A Trade that Sounds Like an Auteur Trying to be a Trade

This doesn’t happen in the NFL.

On Sunday, two number-one overall picks from past drafts were traded… for each other. Both are quarterbacks playing under “you’re my franchise guy” contracts. Combined, $40 million in dead cap space was added to 2021 rosters in a season where the salary cap is expected to significantly drop.

How is this possible?

Let’s look at the players involved. The Lions traded Matthew Stafford, the first pick of the 2009 draft. He will be 33 entering next season. He has a career record (74-93-1) well under .500 in a world where under .500 guarantees you’re holding a clipboard, at best. Quarterbacks are paid to win. Nevertheless, Stafford’s physical skills are quite impressive and his statistics, for a team that is usually inconsistent or worse at running the ball, reveal someone who is a little above average for an NFL starter.

After another season of Lions failure and another fired head coach, Stafford, who has never won a playoff game, was reportedly seeking a new home.

The Rams traded Jared Goff, the first pick of the 2016 draft. He will be 26 entering next season. His career record is very good (45-30), and he has even played in a Super Bowl (though the Patriots solved him quite effectively in that game). He did not play that well down the stretch this season, broke his thumb in week 16, came back two weeks later, despite playing with pins in his thumb, to relieve injured replacement John Wolford and won a wild card game. The Rams then lost in Green Bay a week later.

Goff was rumored to be out of favor in Los Angeles.

Statistically, both quarterbacks are good. Aside from the record, which depends on team situations, I look at interception percentage and yards per dropback when first evaluating performance. Goff is at 2.2% interceptions and 6.72 yards for his career. Stafford, 2.3% and 6.49 yards. The average for the top 50 winning quarterbacks in modern NFL history is 3.1% and 6.65 yards. Both quarterbacks have earned starting roles.

With Stafford, you see the potential for improvement in a better situation. Goff has played his entire career under an ideal situation. Analysts seem to agree that Goff is not developing as he should and Stafford is still elite.

In the NFL, the money you receive after your first contract expires generally demonstrates your perceived value. Goff signed a contract worth $33.5 million per year before the 2019 season started. Stafford signed a contract worth $27 million per year before the 2017 season started. Both were elite contracts at the time.

I can’t say what changed between then and now. Goff did not have a great 2019 season and while he improved a bit this season, the key numbers are still a bit low. My quarterback metric, which correlates quarterback stats with winning in the NFL, shows that both are solidly above average and that Goff’s 2020 season was pretty good despite a large drop in yards per attempt.

At any rate, the Rams are also giving up first-round picks in 2022 and 2023 (they already traded away this year’s) and a third-round pick this year. This suggests Goff is being valued as a premium backup or stopgap starter. That could be because he objectively is playing at that level or the Rams, initiating trades, are signalling to the rest of the league that his value, internally, is near zero.

I don’t agree with this assessment. Goff has always had accuracy issues when pressured, but he protects the ball well and is an important piece of what has made the Rams a decent team the last few years. While Stafford has this unique position in the NFL of being a sub-.500 quarterback who performs very well and is perceived as elite, I don’t know that he’s all that much better than Goff. He also benefited from having one of the greatest talents in NFL history sharing his huddle – Calvin Johnson – for much of his career.

So, what happened? What’s obvious is Rams management gave up on Goff. They committed to finding a new quarterback for 2021. They got one for two years – the first at an extraordinary cost when you consider the dead cap space added from Goff’s contract. They are about $35 million over what the cap is expected to be in 2021 and they’re essentially sitting out the draft. Somehow, they’ll figure out how to fit under the cap, but it will be a challenge to do so without losing any key pieces.

Franchise quarterbacks generally spend most, if not all, of their productive careers with the team that drafted them. Sometimes, like with Joe Montana, Peyton Manning, Philip Rivers and Tom Brady (and Brett Favre with the team that traded for him not that long after he was drafted), they join a new team at the end of a long career. They rarely, if ever, switch teams once they’re established in their prime.

That may be changing. Deshaun Watson just received his “you’re my franchise guy” contract and is demanding a trade. If a better-than-average 33-year-old can garner two firsts and a third, what is the value of a 26-year-old quarterback who isn’t that far from MVP level? But the implied corollary is that contracts are less binding and any guy on your team is only as valuable as what he’s currently paid. No easy re-signings, no worries about a cap algorithm that favors long-term player retention (spreading bonuses to future years). If this trend continues, teams will start focusing more on immediate salaries and less on big bonuses. In the short term, with the cap dropping so much, there will be many teams out there who won’t be able to pay more than minimum salaries for many positions – even starting positions.

Still, Rams management, through its actions, signaled a team that believed it could win championships if only it had a better quarterback. Self-fulfilling prophecy, or reality? Whenever stakes are so high, with coaches and management fired for winning less than it should, perception is paramount. The NFL is set up to reward those who convince the rest of us that they alone can orchestrate a championship.

In the film industry, with movies constructed under budget constraints, a film-maker who creates a unique vision can be called an Auteur. A coveted moniker. But what separates true artistic genius from some Poseur who decides tripods are superfluous and coherent scripts are unnecessary? I don’t know what will happen with the Rams and Stafford. This trade feels very forced, like someone wanting to make a name, rather than a solid team-building strategy. The Rams have serious cap issues and throw away first-round picks like they believe the draft is no deeper than the NBA’s version.

On the flip side, the Lions get to start Dan Campbell’s reign with extra first-round picks and if they can absorb the 2021 cap hit from this deal (no easy task, but they’re in far better shape than the Rams), either they have a decent young franchise quarterback to resign or they can use one of those extra picks to draft one. They’re in building mode, Stafford wanted out, and the Rams just gave them a late Christmas.

