This was originally posted at numberFire.com, home of the most accurate sports analytics and projections.
Let's get my personal biases out of the way in the very first sentence: I was raised in Pittsburgh, and my Dad went to Pitt. I was taught at a very young age that my teams - so far as college football goes - are Pitt, and the opponents of Notre Dame, Penn State, and West Virginia. Now, I'm not going to go say I hate those schools - although frankly, there are truly detestable things about all of them - but I'm not going to pretend that I wasn't trolling the hell out of any WVU board I could find after the 13-9 game, either.
So with my personal biases lingering in the back of my mind, I sat down to look at Tavon Austin, the No. 8 overall pick and the most touted rookie for fantasy purposes out of this year's class. Several reputable sites have come out with their initial feelings and opinions on Austin, but qualitative just isn't the way we do things around here; opinions are meaningless without numbers. Enter READ: the Rookie Estimation Algorithm Derivative.
READ starts with a a similarity scoring algorithm based off of the combine and pro day results, chosen simply because it ignores the variance inherent in different offensive schemes and the subsequent performance statistics. Because not everyone runs the same drills at the combine/pro day, we're choosing to compare them on the ones more commonly run and the ones most predictable of positional value. In this case, we're using 40-yard dash, vertical leap, broad jump, and the shuttle drill.
From there, we bake in the team analytics for their future NFL team from the season prior. You'll note that we use analytics and not statistics; this is because statistics like yards against, points against, etc. can be extremely misleading as they do not take into consideration the strength of opponent, situation of the game, and so forth.
Finally, we bake in a simple sanity check. This check looks at the projected position on the depth chart (in this case, starting WR) and weighs in the average performance from that role on teams similar to their future NFL team. After all of this, we have an estimation of performance which is much more than just someone's anecdotal opinion, but rather a projection that is rooted in mathematical modeling. Whew.
After that, it gets a little..how shall we say, CFL-ish. Of course, it's worth noting that this is based off of combine statistics, so two issues bubble up: one, not everyone completed the same drills at the combine and two, not everyone was drafted and as such, we'll have to be mindful of expectations in relation to where they were drafted. As we go deeper into the comparables, we find a few more interesting players: Lee Evans, Kevin Ogletree, T.Y. Hilton, just to name a few.
|DeSean Jackson||99.31%||2nd round, 2008|
|Calvin Johnson||98.42%||1st round, 2007|
|Louis Murphy||96.83%||4th round, 2009|
|Cliff Russell||96.82%||3rd round, 2002|
|Reggie Germany||96.04%||7th round, 2001|
|Demetrius Byrd||95.55%||7th round, 2009|
|Roscoe Parrish||95.24%||2nd round, 2005|
|T.Y. Hilton||95.22%||3rd round, 2012|
|Corey Fuller||95.09%||6th round, 2013|
|T.J. Graham||94.13%||3rd round, 2012|
|Sinorice Moss||93.66%||2nd round, 2006|
|Lee Evans||93.29%||1st round, 2004|
|David Clowney||92.83%||5th round, 2007|
|Andre Caldwell||92.79%||3rd round, 2008|
|Percy Harvin||92.07%||1st round, 2009|
|Terrence Murphy||92.05%||2nd round, 2005|
I don't think I'm shocking anyone when I say that last year, the Rams weren't all that good. To get an idea of how Tavon Austin might fit into the Rams offense, we'll take a look at the Rams as a team - specifically their passing analytics - as well as the career progression of Sam Bradford. Next, we'll look at the emergence of Daryl Richardson as a replacement for Steven Jackson. In the end, we'll have a sense of what the Rams might look like analytically in 2013 and then bake that into our set of comparables above.
See related article: Steven Jackson to Atlanta: One Giant Step Sideways
First, the team.
|Net Adjusted Passing||+73.85||#15|
|Net Adjusted Rushing||-32.09||#16|
The Rams were probably a little better last year than you thought they were, although finishing No. 20 in overall offensive efficiency - to say nothing of the defense - isn't exactly a world-beater either.
