I'm talking about someone running a 3.5 forty. It will never happen. And I seriously doubt someone will run a sub 4.0 forty.
It does seem that a 3.5 just doesn't make sense. Let's take the fastest splits I can find taken from various 100m races:
0-10m 1.69 (shared by several)
10-20m 1.00 (shared by several)
20-30m .89 (M, Greene '99)
30-40m .85 (J Gatlin '06)
OK, that's a 3.58 for 30 meters or 33yds. Now, if we are charitable and actually give the remaining 7 yards the same split rate as Gatlin's 30/40 split that would give you a raw 40yd time of 4.12.
Now, this is the part where a lot of people lose their way, like that moron Mark Zeigler that wrote that atrociously misleading article about Ben Johnson running a 4.38 40 as though that's what he would have run if he was lined up with a bunch of football players running 40's at their camp or the combine. Assuming for the moment we had a fully electronic 40 timing setup that 4.12 above
includes the runner's reaction time. Football 40's do not have runner reaction times. Maurice Greene's 1.69 10m split included a .13 reaction time, so a fully electronically timed "football" 40 based on the fastest splits available (or that I can find) would give you a 3.99 40.
Now, what if we don't have FAT (fully automatic timing) available? In that case we have to settle for a stopwatch. The longtime standard rate of conversion from manual to FAT has been to add .24 to the manual time. Now remember, that 4.12 represents FAT splits so if we were timing that with a stopwatch you would
subtract the .24 (FAT to manual vs manual to FAT) so that would give us a 3.88.
Some things to consider with that time. First, that's an honest hand time, which is to say .24 is a track accepted conversion, not some mysterious number that some anonymous coach clicked off. Second, remember that last 7 yards included Gatlin's full 10m split so that is going to include a bit more acceleration time. The point of doing this was that even giving that little bump you're still nowhere near a 3.5 40 even if hand timed.
I guess it's worth pointing out that most 40 times are not in track cleats on a track surface with starting blocks so REAL football 40 times would likely be a bit slower. Having said that I hope my raging insomnia (it's 2:10 AM) puts some perspective on the fact that while a 3.5 40 simply doesn't seem possible having very low 4's certainly is, at least for the truly fast.
Since the above times are "fastest on record" splits as opposed to what one person has done I thought I'd throw in a real person. At first glance it would seem that Maurice Greene's '99 9.82 100m would also include the fastest 40 yards. Splits were 1.69/1.00/.92 so it is a matter of record that a person ran a raw FAT 30 meters (33 yds) in 3.61. As opposed to being charitable with the above time we are going to use his own .92 through 30 meters to determine his time through that last 7yds which gives us a 4.2. That's going to be a couple hundredths slower than reality since it gives him zero credit for his actually still accelerating but hey, we'll say that evens out his having starting blocks a little. Since he had a .13 start reaction his time would be AT LEAST a 4.07. If a person were manually timing that (4.2 -.24) that'd give us a 3.96. Again, surface conditions and starting blocks etc would make this time impossible at, say, the combine but it demonstrates that this myth (Zeigler, wherever you are, you're an idiot) that the fastest of humans (well, at least fast starters) can absolutely run very low 4's when timed by football criteria. (which isn't to say the vast majority of so-and-so's out there supposedly running 4.25 40's isn't pure bunk)
FYI, I found out researching this that Usain Bolt ran his 9.69 with the second slowest RT at .165. So with a slow start and cutting off the engines before crossing the finish he STILL runs a 9.69. Simply astonishing.