Bradford Doolittle: Chiefs shouldn't discount Thigpen

[FringeNC]

Chiefs shouldn't discount Thigpen

While Chiefs fans are stuck in limbo waiting for new coach, let's talk some actual on-the-field football and address a little personnel matter. Actually, it's not so little. If it weren't for all the turmoil surrounding the leadership of the Chiefs, this would be the No. 1 question on the minds of their fans right now. The question is simple and its answer will no doubt have serious ramifications in regards to who KC selects with the third pick in the upcoming NFL draft. Is Tyler Thigpen a legitimate NFL quarterback?

Let's start with some scouting observations. First, from his rivals.com profile based on his college days:

* "Has a live arm with a compact delivery and has developed a nice understanding of defenses. He is quick in his setup and has the mobility to avoid the pass rush. ... He needs work on the finer points of the position to maximize his fine arm and athletic ability. ... Could become a regular in the right setting with development."

That sounds about right. Next, let's turn to Pro Football Prospectus, which bases its observations on statistical analysis and also features a nifty system for projecting the pro performance of college quarterbacks. In 2007, after Thigpen had been drafted by Minnesota, PFP wrote:

* "He's an undersized scrambler with bad mechanics."

Not a glowing report and clashes with the Rivals piece. However, in 2008, PFP dug a little deeper:

* "Apparently we didn't know what the Vikings had, and neither did the Vikings. ... relies too much on physical gifts, trying to make plays that worked fine against the Division I-AA or Division II stars, but will get him destoyed in the NFL. ... In a perfect world, he turns into (the Chiefs') version of Derek Anderson."

This was in the last offseason, when becoming Derek Anderson seemed like a good thing. PFP also offered its first statistical projection of Thigpen: 56.7% completion, 17 TD/16 INT in 493 attempts and a -32.1% DVOA (read about that stat here). His actual numbers: 54.8%, 18 TD/12 INT, -6.1% DVOA -- much better than projected.

So, as you would expect with a seventh-round pick, the scouting reports aren't glowing but there were some things to like, some upside potential and the Chiefs did well to grab Thigpen when the Vikings tried to sneak him through waivers. Could they have expected Thigpen to emerge as their starting quarterback in his second year in the NFL? Certainly not. It's exceedingly rare for a seventh-round quarterback to play -- ever. But it happens. Since 1990, there have been eight QBs drafted in rounds seven or later that eventually became the primary starter for their team for at least one season:

1. Brad Johnson
2. Trent Green
3. Gus Frerotte
4. Elvis Grbac
5. Kent Graham
6. Ty Detmer
7. Tim Rattay
8. Matt Cassell

Not exactly a murderer's row of signal-callers and a couple of guys KC fans are quite familiar with. There are six pro-bowl seasons mixed in there and Johnson was the starting QB on a Super Bowl champion. Of these, Green and Frerotte were both starters in their second season. Those two have combined to throw for nearly 50,000 yards in the NFL. So it's not out of the question that a guy like Thigpen could turn into a good starting quarterback. At the same time, guys like him are the exception not the rule: There were 67 quarterbacks taken in the seventh round or later in the time frame I examined.

COMPARABLES

The upshot of Thigpen getting so much field time last season is that you actually have some empircal data with which to work. That's always good. What we're interested in finding out is how quarterbacks with similar statistical profiles to Thigpen eventually developed. The hope is that we can hazard some sort of guess at what the chances are that Thigpen develops into a legit NFL starter.

Technical explanation for what I've done: I calculated a "similarity score" for Thigpen and every quarterback that's thrown at least 100 passes in an NFL season since 1995. I've chosen that time frame because that's how far back the data at Football Outsiders dates back and I wanted to use that as part of my criteria. After doing that, I narrowed the list down to seasons for first- and second-year quarterbacks. There are 109 such seasons. I've only looked at passing categories.

Here are the categories I used in my sim score calculation: completion percentage, yards per attempt above/below league average, first down percentage, TD/INT differential (TD-INT divided by pass attempts), sack percentage, percentage of completed passes for short yardage (1-19 yards), percentage of completed passes for medium yardage (20-39 yards), percentage of completed passes for long yardage (40+ yards) and DVOA.

