Wednesday, September 12, 2012

Dogs v. Cats 2: This time its classical political philosophy

Dogs are born free and everywhere they are on leashes. You would never see cats tolerate such a situation. The feline intuitively grasps Rousseau's social contract, then has the sheer gumption to reject it. Meanwhile, the canine has managed to fall out of the state of nature into something even worse: the state of captivity. The domesticated canine has bought pee breaks and 2 meals of dry food per day at the impossibly high cost of its initiative.


Sunday, September 9, 2012

Treachery! Or, the case for 2012 Buffalo Bills Pessimism

I'm a big Buffalo Bills fan but the team's extended string of incompetence has strained my affections to the point where it is now possible for me to bet against them without feeling like I am betraying a cherished friend. This is they year that I take advantage of that opportunity.

There is a significant amount of hype around the team. At the end of last season there was little reason to be excited about the future when the team had just ended the season on a 1-8 run. But then the Bills "won" the off-season: they signed the biggest name on the free-agent market (Mario Williams) and a complementary piece (Mark Anderson) then added some promising pieces in the first two rounds of the draft. When brand name teams like the Dallas Cowboys or the Washington Redskins sign the headline free-agent everyone assumes that they spent too much money in a myopic but misguided attempt to "win now" but when Buffalo spends more on a defensive player than anyone ever has it is a brilliant move that makes pundits take the team seriously. In my opinion, all the Mario Williams excitement has combined with a soft schedule to propel the Bills' over/under win total number too high with good enough odds on the under to make betting against their season have an expected return above zero.

Excited Bills fans before losing to Dallas in the Super Bowl for the first time, 52-17

I grabbed the under on the Bills win total this season on August 22 after it went up to 8 with 6/5 odds meaning that for every dollar I bet on them I get the dollar plus another $1.20 back if they finish with a record worse than the league average. If they finish with a .500 record then the bet is a wash and I get my initial bet back. They need to get to nine wins for me to lose money. (Ties favor me because they count as another game without a win but they are so rare I am just going to ignore their possibility so if they happen it is a surprise bonus).

The odds actually improved further on the Bills under bet to 5/4 (or +125) after I made it but that was brief and the odds have since come back down to 6/5 (+120) and settled there going into the season. Oddly enough, the odds spiked to 5/4 the day after the Bills got slaughtered by the Bills 38-7 in the offseason

For reference, the team's over/under opened at 7 wins back in May. The over was -110 and the under was -120 so the bookies were actually trying to encourage bets on the over relative back then. If you follow that link, you can see that there were no strong feelings at that time that an over/under of seven was too low for the Bills even though Mario Williams was already with the team.
 
The fact that the over/under is 8 makes the analysis of the bet easy because it is essentially a question of whether I think they are going to be better or worse than the average team they play. 50% is the balancing mark. And since I am getting 6/5 odds, even with the assumption that the Bills are as likely to be above .500 as below it this season a bet made on the under still has a positive EV.

Let's go through the reasons I think the Bills are at best a . 500 team:

Superstition

As a general rule, you should never feel good about a Bills season based on their track record of disappointment, but it is particularly dangerous to be optimistic about the Bills when everyone else is. Bills fans have a habit of getting sucked in. I know, I've been there. When they won that season-opener against New England 31-0 back in 2003 I thought we were going to dominate the Patriots for a long time; it was the other way around. When the Bills had a chance to qualify for the playoffs in 2005 by beating the Steelers third-stringers, I was confident we were finally going to break a 4-year playoff drought; it has lasted another 7 years. When the team played good teams close in the opening half of 2010, finished the year strong, then started the 2011 season with a 5-2 record and a long-awaited victory against New England, I thought that we destined for the playoffs at long last; we finished the season 1-8. When fictional head coach Chuck Dichter (played by the great Jon Voigt) thought we were a lock to win the Super Bowl in direct-to-tv TNT movie 'Second String' I was right there with him; then the whole first string offense and all the backup quarterbacks caught a bad case of food poisoning and had to miss the playoffs. In all these cases (except the last one because I had already read the back of the VHS box which said what was going to happen) I thought the Bills were turning a corner, but I was sorely mistaken.

