Not optimal wrote:

The physiology of the event backed up by research is non-scientific to you?

Here is some research and math modelling of the event for you:

A simple model of entropy accumulation suggests that for any reasonable relation between entropy generation and speed, the optimal race strategy is to run the rst half of the race faster than the second. Such a strategy causes the entropy density gradient across the boundary of the working muscle to take on a large value early in the race, which increases the entropy eux. That this is in fact the optimal race strategy is proven by the overwhelming preponderance of positive-split races in the set of world records for one-lap and two-lap races.

Source: Optimal Pacing for Running 400 m and 800 m Track Races James Reardon University of Wisconsin{Madison, Department of Physics, Madison, WI 53706 (Dated: April 3, 2012)

It is not annecdotal to say that every world record ever set in the event has a slower second lap. That is called evidence.

It might also be of note to you that this calculator trackbot has hijacked was never designed for the 800m. So the owner and creator knows it does not work for the 800m. So don't bother using it for that.

Did you actually read the quote you posted? He used anecdotes as proof. That is not physiological research. That is confusing correlation and causation. Just because most records are set using that pacing strategy does not mean that is proof. Again, that is like correlation not causation. Yes, that means it is not scientific. Yes, it is anecdotal. Yes, anecdotes can also be evidence.