Hmm, after some very interesting experiments it seems that there is less deviation between a player with 99 stats when he has the ball and a player with 30 stats when he doesn't have the ball.
Here's how I set up the experiment: The two players (99 and 30) each had to run from the goal line to the half line, both with the ball and without the ball. I took 5 separate measurements with each player, then calculated the average (among other things).
Here are my findings:
99 Stats Player
WITH BALL
Dead Sprint (Hold R1)(secs) – 7.23, 7.25, 7.20, 7.27, 7.23
Average Time = 7.24 secs
30 Stats Player
WITHOUT BALL
Dead Sprint (secs) – 7.16, 7.13, 7.16, 7.12, 7.14
Average Time = 7.14 secs
This was very surprising. According to my observations, the 30 stats player, when chasing without the ball, is actually
faster.
Here are more results:
99 Stats Player
WITHOUT BALL
Dead Sprint (secs) – 7.0, 7.03, 7.0, 7.02, 7.03
Average Time = 7.02 secs
30 Stats Player
WITH BALL
Dead Sprint (secs) – 9.04, 9.0, 9.01, 8.98, 9.01
Average Time = 9.01 secs
As you can see, the deviation caused by having the ball for the 99 stats player is only ~0.22 secs, while the deviation for the 30 stats player is a whopping ~1.87 secs.
These results suggest that ball possession makes all the difference. I'd wager that as Jimmy said, this system was put into place to stop sprinting abuse. The CPU isn't rubberbanding, when chasing you down, they are simply faster.
Gotta love good old empirical data.