Eagles vs. Cowboys: Breaking down this week’s betting lines.

Posted on December 27, 2013 by Brent

Split the action again last week, so let’s try to finish strong. I will do a breakdown for any playoff games, but of course, that’s not guaranteed.

I get the sense that a lot of fans are nervous about Sunday night’s game. Things are going a bit TOO well and everyone’s scared of getting the rug pulled out from under them. That’s fair, but really, Eagles fans should be nothing but ecstatic and excited heading into this week.

Before I start breaking things down, I want everyone to ask themselves one question:

What were your preseason hopes and expectations for this team?

My guess is, if you’ve answered honestly, then the team has already surpassed them. My personal projection was for a 9-7 record, with a playoff spot depending on whether one of those nine wins came from the last game of the season. That was spot on, but I’ve still been surprised by the level of play. The fact is that we’re now playing with house money. A loss on Sunday sucks, but it doesn’t change the fact that the season has been successful in every sense.

The goals for this season were:

– Prove Chip Kelly can be successful.

– Make the transition to the 3-4 and find a couple of young players who fit the scheme.

– Identify one or two “cornerstones”.

– Give Nick Foles a chance and see what he can do (not a popular one, but a big one for me).

In every aspect, the team has met those goals. Chip Kelly’s “college” offense is tearing up the league, even after teams have seen it once and supposedly had the chance to make adjustments. So far, the only thing that has really defeated it is having a 3rd string rookie QB at the helm or a Nick Foles body-snatcher. The 3-4 is installed and functioning much better than I expected. To be fair, I wasn’t a fan of the switch. However, it’s been decent and, more importantly, Mychal Kendricks and Fletcher Cox have made successful transitions. In that vein, while it’s too early to call Kendricks a “cornerstone”, the team clearly has a few guys with the potential to be impact players (Cox, Kendricks, Boykin). Moreover, there have been several big surprises that bode extremely well for the future (Logan, Thornton, Wolff in particular).

Lastly, Nick Foles is having one of the greatest seasons ever for a QB. Read that sentence again, this time emphasizing NICK FOLES. He leads Peyton Manning, having perhaps HIS greatest year ever, by 5.7 points in the Passer Rating title race. Safe to say Nick Foles has done better than anyone thought possible. Ok, one more thing:

Nick Foles’ current Rating is 118.7. Only TWO QBs in history have had higher single-season Ratings (Aaron Rodgersand Peyton Manning). Remember Tom Brady’s 2007 season? The only with Randy Moss and the undefeated record? Well Foles’ rating is currently 1.5 points higher than Brady’s was that year.

So, relax. Win or lose, this season has been an unqualified success, and the Eagles look like they’ll be the class of the NFC East (again) for the foreseeable future.

Now, this week’s game:

The Breakdown

The Eagles are 7 point favorites. The Over/Under is 52.5.

The comparison:

– The Eagles are ranked 8th overall by DVOA. The Cowboys are ranked 18th.

– The Eagles are ranked 3rd overall by Weighted DVOA, which is just regular DVOA weighted more heavily to recent games. The Cowboys rank 22nd.

– The Eagles offense ranks 2nd overall by DVOA. The Cowboys defense ranks 30th.

(If you didn’t just tee-pee your hands and say “Excellent….”, go relive the 90s)

– The Eagles defense ranks 23rd by DVOA. The Cowboys offense ranks 11th. Note that’s with Tony Romo at QB. We can safely assume that its AVERAGE performance level would be lower with Kyle Orton at QB.

– The Eagles STs rank 26th overall, Dallas’ rank 6th. This is the only matchup where Dallas has a significant advantage.

– The Eagles have a point differential of +58. The Cowboys’ is +9.

– The Cowboys last five games:

Beat the Giants (27th DVOA) by 3 points.

Beat Oakland (31st) by 7 points.

Lost to Chicago (12th) by 17 points.

Lost to GB (21st) by 1 point.

Beat Washington (29th) by 1 point.

