Here, btw, is a copy of my 2001 post to this forum on the subject:
===
During the recently held FFF Conclave (Livingston, MT, August 2001) there was a program/session called:
Computer Simulated Fly Casting and a Device to Analyze the Casting Stroke - Bruce Richards and Noel Perkins.
(To place the speakers: Bruce Richards is working for SciAng in developing flylines/tapers and a member of the FFF Casting Certification program Board of Govenors, Noel Perkins, the engineer, is less familiar to me.)
While I was unable to attend the session as I was demo tying at the show that same afternoon, I did take part in some experiments held outside the complex later that afternoon. Facinating! 
While the project is still in its early days, here is a general description of what was/is in place at the moment.
A small/light sensor is mounted on the rod, just below the reel. The purpose of the sensor is to register changes in rod angle over time.
Being small/light it does not influence or hinder the caster in the slightest. A known amount of line is outside the tip guide. While the caster makes a series of false casts, part of the sequence is recorded on a laptop and saved in Excel. One can subsequently plot the data in a graph for analysis.
Some twenty people took part in the experiment, many of whom are high profile casters/instructors, such as Mel Krieger, Joan Wulff, Bruce Richards, Tim Rajeff, etc., as well as some lesser mortals such as myself 
While the ultimate aim of the project is to plot rod tip movement (a next step), even this first step yields some very interesting and informative results. Clearly seen on the graphs is:
- How each caster applies power over timeline.
- How well the stop is executed on both forward and back cast.
- Whether same amount of power is applied during both casts.
- How well the caster manages to reduce/avoid rod occillation.
- Whether the caster introduces unwanted ‘creep’ between casts, and if so when and how much.
- Demonstrate cause of any tailing loops.
- Etc etc.
Some people I have talked to were somewhat sceptical, or even downright negative “Casting should never be reduced to an exact science…”, but I believe that this project may offer wonderful opportunities to both students and teachers of flyscasting. Even in its current, very early stage, so much information can be gleaned from the data.
One of the real eye openers for me was the confirmation/visualization that while two very competent casters, such as Mel Krieger and Tim Rajeff, but with vastly different styles of casting, both will load the rod in an almost identical manner, with their graphs being very closely matched.
But also that a distinct ‘alternative’ cast, such as Mel Krieger executing the so-called Belgian cast, will show up very different on the back cast stroke and identical to the ‘normal’ cast graph for the forward cast.
Regards,
Hans W
=== You have a friend in Low Places ===
http://www.danica.com/flytier