It's precisely when the fans get their hopes up that the team lights itself on fire. (Speaking of which, look for some Bills fans to light Mario Williams' lawn on fire after a game where he doesn't record a sack and the opponent -- probably New England -- scores on a late game-winning drive that he extends with a roughing the passer penalty.)


The second superstitious worry I have is that, in general, teams that go out and spend big money and get people excited, whether on purpose or as an innocent product, tend to fall flat on their faces. This could be just something I made up in my head (because of the Sabres recent extravagance and failure under Terry Pegula), it could be coincidence (unlikely), it could be because the new pieces struggle to fit in (more likely), or it could be because of the extra pressure that comes with people having higher expectations for the team as a whole (most likely). Regardless, it just doesn't give me good vibes to be the team that people are selecting to make the leap into the playoffs.

Additions

Mario Williams status as an elite player is undisputed, but it is a little concerning that Houston let him walk. It is also a little concerning that he has not had double-digit sack totals since 2007 and 2008. He needs to produce more than just quality rush defense to justify such a massive contract.

Picking up Mark Anderson from New England is the signing that seems much more likely to he a waste of $20 million over 4 years. That is a lot of money for a situational pass rusher. In Gailey's own words, "You don't pay that kind of money for a specialty player." Belichick grabbed Anderson on the rebound last year and found a niche for him that produced 10 sacks. New England promptly let him walk. This is classic Belichick: buy low, sell high. I guess he thought that Anderson's 10 sacks had made him overvalued. Given Belichick's track record, I don't like being on the other side of a bet on a personnel decision with him.

The other two key additions are 1st round draft pick Stephon Gilmore (CB) and 2nd round draft pick Cordy Glenn (LT). Gilmore sounds great, by all accounts, and Glenn could very well be competent, but I hate having rookies in such key positions. Protecting Fitzpatrick's blind side in a pass heavy offense is too important to trust to a rookie of so-so standing. And first-year cornerbacks almost never have an easy ride. I would feel much better if Gilmore was not being asked to come in as the new #1 CB for the team and could instead ease into the league as the #2.

Subtractions

I have not seen anyone mention that the Bills lost some adequate players this offseason. Roscoe Parrish could break games open while Demetress Bell had proven himself as a capable starter at the difficult and crucial position of left tackle, something that cannot be said about Glenn. Drayton Florence I could do without but CB depth is valuable in a league that is becoming so pass-friendly.

DVOA

Football Outsiders' has a DVOA statistic which measures every team's performance on a play-by-play basis against how an average team would expect to do in the same situation. Buffalo finished with a DVOA of -8.2% last year. Anything below 0% is a below-average team so right there you have macro-level statistical evidence that the Bills are a below-average team. And the year-long DVOA number hides just how bad the Bills finished last year; their weighted DVOA, which emphasizes the significance of more recent games, was -25.3%, 28th in a league with 32 teams. That is the DVOA of a 4-12 team.

There is a reason that Football Outsiders calculates a weighted DVOA and that is because recent performance is a better indicator of future performance than data from further in the past. The most recent Buffalo Bills play is the best measure we have available of how good they are and they were terrible. You don't even need advanced statistics to know that. Doesn't it strain credibility to believe that could be so much better now?

You could argue that their weighted DVOA is so much worse because of injuries but injuries happen every year to every team, especially the Bills. The other, more worrisome, explanation is that the team got figured out. It's probably a little of both, but Bills are already so up against at to beat 8 wins that they can't afford to be figured out even to a minor degree.

Plexiglass Principle

The Bills are 6-10, 4-12, and 6-10 over their last 3 seasons. This is not a team that had a sudden drop-off in performance last season which usually foretells a bounce back to some degree. In the Bills case it's the other way around. By itself, their jump of two wins from 2010 to 2011 suggests a small drop-off this year.

Turnover Luck

Luck with turnovers artificially boosted the Bills performance at the start of last season when they opened 5-2 and got all our hopes up. They were plus 8 in the turnover department after seven games, good for third in the league. Then the worm turned and their turnover margin for the full season fell down to 1. This was probably better than the Bills deserved because their their fumble recovery rate was an unsustainably high 59.46%. A team's ability to recover fumbles typically regresses to 50% because recovering fumbles is not a skill, it is a matter of lucky bounces. When the team was 5-2 it was recovering 2 out of every 3 fumbles. 