A New Quarterback Rating

Over the years, I’ve focused a lot of my original statistical analysis on the quarterback position. The idea behind this is the more I understand about the most complex job in professional sports, the more I understand about sports. Most of this analysis isn’t ever part of a game. It’s more like a hobby. If I gain an insight that helps my work, that’s great.

I’ve put together a fairly extensive spreadsheet of quarterback performance going back to 1974 (insert oft-repeated explanation about NFL passing rules changes here). It was a different game before the rules changes. None of the timing routes or combination routes that have defined modern NFL offenses would have been possible under the old rules.

Still, the game constantly changes. Back in the late ’70s, the league’s interception rate was around 5.2%. Then 4.2% in the ’80s, 3.4% in the ’90s, 3.2% in the ’00s and 2.6% in the ’10s. Completion percentage has risen in a similar manner and yards per catch has dropped.

The most interesting question I’d like to answer is what makes an NFL quarterback? I’ve written a lot about this as well, but in 1998, the year I left the corporate megalith world and began Front Office Football, there was a critical question for Bill Polian, as General Manager of the Indianapolis Colts. The Colts were drafting first overall and needed a quarterback. Two quarterbacks were scouted as being worthy of a number-one pick.

Polian picked correctly, and he’s in the NFL Hall of Fame. Much has been written about Polian’s decision, and I think it would be a fantastic game to recreate that process and have people run drafts and make similar decisions. The problem there is that while we know the answer to Polian’s question, without any trace of doubt, we don’t know exactly what, in all the information he had available, truly determined the outcome.

In order to create a simulation, you have to model something. The more you incorporate into your model, the more engrossing the simulation. But that means you have to make more decisions about what you’re modeling. Some random chance is necessary – you can’t have a sports simulation without some elements of randomness. But somewhere in the data you present to GMs is a piece of information that Polian would see and would lead him to the correct decision.

And once you know, you know. The uncertainty of knowing whether a certain player fails is no longer a mystery. You see attribute X on a rookie’s card, and you know that there’s a high probability he will never develop into a good player. So I try to model that part of the decision process as little as possible. There’s no “well, he’s checking his phone and wearing headphones instead of interacting with his teammates” rating. Scouting error is higher for draftees and there’s a known variable, volatility, which, when triggered, forces a huge change in a player’s ratings. It sure doesn’t feel great (or realistic) when your top pick suddenly gets the volatility drop, but the alternative is a map of Polian’s black box which could never be unseen once revealed.

A while back, I created my own quarterback rating system. I took the major statistical categories, and determined how much they correlated with “winning” football games. This varies a bit from year to year. Since the ultimate intent of this data is to study career paths, I wanted a data set that’s normalized from year to year. This is unlike the oft-published quarterback rating, which has risen considerably as offensive strategies improve.

The system I came up with seemed pretty good, but it doesn’t offer enough credit for the quarterbacks who are more of a threat to run the ball. Now part of that effect takes care of itself. If defenses have to keep a linebacker or a nickel in a short zone as a “spy” in order to protect against a long quarterback run, presumably receivers will have an easier time getting open. But that doesn’t take into account the rushing yardage the quarterback gains, which may be more valuable because quarterback-rushers have a much higher yardage per carry than running backs. Just the nature of these plays.

I’ve spent a lot of trial-and-error time determining what stats correlate best to winning. I found that quarterback rushing yardage was enough of a positive to warrant breaking out on its own. I also hadn’t built sack numbers into the system, and getting sacked is also pretty bad for winning percentage.

I needed to improve my quarterback database, so I decided to take sack and quarterback running numbers back to 1998. I also tried to remove plays where the quarterback takes a knee at the end of a half, since that muddies the quarterback rushing picture. I also removed quarterback spikes from passing attempts.

From there, I put together a list of quarterbacks who have made about 50 or more starts since 1998, adding in top draft picks since 1998 and quarterbacks of significance who were still active in 1998, but had fewer games played later. I ended up with a list of 109 quarterbacks and I included their full careers, modifying the formula to remove fumbles from the rating before 1994 (it’s hard to find consistent fumble data before then).

Each performance is then normalized based on the correlations for that season. I use rolling seven-year averages to smooth out the correlations, because there are wide year-to-year swings. These allow the rating to change gradually, on the hypothesis that each season is one trial – one that could be skewed to one side of the expected result distribution – on an unknowable league mean.

As it turns out, Peyton Manning is ranked #1 of 109 in average quarterback rating over his career and Ryan Leaf is ranked #109. Polian not only made a great decision; it might objectively be the best decision ever made in sports drafting. Imagine – pick correctly and you get the best quarterback of his generation. Pick wrong, and you not only waste the pick, but you’re starting the worst quarterback of his generation for a couple of years.

Another reason I wanted to make sure I was more effectively crediting quarterbacks who run the ball a lot was that I wanted to separate running quarterbacks from those who rarely run the ball and see what I could learn. So I divided the 109 quarterbacks into quintiles based on the percentage of plays they either run the ball or get sacked, or throw the ball.

I found that the rating was almost flat through the top three quintiles – those who ran/were sacked the most. In fact, the top quintile saw a bit of an improvement over the next two. But the fourth quintile was considerably higher than the first and the fifth quintile (those who run/were sacked the least) a huge improvement over those numbers. This was reflected in winning percentages as well.

Keeping in mind that there are only about 22 quarterbacks in each quintile, and Manning and Tom Brady are in that fifth quintile, that might not mean all that much.

I noticed two other important pieces of data. First, quarterbacks in the first quintile reached their performance peaks after less than two years of starting. That’s about a year ahead of every other group. And second, quarterbacks in the first quintile had the shortest careers on average – about seven years. That rises to eight for the second and third quintiles, eleven for the fourth and twelve years for the fifth.