Here are the 10-most similar QB seasons for first- and second-year players to Thigpen's 2008 campaign:

1. Patrick Ramsey, 2003
2. Damon Huard, 1999
3. Trent Green, 1998
4. Bobby Hoying, 1997
5. Shaun King, 1999
6. Donovan McNabb, 2000
7. Patrick Ramsey, 2002
8. Aaron Brooks, 2001
9. Jason Campbell, 2006
10. Chris Simms, 2005 (When he had his spleen, Gene)

That's a pretty good range of quarterbacks, both in the positive and the negative. McNabb is the name that jumps out at you, but let's not forget our scouting report/draft position knowledge. McNabb was very highly rated and was the second player picked in the 1999 draft.

However, Green, as we've pointed out, was a seventh-round pick. So if you agree that draft position is a good proxy for consensus scouting opinion, then in many respects, Trent Green represents the ceiling for Tyler Thigpen. The likes of Ramsey, Hoying, King, et al, represent the floor, at least as it pertains to our sample of young starters. Thigpen has already exceeded his literal floor -- those dozens of late-round QBs that never saw the field in an NFL game.

CAREER VALUE

To recap, what we've done to this point is to identify the 109 quarterback seasons posted by players that, like Thigpen, played extensively by their second year in the league. Then we calculated which of those players were most similar to Thigpen. Now, our next step is to look at the development of these similar players.

I sorted my list by similarity scores, then cut it off at the 22 most similar QBs. I wanted to pick the top 20, but Patrick Ramsey and Shaun King, who started in both their first and second years in the NFL, are listed twice, so I added two more. As a proxy for career value, I used Football Outsiders' DYAR (defense-adjusted yards above replacement) metric, prorated for 500 attempts. Sounds complicated, but the scale works just like raw yardage: the higher a player's DYAR, the better.

In this instance, a figure of about 400 DYAR per 500 attempts signifies a playoff-caliber quarterback. There are six players on Thigpen's top-20 comparables that have posted a figure that high, led by Trent Green. So, by my reasoning -- and this is the guts of this whole piece -- you could peg Thigpen's chances of developing into a legit (ie, playoff-caliber) starter at about 30 percent.

Those are actually pretty good odds in player-development terms. The percentages aren't nearly high enough -- and the sample isn't large enough -- to discourage the Chiefs from bringing in another quarterback in the offseason, be it through the draft, a trade or free agency. However, they do suggest that Thigpen has some upside.

Personally, I would be concerned about Thigpen's completion percentage, which ranked 30 out of 32 qualifying quarterbacks in the past season. Among the many bits of information that have resulted from the research by the Football Outsiders/Pro Football Prospectus crew is this: Quarterbacks tend to be accurate or not. It's hard to "coach up" this aspect of a signal caller.

However, when I ran regressions of the different categories in my similarity score scheme against career value, two categories stood out as most important: DVOA and PAA (yards per pass attempt above or below the league average). In those two crucial categories, which track very closely, Thigpen ranked 48th and 49th out of the 110 players ranked. That's fairly encouraging -- he's in the top half. However, it underscores the overall notion that while Thigpen may have some real potential, no one, including the Chiefs, really know what they have in him at this point.

CONCLUSIONS

If the Chiefs bring in another QB, what are the odds that that player develops into a playoff-caliber performer? Fifty-fifty? Worse than that? That's why it doesn't make any sense to me that people approach this Thigpen/QB situation as an either/or proposition. The guy makes squat in salary-cap terms -- keep him and bring in someone else to compete.

Young quarterbacks are rarely a sure thing and there doesn't appear to be a sure thing in this draft. But there will be quality quarterbacks that emerge from this draft class. Whether or not the Chiefs use their first pick on a QB, they do need to take one on the first day of the draft. That player would then battle it out with Thigpen next summer, presumably with a veteran quarterback -- Damon Huard? -- waiting in the wings.

Tyler Thigpen may not pan out as a starting quarterback in the NFL. However, at this point, based on what we saw last season, you can't rule him out just yet.

THE LIST

I'll leave you with the full list of the 109 quarterback seasons used in my Thigpen analysis. The seasons are ranked by DYAR/500 attempts, however the left column lists the similarity ranking to Thigpen's 2008 season. In other words, the #52 ranking next to Peyton Manning's rookie season means that it was the 52nd most-similar season in this sample to Thigpen's 2008 season.

[Full list in a table at the link above]

[Chiefnj2]

Bring Thigpen back to compete. If he can play under center and continues to improve the Chiefs made a good find. If he can't improve then not much wasted.

[FringeNC]

Originally Posted by smittysbar

Factoring in the staying healthy, he would lose, but if you are just going on QB performance during TC between the two, it isn't close. I have watched them both in TC.......but Brodie ended up having the body of a glass jar

I don't think how a player, especially a quarterback, looks in training camp predicts how they well handle the pressure of an NFL game. There's really no evidence at all the Croyle can read an NFL defense, or make throws under pressure, so even if Croyle could stay healthy, he hasn't proven anything other than he has a good training camp arm.

