had to belong to a sufficiently complete decathlon.

The pole-vault progression was tailored to the record attempt, with the heights ending in seven centimetres. Lavillenie entered at 5.27 metres, cleared 5.47, and then failed at 5.67. Had he continued successfully, the intended route led to 5.77 metres – one centimetre above Lobinger’s mark.

This makes the case unusually revealing. Lavillenie was not a reluctant specialist preserving himself for another competition. The vault was the central objective of his appearance. He had exceptional motivation, the competition was arranged around the attempt, and he knew exactly what historical mark he was pursuing.

Yet he finished with 5.47 metres and a total of 6676 points. Earlier in the same season he had cleared 6.02 metres in an individual pole-vault competition. The comparison deliberately uses his season best, not his lifetime personal best: it contrasts two performances by essentially the same athlete at the same stage of his career, but in two different athletic contexts.

The entire 55-centimetre difference cannot simply be attributed to fatigue. Pole vault varies with conditions, attempts, technical rhythm and the competition itself, and Lavillenie was not an experienced decathlete. But that is precisely why the example matters. Even with special motivation and a progression designed around the record attempt, the eighth event of a completed decathlon did not reproduce his specialist level from the same season.

If all ten world-record holders accepted the completion rule, they would not recreate the record book inside one stadium. They would create a new set of performances conditioned by event order, cumulative fatigue, unfamiliar techniques, risk management, strategy and the obligation to finish. Those conditioned performances – not the isolated records they brought with them – would be the relevant evidence for a decathlon scoring table.

The specialist record answers one magnificent question: how far can human performance be pushed in this event? The decathlon table must answer another: what is this performance worth when it is produced here, at this point in the sequence, by an athlete who must complete the whole?

Is the 400 metres in a decathlon the same event as an isolated 400 metres?

This leads to a philosophical question that every scoring system should answer: what is the 400-metre scoring table in a decathlon supposed to measure?

Is it measuring the abstract quality of a 400-metre performance relative to the limits of human sprinting? Or is it measuring a 400-metre performance achieved as the fifth stage of a decathlon?

Wayde van Niekerk ran his world record after preparing exclusively for the 400 metres. He did not run it after completing the 100 metres, long jump, shot put and high jump. His result represents the extraordinary upper limit of a specialist. It does not directly tell us the correct exchange relationship between 400 metres and the other nine events of a decathlon.

The same question applies to every event. Once a mark is produced inside a decathlon, it belongs to a different context from the same mark produced in an isolated specialist competition. The issue is not whether the specialist performance is superior; it is whether specialist limits provide the right exchange values for a combined event.

These are not equivalent questions.

The forest and the trees

A scoring philosophy built around specialist world records naturally sees ten separate disciplines. Each event has its own history, its own record, its own technical culture and its own limit of human performance. From that perspective, it may appear logical to anchor the scoring table to the best specialist performance ever produced in each event.

But that approach risks seeing the trees while missing the forest.

The decathlon is not a competition asking who can come closest to ten specialist world records. It asks who can create the strongest complete performance across a fixed sequence of conflicting demands.

The athlete must develop speed without becoming only a sprinter, strength without becoming only a thrower, endurance without sacrificing explosive ability, and technical competence across events that normally require years of specialisation. He must also distribute his energy across two days.

The scoring table should therefore describe the internal relationships of this complete system.

From zero thresholds to exchange relationships

Part VIII made one feature of the classical formulas especially visible. The location of B can move substantially while the high-level part of a scoring curve changes very little, provided the remaining coefficients are recalibrated to the same upper structure. The zero point and the ordinary high-level exchange relationship are therefore not the same thing.

That matters because every scoring table is, underneath the formulas, an exchange mechanism. It decides how much improvement in one event compensates for a loss in another. A tenth of a second in the 100 metres, a centimetre in the high jump, a metre in the discus and a second in the 1500 metres become comparable only because the table assigns exchange values between them.

The traditional power functions make those exchange values change with performance level. When the exponent is above one, differences are compressed near B and become more valuable farther from the zero point. When the exponent is below one, the opposite happens. Part VIII showed that changing B can change that curvature considerably even while familiar elite scores remain almost untouched.

This does not prove that the curves should be linear. Nor does it prove that every changing exchange rate is artificial. It does, however, clarify the question. If the table is meant to describe one decathlon, then any non-linearity should earn its place by describing relationships that actually exist within the decathlon – not merely by producing an elegant mathematical shape around an inherited zero.

Other combined sports sometimes make the idea of exchange more explicit. In Nordic combined, ski-jumping points are translated into a time relationship for the cross-country race. The sports and scoring problems are very different, but the analogy is useful: the numbers are not independent statements of absolute athletic worth. They are a mechanism for combining unlike performances into one competition.

Decathlon scoring hides that exchange inside ten separate formulas. Thinking in terms of exchange relationships does not yet tell us what the best formulas should be. It does tell us what question those formulas must answer.

Why FDM may calibrate so well

This may help explain why the Fair Decathlon Model calibrates surprisingly well.

FDM does not explicitly calculate the fatigue, strategic trade-offs or technical disruption that accumulate from one event to the next.

It does not need to assign separate theoretical values to those effects. They are already present in the empirical data.

It does not have to reconstruct that context afterwards, because the marks in its calibration data were produced inside actual decathlons, in the prescribed sequence and under the obligation to complete the whole competition.

The athletes represented in the data also made real strategic choices: how much energy to spend, when to accept a safe result, when to take risks, when to chase a position, and when simply to do enough. The reversed-100-metre thought experiment does not introduce a new principle. It reveals one that has always been present in the data.

This does not make the data pure. They also contain changes in training, athlete selection, national traditions, tactical behaviour and population structure. But these are not external contaminants added to an otherwise context-free event. Many of them are part of what the decathlon actually is.

Perhaps FDM calibrates so well because it asks the data a question the data are capable of answering. Instead of importing ten external specialist ideals, it learns from performances generated by the exact system the scoring table is meant to describe.

What should the table reward?

A decathlon table should not reward proximity to ten specialist ideals. It should reward the quality of the complete decathlon performance.

That means recognising that the value of a mark depends not only on the mark itself, but on the system in which it was achieved.

That context changes continuously through the competition. An opening sprint, a technical event performed after accumulated work, and the closing endurance test are not interchangeable with isolated specialist performances; reverse the order, and their physical and strategic meaning changes again.

The scoring table is therefore not merely a converter from seconds and metres into points. It is an attempt to describe the exchange relationships among ten successive parts of one demanding athletic whole.

One decathlon

The experiment with untrained men provides a simple way to see the issue.

Do not ask them for ten isolated maximum performances. Ask them to complete one full decathlon. Tell them that success means returning for the second day, recording ten valid marks and reaching the finish of the 1500 metres. Then observe how their performances change as fatigue, fear, technique, pacing, information and strategy accumulate.

That is the environment the scoring table is supposed to describe.

The same principle applies from beginners to world-record-level athletes.

The ten events are the visible components. The decathlon is the thing being measured.

APPENDIX TO FAIR DECATHLON MODEL - PART 9
First valid marks reaching each 100-point threshold.

Rafał Snoch for Decathlon 2000