Starting is about opportunity and opportunity comes from winning. With small samples, just a couple of players can make a difference. It doesn’t mean that running quarterbacks can’t win. The first quintile includes Hall of Famer Steve Young, and Russell Wilson seems to be having a very similar career (he’ll be in Canton, I have no doubt). Of today’s young stars, Deshaun Watson and Lamar Jackson are in the first quintile – Watson and Patrick Mahomes (third quintile) have average ratings already at a Hall of Fame level.

Interestingly enough, college performances from the first quintile were by far the best in terms of college quarterback rating. College defenses simply can’t handle a fast quarterback with a good arm who is always a threat to run the ball. The rest of the group had similar college performances, but the fifth quintile included quarterbacks who had a lot more experience in college throwing the ball. This group threw about 25% more passes than the other four quintiles, on average, even though Brady himself was rather inexperienced coming out of college.

My takeaway? It’s hard to find a good quarterback in the NFL, so you have to take the best one when you need one. A good runner is a bonus at the position, and they tend to develop faster but they have shorter careers. As quarterbacks develop, if they need to stay in the pocket, they need to learn to get the ball out quickly. You’ll find all five quintiles represented among the twelve of the 109 quarterbacks studied who are either in the Hall of Fame or definitely Canton-bound.

However, five of the twelve are in the fifth quintile, as are the next three most likely to receive serious consideration. Ideally, if you want your number-one draft pick to develop into a 16-year franchise guy, you want someone who makes quick decisions and not end plays with the ball in his hands.

From the Home of Small Sample Sizes

Continuing with my theme of wondering about the effects of playing in empty or reduced-crowd stadiums, I took a look at late-game comebacks in the NFL.

The base, as I’m using for these studies, is games played from 2002-2019. That’s 18 seasons of data with the current eight-division format. There were 4806 games played, including 57 at neutral sites. Through week 9 of 2020, there were 133 games played, which is a fairly low sample size.

For the purposes of providing a consistent metric for study, I consider a late-game comeback to be a score that puts the winning team in the lead in the last eight minutes of play, where the quarterback, on offense, has a drive that includes at least one first down. This includes most late comebacks, but wouldn’t include, for example, a pick-six that ends overtime.

From 2002-2019, there were 1146 comebacks at non-neutral sites. This encompassed 24.1% of all non-neutral-site games. Of those, 596, or 52.0%, came at home.

If you have perceived 2020 as providing more exciting finishes in the NFL, you’re not wrong. So far in 2020, there have been 35 of these winning drives in 133 games (26.3%). And 18 have come at home (51.4%).

The 26.3% is interesting. But it’s probably insignificant (if the 2002-2019 average held, we’d expect 32 of these drives), or it could be that quarterbacks are more efficient when it’s relatively quiet, or it could be that scoring is up a lot in general. There isn’t enough data to conclude anything at this point, but I thought it was interesting enough to mention.

Again, to non-update FOF9 News, I am still waiting for information related to the project that will determine what I do next. When I have that information, I will let everyone know what will happen. I thought I’d have an announcement back in August, but some things are a lot harder to finish in our new world and I think that’s what causing this delay. Can’t say for certain, only that I know how I’ll proceed either way – it just makes no sense to announce anything under these circumstances. Normally, I’d just keep quiet about everything until I had something to say, but I don’t want people to think I’m holding back on information given that FOF9 was originally announced more than two years ago.

Home Field and Scoring in the NFL, an Update

After four weeks, I noticed (well, everyone noticed) that scoring was increased in the NFL. In addition, the home field advantage didn’t appear to be what it used to be.

We’re now 105 games into the 256-game season. I wanted to apply some basic statistical concepts to the observations.

After week seven, home teams are 53-51-1 (50.9%). The historic winning percentage for home teams is 57.4%. If we assume that home teams win 57.4% of their games, what we’ve observed would give us about a 91% probability that we’re looking at a different home win expectation. That’s a complicated concept. What it means is that given 105 games at a 50.9% winning percentage, there’s a 91% likelihood that the 105-game sample came from a football world where we can no longer assume home teams win 57.4% of the time.

That’s not a very high threshold, and I’d say it’s too early to make that conclusion even though it looks like playing in empty stadiums is removing the home-field advantage.

What about scoring, then? After seven weeks, teams are averaging 25.4 points per game. The league record, set in 2013, is 23.5. The assumption that the 105 games played so far could come from the same football world as 2013 has about a 0.2% likelihood. That’s well above the threshold statisticians would use to test that kind of question. We are very likely in a new world of offense here. Whether that’s empty stadiums, or new offensive innovations or defensive rust is impossible to determine right now.

Meanwhile, I had promised people another Front Office Football update by the end of October. Again, I apologize, but I am still waiting to hear information that would allow me to make the decision whether to proceed with FOF9. Development remains on hold and I continue to learn new techniques. We’ll try again next month. As we all know by now, I am really, really bad at adjusting to the obvious. I beat myself up all the time about it, but unfortunately, it’s just who I am.

An Unusual Signing

Yesterday marked one of the more unusual signings of the NFL year. It isn’t one that’s likely to have any impact on any team’s future performance.

Dallas signed journeyman Garrett Gilbert to a one-year deal worth about $750,000. Presumably, injured starter Dak Prescott, who is definitely out until around the start of next year’s training camp, will go on Injured Reserve.

Gilbert won’t start – that job goes to veteran Andy Dalton, who is third only to Joe Flacco and Alex Smith in career wins by an active quarterback who wasn’t starting prior to last week. In fact, Gilbert has never started an NFL game and has attempted only six passes in his pro career. He’ll likely move ahead of project Ben DiNucci on the QB depth chart. DiNucci was drafted in the seventh round this year, out of James Madison, where he lit up scoreboards but didn’t have to face FBS defenses.

