http://www.trackandfieldnews.com/results/newsletter/200105/reaction_times.html
Did Sydney Blocks Rob Mo Greene Of Olympic Record?
Most people are content to either watch a track meet or read about it in the paper and then accept the results at face value. Not Sweden’s A. Lennart Julin.
One of the sharpest and most inquisitive minds ever to grace the sport, Julin analyzed the Olympic men’s 100 final and came to the startling conclusion that if different starting blocks has been used in Sydney, Maurice Greene would have broken Donovan Bailey’s Olympic Record of 9.84 and finished very close to his own WorldRecord of 9.79.
In other words, his 9.87–already a spectacular time considering the light headwind–was more like a 9.82.
by Lennart Julin
Most expert observers in Sydney considered in view of weather conditions Maurice Greene’s 9.87 in the Olympic final as one of the finest 100m-performances ever. Especially his impressive winning margin over the rest of the world’s fastest humans highlighted the exceptional quality.
And still those experts then did not weigh in that the starting procedure used in Sydney added some five hundredths to Greene’s time as compared to what it would have been if the start had been carried out as e.g. in the 1999 World Championships (where he ran 9.80 in much, much sprint friendlier weather)!
Yes, you’ve just read that the starting procedure was slowing the Sydney 100m-times by approximately half a tenth!
This statement appears almost unbelievable in an era where we through the media are regularily fed claims (mostly unfounded …) by manufacturers to have developed new tracks, shoes or uniforms that would shave the occasional hundredth off sprint times. How could it then be that much more time is allowed to be lost through the starting procedure at the most important meet of the quadrennium?
The answer: “The mechanics” of the sprint start the most common occurance at all track meets is paradoxically still today the most unknown and misunderstood factor within our sport. And this not only to the general public and the media but also to athletes, coaches, officials and yes - even Starters! (To sort out all the abundant misconceptions would probably be the subject of a book rather than an article. )
Very few seem e.g. to be aware of the fact that the “reaction” times now issued at every major meet actually NOT at all are direct measurements of the human reactions. Rather the times registered are either
the sum of the “true” human reaction time ® and the time (D) it takes for the start stimuli (the “bang” sound) to reach the ears of the athlete, or
the “guessing time” (G), when the athlete rather than waiting to hear the “bang” tries to anticipate the firing of the gun.
If we focus on the a) alternative the theoretical ideal of course is that D = 0.00. But in the real world that is not achievable. At the normal meet with the Starter using a normal gun D is determined by the distance between the Starter and the athlete. As the speed of sound in air is approximately 330 metres per second every 3.3 metres (11 feet) of distance between gun and athlete means a delay of the stimuli of one hundredth (0.01).
In a typical 100m-scenario, when the Starter is approximately 10 metres from the runner in lane 1, D = 0.03. It should also be noted that in the same race D = 0.05-0.06 for the runner in lane 8 who is some 8 metres further away from the Starter and his gun. (The lane dependance of D of course becomes really significant in staggered starts. In the 200m the difference between lane 1 and 8 corresponds to 0.07, and in the 400m to 0.14!)
But isn’t that problem taken care of by the use of individual loudspeakers behind the blocks of every lane? This seems to be a doctrine almost universally subscribed to and even manifestated in the IAAF Handbook (Rule 128.2 Note). However, it is a total misconception!
What the loudspeakers relay to the athletes is just the Starter’s verbal commands, not the “bang”. That sound still travels through the air from the gun to the ears of the athletes. And it even has to be like this! Because if the “bang” would be coming through both loudspeakers and air the athletes would be experience two distinct shots, which would then be interpreted as a starting “bang” followed by a recall!!
But there is one obvious solution to this problem: To have the “bang” only via the loudspeakers, i.e. to use a “silent gun”. A silent gun is a “gun” that doesn’t fire a conventional shot but one that when the trigger is pulled sends not only a starting impulse to the electronic clock of the timing device but also an impulse that manifestates itself as a distinct “bang” sound coming out of the individual loudspeakers in each lane.
With such a system not only is D the same for all runners regardless of lane draw but it is also independent of the position chosen by the Starter. Actually D is reduced to an absolute minimum as the setup is equivalent to having the Starter standing right behind the blocks of each and every runner!
Sounds quite convincing when you read it like this, doesn’t it? But isn’t it just a beautiful theory made from a comfortable position behind a desk far removed from the athletics arena? Would it work out like this in the real world of competition? Well, actually the theories have over the last half decade been successfully field tested on the very highest level of competition several times:
Because after performing an analysis like the one above the organisers of the 1995 World Championships decided to request to achieve maximum fairness to the athletes - a “silent gun” type of starting system for their meet. The request was granted and since then “silent gun” starting has been used at all following World Championships, outdoors, indoors as well as juniors! And it has indeed worked precisely as intended!
