Why on earth does IAAF considers Bolt's 9.58 equivalent to a 12:21.34 5k / 7:12.10 3k?

Thats not how it works though, they didn't determine these numbers by the ratio of distance to time-improved. I only took one statistics class ever, but its pretty obvious how they came to this conclusion. They calculated this presumably by taking the data set of every elite 100m race in recent history, then they calculate the standard deviation of all these times, meaning the average distance that a result is away from the mean of all the results. This is a measure of how much times vary in the event. If there is very little deviation from the mean and all the times are very close, that means that being just a little bit over the average is very statistically improbable and in this context very impressive. In a data set that is very spread out, with high average distance from the mean, it is not as impressive to be far away from the mean. The differences in the two events are completely accounted for by finding the standard deviation of the data set (all race results ever) and then finding how many of that standard deviations a particular result is away from the mean. So if you can imagine how far kenenisa's result towers above other results, now you can realize just how much Bolt's result towers above others results. Or you can reason that drugs improve sprinting times to a greater degree than they improve distance times, and that explains why Bolts result is much more impressive.

You are probably right to say it's a bit strange.

If we compare Bolt's 9.58 to, for example, the 50th fastest 100m time ever run, we get 9.81 vs 9.58, roughly a 2.4% difference.

The 50th fastest 5k time is 12:52. Compare that to 12:21 and you get a 4.2% difference.

Using this maths, that would mean Bolt's time is the equivalent of something more like ~12:33.

It would also mean that 12:21 is actually the equivalent of running a 9.41.

Obviously this is a pretty flawed system to work out 'equivalents' but it does suggest that they're exaggerating the level of Bolt's performance.

So the IAAF points formula is

points = A * (B - T)^C

where T is the time, B is a baseline time, and A and C are parameters.

Can anyone tell us what the current values of A, B, and C are for the 100m, 3k, and 5k?

Why do we care at all? Bolt was doped up bigtime.

The Numbers wrote:

If we compare Bolt's 9.58 to, for example, the 50th fastest 100m time ever run, we get 9.81 vs 9.58, roughly a 2.4% difference.

But how many of those top 50 times (especially top 10-20) are actually Bolt? Might skew things a bit?

mathman wrote:

So the IAAF points formula is

points = A * (B - T)^C

where T is the time, B is a baseline time, and A and C are parameters.

Can anyone tell us what the current values of A, B, and C are for the 100m, 3k, and 5k?

I cannot find actual published values anywhere. However, the top answer in the stackexchange thread i posted linked to a PHP calculator he had written to calculate IAAF points. its on github. the file containing the parameters is here:

scroll down to line 117 to see the parameters. The way he wrote the formula, A=conversionFactor, B=|resultShift|, and C=2 (for running events).

to save people the trouble of clicking:

100m- A=24.63, B=17

3000m- A=.00815, B=840

5000m- A=.002778, B=1440

B (the reference time) is in seconds. There's no source of that mentioned, nor any indication of how those parameters were arrived at, but those seem to be what was used as of 2017 at least.

a quick spot check against a 9.46 100m (1400 points in the tables) seems to confirm that these are correct:

24.63*(17-9.46)^2 = ~1400.25

Bekele's WR is 2.87% better than the 100th best performer.

Bolt's WR is 3.91% better than the 100th best performer.

This despite the 100m being run "all out" more often.

I don't think some of you are understanding how much of an outlier 9.58 is.

9.93 vs 3:54.9

Masters 100m - Masters mile record.

No way the 1231 points for the 100m is equal to a 3;47.72 mile performance.

Something is wrong with the points - they overestimate the value of the 100m.

If more world class runners were able to stay in shape - 9.93 by Collins would not be the record today. Lagat was one of the best 1500/milers ever and there is a world of difference between 3:54.9 and 3:47.72. I think 9.8 is closer in impressiveness to a 3:47 mile.

fchj wrote:

Because they have no idea what it takes to run a 7:12. If anything, a 7:20 is beyond anything Bolt has run.

Perhaps you don't have any idea what it takes to run 9.58?

finewithme wrote:

9.93 vs 3:54.9

Masters 100m - Masters mile record.

No way the 1231 points for the 100m is equal to a 3;47.72 mile performance.

Something is wrong with the points - they overestimate the value of the 100m.

If more world class runners were able to stay in shape - 9.93 by Collins would not be the record today. Lagat was one of the best 1500/milers ever and there is a world of difference between 3:54.9 and 3:47.72. I think 9.8 is closer in impressiveness to a 3:47 mile.

You're using gut feelings and just two race times as opposed to taking a more in depth look at the statistics as my post did.

You also need to consider the 100 is the ONLY race on the track you can truly go 100%. Every other event you have to hold back a little at some point in the race (even if it's only a few strides).

This is important because in a 3000 or 5000 you will never truly come close to going "all out," 100% the whole race. These scoring tables can be understood as IF someone were to run a 3000 or 5000 at the maximum potential the human body can go, then they'd run 7:12 and 12:21.

sbeefyk2 wrote:

You also need to consider the 100 is the ONLY race on the track you can truly go 100%. Every other event you have to hold back a little at some point in the race (even if it's only a few strides).

This is important because in a 3000 or 5000 you will never truly come close to going "all out," 100% the whole race. These scoring tables can be understood as IF someone were to run a 3000 or 5000 at the maximum potential the human body can go, then they'd run 7:12 and 12:21.

+1000000, finally someone said it. It's just improbable if not impossible to run a "perfect" race over long or even middle distance.

seasoned ranker wrote:

Bekele's WR is 2.87% better than the 100th best performer.

Bolt's WR is 3.91% better than the 100th best performer.

This despite the 100m being run "all out" more often.

