Draft theories and systems: a new trend?

The Option of Drafting a Later Round QB

I was wandering through the day's information, casual as you please, when this next bit from walterfootball.com jumped up and bit me on the imagination. As most of you know, I tend to be a big Bill Walsh fan, and Mike Holmgren was his star QBs pupil. That made me think about the following blurb:

Holmgren has never spent a first-round pick on a quarterback. There's a reason for this. Like Bill Walsh, Holmgren firmly believes that he can take "inferior" quarterbacks and make them into really proficient passers. These "inferior" signal-callers don't have to possess a great or even a very good arm; they just need to make quick decisions, and be accurate in the short and the intermediate passing game. Clausen could absolutely thrive in Holmgren's offense, but I don't think Holmgren would be willing to spend the No. 7 overall pick on the Notre Dame product - even if he is the top quarterback prospect in this class.

What caught my eye was that this article believes that there is a trend that Walsh started - supposedly - and that Holmgren is supposed to be maintaining. The evidence is sketchy, but it's possible. What interested me more was that Bill Belichick and his pupil, Josh McDaniels, may have also gotten into what may be a growing trend regarding looking for later round QBs. So far, Josh McDaniels has followed Bill Belichick in this approach, just as Holmgren has (if the article is to be believed), follows Bill Walsh. What interests me the most is that these four men - Holmgren, Walsh, Belichick and McDaniels - are all considered highly knowledgeable about QBs. Walsh, Holmgren and Belichick have had a chance to prove it. This poses a question: Do they just believe, that with their backgrounds, they can know something that others don't? If so, is that because they believe that they can find, or see, things in quality QBs later in the draft that others will miss, or is there more to it? I decided to follow the data - once I'd found it.

Looking at the research on 1st round QB success, also from walterfootball.com, you can see why they'd at least want to try. A 70% failure rate which is what is certainly a good impetus toward change. That was the number that a recent analysis of data published by Walterfootball.com put on it - 70% were poor picks. You can look it up, and some folks will probably argue the precise accuracy of the data. I'm fine with that - even with some discussion of methodology, the number that were poor picks is, "A lot." What's far more interesting is that there may be an identifiable, repeatable skillset involved here that might identify good risk later round QBs. Wouldn't that change the face of the game a bit?

By the way, at least four factors go into most accurately predicting whether a  drafted QB will be successful. One is completion percentage - it has to be 60% minimum, and you'd like to see it around 64-67%. The second is games started - the cut-off point seems to be about 37 games. Finally, you want a QB if at all possible who throws out of a pro style offense. It doesn't mean that the QB isn't throwing out of the shotgun - all programs do, in part. But a gadget offense where no skills under center are taught leaves the QB behind the curve and more likely to fail at the next level. . A fourth factor for both approaches has to do with mechanics and footwork - every college player needs some degree of work, but the closer you can find to a QB with a smooth, fast release with good balance and footwork goes a long, long way.

A lot of people might see these four factors as self-evident, but the truth is that they are just recently becoming more commonplace in terms of the NFL draft. Hype is an amazing thing - QBs with very good mechanics and footwork, the appropriate number of starts, intelligence, a good history in a pro-style offense and a nice completion rate still drop on the basis that they haven't played against 'elite talent'. It's not an unfair criticism, either, but the belief that Walsh had, and Belichick and McDaniels do gravitate to, is that the QB has to show the ability to compete against better talent - not necessarily that they should have done so in a majority of their games. If a player likes a certain school and is concerned about the quality of his education, some coaches will grade them down. Those coaches are looking for a player who wants to do nothing but play football. In the modern game, with a 3.8 year average lifespan in the league, that may be asking more than is sensible, but things like that will differentiate the beliefs that different coaches, GMs and teams go into the draft with.

