By Max Partlo
This Sunday, the Browns made their most recent attempt to find stability at the quarterback position, as Johnny Manziel made his long anticipated debut against the Bengals. The results weren’t great, as Manziel went 10/18 for 80 yards with no TD’s and two picks. Combine his dismal passing numbers with the three sacks he took for 26 yards, and you get his adjusted net yards per attempt at -1.71. To put that number into perspective, only 10 other active passers had home games as rookies where they threw at least 10 passes and had negative ANY/A. Of those ten, only Alex Smith and Derek Anderson have made the pro bowl, each going once.
It would be unfair, however, to judge Manziel’s career after less than 20 passes, and the Browns intend to let him start the last two games of the year, and “see what they’ve got”. The issue, is that first, three games is a small sample size to use to make an informed decision on who will start at Quarterback in Cleveland next year, and second, that rookie passing performance really doesn’t tell us that much about how a layer’s career will unfold. If you were to take all of the players drafted since 1998 who started at least six games at QB as a rookie, and at least six games for their original team after their rookie years, and plot their rookie ANY/A against their non-rookie ANY/A, you get this graph:
The trend line does show a slight positive correlation between the x axis (ANY/A as a rookie) and the y axis (ANY/A in later years), but it is weak to say the least. In fact, the correlation coefficient between the two is .37, noticeably below the .5 barrier between strong and weak correlation. The positive trend tells us what we already know, good rookies tend to be good veterans, and struggling rookies tend to struggle as veterans. However, the weak coefficient and the middling slope of the trend line suggest that the two are not as closely tied as our assumptions would suggest. In fact, if we cut this group of 38 players into three tiers, strong (with a veteran ANY/A> 6), average (with a veteran ANY/A between 6 and 5), and poor (with a veteran ANY/A< 5), then we see that there are 9 strong passers, 12 average passers, and 17 poor passers. When we take a closer look at the top 9, the franchise QB’s if you will, we can learn something about their rookie performance, and see the relationship in this more precise, more desirable cross-section. For these passers, the coefficient R is actually -.022, which, while it technically implies a negative correlation, mainly tells us that there is little to no relationship between an elite passer’s rookie stats and his ultimate performance. Using the tier system again to look at the rookie seasons of the elite quarterbacks, we notice that three were considered elite, while two were average, and four had seasons that would be considered poor. So ultimately, we really can’t use one year’s worth of data to predict whole careers, because there are just too many things that can change from season to season, and they aren’t the same players from year to year. So cheer up Browns fans, better luck next week, and maybe in a couple years we’ll be able to tell what type of player Johnny Football is going to be.
 In order from lowest ANY/A in their respective games to highest: Alex Smith, Mark Sanchez, Derek Anderson, Kyle Orton, Brandon Weeden, Christian Ponder, Geno Smith, Josh Freeman, Matt Barkley, and Curtis Painter.
 The last year that an active NFL QB was drafted (Peyton Manning).
 Keep in mind that these are players who started six or more games as rookies. There have been a lot more than 38 quarterbacks who have started over this time span.
 From oldest to youngest: Peyton Manning, Donovan McNabb, Carson Palmer, Ben Roethlisberger, Matt Ryan, Matt Stafford, Andy Dalton, Russell Wilson, Andrew Luck. Players from this time frame with ANY/A’s over 6 who didn’t start enough games as rookies to qualify include: Aaron Rodgers, Tony Romo, Philip Rivers, Tom Brady, Drew Brees, and Kurt Warner.