Gilbert was as heralded a high-school signing as anyone, and chose Texas. He didn’t perform well for the Longhorns, and transferred to Southern Methodist, where he had one bad season and one decent season. He has all the obvious tools, and that merited a sixth-round pick from the Rams in 2014. But he was cut before the season even started. He had stints with four other teams, spending almost no time on active rosters, until last year. While it’s impossible to call a sixth-round pick in the NFL a bust, it’s rare that one remains in the league six years without ever seeing meaningful action.

Gilbert joined the Alliance of American Football in early 2019 and proceeded to dominate the league. Since there aren’t many opportunities for players to compete in live games against decent competition once college is over, this was notable and Gilbert earned another look in the NFL. The Browns signed him when the league went under, and they kept him as Baker Mayfield’s backup last season.

Still, Gilbert didn’t impress enough to prevent the Browns from bringing on journeyman Case Keenum in March, and putting Gilbert on the newly expanded practice squad after training camp. That’s where he was until yesterday, when the Cowboys signed him to their active roster.

OK, then. Why is this signing unusual? The NFL rules surrounding practice squads are a bit confusing. Players receive a small portion of the minimum salary and can practice with the team. Any other team is free to sign them at any time, as long as they are signed to an active roster. Each week, teams can protect up to four of their practice squad players. Those players, too, can be signed away, but there’s only a limited window for these transactions – essentially until early Tuesday.

Gilbert had been protected in this fashion by the Browns, but the Cowboys made the signing during the window. Why? This has to greatly annoy the Browns because he’s a quarterback who has spent 18 months in Cleveland and has learned the system as Mayfield has learned it. That has some value. But he’s in that magic bubble between succeeding in the NFL and being so hopeless in practice that teams won’t invest the time. That probably describes about 40-50 quarterbacks right now. Guys who are in shape, healthy, can learn a system, but you don’t want them on the field.

I don’t know of any feuds between the Cowboys and Browns. They played each other a couple of weeks ago and Cleveland embarrassed the Cowboys on their home field. The following day, the Cowboys signed lineman Greg Senat off of the Browns’ practice squad. Generally, feuds don’t begin because one team is historically bad on defense in a game, but one has to wonder if something happened to warrant this reaction. In what’s looking like a lost season in Dallas, stirring stuff up has only the downside of attracting some negative attention. But, for Cleveland, enjoying a solid start that could lead to their first playoff victory since their opponent was New England and Bill Belichick was the head coach on their sidelines, this is an unwelcome distraction.

There is clearly a need for more quarterbacks in the NFL. Anyone who sees any level of success on the field will continue to see new opportunities. Just being able to learn an NFL offense and run it on the field without being overwhelmed by the speed of the game is one of the most challenging tasks in professional sports. Teams like to carry three quarterbacks, and most, unless they’re protecting a young prospect who isn’t quite experienced enough for second string, stash the third one on the practice squad. Dallas could have chosen from the 20-or-so quarterbacks recently released by other teams. They didn’t, and Cleveland will have to find a new third quarterback.

There’s no rule against poaching practice squad players or targeting a particular opponent. And there’s no proof this was anything other than the Cowboys wanting Gilbert because their scouts saw something in those AAF performances. But it’s unusual, and especially with Cooper Rush, who spent 2017-19 with the Cowboys in this capacity, available after being released by the Giants, it’s a situation worth watching.

On Cardboard and Pinball

It’s a different NFL in 2020.

I could fill a good-sized novel with all the obvious changes in the world. When we look back at 2020 decades from now… well, no one will forget this year.

As we look at the empty stands, perhaps decorated with cardboard cutouts paid for by fans (I have to hand it to NFL marketing on that one – a lot of ticket holders will spend an additional $100 to have their pictures placed in the stands and I would have guessed the over/under on this one at somewhere around 10 people), subconsciously policing coaches for mask violations, wondering if Pete Carroll can still engulf 37 sticks of bubble gum without accidentally devouring his mask… what a world we live in. There’s an election coming up, did you know?

Masks are funny; I don’t quite know why because nothing else is funny about COVID-19. That poor sideline reporter at the start of Sunday’s night game… just not on quite right, slipping, slipping, slipping at she talked away… there’s the nose! I have no idea what she was talking about. Being hearing impaired, I might as well just stay home 24/7 until we’re through this because it’s really hard to pick up on when someone is talking to you. I don’t read lips, but I depend on these cues, I guess, to tune into a stranger’s voice.

I hope you’re all staying safe and healthy.

Aside from COVID-19, the NFL feels like a different game. I think it’s full-speed, but it seems more controlled. Injuries seem way down, except in New Jersey with its killer turf and Inglewood, California with the team doctor doing his best Norman Bates impression. College-like scores and comebacks. Mitch Trubisky and Bill O’Brien out, the Browns turning those 30-year frowns upside down. Josh Allen, huh? Didn’t expect that one.

I thought I’d answer a couple of questions I had about the game this year. Nothing definitive, because we’re only four weeks into this brave new world. The obvious… how are home teams doing without fans in the stands? How much of the home field advantage is the crowd, and how much just being familiar with the stadium and not having to take that plane ride and stay in a hotel?

The baseline: all regular-season and playoff games from 2002 (the start of the 8-division format) through 2019 played at a non-neutral site. That’s 4749 games. Home teams are 2723-2017-9 (57.4% win percentage) over that span, outscoring opponents 23.3 to 20.9.

We are now just 63 games into the 2020 season, so it’s impossible to make conclusions. That 57.4% would give the home team a 36-27 edge right now. Instead, home teams are 31-31-1. They are outscoring visitors, 26.1 to 25.2. Just throwing that out there – taking a sample of 63 games is not going to give you enough, statistically, to say that 31 wins versus an expected 36 means all that much.

