http://views.washingtonpost.com/theleague/smarterstats/ 15 Doug Farrar crunches the numbers that really matter en Copyright 2009 Wed, 18 Nov 2009 09:42:51 -0500 http://www.movabletype.org/?v=4.2-en http://blogs.law.harvard.edu/tech/rss When forecasting upcoming prospects for teams, strength of schedule is often used as a benchmark for the ease or difficulty with which a team will navigate a full season, or an important run down the stretch. But simple won-loss records tell an incomplete story, at best. The Baltimore Ravens currently stand at 5-4, but they've outscored their opponents, 222-154, and they've lost their four games by a total of 21 points. The Jacksonville Jaguars also stand at 5-4, despite being outscored, 220-181, and losing their four games by a total of 74 points. But these teams look the same in the won-loss column, which is why we need a better measuring stick. Enter DVOA (Defense-adjusted Value over Average), Football Outsiders' per-play stat that measures drive and scoring efficiency, adjusted for opponent. The idea with our statistics is to weigh the real value of plays, and the wins and losses that http://views.washingtonpost.com/theleague/smarterstats/2009/11/smarter-stats-strength-of-schedule.html http://views.washingtonpost.com/theleague/smarterstats/2009/11/smarter-stats-strength-of-schedule.html Doug Farrar Wed, 18 Nov 2009 09:42:51 -0500 Last week, we looked at Adjusted Line Yards, Football Outsiders' proprietary statistic, for different NFL offenses. ALY assigns responsibility and value to offensive lines based on the length of a run, the situation, and the opponent. We then drilled down with situational stats that reveal each team's ability to break long plays, extend drives, and avoid negative plays. When you put all those stats together, you get a more complete picture of a line's value to a team. In that same spirit, we have ALY and other numbers for defensive lines, and the principle is the same. The first step is to separate actual yard per carry from line yards by factoring in play value by situation. Then, we add the percentage of yards on running plays that are 10 yards or more past the line of scrimmage. Power and Stuffed stats illustrate how well defensive lines prevent opposing offenses http://views.washingtonpost.com/theleague/smarterstats/2009/11/adjusted-line-yards-defense.html http://views.washingtonpost.com/theleague/smarterstats/2009/11/adjusted-line-yards-defense.html Statistics Wed, 11 Nov 2009 08:16:18 -0500 One of the primary goals of football sabermetrics is also one of its main challenges -- separating individual efforts in the ultimate team sport. Even skill position players depend so much on the efforts of others -- quarterbacks on receivers, running backs on offensive lines, linebackers on the defensive tackles who soak up double-teams and let them shoot the gaps. At Football Outsiders, we've put together a number of proprietary statistics to focus on the importance of line play on both sides of the ball. Adjusted Line Yards, which takes all running back carries and assigns responsibility to the offensive line based on yardage gained, is the primary stat. We then separate line yards per carry from running back yards per carry. The 10+ Yards stat separates yardage on plays 10 or more yards downfield -- this places more credit on the shoulder of the running back as he's making http://views.washingtonpost.com/theleague/smarterstats/2009/11/adjusted-line-yards-farrar-nfl-offense.html http://views.washingtonpost.com/theleague/smarterstats/2009/11/adjusted-line-yards-farrar-nfl-offense.html Doug Farrar Wed, 04 Nov 2009 00:36:00 -0500 When the Chicago Bears sent a 2010 second-round draft pick to the Tampa Bay Buccaneers for defensive end Gaines Adams, there were many justifiable questions about the price the Bears paid. Adams was the top prospect at his position in the 2007 NFL Draft, having amassed 28 sacks at Clemson, including 22 in his final two collegiate seasons. But he became a disappointment over his early NFL career, with 12.5 total sacks in 2007 and 2008, and only one in 2009. Bucs head coach Raheem Morris intimated that Adams would be a "bust" if he didn't put up double-digit sacks in his third season, but a quick look at the NFL's sack leaders over the last decade tells us that it usually takes time to develop a great pass rusher. Nine different players have led the NFL in sacks in the last decade -- only Michael Strahan has taken the http://views.washingtonpost.com/theleague/smarterstats/2009/10/sack-leaders-pass-rushing-farrar.html http://views.washingtonpost.com/theleague/smarterstats/2009/10/sack-leaders-pass-rushing-farrar.html Doug Farrar Wed, 28 Oct 2009 10:38:59 -0500 Does it seem that there have been more roughing-the-passer penalties in 2009 than in recent seasons? That has indeed been the case, but everything is not what it seems. Through the first six weeks of this season, there have been 35 roughing calls, or 5.83 per week. That stands in sharp contrast to the 59 overall in 2008 (3.47 per week over 17 weeks), 61 in 2007 (3.59), and you'd have to back to 2006 to find an equivalent average (110 overall, or 6.0 per week). The 2005 numbers prove that it's hardly a "new focus" -- there were 127 roughing calls in that season, or an average of 7.47 per week. The total was even higher in 2004; 135 overall and 7.94 per week. So, the "new emphasis" is less