The Bills recovered 11 of their 16 fumbles on offense and scooped up 11 of 21 fumbles on defense, for a total of 22 out of 37 fumbles. If they recover only 18 or 19 of those fumbles this year, as should be expected, then those 3 or 4 lost balls could cost them a win or even two.

Fred Jackson

Fred Jackson was awesome for the Bills last year. He was having his best season ever. Before he got injured he was third in the league in rushing yards on a team that threw the ball more than anyone else as a percentage of their plays. He ran for almost 1,000 yards in less than 11 full games and had a yards-per-carry average of 5.5. He was great catching the ball out of the backfield but also when lining up as a traditional receiver on the outside or in the slot. He even pass-protected very well when he he was kept in to block. All this gave Chan Gailey a lot of flexibility with his formations in a single personnel grouping, which was key to his offensive philosophy and the team's success. Jackson was a legitimate MVP candidate when the Bills were rolling and the offense was putting up gaudy numbers.

Then Jackson broke his fibula in week 11. This is a big red flag for me because I don't think there is a great track record for elite running backs maintaining their elite status when they are 31 years-old (actually closer to 32 now) and coming off a broken leg. Running back performance is volatile to being with -- in 2010 Jackson's yards-per-carry average was 4.2, 1.3. yards lower than last year over a bigger sample size -- and typically declines after age 28 and after serious leg injuries. Once a back loses a step their performance hits a wall and I really worry that will show up in Jackson's game this year. I don't know what his recent 40 times are and the pre-season was too small a sample size to judge, so Fred Jackson is just a huge question mark right now. That's not what you want to be saying about your team's best offensive player going into the season.

Speed has never been the lynchpin of Jackson's running style but you need to have it to succeed as an NFL running back, let alone be one that carries a subpar quarterback to a winning season. I suppose the silver lining for the Bills is that between C.J. Spiller running too fast into the line before the hole opens up and Jackson hobbling along on beaten-up, broken-down 31 year-old legs the Bills may have the ultimate change-of-pace in their running backs.

Strength of Schedule

A lot of what has been driving optimism about the Bills season is their soft schedule, but I think a chunk of optimism is placed on mistaken assumptions about the team's they are goi. Last year the team faced opponent's that ended the season with a combined .520 winning percentage. This year, the teams they face had a .473 winning percentage last year. However, that number doesn't take into account how good the teams they face are this year, only last year. Indianapolis, for example, is going to be better in week 12 when the Bills face them because they will have Andrew Luck behind center and not the Curtis Painter/Dan Orlovsky losing combo the Colts were stuck with last year. That's a big reason why the Colts over/under this year is 5.5 when they only won two games last year. Nor does looking at the average of your opponents' winning percentages last year take into account how lucky they were.

So, instead of just relying on last year's records to judge the strength of the Bills schedule this year, I look at the average of their over/unders. As of August 22, the over/under win totals for the Bills opponents were as follows:

Team                          Over/Under
NYJ 8.5
KC 8
CLE 5
NE 12
SF 10
ARI 7
TEN 7
HOU 10
NE 12
MIA 7
IND 5.5
JAC 5.5
STL 6
SEA 7.5
MIA 7
NYJ 8.5
Average 7.9
Win % 0.494

As you can see, Vegas thinks that the Bills opponents are going to be a lot closer to average than their records last year would suggest. The difference between their opponents winning percentage last year and their opponents' expected winning percentage this year is 0.026, which is equivalent to about .416 wins over a whole season. If you add that to the Bills Pythagorean win total of 6.416 last year (which is the number of wins you would have expected them to have based on their point differential for all of last season), you get 6.832. Note that this number is still well below 8.

The other thing to keep in mind is that a weak schedule only helps if you can beat up on the weaklings in the league. If you also suck then games against St. Louis, Miami (x2), Arizona, and Jacksonville won't be gimmes. This is important because a lot of people have been going through the Bills schedule week-by-week and assigning a win to those five games, but if they are about as good as those five teams on average then they can only expect 2.5 wins from those five games.

