In just 15 games in 2008, Brandon Marshall caught 104 passes. The feat ranked third in the NFL for the season (behind Andre Johnson and Wes Welker) and second in the all-time history of the Denver Broncos. With our own eyes, we saw the remarkable degree to which he keyed the offense. We marveled at his acrobatic catches (think of the rising, back-of-the-endzone grab in Kansas City) and his often astonishing ability to stay on his feet, describing a perfect circle behind the defense as he hauls in the ball on the sideline, shakes off a defender, and cuts back behind the grain. It was wonderful, and as much as we all loved Rod Smith's dedication to the game and technical expertise, Marshall's physical gifts captured our attention in a unique way.

But our eyes often deceive us, and the highlight-reel memories we assemble of a player can misled - often badly - when it comes to assessing his impact on the team. This article looks at the considerable cost at which Marshall's accomplishments came, and asks what a fair appraisal of his production (as opposed to his talent and potential) looks like. Two points before we jump in. First, a hat tip to Jaffe28, who posted a more concise form of these statistics earlier today. I've been working on this for awhile, and don't want it to be seen as plagiarism. Second, I want Marshall to stay on the team - this is not a hit job. My point of departure is the observation that he has asked the Broncos to evaluate his potential. Given all the if's attached to this young man, I want to ask what his production has looked like once we account for the disproportionate number of times Jay Cutler threw to him last year. Every throw on the football field comes at a cost - of another receiver not thrown to, or of a run not taken. To me, the dominant question is: how efficient were the Broncos when they directed the action Marshall's way? The analysis below doesn't try to answer the quesiton of whether Brandon Marshall or Jay Cutler or Jeremy Bates is at fault for Marshall's underwhelming numbers. It just asks what those numbers are, and contextualizes them via comparison to the league's other frequently targeted wide receivers.

Putting aside his injuries, legal troubles, contract demands and attitude, a closer look at the numbers suggests several important qualifications to Marshall's on-field performance. We all know about his relative lack of touchdowns. Marshall reached the end zone just six times last year, and just seven times the year before that. But a bigger concern to my mind is the relative inefficiency of the passing game when it has been directed Marshall's way. Rather than dance around the point, I'll cut straight to the table:

Table 1: Reception Percentage (Receptions/Targets)

On his way to completing those 104 throws to Marshall, Jay Cutler threw the ball his way 182 times. Andre Johnson, the league's second most thrown-to receiver was targeted "just" 170 times and finished with 310 more yards. Putting aside, for a moment, the question of what kinds of throws the quarterbacks in question made, we have an easy explanation for Marshall's comparatively low yardage production: Compared to the league's best wide-outs, he caught a small percentage of the balls thrown his way. To be specific, Marshall's catch-conversion rate (that is, the share of times he caught balls thrown his way) was 57.1% -- the thirty-first best rate for wide receivers targeted at least 80 times.

Pro football has many kinds of receivers, routes, and plays, so I'll try to put the number in context. First of all, look at the top three receivers in the table. Wes Welker and Eddie Royal (OK, I buried the lede on that one) run different kinds of routes - shorter, lower-reward, and far more likely to result in completions. Welker caught nearly three out of four balls thrown his way, and Royal caught seven out of ten. Anquan Boldin, while more obviously athletic, plays a similar-ish role on the Cardinals' offense. As far as I'm concerned, he's a revelation, and if I were a desperate team targeting a disenchanted wideout with iffy health and dollar signs in his eyes, I know who it would be.

OK, let's take another cut at the numbers. Draw an imaginary line below Michael Jenkins. That line marks the league-wide completion percentage of 61% for last year; every Wide Receiver above the line had a catch-rate (or, if you prefer, completion percentage) above the league average. That's a noteworthy achievement for a wide-out because the league-wide completion rate includes lots of low-risk, check-down throws to running backs, #4 wideouts, and other players who make good targets precisely because they're such lesser threats that the defense often leaves them open. Look closely at the twenty players above the line, and you see many (although not all) of the truly elite #1 wide-outs: Andre Johnson, Larry Fitzgerald, Lee Evans, Greg Jennings. What's notable here is that Marshall comes in behind not only these highly accomplished monsters, but behind a number of receivers - Matt Jones, Tedd Ginn, Jerico Cotchery, Kevin Walter, Antonio Bryant - who earn far less acclaim.

OK, fine, he didn't catch a lot balls, you say, but catching the ball is not a wide receiver's only job. He's also supposed to make big plays. Sure, the Walsh offense and increasing sophistication of the passing game have made it a lower-risk, lower-reward endeavor, but a deep threat that stretches the defense and moves the team down the field in big leaps still has a lot of value. Despite his physical gifts, Marshall is not that player. To the contrary, despite his low reception rate, he performs better as a short, possession receiver - drops and all - than he does as a deep threat.

Table 2: Yards per Reception

More interesting than Marshall's numbers are the overall numbers for the Broncos' passing game. Marshall ranks 33rd in yards per reception among the highly-targeted players here, but first among

 Broncos. Eddie Royal and Brandon Stokely followed, each at 10.8 yards per catch. (Tony Scheffler is another story, but he plays another position). This fascinates me because I believed that the Broncos employed a high-risk, high-reward passing game last year. The numbers here suggest it was more like a medium-risk, low-reward game.

OK, before wrapping it up, let's look at the most commonly cited mitigating factor in Marshall's production. The guy racks up a lot of first downs. In fact, it was posited - quite smartly - early today that his low Yards per Catch rate owes to the fact that the team asked him to run lots of routes only as far as the first-down marker. On this count, most of us feel that he succeeded. But once again, the numbers tell a different story.

Table 4: First Downs per Times Thrown to on Third Down

In 2008, 44 NFL receivers were targeted 25 or more times on Third Down. Collectively, Quarterbacks threw 1,509 passes their way. They converted 687 into first downs, for an average conversion rate of 46%. Brandon Marshall earned 18 first downs on 43 third-down targets - a success rate of 42%. That figure places him 28th in the group.

Overall, Marshall is in the middle or the bottom of the pack for every statistic that describes a specialized wide-receiver skill. He did not catch a particularly large number of the balls thrown his way. His receptions went for short yardage. And on third down, he moved the chains at a below-average rate. On none of the receiving statistics above did he rank above 28th in the NFL.

So am I saying he's a lousy receiver, and that the Broncos should kick him to the curb? No. I think he can flourish, and mightily, in the McDaniels system. My analysis is not the most charitable. He probably should not be blamed for Cutler forcing the ball to him as often as Cutler did. The problems of the Jeremy Bates scheme are not his fault. And yet those problems are the reason why Marshall, despite the many what-ifs surrounding his future performance - can he stay healthy? Will he give his all to the team? Will he stay on the roster? Will he stay out of lock-up? - is demanding his payday. If Cutler had spread the ball around more or the Broncos had called more running plays (as they should have), his numbers would not be so gaudy. As the tables above show, the single most impressive thing about Marshall's on-field performance was the number of times Jay Cutler threw him the ball.

Marshall's talent captivates me. I want him to succeed. And I want him to do it in a Broncos uniform. But when weighing the many risks of his future against the documented inefficiency of his production with the Broncos, I don't like the results. I hope the team can keep him this year under his current contract. Barring that, I hope that some other team - hello, Dan Snyder - focuses on his upside rather than his production. Because I fear that's the only way the Broncos will get maximum value out of this gifted, frustrating athlete.