that and more a return to form. It also puts a hole in the theory that the NFL has http://views.washingtonpost.com/theleague/smarterstats/2009/10/qb-protection-farrar-brady-rule.html http://views.washingtonpost.com/theleague/smarterstats/2009/10/qb-protection-farrar-brady-rule.html Quarterbacks Tue, 20 Oct 2009 11:31:55 -0500 Quarterback protection has been a hot topic in the early part of the 2009 season. The Baltimore Ravens' October 4 loss to the New England Patriots was marked by two ticky-tack roughing-the-passer penalties -- one against nose tackle Haloti Ngata and one against end/linebacker Terrell Suggs. In both cases, the contact appeared to be borderline incidental, and Suggs looked like he was trying to avoid contact with quarterback Tom Brady's knees when he dove to Brady's right. Each call gave the Patriots 15 extra yards and helped facilitate two touchdown drives. The Ravens were not happy after the fact. Last Sunday night, the Tennessee Titans were flagged for two roughing calls in the same second-quarter drive as they tried to defend Colts quarterback Peyton Manning. The first call against end Kyle Vanden Bosch came after Vanden Bosch beat a double team and grazed Manning's legs with incidental contact. And the http://views.washingtonpost.com/theleague/smarterstats/2009/10/brady-rule-patriots-football-outsiders-farrar.html http://views.washingtonpost.com/theleague/smarterstats/2009/10/brady-rule-patriots-football-outsiders-farrar.html Doug Farrar Wed, 14 Oct 2009 08:00:12 -0500 Last week, we detailed the problems associated with the NFL's most disappointing teams, and the real (or at least secondary) reasons for their failures. But for every team that's fallen off the map, there's a surprise team, or a team that has outperformed their expectations. One team in particular has left the league amazed. Denver Broncos (4-0) Preseason Expectations: Chaos, destruction, and pain. Pat Bowlen surprised the NFL by firing longtime coach Mike Shanahan and hiring Patriots offensive coordinator Josh McDaniels, and McDaniels surprised most everybody by going on a bit of a rampage. After trying unsuccessfully to engineer a trade for Matt Cassel, McDaniels and Bowlen were left with a very unhappy Jay Cutler. They traded Cutler to the Bears for Kyle Orton and several draft picks. Little of note was done to restock a defense that finished 31st in DVOA in 2008, and the first-round pick of running http://views.washingtonpost.com/theleague/smarterstats/2009/10/smarter-stats-surprise-surprise.html http://views.washingtonpost.com/theleague/smarterstats/2009/10/smarter-stats-surprise-surprise.html Doug Farrar Wed, 07 Oct 2009 04:15:02 -0500 When teams with great expectations come out of the gate in true Three Stooges fashion, the subsequent finger-pointing is usually rampant, and not always directed at the real culprit. Frustration tends to trump analysis when the team you love can't buy a win and seems out of the playoff race before the season's first month is complete. In the interest of acknowledging the obvious when it applies to the situation and looking beyond the obvious, here are the real reasons that eight disappointing teams can't find their way. Washington Redskins (1-2) Surface Problems: Jim Zorn's coaching, and Jason Campbell in general. Real Problem: The defense can't stop anyone on third down. As under fire as they've been, you wonder if Zorn and Campbell dream of leaving their current team for an Albuquerque expansion franchise, "Any Given Sunday" - style. Some of the angst is justified - Zorn appears in over http://views.washingtonpost.com/theleague/smarterstats/2009/09/smarter-stats-the-blame-game.html http://views.washingtonpost.com/theleague/smarterstats/2009/09/smarter-stats-the-blame-game.html Doug Farrar Wed, 30 Sep 2009 02:04:56 -0500 Two Sundays ago, New Orleans Saints quarterback Drew Brees set an opening day record with six touchdown passes against the Detroit Lions in a 45-27 win. There haven't been too many single-game performances better than his. Brees followed that up with an outstanding game against the Philadelphia Eagles - unlike Detroit's, a defense you can take seriously - and he looks very much like the quarterback who came 15 yards short of Dan Marino's single-season yardage record (5,084) last year. But do incendiary single games lead to outstanding seasons at a rate that would lead us to take them seriously as indicators? Yes, according to Football Outsiders' DYAR (Defense-adjusted Yards Above Replacement) metric, which gives a quarterback's value based on situation and opponent, compared to replacement level, and converted into yardage. Brees' 2009 debut gave him a DYAR of 300, which would put him in the all-time top ten. What http://views.washingtonpost.com/theleague/smarterstats/2009/09/drew-brees-quarterback-dyar-saints-farrar.html http://views.washingtonpost.com/theleague/smarterstats/2009/09/drew-brees-quarterback-dyar-saints-farrar.html Statistics Tue, 22 Sep 2009 03:18:59 -0500 When it comes to measuring quarterbacks, the old and confusing rating system has been the standard for a long time. San Diego's Philip Rivers led the NFL with a 105.5 quarterback rating in 2008, but what does that mean? It's an arbitrary number that only means something in the context of other, similar ratings. Does it mean that Rivers was 8.1 "rating points" better than Miami's Chad Pennington, who finished