Below is an excerpt from: http://www.iaaf.org/mm/Document/Competitions/TechnicalArea/ScoringTables_CE_744.pdf

Developments in the Theory of Scoring Tables

From 1920, three concepts became prominent in the theory and development of

scoring tables. These have, in varying degrees, influenced all subsequent tables.

1) The fact that each unit of improvement in an athlete's performance gets

increasingly harder as the athlete approaches his ultimate. This can be

expressed statistically as follows: the probability of any athlete achieving or

exceeding a given performance rapidly gets less as the performance rises

towards the record. The score for a performance can be derived as the

inverse of that probability. The resulting scoring table is progressive but,

applied simply, this leads to an exceedingly progressive scoring table, and the

main challenge has been to control this excess.

2) The need to be able to compare the performance of an athlete in one event

with that of another in a different event or, indeed, in a different individual

sport.

3) The wish to have a really "scientific" basis for any scoring system. With the

growing research into human physiology and sports science, it seemed

possible that a basis could be found in physiological parameters, such as

heart beat, breathing rate, oxygen uptake or oxygen depletion and so on.

The interplay of these and other interests in the development of the scoring tables

over the past 65 years is a fascinating study.

1934 IAAF Scoring Tables

At the end of the 1920's the Finnish Federation set to work on a new set of

national scoring tables. An early decision was made to drop all fractional points,

the score in each event to range from 0 to 1150 points. The aim of the new tables

was that a performance in any event should score the same as an equally good

performance in any other event. To this end, seven standard performances in

each event (labelled A-G) were selected by experienced judgement. All the

performances scoring 1000 points would only be reached rarely by combined

events athletes. All the G performances would be reached occasionally by leading

boys. The range of performances in each event between A and G was subdivided

into 20 equal steps. The number of steps between the standard performance was

divided A, 1, B, 3, C, 3, D, 3, E, 3, F, 7, G, and a progressive curve was employed

such that the slope of A was twice that of G. The whole scheme clearly works

directly for field events, but not track events using time as the performance figure.

However, if the times are converted into average speeds for the race, these can

be used equally as well as distances in developing a scoring table.

The new scoring tables were calculated by J. Ohls from Finland in 1931. These

tables were progressive and corresponded to the formula P = f (eM), where P

means the points, e is the base of natural logarithms and M corresponds to the

performances. The tables were calculated for sprint events up to the hundredths

and the performance were evaluated only by full points. A zero point value was

allotted to average performances of pupils and the 1000 point value was near the

then world records. The tables were calculated up to 1150 points

The new scoring table was such a success when introduced in 1932 in Finland

that it was adopted by the IAAF at its next Congress in 1934. The main difference

consisted in the progressive character of the Finnish evaluation as against the

linear evaluation of decathlons at the Olympic Games in 1936 and at the

European Championships in 1938, 1946 and 1950