What about the scoring in general? Here we are at 25.7 points per team per game. The league record is 23.4, set in 2013. Last year, teams scored 22.9 per game. So that’s an increase of 2.8 points per team per game over last year – almost a full touchdown between two opponents. Four weeks obviously does not make a season. Scoring isn’t all that weather-related. In fact, it tends to go slightly higher later in a season. Week 2 is the lowest-scoring week, on average, while most of the highest-scoring weeks bunch at the end of the season.

Let’s say the 25.7 holds up. That 2.8 points per team per game would be the largest season-to-season change since 1947 (2.9). Generally, changes are less than 1 point, if that. There was more volatility in the 1940s (free substitution rules changed quite a bit) and the 1970s (contact with receivers was gradually made illegal). Yet there are no major rules changes going on right now – it’s just that defenses seem far less able to stop the pass. There are still 13 teams averaging more than 7.5 yards per pass play. Last year, 13 teams averaged more than 7.0 yards per pass play. It is a bit pinballesque, like the college game these days.

Cause for concern? With everything else going on these days, there’s comfort in the familiar. Most sports enjoy relative consistency from year to year. Sudden changes in scoring or the pace or the rules leave us less able to put what we see in proper perspective.

Again, it’s far too early to draw conclusions, but so far I’m putting this in the long list of things I’m not really liking about 2020.

The Proverbial Non-Update Update

I promised to update everyone about the future of Front Office Football by the end of September. As we’re closing in on this date, I’m providing that update, but I have not made any decisions.

I was fairly certain I would have those decisions made by now, but they’re somewhat dependent on events which are out of my control.

I have been evaluating what I have in what we were calling Front Office Football 9. It needs a lot of custom artwork and it probably needs to be converted from Qt (which we were using as a framework base) back to a more familiar framework. Qt can do a lot of things, but every little change is time-consuming as all get-out and there are some serious incompatibilities with how it serves tables and what I need it to do. I would also have to redesign the framework to back out of using FaceGen since we never did the pieces that require our own custom artwork.

In the meantime, I’ve been working on my drawing skills and also how to better use custom artwork. The results look nice, but I’m stretching what that familiar framework can do. Plus, I need to redo the menu system and convert the framework to do paging the way pretty much every other game in the universe does it (spreading windows all over dual monitors is a lot of fun, but it destroys your review scores). So I’m not certain I could have FOF9 done by next fall.

I hope to have a real update by the end of October, but, then again, I thought I was being fairly conservative by saying end of September.

College of Quarterback Origin

From time to time, I like to take a look at data I’ve collected about quarterbacking in the NFL. Examining these statistics is like a history lesson. You can view league totals and averages and see exactly how play-calling has evolved over the years.

I’ve put a lot of that work into a spreadsheet that has bits and pieces of demographic information. I haven’t proofed it against the major databases, but it is a fairly complete list of every game played in the NFL (regular season and playoffs) since 1974, when the rules were changed to open up the passing game.

This post is just a piece of that data – an attempt to show the last college attended by quarterbacks who have won games as starters in the NFL. This includes 144 colleges and more than 11,000 wins. No analysis here – it’s just a list, heavily skewed by the Hall-of-Fame level quarterbacks who add the most to the totals. I thought it would be interesting as discussion continues as to whether there will be a 2020 college football season.

QB Wins, by College
** – Quarterback had a significant portion of his career prior to 1974, which is not included in the total

I know people will ask, but I have no news regarding the future of Front Office Football. I believe I am close to a decision, and will post this information soon. If I don’t have anything to update by the end of September, I will post again at that point regardless.

Changes Ahead

“The only certainty is that nothing is certain,” – translated from Pliny the Elder.

Roman author Gaius Plinius Secundus was a noted philosopher of his time. One of the reasons so many of us endured Latin in high school was so we would have the pleasure of reading various motes of philosophy in their original form, then struggle to make sense of them since it has been a long time since anyone has actually said anything in Latin.

This quote of his, and I lost whatever meager ability I ever had to translate Latin about ten minutes after tenth grade ended, has endured, often accompanied by Benjamin Franklin’s exceptions of death and taxes, which he was far from the first to notice.

It endures because the older you get, the more you understand that even when you make decisions based on solid reasoning and expectations, life often throws you a proverbial curve-ball. All you can do is embrace the uncertainty of being and do your best to find a new path, not get into ruts, not dwell over things you can’t control.

I find it interesting that Pliny the Elder died at the same age I am now. Would he have found certainty if the Romans had better health care available?

“Carpe Diem,” – Horace.

Roman poet Horace is best known for this simple, two-word phrase that doesn’t require high-school Latin to translate. Seize the Day (though carpe is not exactly Latin for seizing). When opportunity arrives, recognize it and go for it, because you don’t have that many genuine opportunities in life.

Twenty-three years ago, I was sitting in a restaurant on Bourbon Street in New Orleans with a few of my co-workers, enjoying the French Quarter’s signature drink – the Hurricane. It’s often served in a Mason jar stamped with the name of the restaurant, which you can take home as a souvenir to prove you have experienced a significant milestone in the life of a well-traveled tourist.

Faded lettering, still inspirational?

I enjoyed spending time with my co-workers. It was a good group of people. Our office was in Bellevue, Washington – I believe Steam, today, is headquartered in the same building, though I could be off by one high-rise. The 1990s was a great time to be a computer programmer, and the “Eastside” of King County was programmer central. But the work was uninteresting and this was my third job in the field after graduating college.

