The 800 begins as a race at 150-200m. There anaerobic debt starts to rise in the next 100m, and by 400m we are officially in substantial anaerobic debt.
By "acceleration" I suppose I mean a more nebulous "increase in effort / selling out". At 500m in Rudisha's race, you can see the change in the leaders as everyone begins to try to exert effort above and beyond what their mounting anaerobic debt would sanely allow. I think what happens is that racers try to use the last mental and physical reserves at this point to gain momentum and transition to a sprinters stride in order to gain more RELATIVE speed over the last 300m.
The massive amounts of anaerobic debt in the last 300m, relative to the phosphate and no-debt first 300m of the race, will invariably be slow by the clock. But the speed produced relative to pain and physiological debt, it is much more stark.
Every runner will perceive "kicking" at the end in this manner. The sense of acceleration, real or otherwise, is in relation to an inability many runners don't have in these moments. Physiologically middle distance "kickers" have some sort of hybrid fast twitch muscles they can employ in situations like this that produce higher speeds under extremely high anaerobic debt than other runners. Whether this is also related to higher pain tolerance or other physiological aspects (reduced "lactic acid" release like they claimed Armstrong and Phelps had, etc).
Dave Wottle's victory, which I recall vividly described by Jeff Galloway in one of his books, was a fluke. Wottle was injured, and basically was limited to that speed for the first 400. It was almost as big a fluke as Rudisha's race. He was trying to simply not embarrass himself in the race. So he ran a careful first lap just to try to finish.
The 800m is such a painful event that relies so much on anaerobic tolerance and anaerobic energy stores, that essentially it is a crapshoot under most circumstances. You can't know whether or not that moment of truth: whether you still have kick/speed at 500-600m when everyone else starts also goes "all in" to try to win, whether you'll have the speed that day.
Wottle that day got lucky: all the people in front of him didn't have the stuff, Wottle was able to survive the first 400m and then adrenaline the last 400m. And perhaps the wind that day was particularly graceful towards someone running through the pack, since wind resistance is a factor where you are running 17-20 mph.
His splits were 49.28, 51.63. He didn't accelerate as much as he slowed down less than the other runners who faded trying to keep up with him. This is typical of the 800M.
From https://www.runnersworld.com/races-places/a20800858/800-mete...
> Okay, so the 800 is more of an extended sprint when run at the highest level. But does that mean positive splits are the only way to achieve excellence? Yes, says exercise physiologist and Science of Sport blogger Ross Tucker. In 2007, Tucker analyzed the 26 men’s world records set for 800 meters in the modern era. He found that 24 of the 26 were set with positive splits. In fact, the last time a world record was set with a negative split was in 1972, when Dave Wottle ran 1:44.3 to bring gold home to the United States. His last lap of 51.4 is also the fastest ever in a world-record race, despite the fact that record has fallen more than 3 seconds in the last four decades.
Here's Wottle's 1972 Olympic 800M gold-medal race in 1:45.86 where he won with negative splits of ~53.3, ~52.56:
https://www.youtube.com/watch?v=5LHid-nC45k
Dave's competitors went out too fast and faded badly. It looks like he's running them down, but mostly they are coming back to him.