I don't think some of you are understanding how much of an outlier 9.58 is.

Agreed here. I think people are HEAVILY underrated how much of an outlier Bolt's 9.58 is. It's far, FAR more of an outlier than Bekele's 12:37. Barega (I think it was him) ran 12:43 a year or two ago, and he isn't exactly Haile or Bekele. If he had been paced better, that could have been close to the WR. And the 3000m is so infrequently run that it's hard to get a good grasp on how fast 7:20 is. El G ran one serious one and got 7:23. How many serious 3000s have even been run, period, especially when you compare them to all-out 100m races (i.e., every single 100m race ever, basically). So many distance races aren't run with the intention of going after a PR or world record, whereas EVERY 100m is run in that way. Even by those standards, Bolt's PR is STILL a full tenth of a second faster than anyone else has ever run. Imagine every marathon or distance race was run with the kind of conditions that Kipchoge had for his Breaking2 project. That is basically what the 100m is like, which is why it's so staggering how much faster Bolt is than everyone else.

Yes, I do think there's a possibility that the IAAF charts are perhaps overrating Bolt's performance(s), but I agree with what someone said on the original page, that so much math and so many performances have gone into the algorithms that create the IAAF charts that they are far closer to correct or objective than anything we could come up with on these LetsRun boards. I think it's disheartening as a distance fan to realize that Usain Bolt was so much better at his event than any of our distance idols were at theirs. I can relate to that. I far prefer distance events, but Bolt was just an absolute freak. No one on the distance side has come ANYWHERE close to his level of dominance, and he happened to have it in the most prestigious and most-frequently-ran events in the world.

seasoned ranker wrote:

Bekele's WR is 2.87% better than the 100th best performer.

Bolt's WR is 3.91% better than the 100th best performer.

This despite the 100m being run "all out" more often.

I don't think some of you are understanding how much of an outlier 9.58 is.

This is the best response. The tables are not a wonky Cool Running calculator that is trying to gauge equivalent times, its a statistically based approach. As some people are saying, there are nuances (how often is an event run, is it run all out, etc).

With all that being said, the scores show that Bolt's 9.58 is an incredible performance, head and shoulders above other 100m performances, and comparatively head and shoulders above any other running performance. The equivalent statistical performance would be 7:12. That just shows how crazy Bolt's time is. I remember seeing it and just being shocked.

cyclista wrote:

seasoned ranker wrote:

Bekele's WR is 2.87% better than the 100th best performer.

Bolt's WR is 3.91% better than the 100th best performer.

This despite the 100m being run "all out" more often.

I don't think some of you are understanding how much of an outlier 9.58 is.

This is the best response. The tables are not a wonky Cool Running calculator that is trying to gauge equivalent times, its a statistically based approach. As some people are saying, there are nuances (how often is an event run, is it run all out, etc).

With all that being said, the scores show that Bolt's 9.58 is an incredible performance, head and shoulders above other 100m performances, and comparatively head and shoulders above any other running performance. The equivalent statistical performance would be 7:12. That just shows how crazy Bolt's time is. I remember seeing it and just being shocked.

Agree with this completely. On any given year, on the entire planet, how often is a 100m race ran ALL-OUT versus a 3000m race? I could 100% see 9.58 being equivalent to 7:12. I honestly think the IAAF table is accurate. The 3000 is almost never run.

These are the eras in the 100m:

Prehistory - Owens, Peacock, 10.3

Old - Hayes 10.06

Modern - Hines, Smith 9.95

Johnson - 9.7

Dolt - 9.5

Johnson’s 9.83 began his era, it was just as huge as Dolt’s 9.69...but while BJ may have ultimately gone 9.75, Dolt actually did go 9.58.

While other guys were going BJ’s times within 10-15 years, here we are 11-12 years later and nobody is in sight of 9.58.

A not-totally-unbelievable argument is made that BJ would have gone high 9.6’s, which would mean that everybody since BJ has been in the BJ era—except for Dolt. And he is BEYOND the era that was BEYOND all other eras.

IDK about distance eras, but think about true revolutions. 9.69/9.58 is ridiculous. It has ruined what was left of the 100m after it was first ruined by 9.83/9.79

Bolt's time, when you consider the doping links up til outside the top 10, is a crazy outlier. Maybe it does deserve something close to these numbers.

I often think it's easier to say how fast someone can go if they run longer but a little slower vs telling someone how fast they can go in a shorter distance, shorter but quicker. That's why we're always pushing athletes to move up an event distance even though they may already be in their prime event.

Sprintgeezer wrote:

These are the eras in the 100m:

Prehistory - Owens, Peacock, 10.3

Old - Hayes 10.06

Modern - Hines, Smith 9.95

Johnson - 9.7

Dolt - 9.5

Johnson’s 9.83 began his era, it was just as huge as Dolt’s 9.69...but while BJ may have ultimately gone 9.75, Dolt actually did go 9.58.

While other guys were going BJ’s times within 10-15 years, here we are 11-12 years later and nobody is in sight of 9.58.

A not-totally-unbelievable argument is made that BJ would have gone high 9.6’s, which would mean that everybody since BJ has been in the BJ era—except for Dolt. And he is BEYOND the era that was BEYOND all other eras.

IDK about distance eras, but think about true revolutions. 9.69/9.58 is ridiculous. It has ruined what was left of the 100m after it was first ruined by 9.83/9.79

Both Blake and Gay ran 9:69..... But I agree that 9:58 was Bolt astonishing moment. He did run 9:63 once more , and after that it's one time 9:69 and over 9.7 after that.

What is the most surprising is that even when he was in his prime and running 100m all the year, he always came up with a huge record or huge time in the final of the olympics and world championship despite running all out the 100m everytime.

Such a clutch racer.