Because the Carvalho model can give you a pin point split for every hurdle for a given end-time it is easy to compare these with the actual touch down's taken from a race. Here is a sample:
This athlete is losing double the time Carvalho suggests, but on the other hand his deceleration is distributed evenly over the race. If you look at the chart you probably notice the split half time te race, he starts off as a 49-er and then finishes as a 54-er. If you'd seen the race you would have noticed this athlete changing from 13 to 14 at hurdle 6. With the TD's at hand there is no other conclusion that this ahtlete has problems with switching from amplitude to frequency at the change over.
So there's no point letting this athlete go off more slowly, then he will only have to change over more quicky loosing speed rappidly. He will have to work on his 14 stride running 4.2-4.3 in that pattern as well. Look at "carvalho training" how that's done.