A Follow Up on Tim Tebow's Release: How Scientists Measure QB Throwing Motion

Last week I put up a post about why people say Tim Tebow has a slow release. It used frame-by-frame analysis from the telecast of the national title game. I lamented the fact that I don't have professional, high speed video equipment to make more precise judgments.

Little did I know, someone who was reading does.

I got an email from Bryan Conrad, an engineer at a biomechanics lab in Gainesville. Among many other things, his lab studies sports-related topics such as golf swings and quarterback throwing motions. He had some insights from that side of the business and graciously allowed me to reprint them here.

I am an engineer in a motion analysis lab at UF and one very small part of my job is to evaluate the throwing mechanics of QBs (most of the time we study orthopaedic diseases and injuries).

Our lab has the ability to film any type of motion with 14 high speed video cameras. This allows us to capture motion from a true 3D perspective and at a frame rate of up to 500 images/second. This is the same type of equipment that is used to measure Tiger Woods golf swing for video games or used to animate computer generated graphics in movies like The Matrix.

When I first started analyzing QB's motion, one of the first questions that a QB coach asked me was, "could I measure the release time of QBs?"  From what I can tell this is a parameter that many people discuss but few people actually measure. I applaud your effort to quantify this measurement during game play using TV film.

Three years ago I established a protocol so that we could try do the same thing in our lab. What we came up with was a scenario where we have the QB stand in the lab in a ready position (both hands on the ball, similar to image1 from your sequences) and asked the athlete to throw the ball as soon as he heard an audible signal emitted from our computer.

Since we have a controlled environment and very good temporal resolution of our cameras, we were able to break down the motion into two components which we call Reaction Time and Release Time.  Reaction time is the time from when the sound goes off until the first movement from the QB (when the non-throwing hand separates from the ball).  Release time is from the first movement until the ball leaves the hand.  What we found surprised me a little.  The reaction time is typically around 500ms and often times longer than the actual release time!  Admittedly, our lab setting is not the same as the game environment and responding to an audible cue is quite different than decision making process required on the field.

I agree that Tebow has a less compact motion than Bradford (as your recent article on Mullen suggests, this might change under Loeffler). However, I believe that he might be capable of faster release time than what you have measured.

I then asked him about the issues that Rocky Top Talk's hooper brought up in regards to how differing throwing motions could affect the reaction times of defenders. Having grown up a John Elway fan, he'd seen how devastating a top shelf pump fake can be.

We weren't quite sure how to go about measuring defenders' reaction times. Bryan said the lab hadn't tested that before, but he had some ideas on how to do it:

One of the things that we have considered studying is reaction time when the athlete must make a decision. I have always thought about it in terms of the passer. For instance instead of always throwing as soon as the signal is given, we could have two signals, one for 'go' one for 'no go'. They would be triggered randomly and the passer would have to decide whether the 'go' signal was given and then begin the throwing motion.

Even better would be if we could use visual signals instead of audible signals, since that would be representative of a game situation. I suppose this kind of test could be applied to defensive players as well, perhaps have them run in one direction and then give the signal, depending on what signal is given, they would have to break to either the right or left.  We could then measure their reaction time and acceleration out of the break (maybe some guys have a slower reaction but compensate with quick acceleration and vice versa).

That sounds like a pretty good method to me, given the constraints that lab conditions enforce.

Big thanks to Bryan for providing some insight on how the pros take a look at quarterbacking mechanics. The bit about reaction time was particularly good. Hopefully this gives you an idea of the process and considerations that coaches and engineers alike go through when trying to improve their signal callers.