I want to be upfront about this - I, too, question whether this really is the beginning of a trend, and I'm really not sure of that. This is more about bringing up the issue than trying to 'prove' it. I do think that there are issues that are arising due to the lack of a rookie wage scale, and the financial cost of drafting a QB early that would have come into this possibility more recently than, say, Bill Walsh's time. Yet there are good reasons to try this approach of finding a QB later in the draft, even if the rookie wage issue gets resolved. Walsh said that he knew as soon as he looked at a QB for more than a few seconds what that player needed and whether or not he was a candidate to be a "Walsh QB." Walsh never, to my knowledge, gave details, although I've read on his work extensively. It might be out there, and perhaps I haven't come across it yet. But Walsh took his share of criticism for some of his QB choices who are now HoFers.

It's easy to tell how Bill Walsh felt in general about this. I wrote this last year:

One thing that really stands out to me is the similarity between the criticisms thrown at Bill Walsh and those aimed at Josh McDaniels. Walsh believed that the system was the real star and that QBs were actually fungible. This may seem heresy to the fan of football, but Walsh backed it up with an incredible depth of knowledge, precision of application and unmatched brilliance in putting his money where his mouth was. For that, he was decried as being arrogant, too wrapped up in his own ego to see that he was wrong; a fool who simply got lucky a few times.

Walsh said on one occasion:

"The performance of a quarterback has to be manipulated. To a degree, coaching can make a quarterback, and it is certainly the most important factor in his success."

That's a remarkable assertion, and I doubt that a lot of fans would agree with it. But you have to take into account the cold fact that Walsh backed up his statement over and over again. He was, in fact, a genius when it came to QBs. He took them in the 1st round and the third, took undrafted QBs and made them into good players. He believed that it's the coaching that makes a QB sensational or otherwise.

He passed a lot of that knowledge on to men like Mike Shanahan (who clearly polished John Elway's storied career - I don't know if you can argue that Shanahan 'made' it, but he made it greater, certainly) and Mike Holmgren, but I never read or heard of Walsh setting down on paper the formula for choosing a QB. I believe that Bill Belichick has done so, and that Josh McDaniels has follow suit. It's a rare thing to be in on a change in the league this early. Quite possibly, this will only be tried by a handful of coaches and GMs - for now. It's also possible that this marks the first stages of a major change in how QBs are chosen - and paid.

My next step was to trace Holmgren's career and to see if I could find any evidence of what the writer is claiming. What I found is pretty shaky, at best. You see, Holmgren came into the 49ers organization after being the QB coach and mentor for Steve Young at BYU. Holmgren didn't recruit Young - LaVell Edwards was said to (I could only find a single reference on this and am open to other sources). I'm sure that he had a lot to do with Walsh trading for Young, who was first taken by the old LA Express and them went to Tampa Bay with the 1st pick in the 1084 supplemental draft. TB was unable to use him well, and the knock on him became that he was an athlete but not a QB. Young cost the 48ers a 2nd and a 4th pick, when the 49ers decided to replace Joe Montana with Young, the fans were incensed until they saw Young play for a season or two. Young became a 7 time Pro bowl QB with 3 1st team All-Pro berths.

Holmgren was then the QBs coach under Bill Walsh, which is about as fine a recommendation for knowing QBs as there is in the game. It meant that he coached players like Joe Montana and Steve Young, but that he didn't choose them. He may have been involved with the choice of Steve Young, but the degree is unknown. Holmgren had coached Young and mentored him at BYU before Walsh took him in the trade. Given the pick involved, Holmgren must have given a heck of a recommendation. We can assume that Steve Young is a Holmgren kind of guy. Young was also a variation for Walsh - a 1st round QB, second pick of the draft, an area that Walsh usually eschewed. You can see that Walsh wasn't hidebound by his theories - sometimes you see the perfect option and he's in the 1st round.

