Guest column: Which event will achieve 20 seconds under 60 second 400m pace first?

You're right to pick out 1000, 1500, and mile. In addition to having world records the closest to the current time, these are some of the older records for men. My pick is 1500 because of how competitive the event is right now and because it's raced more often.

We'll never see a sub 40 400. I might bump 800 up to be on the same level of proximity as 1500 for similar reasons. Mile and 1000 are tier 2.

I think that the minimum lies between the 800 & 1500; The 800 record is off by 0.91 sec and the 1500 record is off by 1.00 sec. Of the three records in contention, I think the 1500 is most likely. The 1000 might possibly be more doable, but it is much less contested.

I think the "sweet spot" might be around 1200m. If highly contested, I think the record would be close to 2:40

A second in a 1500m is easier than 0.91 in the 800,

; percentage wise it's about half

I don’t think the 1200m in 2:40 is doable by anyone today. You would have to come through, if running even pace, in 2:13.33 at 1000m and not slow down. This isn’t going to happen anytime soon.

I disagree that a sub 4:40 2000m is the same level as a sub 4 mile for women. Jakob got the WR for it on his first real attempt at 4:43.1. I would say its on the same level as a 3:25.00, maybe even slightly easier.

no event can ever get that far under 60 seconds per lap.

The universe designed tracks on a principle of one minute, one lap. You can violate that principle to a degree, but only so much. And for everyone going under 60 second pace, there has to be an equal and opposite someone going slower.

Very interesting piece! As a fellow track stat nerd this was a great read and a novel way to analyze time over distance

I agree with the poster above — 1200m is probably the ideal sweet spot, and 1000m for an actual contested event.

I'm another need, though not as as accomplished. It was decades ago when I considered a parallel issue, how far up in distance do we get where the next 400m takes more than one second.

When I saw this piece I thought of another question: who would have the biggest lead over 60-bot? The author here deserves a claim possibly the first to determine that query (whereas others have probably done an informal version of the biggest gap). This metric favors the shorter distances with faster pace.

Finally, the "standard" 0.24 hand timed conversion is for events like the 100 and 200 where the start and finish are separated by significant distance. I have found a smaller adjustment for races that don't have such gaps (while you can time off the smoke from a real pistol rather than the distance delayed sound, 0.33 seconds per 100m, for me that reaction takes longer than the sound). Thus a differential of 0.14 seconds is a better adjustment. On a related note, at big XC races I put about half a second on my watch when starting it at a distance from the start. That timing is not going to be used anywhere but I'm a nerd...

Another Finally. Thank you for your efforts!

Thanks for this interesting explanation of reaction times!

Perhaps 0.14 seconds would have been better. Luckily, it wouldn't change too much in the article as I think every run in there aside from Snell's had an electronic time (so maybe the note at the end was unnecessary). For instance, that Rome 400m had both a hand-time (44.9 for Davis and Kauffman) and an electronic time (45.07 for Davis and 45.08 for Kauffman). Adding 0.14 to the hand time does indeed come closer to the electronic time than 0.24 would! But it's still over 45 seconds so luckily (for the purposes of the article) it doesn't change the conclusions.

Both the James and Evans records were similar: hand times and electronic times were both available, and I used the latter.

The one other place where this might matter is for the (much more important) question of when Snell's world record was actually broken. This doesn't matter for the article since it wouldn't change the progression of the "seconds-under" record, but here is the 800m progression starting with Snell:

Snell (1962): 1:44.3 hand-timed
Doubell (1968): 1:44.3 hand-timed, 1:44.40 electronic
Wottle (1972): 1:44.3 hand-timed
Fiasconaro (1973): 1:43.7 hand-timed

Hence the phrase: "no one would run materially faster for over a decade" since Doubell's electronic time is faster than the conversion (either 0.24 or 0.14) of 1:44.3 hand-timed.

Yeah, if all distances were a multiple of 400m this would be much more straightforward! We could imagine standard distances and records that looked like this:

400m: 43
800m: 1:40
1200m: 2:40
1600m: 3:41
2000m: 4:43
3200m: 7:48

Then the whole concept would be pretty obvious since all the records would end with :4x.

Agree with you on the sweet spot. If you wanted to get *really* in the weeds, perhaps the sweet spot would be in between 1000m and 1200m. Here are the 800m paces one would need to carry to go 20 seconds under:

1000m: 1:44.00
1020m: 1:44.31
1040m: 1:44.62
1060m: 1:44.91
1080m: 1:45.19
1100m: 1:45.45
1120m: 1:45.71 (3:18.21 1500m pace)
1140m: 1:45.96 (3:18.68 1500m pace)
1160m: 1:46.21 (3:19.14 1500m pace)
1180m: 1:46.44 (3:19.58 1500m pace)
1200m: 1:46.67 (3:20.00 1500m pace)

Personally, I like the look of 1140m, but any would be extremely hard!

Clarification and comment. Since the Bot is going at constant speed (0.15 seconds per meter), the Time/Distance is not dependent on speed. To bring speed in, you could calculate how much further would the runner get at their average speed. For example at 800m per 100.91 seconds, but that is a bit artificial since you can maintain that speed beyond the finish. Another question is about trends over time and that deserves a new post

Nice addition adding in quasi times based on pace. A further refinement is to use a fitted value for the distances rather than 800 then 1500. This can be appoximated well enough by interpolation between 800m and 1500m. So go one 7th of the way for 900m, five 7th for 1300m, etc.

Another interesting look is to have a series of curves for 400, 800, 1500, mile, 3000m with the horizontal axis is time, Years, plotting all of the differentials. The slopes give the rates of improvement for each distance. If course this takes getting the data but you might have much of it already

i pointed out this numeric phenomena back in 2007 ish on track and field news under either figo or gibson handle.

it is an obvious trend if you studied numeric series, and interpolation.

wobderful it is the subject of fun, and there is an underling physical phenomena going on, the relationship of anaerobic ability coupled with a near hard set limitation in aerobic ability, where a 3 rd phenomena kicks in at 3 km.

this third system needs to be defined.

i became loosely aware of the different energy system for 5km when aerobic monsters John Walker , Steve Ovett could not emulate world record pace.

and then there is the marathon, where gas tank energy system must be paired with monster aerobic.

thai said, PEDs can change it, where the human becomes a slightly dirrerentvanimal.

Enough said

This is incredibly poorly written. What you stated would mean an 800 would be run in 1:20. I don’t even get what you’re asking. Are you saying one 400m would be run in 40 seconds? 1:40 would be 50sec 400m pace and surely would be run without a 400 run in 40 seconds. What are you even asking?

Minute Bot is given a running start; it is already going full speed at time = 0. If you gave 800m runners a running start, they might already be sub1:40.

This is really fascinating. I wonder if it could be the ability of muscles to utilize stored glycogen, including glycogen in the liver, as a fuel source coming into play.

I wonder if we go from anaerobic to aerobic (sheet ability to use oxygen) and then to glycogen / fuel usage (ability to sustain and maintain high oxygen usage).

Does anybody else have any thoughts or ideas about this stuff?