Something else that ties into the emphasis on the Bills' weak schedule and a week-by-week prediction of wins and losses is the weak crop of QBs they hypothetically face if the starting QBs of their opponents don't change. Specifically, a lot of talk I have seen has focused on how they face several rookie and sophmore QBs. First of all, I don't think it is fair to assume that nothing will change between the starting QBs of their opponents now and in the future. I think that those QBs will be better by the time the Bills face them than they are now and if they aren't they will be replaced by their backups. For example, it is reasonable to assume that Matt Moore will be back in the saddle for the Dolphins by the time the Bills first play them in week twelve, for example, because Ryan Tannehill will be out either because of poor play or because of injury. If he is still playing by that point in the season it will be because he is at least semi-competent.

Pre-Season

Shortly after I made my bet against the Bills on August 22, the team played against the Steelers in their third preseason game, which is the game that the starters play the most and their is a modicum of game-planning. The starters did OK for the most part but got outscored and the backups got obliterated, leading to a final score of 38-7. So, in the game that the Bills put the most into during the preseason they got absolutely destroyed.

The rest of the preseason was not much better. The team went 0-4.

Coaching

Now, the retort to the previous section on the Bills' poor preseason play is that preseason is a joke because the teams do not game plan and you cannot extrapolate the team's future performance based on how its base offense and defense did without game-planning. I disagree. I think it does matter how the base defense and offense play because they are the foundation of those units. If the base is not stable then it does not matter what window dressing you put on top.

It's true that adjustments will be made to close holes and exploit match-ups when the regular season rolls around. The thing about adjustments and game-planning, though, is that both teams get to do them. Their existence, therefore, only benefits the team with the coaches that are better at it. It is a zero-sum game.

So, how good a coach is Chan Gailey? Well, he is 10-22 as the Bills head coach. He wasn't good enough for Todd Haley with the Chiefs. He was very successful one season back in 1984 with Troy. Since then he has flunked out of the NFL, had several middling years in the ACC, and then got called back up to the NFL by Ralph Wilson to unanimous cries of, "Who?" 

Look, I think Gailey's preference for the pass is the right one in this day and age with the new rules protecting quarterbacks and receivers, but it is not the right approach for this team. The Bills passed on a higher percentage of their plays last year than every team in the AFC except the Titans despite having a great running-back, a massive offensive line, a mediocre quarterback, and one receiver. How can you lead the league in 3+ receiver sets when you lack a reliable No. 2 receiver, let alone a quality 3rd or 4th receiver? The Bills ran more than twice as many four-wide receiver sets as any other team in the league. That doesn't sound like a coach who is going to win the gameplanning battle every week to make up for a poor base offense.

Well, you might say, Gailey only gameplans on the offensive side of the ball. What about on the new defensive coordinator we have, Dave Wannstedt? Wannstedt is like Gailey in the sense that both of them of failed in the NFL as head coaches before struggling in college as head coaches in second-rate conferences then getting called back up to the NFL by the brain trust Bills. And Wannstedt is not really new; he was around last as a defensive coach when the team had the worst rush defense in the league.

If you think either Chan Gailey or Dave Wannstedt is going to outcoach their counterpart on the other side of the ball then I have to wonder where your confidence in them comes from.

Let's not put too much faith in this man.

Wide Receivers

Except for Stevie Johnson, they are all replacement level.

QB Play

Which is a better sample size for Ryan Fitzpatrick's play? The six games that opened the season last year (before he got his big contract extension) when he threw 14 touchdowns and only 7 interceptions or the next ten games (more recent) when he threw only 10 touchdowns and 16 interceptions. You can blame his ribs if you want, but I doubt his ribs were injured in his previous seasons when his play was roughly equal in quality to last season as a whole. His career QB rating is 75.0. and has been within 7 points of that number every season since he started playing significant snaps in 2008. He is not terrible, but he is not great either. He probably is a top-32 quarterback in the league so he deserves to be starting somewhere but the team that has him as their starter is not going to ride him to a winning season with David Nelson as your number two receiver.