I remember musing about how nice it would be to try something new and original. Many of us were attaching ourselves to new opportunities. That summer, about half of our office left to join a tiny start-up in a Seattle warehouse. Some guy named Jeff was trying to build an on-line bookstore and was hiring almost anyone who could code. Many of those start-ups failed, but some of them didn’t. That one didn’t.

The previous day, we had gone through our routines demonstrating our company’s products in the New Orleans convention center. Dressed in identical logo-stamped tee-shirts, showing off technology that did nothing more than allow our customers to connect their computers and handle their businesses seamlessly. From time to time, the company CEO, impeccably clad in a designer suit and sunglasses, would travel the center’s halls in a golf cart with a personal driver, flanked by security guards on foot. I have no idea how they handled escalators or elevators.

Was he the Rock Star we all wanted to be? He was rumored to be in the process of purchasing an NHL franchise, of all things. I didn’t particularly like the image, but I appreciated the display of primacy. That night, the highlight of the experience – he brought in Penn and Teller to perform their routine for the thousands of employees who had flown in for the conference.

As we sipped our Hurricanes and waited for our Blackened Redfish, we complained about the routine. Someone had heard a story about the CEO noticing an employee sitting by the pool at their hotel when he should have been on the convention floor. Allegedly, fired on the spot. Stories like these are often twisted or invented to keep the rank-and-file scared and focused. It’s not an approach that works well with programmers, so it was having the opposite effect on our group. I remember talking about sports gaming and my side-work at the time, which was reviewing games for Computer Gaming World. It may have been the first time I spoke aloud about wanting to leave the cubicle world. At least not in Latin.

The Redfish was spicy and a bit uncomfortable, but new and delicious.

A few months later, after discussing my ideas with editors at CGW, I decided to take the plunge and got a business license for Solecismic Software. I left that job on February 20, 1998. Freedom Day. I carpe-d. It worked out well.

Obviously, the gaming industry has changed a lot since 1998. I have been saying for years that the days when a single person could do everything for one product are over. There were several of us who got into sports gaming in the ’90s. We all had different approaches, and some of us were able to sell enough games to do this for a living. I’ve been lucky to be one of that group. Markus Heinsohn and Andreas Raht have done the same with OOTP Developments.

One thing that might surprise you is that we don’t really view each other as competitors. We’re all people who have had similar dreams, interests and abilities. The success of one of us only helps the genre. It’s not an either-or. Our products are, and remain, different enough that those of you out there who enjoy the simulating side of sports gaming have many options.

Markus and Andreas have focused their efforts more on growth. I have tried to avoid growth. But their vision fits more into today’s reality than mine. We’ve all remained friendly over the years. In early 2017, I contacted Markus and asked if he had any interest in joining forces. My initial thought was to write the sequel to TCY. But the timing was right for OOTP to start over with pro football, and a new FOF would be the better choice for the marketplace. A few months later, we formed a partnership to produce FOF9. I would re-write my game with expansion in mind and to fit an entirely new GUI system. They would provide graphics, a framework and marketing.

Any good business arrangement requires timing, opportunity and hard work. In 2017, we had all of that and were making great progress. We felt it likely we would have a great new product out in the fall of 2018. We felt good enough about it that we announced the partnership – not coincidentally on the 20th anniversary of Freedom Day.

Fast forward to today, as we announce that we’re ending the partnership. What happened?

The opportunity remains. The game we were creating would be a good fit for the current marketplace. Certainly, we’ve put a lot of hard work into it. What we lost was timing.

I can’t speak for Markus or Andreas. What I can say is that after OOTP 19 came out, with the promise of the Perfect Team 1.0 beta a few months later, OOTP had to carpe diem. We tried to continue with the football, but once momentum is lost with a project, it’s hard to find a new rhythm. We still hoped that 2019 was a possibility, but we didn’t do a good job redefining responsibilities. That was straightened out, and we had a good road map for 2020. We only needed more resources. And when it comes down to it, this is OOTP’s opportunity, with their primary product, to become a much bigger player in the sports gaming business. What made perfect sense in 2017 no longer fits their company.

I understand and respect their decision. It’s tough. I can’t lie about that. I have most of FOF8 and a good number of exciting new features completed and a new framework in place. Almost all of the UI is completed. But we don’t have anything in the way of graphics, so it’s not something I can show off or finish in time for a 2020 release. I’m still using place-holders, mostly from FHM as the UI is the same base as theirs, not OOTP’s.

Markus and Andreas remain friends. We are not ending this out of anger or because we’re not making progress. It simply doesn’t make sense for OOTP to have a new football product right now and they have to seize the opportunity they have now, without distraction. They are letting me use what we’ve created without restriction, and that is because all of us in this business root for and try and support each other. I’m genuine in my hopes that OOTP and Perfect Team and the upcoming mobile product help their company reach new heights. I’m sure they will feel the same way if I return to FOF development and am able to get that project together.

Will that happen? I don’t know. I am taking the time right now to evaluate Solecismic Software. Does growth make sense? Is it feasible to continue work within this new framework or would I need to start over? Do I leave the business entirely and get a “real” job. Everything is on the table right now and I honestly don’t know what the future holds.

Thanks for reading. I will post again once I have a better idea of what form Solecismic Software will take.

Drafting Quarterbacks

Inspired by a brief discussion and some trash-talking among friends over the Jordan Love selection, I’ve been taking a long look at quarterbacks in the NFL draft. How are these selections made? When do you take a quarterback and what should you expect?

I haven’t learned anything earth-shattering. I thought I’d share my process and use this to potentially define future discussions, should a new way of looking at the data come to light.

The scope of this study is quarterbacks who were drafted or came into the league and played from 1998 onward. Dividing the NFL into time periods is somewhat arbitrary. I generally don’t study data before 1978. That was the first year of the 16-game schedule. It was also the year a major adjustment was made to the 1974 rule change defining pass interference/defensive holding. The game was completely different before 1974.