Brett Favre was already at GB when Holmgren moved there. Holmgren went to GB the year after they obtained Favre, according to what I could find on this, and then he moved to SEA, working a trade to obtain Matt Hasselbeck. In other words, once you're past Holmgren's work with Steve Young (and you have to note that LaVell Edwards did the recruiting of Young, not Mike H) until you get to the Hasselbeck pick, all that you want to know about Holmgren's approach to drafting QBs is obscured by the fact that he kept inheriting brilliant QBs chosen by other people. Having great karma with regard to QBs isn't going to answer this question - but the next QB obtained for Cleveland just might. Hasselbeck, though, is also a big (6'4", 225 lb) highly intelligent QB with limited mobility and a limited arm. This is Hasselbeck's scouting report from ESPN Insider:

Hasselbeck has developed into one of the better quarterbacks in the league since coming to Seattle from Green Bay in 2001. He is a smart quarterback who knows how to prepare for games and how to manage his team on the field. He is not a dynamic presence on the field, but does play with a certain amount of calmness and confidence that translates to his teammates. He has decent footwork and foot quickness to set up in the pocket. He has the ability to slide or step up in the pocket to buy some time, but will never be mistaken for a scrambling quarterback. He can read coverages and understands what the offense is looking for depending on what the defense shows and will generally go through his progressions and deliver the ball to the correct receiver. He has an above-average arm, can throw the ball through relatively small windows with accuracy and has enough of an arm to make the intermediate out throws. But, he does not have the canon necessary to hit the longer outs, without giving defenders time to close on the ball. He generally throws the ball from a three-quarters delivery and can be extremely sharp when he gets into a rhythm. He has no glaring weaknesses, but also does not really jump off the screen with any outstanding physical qualities.

In other words, he sounds a lot like Orton. Hmmm...it's also worth noting that when Hasselbeck had good completion rates, SEA won. When Hasselbeck struggled with accuracy, they lost. I don't have enough information to know exactly why he struggled - the OL, receivers, injuries, or just poor play - but there was a connection is obvious between completion percentage and wins when you look back over the years is obvious. You will also note that his sack numbers go way up when he loses and down when he wins - maybe he needs a better OL, just as Orton does and Cassel did. Brady needs one, too, for that matter. Frankly, they all do. It helps to have folks to catch the ball once you've thrown it, too. But it's fair to note that Hasselbeck is the only QB that Holmgren went out and obtained as a starter and that he fits the mold that Belichick and McDaniels are now trying to use

Both Montana and Young were 6'2 and were on the skinny side at first. That didn't matter for the WCO that Walsh had designed and was implementing, but it could be a second type to look for in the later rounds. After all, Montana was a third round player (ironically, Walsh wanted a different QB - Phil Simms - but 'settled' for Joe Montana). Young came in the first round, but of the supplemental draft. Both players were brilliant in terms of their understanding of the game - that seems to be a constant in both the Walsh/Holmgren line (if it exists) and in the Belichick/McDaniels line, for which there is evidence. Both approaches require that the QBs be extremely smart.

What is the presumed Belichick/McDaniels mold? Everyone has probably seen this, but it's a compilation of factors that I put together on what it is that BB and JMD seem to be looking for in their QBs. To be clear, this doesn't go back further than the New England head coaching job for Belichick, because that's when he put all of his theories together and designed the NE System. Since them, as nearly as I can find, 7 of the 8 QBs drafted or obtained between Belichick and McDaniels are:

6'4+ if possible, must show the ability to read coverages and blitzes pre-snap, must show the ability to routinely check down through his receiving progressions post snap, must show above-average accuracy and, more importantly, timing on short and intermediate routes - the routes his offense is based around. 'Arm strength' is a must but not as far as throwing lots of deep passes. Rather, what's needed is more regarding velocity on the short and intermediate throws. Athleticism in terms of foot speed and quickness are a low necessity... you can teach a guy through improved mechanics and via time in the weight room how to drive the deep ball better. Through drills and practice you can get a guy who runs the 40 in 5+ seconds (yes, that's an exaggeration) to feel pressure well and maneuver in the pocket. The players who have been chosen so far have also had excellent leadership qualities . . .

What you can't do (or at least, what is much harder to do) is to change how a guy thinks and processes information. Guys who can run around and make plays, throw the ball 70 yards like it's nothing - they get drafted early because coaches think they can teach them the mental aspects of the game. And let's face it - many times, it's worked, it's just that it has only worked in a too-small percentage of cases.