This knowledge makes it even more astounding that “silent gun” starting was not used in either the 1996 or the 2000 Olympics! And as the figures below, where the men’s 100m in Seville 1999 and Sydney 2000 are compared, prove: The choice of starting system really makes difference for the “reaction times” recorded! A difference in this case of as stated above approximately 0.05!
Table 1 presents the distribution of reaction times in Seville and Sydney. For each meet not only the totals are presented but also the numbers broken down into 1st round (including some very inexperienced runners that might “blur” the picture) and 2nd round and onwards (only experienced runners). For each column the median (i.e. not the average which could be distorted by an occasional odd registration - but the “middle” value) has been calculated.
You don’t have to be an experienced statistician to see the enormous discrepancies between the two meets in Table 1. E.g. looking at “2nd round and onwards” only 3 out of 61 (4.9%) in Seville exceeded 0.18 while it was 39 out of 65 (60.0%) in Sydney! And at the other end of the spectrum: 40 out of 61 (65.6%) in Seville were under 0.16 while it was only 18 out of 65 (27.7%) in Sydney!
The medians tells the same story: The athletes in Sydney were at a disadvantage of approximately 0.05 timewise compared to those running in Seville, which to a large extent where the same individuals! One of those running at both meets were Maurice Greene which brings us back to the original assertion.
In the final in Seville Greene’s reaction time was 0.132, in Sydney it was 0.197. Some observers have unfairly interpreted this as Mo “missing the start” in Sydney. Actually rather than being the fault of the athlete it was the inevitable consequence of not using a silent gun starting system in Sydney, enhanced by the Olympic Starter chosing to position himself extremely far away (from the TV it could be estimated that he was perhaps 20 metres from the nearest runner)! Something that resulted in an exceptionally high D (as in delay) value!
Actually the “non-silent-gun and Starter-very-far-away” setup in Sydney had another very serious disadvantage from the fairness perspective. The numbers in Table 1 not only show that the “reaction” times are considerably slower in Sydney but also that they are not as tightly bunched as in Seville. In Seville (looking at “2nd round and onwards”) 54 out of 61 (88.5%) were crammed at 0.12-0.17 while in Sydney you needed a 50% wider time spectrum (0.14-0.22) to get a similar share.
Translated into reality these numbers tells us that the starts in Sydney were not as tight and even as those in Seville, rather they were quite ragged. One partial explanation was that the Starter actually was too far away to be able to really ascertain that all runners were motionless in the Set position before he fired the gun. There were lots of twitches and other movements that got by without notice.
However, the main explanation to the ragged starts was that several athletes quickly realised that it would be worthwile to “guess” rather than to patiently wait for the sound of the gun. With a successful “guess” you could eliminate the D disadvantage without getting punished with a formally illegal “reaction” time.
Actually the limit of the “acceptable” reaction times given by the rules as 0.100 of course ought to be adjusted according to the D value of the meet. If it is 0.100 with a silent gun starting system it should have been at least 0.150 with the starting system used in Sydney. I.e. all registered times under 0.150 in Sydney could not have been achieved by a runner reacting to the “bang”, but are rather illustrations of what above was characterised as the type b) “reaction times” (guesses).
One good example was given in Sydney by Ato Boldon in the final. Starting in lane 8 which means that he was furthest away from the Starter he got away after 0.136. A good but not exceptional reaction time in a “silent gun” setting, but an in Sydney unattainable reacting-to-the-gun time. It could be equated to 0.05-0.07 in a Seville setup.
That the unevenness in the Sydney starting not only affected the inter-athlete relations but also the athletes as individuals could be seen from Table 2 below. In that table the reaction times registered through all the four rounds of competition are listed for all finalists.
Before commenting on the numbers in Table 2 it is important to stress that what we are talking about are times measured in thousandths of a second. When we see a “spread” of 31 that might appear sizeable but actually it means a mere 0.031 second!
In Seville the maximum spread was 41 and the minimum 9, in Sydney the minimum was 24 and that was the only value not exceeding the maximum in Seville! The effects are very clearly illustratred by Obadele Thompson:
- In Seville he was remarkably consistent with all four reactions times crammed into less than a hundredth despite starting in lanes 2, 6, 4 and 1.* In Sydney he had a spread of 52 despite running his four races in lanes 2, 3, 4 and 4!