The NFL gradually evolved in adjustment. I often choose 1998 out of convenience as the beginning of the modern NFL. Peyton Manning was drafted, signalling the impact of having a true franchise quarterback. Statistical analysis was gaining sophistication, so the importance of keeping turnovers down and the chains moving was no longer a secret. It was also the year I released Front Office Football, so my own collection of data is more robust.

The game has continued to evolve, and game-planning is giving way to play-planning with the prevalence of quarterback/receiver reads and the impact of the run-read/pass option. In the future, I’m guessing I’ll use 2021, if it’s the start of the 17-game schedule, as the definitive mark of a new era. Then again, predicting the future is often a losing proposition.

I have 322 quarterbacks in this list. This constitutes the entire set of quarterbacks who entered the player pool from 1998-2019 and were either drafted or have attempted a pass in a game. For statistics, I’m primarily using wins in games they’ve started, with a look at losses for those who have a higher number of starts. It’s hardly a perfect measure, but game planning and team philosophies lead to far more variation when using numbers like passing yards or yards per attempt. Those are more useful when evaluating impact across tighter periods or in specific comparisons between players. But for something this broad, they would only be a distraction.

Since this is a draft study, I’ve tried to collect information about college experiences. I found total college passing attempts and passer rating for 305 of the 322 quarterbacks. This includes anyone drafted in the fourth round or higher and anyone from what we call the FBS today. Only two quarterbacks who started games (each with four) were not included – Quinn Gray, undrafted from Florida A&M in 2002 and Keith Null, a sixth-rounder from West Texas A&M in 2009.

One piece of data I would have liked to include is number of wins in college. But that would have taken too much time to calculate, and with less inclusion. So I broke colleges down into four categories – power-five (194 of 305), mid-major (88), non-FBS (23) and a subset of arbitrarily the 16 most successful programs within the power-five (77 of the 194). Is the expectation of winning a good substitute for winning?

The next step was to define expectations for quarterbacks drafted at a certain level. Of the first-round picks (63), all of them have started and won at least one game and 16 of the 63 have won more than 50 games. Second round: 16 of the 22 have won a game (two more than 50). Third round: 20 of the 32 have won a game (one more than 50). Fourth round: 14 of 30, 0. Fifth round: 13 of 35, 0. Sixth round: 14 of 48, 2. Seventh round: 8 of 42, 1. And undrafted, inclusion being dependent on having thrown a pass rather than being drafted, 22 of 50, 1.

A legitimate question you can ask at this point… do quarterbacks start because higher draft picks were invested, or because they are better quarterbacks? I’m not certain how to answer this question. Exceptions exist – there have been three true franchise quarterbacks in the last 22 years who were not taken before the fourth round. Those include Tom Brady (sixth round, 2000), Matt Hasselbeck (sixth round, 1998) and Tony Romo (undrafted, 2003). The seventh-rounder with 50-plus wins is Ryan Fitzpatrick (seventh round, 2005).

Calling Fitzpatrick a franchise quarterback, since he has a 55-83 career record, is difficult. But he’s also the only Ivy Leaguer in the study and had the best Wonderlic test score (48). Which brings up other interesting questions. I found Wonderlic scores for 191 of the 305 quarterbacks. There is no correlation between score and either college pass attempts or when a player was drafted. Coaches at both levels don’t seem interested in Wonderlic-type mental abilities. But there was a 6% correlation between Wonderlic score and college passer rating and a 13% correlation between Wonderlic score and NFL wins. This is probably the most interesting piece of data I found. But beware of applying trends to individuals: Blaine Gabbert (13-35 career record as a starter) scored a 42 on the Wonderlic.

For passer rating, I found that draft position held a strong correlation, and that increased with each tier of quarterback. There was a 38% correlation between rating and draft position for the 77 elite-college quarterbacks. Did that translate to performance? At the mid-major level, yes. But at the major level and the elite level, correlation was only 5% between wins and passer rating. This suggests that coaches are perhaps over-estimating the value of good statistical performance in college when looking at those chosen to lead the top programs.

Some analysts suggest that the most important college statistic is simply experience. Pass attempts do not correlate at all with draft position for mid-major and lower-level quarterbacks, but at 16% for both elite and power-five quarterbacks. Did that translate to more wins? Unfortunately, the correlations with wins are a little bit lower, hardly significant at all.

This is not to say that experience or statistical excellence are irrelevant. The pool of players studied does not include undrafted quarterbacks who never played in the NFL. With about 20-30 quarterbacks eligible from power-five conferences every year, most don’t even get an invite to the Combine. A more extensive study could add those quarterbacks to the pool and draw out more information.

Without that extension, basic scouting has to be trusted. The lists we see from the draft experts invariably include any quarterback who is going to be selected in the top four rounds. How much can you trust scouting? How much room is there to give someone a chance who might otherwise be overlooked. The late Joel Buchsbaum was Pro Football Weekly’s draft expert. He had connections throughout the league and watched a ton of film himself. Every year, he put out a book that I’m sure was even used in some war rooms, at least for reference. This is what he had to say about the number-six rated quarterback in one draft (who ended up being the seventh selected, about where he was graded as a mid sixth-round pick).

“Summary: Is not what you’re looking for in terms of physical stature, strength, arm strength and mobility, but has the intangibles and production and showed great Griese-like improvement as a senior. Could make it in the right system but will not be for everyone.” Griese refers to the Hall-of-Famer’s son, Brian, who had a nice career for a third-rounder, starting just a couple of years earlier.

One could say that this player’s coach had some inkling that he would be more than the system-limited career backup the scouts projected, but any other coach potentially having that inkling would warrant a much higher selection. As we all know, 198 players were selected before Tom Brady, who turns 43 in August and will get his first look at a completely new system in 20 years as he suits up for “Tompa Bay.”