Belichick and McDaniels have made their 'type' very specific, and so far they do not seem to have wavered from it, save for one case, which I'll detail later. Their choices include Tom Brady, in theory the first of this type, Matt Cassel, Kevin O'Connell, Chris Simms, Kyle Orton, Tom Brandstater and Brady Quinn. It's not a very long list, so you can argue with some basis that the sample size is too small. However - when 8 QBs have been acquired, all of whom share certain characteristics, and only one of whom falls outside of those characteristics, to me that is clearly a pattern. I did find a single variant of type, and it was from 2002 and was Bill Belichick's choice. His name was, believe it or not, Rohan St. Patrick Davey and he was a fine, Irish, 6'2, 245 lb bust. The successes and the maybes, so far, are of a specific, consistent type. The single errant example didn't make it. There may be a lesson here - only time will tell.

From what we have seen so far, it looks like Belichick/McDaniels prefer to grab the smarter, less athletic guys in the later rounds and to build an offense around the ability to be accurate and make good reads within the system that has been put together. This idea is similar to Bill Walsh's - but Walsh was a bit more extreme. He considered the best approach was to deal with QBs by taking away as much decision-making as possible. He also felt that they were actors in a production that he was putting on, and considered himself the director and the stage manager all in one (See "The Genius", by David Harris, and "The Blind Side" by Michael Lewis. Both texts deal with this side of Walsh).

The eminently logical 'Spock' stopped by and added this gem about the idea of teaching the mental side of the game:

(C)oaches think they can teach them the mental aspects of the game

Ain't it the truth. People just don't learn (especially Al!) and think they can take a QB "with all the tools" and teach him how to read defenses and think fast. Yet we see QBs with physical limitations - do Tom Brady and Drew Brees ring a bell? - reach the top of the heap because they have what the physical freaks lack: the mental tools. But if the rest of the league doesn't get it and McDaniels and Xanders do, that's to our advantage. And I think the same is true to only a slightly lesser extent at other positions. On defense, for instance, diagnosing rapidly and correctly and taking the first step in the right direction gives an edge that superior athleticism can't overcome, hence the emphasis on players who are not only tough but smart.

As usual, Spock has a point.

Bill Belichick's story is a little more involved. He was an econ major in college, and when he designed the current NE organization from the ground up, he asked for help from his old econ professor as well as those in other fields (like the Laker's GM) who had found success while others struggled. Belichick designed the exact way that he wanted to team to function, and that included how he wanted to draft. If he is, in fact, looking for QBs in later rounds, there are at least two good, money-based reasons for taking on this approach:

  • First, the first round QBs are outrageously expensive and a crap shoot at best. As is common in gambling, the odds favor the 'house' - you lose much more often than you win, and the cap hit can cripple you for years.
  • Second, any funds that are spent on that 1st round QB can't be spent on the OL that protects him, the WRs that catch his passes, the TEs that create both pass protection and receiving options and the running backs that balance the passing game and keep the defense honest. For that matter, those are dollars that you can't spend on your defense, either.

It will take years before we know for certain that this is the Belichick/McDaniels approach. There may well be 1st or 2nd round talents that either man snaps up, just as Walsh took Young. Talent is talent and you need as much of it as you can get. I would never suggest that you draft a QB because he's a later round player. I'm pointing out that increasingly, there is more interest in looking at developing QBs from later rounds.

The real question is, "How do you recognize 'it'?" That's what makes the Belichick 'type' so interesting. It's the first time that I can find so far that a coach has deliberately chosen to pick later round QBs, and the first time that you could potentially delineate a 'type' that was being considered.

