CA Charting Project: Reviewing Phil Costa

On Thursday, the Colts signed former Cowboys center/guard Phil Costa. Colts fans on Twitter didn’t exactly feel good about it, but after seeing the contract numbers and doing some quick charting/film watching, I love the move. And not just because he falls under the category of “anyone other than Samson Satele or Mike McGlynn.”

The Dallas Morning News said Costa’s contract is for two years and $2.7 million and has a “max value” of $3.5 million (apparently the extra money is tied to playing time incentives). Per, Mike McGlynn just finished a two-year deal for $2.65 million. So the Colts are basically replacing McGlynn’s salary slot with a guy who is three years younger.

Also, whereas McGlynn was reliably bad and Satele was . . . Satele, Costa has at least been good sometimes.

The last time Costa played a full game was week 6 of 2012 in Baltimore. Pro Football Focus gave him a remarkable 6.3 grade (subscription required) for the game. I charted his blocks in that game and came away similarly impressed. I’ll get to the results in a bit, but first, a few representative plays.

Costa was pancaking Ravens left and right. Here are a few:





By my unscientific count, he had seven pancake blocks on the day. He did an excellent job of leveraging defensive tackles out of the way and getting them on their backs.

I did notice one play on which the Cowboys were in a bad protection to deal with a Ravens blitz. The Ravens brought six rushers, and the Cowboys had seven guys in to block, but a free rusher came up the middle and rushed Tony Romo into an interception:


As the guy closest to the unblocked rusher, Costa probably should have tried to stop him. While I’m not sure how he compares to Satele in calling protections (Satele seemed to struggle at times last year), this was the only unblocked rusher I saw.

This next one was my favorite play by Costa, though the first time I watched it, I thought he had screwed up. The Ravens are in a 2-4, and the linebacker on the left side is about to loop through the B gap to Costa’s right:


The left guard lunges at the defensive tackle and completely whiffs, leaving him with a clear path to the quarterback. The left tackle is stuck outside and in no position to help:


As he should, Costa slides over to help. He lays a big hit on the defensive tackle, knocking him far enough off course that Romo has time to throw:


Much more impressively, he recognizes the stunting linebacker coming behind and, immediately after blocking the tackle, spins on a dime to stop him as well:




A fine example of awareness and athleticism.

Now for the overall numbers.

DISCLAIMER: Grading offensive line play is inevitably subjective, since it’s impossible to know assignments and how the linemen are coached. These scores are based on whether the lineman appeared to succeed in his assignments, based on his apparent targets and how the plays developed. I assign all blocks a grade of ‘+’ (good block), ‘-’ (bad block) or ‘/’ (not involved, usually meaning the lineman couldn’t find anyone to block); ‘/’ plays are not scored. See here for an explanation of my terminology.


That 89% total blocking percentage is the same figure I found when I charted three games for Alex Mack, the guy for whom Colts fans are desperately pining (with good reason, though it’s probably a pipe dream). Satele, you may recall, checked in at 82% for the season. Costa’s 87% run blocking is Mack-ish, the 92% pass blocking is more than sufficient, and he didn’t give up a sack, hit or pressure in 38 pass blocks. In other words, Phil Costa is as good as Alex Mack.

I don’t mean that, of course. Costa hasn’t sustained anything close to that level of play on a regular basis, and the Ravens game was a clear outlier from his usual standard. He was the Cowboys’ starting center for all of 2011 and wound up with a -9.3 PFF grade (Jimmy Kempski of Blogging the Beast has some grisly analysis here). After that shining moment against the Ravens, Costa played poorly the next week before getting injured and missing the rest of the year, then Dallas drafted Travis Frederick in the first round in 2013 and Costa took one snap all season.

A guy who played one snap in 2013 and 126 the year before isn’t going to come barging into Lucas Oil Stadium and set the world on fire. But the Colts didn’t pay him like he will. They’re essentially paying him to be the offense’s version of Fili Moala: not a first-choice starter, but a player who won’t embarrass you if the better ones go down. Depth has been a major problem for the Colts’ line the past couple years; last year, when Donald Thomas got hurt and Satele, McGlynn and Hugh Thornton all struggled, the only option on the bench – aside from the mysterious Khaled Holmes – was Jeff Linkenbach. Not good.

If Costa can occasionally offer a reasonable facsimile of that Ravens game, the Colts will have a quality depth lineman. If not, $1.35 million a year for two years is a relative pittance for a warm body who can contribute on special teams. I still expect Ryan Grigson to sign another veteran center before training camp, but this was a low-risk, high-reward move that adds a bit of depth and potential to a unit that had been lacking both.