In keeping with the theme of mediocrity, Sam Bradford finished No. 19 in passing efficiency, just ahead of Carson Palmer for those of you scoring at home. This is fairly in-line with his historical curve, as he vacillates between the No. 15 and No. 20 positions regularly, staying just above replacement-level. This means that we can't reasonably expect Sam to suddenly become Aaron Rodgers, so for the sake of progression, we'll keep him static.
What is difficult to quantify at this juncture is the impact that the loss of Steven Jackson and Danny Amendola will have on the team. Even if we take a leap of faith and presume that Daryl Richardson can provide similar value to Steven Jackson - Daryl's VORP suggests a similar efficiency - the loss of Amendola will hurt what was already a fairly pedestrian passing offense, run by a relatively stagnant Sam Bradford in terms of growth. Thus, for the sake of estimating the team situation that Austin will be placed into, we'll project the Rams to be similar, if not a little worse.
The Sanity Check
What we'll need to do from here is look at the comparables and look at the teams that drafted those players, checking to see how similar their team analytics are to those of the 2012 Rams. For the sake of brevity, we'll pare down the list at this junction to players drafted in rounds 1-3, as Tavon is a likely starter out of the gate and players like seventh-round pick Aaron Lockett simply aren't logical comparables for that position on the depth chart.
|DeSean Jackson||81.67%||2007 Eagles|
|Calvin Johnson||78.64%||2005 Lions|
|Roscoe Parrish||91.08%||2004 Bills|
|T.Y. Hilton||80.03%||2011 Colts|
|T.J. Graham||90.27%||2011 Bengals|
|Sinorice Moss||85.26%||2005 Giants|
|Lee Evans||83.09%||2003 Bills|
|Andre Caldwell||79.62%||2007 Bengals|
|Percy Harvin||83.44%||2008 Vikings|
|Terrence Murphy||72.59%||2004 Packers|
Just because this one is a little bit more confusing, I'll explain it again: what I wanted to do is take the list of players who were comparable to Tavon Austin at the combine, and then look at the teams they were drafted onto and see how comparable those teams were to the Rams. This achieves what I consider two-sided similarity; a great player on a bad team will not be as good as a good player on a great team, and so on. Comparing both the player and the team situation will only serve to enrich the accuracy of our final projection for Tavon.
Looking at the similarities, a few outliers stand out: the 2004 Bills, the 2011 Bengals, and the 2005 Giants. Combining this table with the comparable table above shows that Roscoe Parrish is likely the strongest predictor on numbers alone, followed by T.J. Graham, Sinorice Moss, Lee Evans, DeSean Jackson, Percy Harvin, and T.Y. Hilton.
With that said, we also need to take into account the role in which Tavon will play in the Rams offense. You'd be hard pressed to argue that Roscoe Parrish was an integral part of the Bills offensive gameplan; certainly not to the level that Austin projects to be. With this in mind, we'll somewhat artificially nudge Lee Evans, DeSean Jackson, and Percy Harvin up in our array of predictors to give them a little more influence on the outcome. This is where the sanity part of the sanity check comes in.
As the last step, we take all of the similarity scores that we've calculated - and the adjustments for depth chart and team role - and perform a simple calculation where the strongest predictors are given more weight on the final projection.
|Lee Evans||94.22%||48 recs, 843 yards, 9 TDs|
|DeSean Jackson||93.29%||62 recs, 1008 total yards, 3 TDs|
|Percy Harvin||90.66%||60 recs, 925 total yards, 6 TDs|
|Roscoe Parrish||87.81%||15 recs, 148 yards, 1 TD|
|T.Y. Hilton||86.63%||50 recs, 890 yards, 7 TDs|
|T.J. Graham||85.50%||31 recs, 322 yards, 1 TD|
Looking at these numbers and weighing them using the similarity scores composed during each step of the process, the final numbers come in at a combined statline of 59 receptions, 961 total yards, 8 total TDs. This puts him at 144 FP on standard scoring settings, good for a positional rank of No. 20 to No. 25 and an extremely solid WR3 with loads of upside.
For more insight on the Rams and more, please visit numberFire.com.