As far as McDaniels goes, it's interesting that all three of his current QBs, including Brandstater and Brady Quinn, are of the same model (Quinn, of course, was a first round choice, but by Cleveland). Chris Simms was also of that type, and I wish him well in Tennessee. So, of eight choices, 7 match the mold I listed. Neither Belichick nor McDaniels really chose Simms, a former 3rd round player who hasn't so far made it back from a horrible injury: McD took a chance that he was fully recovered, physically and emotionally. It didn't pay off. McD did choose Orton, even though he didn't draft him. That one is murky as well, since some will state that the trade was determined solely by the number of picks. I think that McDaniels' glowing comments about Orton's smarts were heartfelt, but certainly, getting an extra first round pick is a fair impetus to take a given player. Even so - both Orton and Simms fit this mold. Does that mean that McD won't break out of the mold for the right player. Not at all - these are principles, not commandments.

There are going to be plenty of HC/GMs who just don't have the understanding to chose quality QBs later in the draft. Until such a time as a template for developing these kinds of QBs comes to light and can be copied, those types aren't going to show interest. Up until quite recently, the only people who were looking for a later round QB (or an undrafted one, like Tony Romo) tended to just want a functional backup, at best. They expected little, put little time into the QBs, and not surprisingly, usually didn't get anything else. That has tweaked the numbers against such pickups - they weren't expected to become NFL starters, they weren't trained to become starters and they rarely, if ever, got a chance to be one. All of these factors twist the statistics. Many of the HC/GM combos don't have enough skill to look for starters at QB later in the draft. This works in favor of the few who prefer to go that way. However - as more decisions are made, and info gathered, we could be looking at the beginning of new ways to choose the position.

Will it work? Your guess is as good as mine. Tom Brady is a one in a thousand find, and while Cassel managed to go from backup to starter, his long term career is up in the air. Orton has been a winner, but he needs to continue to improve to get beyond the level of a mid-range starter or a top backup and into top-10 range for QBs. It's accurate to say that McDaniels didn't exactly chose him in the draft (My own feeling is that Orton would be much farther along if he had), but McD did acquire him, and he does meet the criteria. Brandstater is too young to know about, the same is true of O'Connell and Quinn was sentenced to three years in Cleveland for some unnamed and unpaid karmic debts - how he'll do with quality coaching is yet to be determined.

By the way, a note on Brady: It came out recently that one of the 49ers coaches wrote a bitter report damning him and it carried so much weight with the staff that they let him go in the draft. Funny how life works, isn't it? If that coach had been hired by someone else, SF would still have one of the two best players at the position in the game.

You may continue to see a newer movement in the QB hunt. I think that a rookie salary cap would change this situation in degree, because it takes away some of the need to minimize rookie QB salary issues. Even so - while there will always be QBs who are well worth the early pick money, an increase in the number of teams that deliberately look to later choices has multiple benefits. You have more money to spend on other players, and even more importantly, it will require teams to look more carefully and specifically at precisely what they want in the QB. Is it the gun, the big arm? Eric Mangini suggested that was a big factor for him - we don't know what Holmgren has in mind for Cleveland yet, and the Delhomme pickup doesn't really tell us much. In fact, looking back over his career, Holmgren only obtained one starting QB as near as i can determine. He brought over Matt Hasselbeck from Green Bay. Hasselbeck was drafted by GB in 1999, the same year that Holmgren moved from GB to SEA. In 2001, Holmgren traded for him, so you can call him a 'Holmgren type'. Ironically, he, too, was drafted in the 6th round. Seems to be something going around...

To sum this up - there's no question that Bill Walsh felt that he could take later round QBs and make them great. He managed it several times. There's Montana, of course, there was DeBerg, who was there when Walsh took over SF. DeBerg really improved under Walsh, but Walsh had the idea that he wanted a QB that he trained from the start. That was Joe Montana, the third-round, skinny, short (6'2") guy with the noodle arm, who's sitting in the Hall of Fame. Walsh taught players as diverse as Montana and Virgil Carter, the scrawny Cincinnati QB who could only throw the ball about 10-15 yards, but was very accurate in that range. The WCO was literally designed for him. Walsh proved his theories when he coached in San Diego, where Dan Fouts has stated that Walsh's training made his career (Fouts then found fame under the Air Coryell system, but was always careful to give most of the credit to Walsh). Phil Simms spent about 30 minutes with Walsh, prior to the draft in which Walsh ended up with Montana, and Simms said that it changed his entire career. Great teachers can do that. Walsh also proved his theories when he was coaching at Stanford. He never, to the best of my knowledge, shared his formula. There may not have been one - some coaches just know, at a level below (or above) thought.