Brady was a draft boom, probably the largest one in NFL history. I also selected booms and busts from the pool. A boom being a player who produced much more than what teams would reasonably expect from his place in the draft and a bust being someone who didn’t produce nearly as much. For quarterbacks, since even expecting a starter in the second round is ambitious, there are fewer busts.

Conversely, you expect an eventual franchise quarterback at the top of the draft. However, over the draft period, those 63 first-round picks break down in a very interesting manner. Picks 15-32 are a lot like second-rounders. Only two have gone on to win 50+ games. A franchise quarterback is so valuable that teams either trade up to the top of the draft or they have suffered so much without a good quarterback that they end up drafting there anyway. Of the 68, 16 were #1 picks, 6 #2 picks and 6 #3 picks. Eight of the 16 top picks have won 50+ games, along with one second pick and one third pick. There are a few more quarterbacks who are clearly on their way to 50 wins and quite a few that are not. Picks 4-12 are less certain, with 4 of the 17 in that group with 50+ wins.

As it turns out, picks 4-6, with more attention on this range given this year’s draft, are more part of the top of the draft. Only one player has been selected for each pick from 1998-2019. Philip Rivers fourth in 2004, Mark Sanchez fifth in 2009 and Daniel Jones sixth last year. Rivers has been a great success, Sanchez not so great, but with a 40-38 record as a starter I wouldn’t call him a bust and Jones is still developing. I’m sure Miami and the Chargers are hoping for franchise talents in Tua Tagovailoa (5th this year) and Justin Herbert (6th). But is it reasonable to expect three franchise-type quarterbacks in one year when only 31 quarterbacks were selected in the top six picks in the previous 22 drafts?

The odds are much worse for Green Bay, which took Jordan Love with the 26th pick. If he ends up being a great quarterback, that’s 25 opportunities to grab him missed. The Packers may be the one franchise that believes most in that story. Aaron Rodgers, well on his way to Canton, was selected 24th in 2005. The other major success from picks 15-32 is Joe Flacco (18th in 2008). Lamar Jackson (32nd in 2018) seems on his way as well. You can also include Drew Brees (32nd in 2001, but in the second round since there were only 31 teams then) in this group, though the Chargers let him go because of an early arm injury.

What about busts, then? If you’re expecting a second- or third-round pick to become a dependable starter and he doesn’t, you’re not being realistic. But you should have someone who can hold a roster spot and play a bit in case of injury. I found 13 of the 54 quarterbacks drafted in rounds 2-3 didn’t reach that level. For the 15-32 picks in round 1, you’re hoping for a starter, but should at least have a solid backup who can hold up a bit longer. Six of the 18 quarterbacks drafted in those positions didn’t fit that category, though by varying degrees. Only Johnny Manziel (22nd pick in 2014) completely failed. The others had more experience as backups. Only Tim Tebow (25th in 2010) was out of the league quickly and the jury is still out as to whether Paxton Lynch (26th in 2016) will be able to get his career going.

With picks 7-12, five of the 14 picks warrant bust consideration. Jake Locker (8th in 2011), Matt Leinart (10th in 2006) and Cade McNown (12th in 1999) were never able to stick as starters. Christian Ponder (12th in 2011) and Gabbert (10th in 2011) saw more starting time, but settled into backup roles. In all, that’s about a 1-in-3 bust rate for first-round picks after number 6. These hurt, because you should be able to find a solid starter at any other position with a pick that high.

The 1-in-3 bust ratio holds true for the top picks, as well. Though these should be franchise quarterbacks because you’re missing out on a true impact player otherwise. Some, like Sam Bradford (1st in 2010), David Carr (1st in 2002), Tim Couch (1st in 1999), Robert Griffin (2nd in 2012), Blake Bortles (3rd in 2014) and Joey Harrington (3rd in 2002), saw a lot of the field, but simply failed to win much. Griffin and Bradford even had a good season or two, and calling them busts may be too harsh.

But the other three on the list were much bigger failures, and are usually mentioned among the biggest draft failures in league history. Ryan Leaf (2nd in 1998, 4-17 career record), JaMarcus Russell (1st in 2007, 7-18 career record) and Akili Smith (3rd in 1999, 3-14 career record) were all out of the league quickly, and those misses had huge impacts on their teams.

Taken as a whole, are there any common links connecting the bust list? They seem very close to the average in terms of college pass attempts and quarterback rating. I found a little over-representation from the elite programs (10 of 33) while there was under-representation from elite programs in the surprise boom list (6 of 44). At these sample sizes, hardly conclusive. Wonderlic scores? Interestingly, for those with scores reported, the overall average for quarterbacks is 27 and both the boom list and the bust list have averages of 28.

This is all meant to be more the start of a discussion than deep analysis. I didn’t come away from this study feeling that I had discovered anything important. Perhaps coaches are a bit too impressed with the quarterbacks of very successful college teams, but, then again, the most impressive boom pick in the history of the league is probably a quarterback with a mediocre scouting report from an elite program. Maybe there’s too much emphasis on impressive college stats at top programs, but that’s not a huge factor. And maybe you should take a flier on an otherwise unimpressive guy who has an great Wonderlic score.

I come away from this brief study, most of all, thinking that taking a quarterback between pick 7 and the end of the second round is almost never a good idea. But understandably, the Ravens and the Packers feel very differently.

The bottom line is this is all about scouting, and whatever Indianapolis did in 1998 when choosing Peyton Manning over Ryan Leaf is what this is all about. You’re not going to make that decision based on numbers alone and you’re not going make a smart choice without spending a lot of time looking at the decision from many angles. The Colts did, and it changed their future.