So not only does a “silent gun” system save a significant number of hundredths (in Atlanta 1996 when the Starter was not so far away as in Sydney the loss was still 2-3 hundredths, i.e. Bailey’s 9.84 would have been a 9.81/9.82 with a “silent gun” setup) it actually also increases the fairness for the athletes within the given race. Because the runners can confidently concentrate on just reacting without harbouring ny nagging feelings that they might be at a significant disadvantage to an opponent trying to guess the gun.
And if runners really concentrate on reacting such consistency as the one demonstrated by Thompson in Seville is really nothing that exceptional. Other examples from Seville: In the women´s 100m Hurdles Glory Alozie had 162/158/154/155 (spread = 8) and Ludmila Engquist 137/136/135/132 (spread = 5) and in the men´s 110m Hurdles Colin Jackson 115/117/110/116 (spread = 7).
Furthermore the reacting ability doesn’t vary that much between individuals either, as could be seen from the Seville numbers in Table 1 above. It is obviously almost completely determined by how the human species is wired. The average male sprinter hits about 0.13/0.14 (the average woman sprinter being 1-2 hundredths slower) with Jackson being the very best with his consistency around and just below 0.12.
No human being has ever showed the ability to consistently hit 0.10’s, not even in a silent gun setting. That is the reason that the limit for the system is set at 0.100. However, that should not be interpreted as it unfortunately has been by most persons in the sport as that a 0.101 (e.g. Surin in Seville semi) is a correct start. Rather that number is also unattainable for a reacting human being.
We must realise that the technical device is there only for throwing out all undisputable false starts, and that the border line cases still have to be handled by an astute Starter. And it will stay like this also in the future because it is impossible to pinpoint the minimum human reaction time down to the thousandths. Any automatic system therefore has to have a limit that incorporates a sufficient safety margin.
By using a silent gun starting system the safety margin is minimized (which means that it is very hard to be a successful guesser because the “window” is just a couple of hundredths wide), by using a “live gun” starting system the chances for the guesser are multiplied at the same time putting the truly good “reactor” who patiently waits for experiencing the sound of the gun at a disadvantage.
Note: The phrase “silent gun” was coined by the organisers of the 1995 World Championships. It was adopted by the Seiko company providing the equipment for the competition. However, any other timing company could although none has done so far of course develop similar systems and then use a name of their own for the device.
TABLE 1:
Distribution of Seville & Sydney 100m reaction times
Click here for a graphic representation of distribution in rounds 2, 3 & 4.
Note that Sydney reaction times are shifted to the right (slower on average).
X = “Reaction” time
A = Sydney 1st round (median=0.197s)
B = Sydney 2nd round onwards (median=0.184s)
C = Sydney all rounds (median=0.189)
D = Seville 1st round (median=0.146s)
E = Seville 2nd round onwards (median=0.143s)
F = Seville all rounds (median=0.143s)
X A B C D E F
0.10 - - - 1 2 3
0.11 - - - 2 2 4
0.12 - - - 12 8 20
0.13 1 3 4 12 16 28
0.14 4 9 13 16 12 28
0.15 9 6 15 5 7 12
0.16 12 6 18 6 5 11
0.17 10 2 12 6 6 12
0.18 5 15 20 2 1 3
0.19 10 7 17 2 - 2
0.20 7 1 8 1 - 1
0.21 9 5 14 1 - 1
0.22 7 6 13 - 2 2
0.23 7 3 10 - - -
0.24 7 1 8 2 - 2
0.25 5 1 6 - - -
0.26+ 4 0 4 5 - 5
TABLE 2: Reaction times in each round for Seville & Sydney 100m finalists
Seville '99 Sydney '00
h q s f spread h q s f spread
M Greene 122 138 163 132 41 195 182 227 197 45
B Surin 127 135 101 127 34 198 130 151
D Chambers 149 116 124 140 33 171 150 164 174 24
O Thompson 146 152 143 145 9 239 187 189 216 52
T Harden 149 155 147 136 19
T Montgomery 167 145 158 136 31
J Gardener 164 127 137 142 37 188 177
K S-Thompson 149 172 169 173 24
A Boldon 170 155 212 136 76
J Drummond 198 145 137 147 61
D Campbell 148 133 143 224 229 161 193 63
K Collins 243 240 222 184 210 56
A Zakari 317 193 236 180 137
median 32 56.5
Graphic representation of distribution in rounds 2, 3 & 4
Copyright© 2001-2002, Track & Field News