second at 97.4? We suppose so, but we don't really know, and it's hard to attach real value to a number without seeing the work, so to speak. Who did he throw to? Against what defenses? With how many or few defenders in the box? Was he helped or victimized by his teammates? At Football Outsiders, we have several proprietary ways of gauging quarterback efficiency. There are our DVOA and DYAR stats, which measure players per play and over a http://views.washingtonpost.com/theleague/smarterstats/2009/09/quarterback-accuracy-beyond-the-ratings.html http://views.washingtonpost.com/theleague/smarterstats/2009/09/quarterback-accuracy-beyond-the-ratings.html Statistics Wed, 16 Sep 2009 07:01:58 -0500 During his first training camp with the New York Jets, former Ravens linebacker Bart Scott insisted to ESPN.com's Tim Graham that the one stat he never wanted to lead the NFL in was total tackles. "What you want to do is limit your reps," Scott said. "You want to lead the league in three-and-outs. You want to play the least amount of plays than anybody in the NFL. That should be your goal, to be No. 1. To have a lot of tackles? That means nothing. I could stand on the field all day and make tackles, but our team isn't going to be very good." Scott's absolutely right, and we're on his side at Football Outsiders. To that end, we keep two specific stats that reveal the actual value of a tackle: Stops and Defeats. Stops are the total number of plays which prevent a "success" by the opposing http://views.washingtonpost.com/theleague/smarterstats/2009/09/bart-scott-tackles-stops-defeats.html http://views.washingtonpost.com/theleague/smarterstats/2009/09/bart-scott-tackles-stops-defeats.html Doug Farrar Fri, 11 Sep 2009 16:38:12 -0500 jkhp982qec http://views.washingtonpost.com/theleague/smarterstats/2009/09/jkhp982qec.html http://views.washingtonpost.com/theleague/smarterstats/2009/09/jkhp982qec.html Thu, 10 Sep 2009 21:26:09 -0500 Last week, we took a look at the NFL's best and worst shotgun quarterbacks. Now, it's time to expand that view and discuss the best and worst shotgun offenses and defenses overall. Certain teams are just outstanding through the air in shotgun sets -- you'll see the Patriots and Cowboys displaying particular efficiency with shotgun draws, where the quarterback delays the handoff to lure the oncoming defenders into over-pursuing, thus leaving open lanes for a back to jet through. According to Football Outsiders' DVOA metric, the following teams were most and least effective with overall offense out of the shotgun formation: Washington's Clinton Portis led the NFL last year with 38 draws, followed by St. Louis' Steven Jackson, Dallas' Marion Barber, New England's Kevin Faulk, and Philly's Brian Westbrook. For the second straight season, the Patriots led the league in shotgun draws -- not surprising for the only team in http://views.washingtonpost.com/theleague/smarterstats/2009/09/smarter-stats-covering-the-spread-shotgun.html http://views.washingtonpost.com/theleague/smarterstats/2009/09/smarter-stats-covering-the-spread-shotgun.html Doug Farrar Sat, 05 Sep 2009 12:01:25 -0500 In the NCAA's last few seasons, the spread offense has become more and more popular. Offensive formations that set enemy defenses on their heels with motion, multiple receiver sets, and wide line splits has taken some programs traditionally impacted in their ability to recruit top personnel to a new level. However, several elements of the spread are not effective at the NFL level. Those wide line splits are caved in by better defensive linemen. The lack of backside protection leads to heaps of quarterback sacks. Shotgun quarterbacks with noodle arms and no feel for play action get drummed out of the pros before they even start. Texas Tech's Graham Harrell, owner of the NCAA's all-time mark in passing touchdowns with 134 and the second-most passing yards in college history with 15,793, wasn't even drafted by an NFL team because those teams are so aware of the adjustment issues. Harrell washed http://views.washingtonpost.com/theleague/smarterstats/2009/08/smarter-stats-covering-the-spread.html http://views.washingtonpost.com/theleague/smarterstats/2009/08/smarter-stats-covering-the-spread.html Doug Farrar Thu, 27 Aug 2009 01:48:36 -0500 A couple months ago, I wrote a long essay on the life of Pittsburgh Steelers defensive coordinator Dick LeBeau for the Maple Street Press Steelers Annual. That piece brought one simple fact into sharp relief - the truly great and revolutionary NFL assistant coaches don't generally get the respect they deserve, and the ones who give decades to their profession and are still at their best now are truly gifted. 2008 may have seen LeBeau's best coaching job. When you watched the way he lined those chess pieces up play after play, adjusted to offensive game plans, and seemed to drill right through the helmets of enemy quarterbacks and steal the playbooks right out of their heads - well, you got the feeling you were watching something special. Something you weren't going to see too often. Now, I'm writing about Pennsylvania's other NFL defensive genius, another man who gave decades http://views.washingtonpost.com/theleague/smarterstats/2009/07/jim-johnson-a-decade-in-review.html http://views.washingtonpost.com/theleague/smarterstats/2009/07/jim-johnson-a-decade-in-review.html Statistics Wed, 29 Jul 2009 01:08:53 -0500