What Does Fastest Mean?

IAAF rules for 100 m sprinting do not recognise a record time if the wind assistance exceeds 2 m/s (a very gentle breeze), and flag it as altitude assisted if run at a location above 1000 m altitude.

How far should this recognition of the assistance or hindrance of wind and altitude be taken ? When events are timed to 0.01 second precision, could records and outcomes be determined by climatic conditions, or other chance events ? Is it fair to equate a time for 100 m race where the athlete had a 2m/s tailwind with one run into a 2 m/s headwind ? If not, how can events run in different climatic conditions be compared ?

Physicist Jonas Mureika Ph D has proposed an approximate model for adjusting times to account for the wind and altitude variables. He has posted a list of best sub 9.90 second performances on the sub-10 discussion forum to serve as an example for how performance might be more fairly compared.

Do Mureika's methods produce results that make more sense than the alternative of ignoring wind and altitude assistance within the wide-span accepted presently by the IAAF ?

Scientific comparison of top 100m performances
Author: JRM, May/13/2003 20:43:40 [+01]

JRM posted the following comparison of the best recent 100m times on the Sub10 Discussion Forum.

JRM includes Ben Johnson - arguing that there are probably other drug-affected performances that haven't been and won't be expunged from the record books.

The table highlights the element of lucky wind in Tim Montgomery's world record. At least on the basis of JRM's assessment, Maurice Greene's performances are superior.

All Time Best 100 m

Figure 1 below is a scatter diagram showing the full list of sub-10 second electronically timed runs (all-time best list from IAAF's site as published May 24th 2003). The scatter of the wind and altitude corrected performances is also plotted:

The scatter diagram plots a magenta + against the date and time an IAAF recognised sub-10 second performance was recorded. There is also a green x plotted for the corresponding equivalent sea-level still wind time (i.e. altitude and wind corrected).

A magenta line connects the official world record marks. The green line connects the best-ever sea-level still wind equivalent performances.

Figure 1 shows World Records for Leroy Burrell in 91, Carl Lewis in 91, and Donovan Bailey in 96 have only been achieved with the assistance of wind and altitude.

After correcting for wind-and-alititude:

1. Maurice Greene's 99 world records would stand. Every other world record had wind assistance that ranks it inferior to the best wind-corrected run of the time.
2. Burrell's 94 run would have been a "wind-and-altitude" corrected world record.
3. Fredericks in 96 and Ato Boldon in 98 would also have rated as world records.
4. Calvin Smith's 3rd Jul 83 altitude wind assisted 9.93 world record would not stand, but his 10.04 into 2.2 m headwind on 25th Aug 83 (at Bruxelles) would rate as a record (corrected to 9.90)
5. Viktor Bryzgin's 1986 run of 10.03 secs into a 2.5m headwind would have been a world record.
   Until 2004, some statisticians had categorized the wind reading for this run as doubtful. But the IAAF records list published Dec 2003 has it as legal. It looks no more or less suspicious than Smith's run
6. Leroy Burrell's 91, Carl Lewis' 91 and Carl Lewis '88 world records would not stand
7. Tim Montgomery's 2002 world record would not make the grade.

The sudden plunge in the world record line around 1991 is often attributed to the introduction of the Mondo synthetic tracks. But after correcting for the wind and altitude, there is not such an obvious effect - the plunge doesn't happen until Maurice Greene hits his straps in 99.

The general pattern in Figure 1 shows that correcting for wind and altitude results in an upwards shift in times (green x are mostly above corresponding magenta +). Figure 1 also shows that most elite performances occur in the Northern hemisphere summer and autumn. A strong increase in sub-10 second runs can also be seen in Figure 1 - but note that until 1995 the top-10 runs in any year were not all under 10 seconds.

Data Table

The data Figure 1 is based on is also presented in Table 2 below and agrees with the list posted by JRM.

The data shows all runs officially timed at under 10 seconds with a legal wind as published on IAAF web site at 20/12/2003. Ben Johnson's performances have been excluded (as per IAAF site). Viktor Bryzgin's 10.03 in 1986, Donovan Bailey's 10.03 in 1997 and Calvin Smith's 1983 10.04, all into strong headwinds, have been included here because they adjust to times under 10 seconds.

In Table 2, the Raw Rank column is the rank as per the IAAF statistics, using the officially clocked times. Wind is in meters/sec, with positive value assisting the athlete. The Adjusted Time and Adj Rank columns show times corrected for wind and altitude using Mureika's back-of-the-envelope formula. The Up column shows how many places up the ranking a performance is lifted by the adjustment (negative values occur typcially where high assisting winds result in a fall after adjustment). The Record Type column shows a W for a world record performance, and an A if the adjusted time would have been a world record (for either W or A, the whole row is emphasised in bold font).

Also shown below is table 3 with the best 20 performances for each year from 1998 to 2002. One of the outstanding features of this recent data is Maurice Greene's domination of the 100 m event in recent years.

Table 2 - Wind & Altitude Adjusted All-Time Best 100 m Rankings