[DJ's left nut]

Originally Posted by Marcellus

Your post got me to looking at career QB completion% (Accuracy). Looks like the standard is about 62.5% give or take .5% for a really good QB. Breese, McNabb, Rothlisberger (sp?), Rivers, and Brady are all about 62%. Payton manning is 64.4% for his career and was only 56.7% his rookie year.

Anyway Thigpen was 54.8% this year. I am not a Thigpen lover nor am I a hater. If he earns a job he earns it, if not he doesn't. We should still draft a QB. I enjoyed his play about 50% of the time this year, 1st half mostly.

My thought is that the difference between 54% and 64% is completing 10 more attempts every 100 or about 3 every game he throws 33 times. Is it really inconceivable that he can't improve that much?

It's not impossible, it's just damn unlikely. That's the point I've been trying to make for awhile now. It's like plate discipline in baseball. One would think it's easy to coach up and work into a player and it just isn't. Guys that don't walk in the minors won't walk in the pros. There are some outliers (Jose Reyes), but not many.

Thigpen may be the outlier, but it's pretty unlikely. And those 3 completions per game are pretty huge when you consider those as 3 more first downs. Additionally, completion % is indicative of more than just how well he completes passes, it also speaks to where he puts them when he does. A QB with a higher completion % is also more likely to put a ball in a position to keep his reciever in full stride for YAC yards. More likely to connect on a deep pass, more likely to keep his guys healthy. It's more than just those 3 completions.

I don't see it happening. I'm still torn on drafting a QB at 3 b/c I'm not in love with either of the options (and Crabtree is a monster). However, there's no way we stop looking for a QB on Tyler Thigpen's account.

[King_Chief_Fan]

Originally Posted by FringeNC

I don't think how a player, especially a quarterback, looks in training camp predicts how they well handle the pressure of an NFL game. There's really no evidence at all the Croyle can read an NFL defense, or make throws under pressure, so even if Croyle could stay healthy, he hasn't proven anything other than he has a good training camp arm.

especially in a training camp led by Herm Edwards.

[Chiefnj2]

Originally Posted by DJ's left nut

It's not impossible, it's just damn unlikely. That's the point I've been trying to make for awhile now. It's like plate discipline in baseball. One would think it's easy to coach up and work into a player and it just isn't. Guys that don't walk in the minors won't walk in the pros. There are some outliers (Jose Reyes), but not many.

Thigpen may be the outlier, but it's pretty unlikely. And those 3 completions per game are pretty huge when you consider those as 3 more first downs. Additionally, completion % is indicative of more than just how well he completes passes, it also speaks to where he puts them when he does. A QB with a higher completion % is also more likely to put a ball in a position to keep his reciever in full stride for YAC yards. More likely to connect on a deep pass, more likely to keep his guys healthy. It's more than just those 3 completions.

I don't see it happening. I'm still torn on drafting a QB at 3 b/c I'm not in love with either of the options (and Crabtree is a monster). However, there's no way we stop looking for a QB on Tyler Thigpen's account.

3 completions can be huge, but it doesn't mean it was always Thigpen's fault. I thought his accuracy was very spotty, but it doesn't help that the #1 WR on team was in the top 3 for dropped passes and other than Tony G there really wasn't a lot of help at the WR position.

[DJ's left nut]

Originally Posted by Chiefnj2

3 completions can be huge, but it doesn't mean it was always Thigpen's fault. I thought his accuracy was very spotty, but it doesn't help that the #1 WR on team was in the top 3 for dropped passes and other than Tony G there really wasn't a lot of help at the WR position.

If you're going to give him a pass for D-Bowe's drops, you have to also take away credit for any # of balls he forced in to TG that wouldn't have been caught by a mere mortal.

Thigpens pass catchers, on balance, did more to help him than hurt him. You take the good with the bad when it comes to that sort of thing.

And I haven't event gotten into the simplicity of the offense he ran.

I think you can say without hesitation that he was one of the 3 least accurate QBs in football last year. The stats can't be taken as gospel when viewed in a vacuum. However, when viewed in concert with his performance, it seems that the stats are pretty accurate. For every ball Bowe drops, TG makes a miracle. Ultimately it's fair to say that at least 1/2 his balls were not well thrown.

That's not acceptable and unlikely to change.