Lots of WCO QBs seem to be around 6'2 - I'm not sure that it means much, but we'll see who Holmgren chooses next and what offense EM is going to employ. Until now, he coached Steve Young in college, found Brett Favre already in Green Bay and left anyway, and he then traded for 6th round Matt Hasselbeck, who fits the Belichick mold exactly, for Seattle.

Then you go to the Belichick line and there are currently 8 QBs in that lineage. I think that there is enough information to establish that a mold is being used, but there isn't enough data to estimate its effectiveness yet. There's a point to be made, though. Both Belichick and McDaniels have been tapping their networks to find these QBs. That's a big reason that New England's approach to the draft has been so successful. It's not the 2nd round 'value' picks - those came later. It's the vast numbers of coaches, trainers, athletic directors and assistants that Belichick and McDaniels know. Both men grew up in coaching households. Both were introduced to coaches from all over the country and both have made a point of developing their contact list. That also plays into it. But at the bottom line?

1st round QBs, especially higher pick QBs, are very expensive. Finding them later in the draft is an art that is not for every team, every coach or every GM. It's not even for those coaches who prefer the approach: if there is a guy at 1 or 2 in the 1st round, second round, who you believe can win it all with, ala Steve Young, you go with him. It will create more financial difficulties in getting better players at every position, but for most clubs, it's a trade-off that they'll take Even so - if a system can be (or has now been) devised that permits a 50-50 chance of success, that will be substantially better than the approaches that are in existence now. Most of them came down to 'saw him, liked him, drafted him and prayed for success'. I'm hoping that we are seeing the beginning of a better way to do this. Only time will tell.

I am comfortable saying that McDaniels will pull the trigger on any player that he thinks that he can make into a winner for the Broncos. It doesn't matter if it's the 1st round or the 6th (or after). This is about putting together a winning team, and nothing else. You won't find Belichick, McDaniels or, in my opinion, (regardless of the opening article) Holmgren, passing up a highly rated QB because of the draft slot, higher or lower. They're just trying to win a...well, you know.

Changes and History

When I look back over the history of the game, it wasn't until the late 40's and 1950's before teams even began to scout small school players, for example. It was rare for even longer before players were drafted from historically black colleges. For a long time, you never drafted a QB who was black. These patterns have dropped by the roadside, and for good reason. In the same vein - right now, choosing QBs tends to be a low-return issue. If there is a way to improve that percentage, expect certain coaches - ones who both have a lot of understanding of the QB position and who aren't afraid to innovate - to look at new and different ways to make this work for the team. They will often be criticized for this, but that's unimportant, really. If the players begin to be seen as working out well at the NFL level, you'll see more of this happening.

Others will always wait and hope that there will be innovations, things that they can copy. I'm comfortable saying that McDaniels will always be open to innovation. From what I've seen of him, it appears to be in his nature, in his makeup. As he said in his Friday presser, he has changed things that he felt weren't done as well as they could be and they are going to keep some things that he felt worked well. When it comes to a QB, no matter what the round or situation, McDaniels will take each situation as it comes and draft players who he feels will be winners at the NFL level.

But there's no reason to expect that preclude later round QBs. Right now, it's almost an unmined resource. in a league that has always changed to find new resources, expect that to change.

Originally posted at MHR

Learn to laugh at yourself. You will be ceaselessly amused. - Sri Gary Olsen

You can reach Doc at .(JavaScript must be enabled to view this email address) or follow him on Twitter @alloverfatman

Doc's Musings

2014 Offseason

Offseason coverage