Does Ball Velocity Tell Us Anything About QB Success?
In an age where medical records, Wonderlic scores and even questions/answers from private NFL combine interviews are leaked, an underdeveloped avenue of information related to the draft process is quarterback ball velocity. But does a measure of how fast a QB can throw a ball have any place in predicting a passer’s success?
With the top four passers in the latest RotoViz Scouting Index all posting notable velocity numbers, that’s what I’ve been tasked to find out.
THE SCOPE OF QB BALL VELOCITY DATA
Unlike the other testing data that is readily available, QB ball velocity data from the combine throwing drills is spotty at best. The best source on the matter appears to be radio host and NFL Draft QB expert Ben Allbright, who has compiled data on every thrower invited to the NFL scouting combine since 2008.
To examine what Allbright’s data tells us, we first need to establish some parameters. The average ball velocity of QBs who threw at the combine since 20081 was 53.7 mph. This year’s average ball velocity was 54.8.
Over that time, QBs have thrown as slow as 42 mph and as fast as 60 mph. Allbright grades ball velocity by absolutes, establishing that at 55 mph or higher a signal caller would earn his grade of ‘draftable’ in the category, and anything below that would qualify as ‘undraftable’.
If there is something here, looking at the hardest, softest, and mid-range throwers would be a good place to start. Of the 111 players whose ball velocities were tracked at the combine or their pro day from 2008-2016, 31 players threw 52 mph or slower, 53 players threw between 53-56 mph and 27 players threw 57 mph or higher.
FASTER, AVERAGE AND SLOWER BALL VELOCITIES
How much do these numbers matter, and what trends, if any, can be determined by them? Perhaps a look at the strongest, weakest and middle throwers by class will reveal something conclusive. To do that, I broke the data into quartiles, with the top 25 percent being the strongest (57+ mph) and bottom 25 percent the weakest throwers (52 or less mph). The rest are lumped in the middle group.
|John Parker Wilson||2009||UD||58|
In this group, the only current or potential starters are Colin Kaepernick, Mark Sanchez, Kirk Cousins, and Teddy Bridgewater. Now let’s examine the average throwers.
There are several starters or potential starters in this range, including Joe Flacco, Russell Wilson, Jameis Winston, Case Keenum, Marcus Mariota, Blake Bortles, Cam Newton and Andy Dalton. Now let’s look at the softer throwers.
|John David Booty||2008||137||51|
Based on Allbright’s data, only one quarterback at or below 50 mph in ball velocity since 2008, Tyrod Taylor, is currently a starter.
But simply dividing those players into ranges and identifying current or possible starters isn’t very scientific, and is biased against older passers who were previously starters or younger passers who have yet to lock down a starting gig. Therefore, we need to correlate that data with a qualifier that is predictive of NFL success. To determine what relationship ball velocity data has to NFL success, let’s set the success criteria as any QB who had at least one season starting at least half the games with an AYA of 7.0 or higher. The eight game starting criteria ensures there’s enough attempts to make the AYA of 7.0 meaningful, and we focus on AYA to focus on the passing component of a QBs game. No criteria will be perfect, and this still could be biased against newer QBs, but both Winston and Mariota met this criteria last year. Upon removing the 2016 and seeing who meets the criteria we get the following hit rates:
- Bottom (1 of 29 = 3.4%)
- Middle (7 of 41 = 17.1%)
- Top (7 of 24 = 29.2%)
Hits in the top tier include Nick Foles, Josh Freeman¸Cousins, Austin Davis, Kaepernick, Bridgewater, and Sanchez. The middle tier had Bortles, Mariota, Newton, Dalton, Wilson, Winston, and Flacco all meet the threshold. Only Taylor met the criteria in the bottom tier.
On the surface, it certainly appears that ball velocity correlated with this particular metric of success2 and the relationship is statistically significant. However, this method of analysis also doesn’t take into account the high number of prospective QBs who were not invited to or chose not to throw at the combine over that time. The number of QBs that were invited and chose not to throw (62) means that 39.7 percent were assigned no value at all, and does weaken the conclusions that can be made about the sample.
THE APPLICATION OF BALL VELOCITY TO THE 2016 CLASS
Using ball velocity as a measuring stick for a QBs passing prospects bodes well for the top three projected QBs. Carson Wentz (57), Jared Goff (58) and Paxton Lynch (59) were the top three in ball velocity at this year’s combine, so we can’t say any of these three are hurt by this metric. On the flip side, Connor Cook (50), recorded the lowest velocity, which should temper expectations for him. Here are how all of the rookies fared by ball velocity.