Ryan Fitzpatrick led the league in interceptions last year. Interception rate is inconsistent but more dependable than fumble recovery rate. Football Outsiders did an analysis of interceptions over the last two years to account for tipped and dropped interception, and they found that Fitzpatrick's interception rate was actually lower last year than it was the year before, he just threw less and had better luck in 2010. By all accounts his throwing motion wasn't fixed by new QB coach David Lee so we can expect the inaccuracy to continue, which is a fundamental flaw in a quarterback. His pedigree is a seventh-round draft pick from Harvard so you can't really expect much. He fits in perfectly with a long line of deficient Bills QBsthat started with Todd Collins and has only been interrupted briefly by the legendary Doug Flutie.

Defense

The team just spent two drafts trying to build a 3-4 front, but now it is switching back to a 4-3 because it is the only defense Wannstedt knows how to run. The problem is, some of the pieces might not fit as well.

The biggest problems last year were the defense's inability to stop the opponent's rushing attack and letting tight ends run free. Has the rush defense improved? On the line, yes, but the linebackers are still lacking and I'm worried that Marcell Dareus has blossomed into late first round talent and not top 3 overall talent. 

In the passing game, the cornerbacks we have are either young or, in Terrence McGee’s case, compromised by injury. If the offense can't score and the other team can run the ball against us, there won't be that many opportunities for the d-line to accumulate sacks, which defeats the purpose of having a pass-rush specialist like Mark Anderson on your team and limits opportunities for Mario and Kyle Williams to shine (although they should still contribute against the run). The Bills will probably still be vulnerable in the middle of the field because their linebackers are not great which is a major problem now that the NFL is a tight-end league.

In short, other than Mario Williams, there is not a lot to get excited about with a defense that was terrible last year and probably still will be. I'm afraid one player can't be great enough to swing things on the defensive side of the ball like a quarterback can on offense. Indeed, a d-lineman doesn't play all the defensive snaps in a game even if he is a star, so his impact is limited to begin with. And are we sure that Mario Williams is that great? Even if he is, is Dave Wannstedt good enough to make this unit even middle of the pack? He wasn't last year and there's not much that makes this year sound more promising.

Depth

The other thing the preseason showed was that the Bills depth is putrid, especially the third game against the Steelers when the backups got outscored 24-0 in the second half and the team realized Tyler Thigpen was not even backup caliber and Vince Young shouldn't be in the NFL. Not to worry though, because the Bills never suffer significant injuries. Oh wait, no, they always have the exact opposite of that.

Already, going into the season, the team's starting center, former first-round draft pick Eric Wood, is limited in practice. That's kind of important because offensive lines need time to gel. It's also disturbing because he was injured last year (and the year before) but still is not 100%. Similarly, Stevie Johnson has a groin injury from last season that is still lingering. He is our best receiver by a long-shot he missed a full practice on Wednesday. And there's still one more important player who hasn't recovered from last year's season: our best proven cornerback, Terrence McGee. I don't think he even played a preseason game. Finally, the other quality CB the team drafted this year, Ron Brooks, is already on IR. That's a big problem when two of your better cornerbacks are injured before the first regular season game.

Having a lack of solid backups is always a bad thing but it is particularly disconcerting in the Bills' case because they always have terrible injury luck. There is no reason why this should be the case, at least none that I can see, but it has gone on too long to be overlooked. Typically this becomes the excuse for the next season (e.g. Fred Jackson's fibula or Ryan Fitzpatrick's ribs). When Fitzpatrick gets injured this year and the team turns to an ill-prepared Tavaris Jackson, we'll get to enjoy watching a Seahawks reject use half the playbook as we march towards 5-11.

Conclusion

There are reasons to believe the Bills will do better this year, especially an improved pass rush and their easier schedule. But I don't think that is enough to improve their record by 2 or 3 games, especially when they got lucky last year with fumbles, got figured out offensively, and there is such good reason to be skeptical about Fred Jackson repeating last year's performance over a full season. I hope that by betting against the Bills this year I will not end up breaking my TV when we the replacement refs are screwing us and helping Miami.