Notice anything? Dallas’s resume isn’t exactly great. In fact, it’s not even good. Looking at the whole season, the Cowboys’ only win over a top 10 team by DVOA (currently) is when they beat the Eagles. Outside of that game, here are the DVOA ranks of the teams the Cowboys beat:

14th, 26th, 27th (twice), 29th (twice), 31st. The Eagles, of course, rank 8th.

Meanwhile, the Eagles’ wins have come over teams ranked:

10th, 12th, 16th, 17th, 27th, 29th (twice), and 31st.

It’s also worth noting that Dallas’ only “good” wins came in weeks 3 and 7. More recently, the Cowboys have lost to both Chicago and Green Bay. Conversely, the Eagles’ “good” wins have come in the past few weeks, against Chicago, Arizona, and Detroit.

So…taking the above as a whole, the Eagles are the MUCH better team. The ONLY significant factor weighing in Dallas’ favor is the fact that it beat the Eagles head-to-head. However, that was in week 7, so that win’s informational value has seriously depreciated since then. Also, Kyle Orton.

That’s why the Eagles are a 7 point road-favorite in a winner-takes-all division title game. (That and the Romo injury).

The Projection

Since the bye week (4 games ago), the Eagles have averaged 35.5 points per game. The Cowboys defense, as I showed above, is bad. It ranks 3 spots below Minnesota’s, and the Eagles put up 30 points against them. There are no significant injuries and weather will not be a factor unless something completely unexpected happens. As a result, I’ve got the Eagles base-case projection at 31-35 points. To make things easier, we’ll take the mid point and say 33.

Meanwhile, the Cowboys have averaged 28.6 points per game since their bye week (5 games ago). The Eagles defense is also not very good, and ranks just 1 spot above Oakland, against which the Cowboys scored 31 points (4 weeks ago). Additionally, the Eagles defense ranks just one spot BEHIND Washington, against which the Cowboys scored 24 points just last week. So there’s our range, 24-31 points. Taking the midpoints, we’ll say 27.5.

HOWEVER, we have yet to account for the Tony Romo injury.

I know a lot of people rip Romo, but the fact is he’s one of the best QBs in the game. In my opinion, not nearly enough blame gets put on everyone else in that organization. Stepping down from Romo, who has a career rating of 95.8, to Kyle Orton, who has a career rating of 79.7 (and just 15 pass attempts over the past two seasons), is a MASSIVE hit.

Also, this “game-manager” stuff is bullshit. Kyle Orton’s career Interception rate is 2.6%, just a tick below Tony Romo’srate of 2.7%. Romo has 55 fumbles in 108 career starts, or .51 per game (that’s total fumbles, including from games not started, if there are any like that). Orton has 32 fumbles in 69 career starts, or .46 per game. So there’s a difference there, but it’s very small (.05 per game). Overall, Orton is really no less likely to turn the ball over, and in fact may be MORE likely, by virtue of the fact that he’s barely played in 2 years.

To account for Orton, I’m taking EIGHT points off the Cowboys’ projection. That sounds like a lot, right? Especially since the line moved just 5 points after the injury was announced. Looking at it practically, though, you’ll see it’s not that drastic. Basically what I’m saying is that, as a result of having Kyle Orton instead of Tony Romo, the Cowboys will have 2 drives that end up in field goals instead of TDs. Of course, that’s just one way the difference could manifest itself, but it’s instructive because of how reasonable it is. Kicking 2 field goals instead of scoring 2 TDs gets you the 8 point difference.

Taking that from the 27.5 points projections, we’re left with 19.5 points. However, there’s a big special teams discrepancy, so I’m going to add another 1-2 points back in, giving us 20.5-21.5. That’s a very convenient number (I swear I didn’t work backwards) because it’s midpoint is 21, or three TDs.

Pulling it all together, I’ve got the base-case projection at Eagles 33, Cowboys 21.

That’s a difference of 12 points. The Spread is 7, meaning our margin of error is 5 (very large). That’s the second biggest margin I’ve seen this year. Take the Eagles -7, and take them confidently.

The Over/Under

This bit is easy, since we’ve already got our scoring projection.

Eagles 33, Cowboys 21 adds up to 54 points overall. The O/U is 52.5. Also, the teams are a combined 18-12 against the O/U this year, and the line hasn’t shifted at all since opening, so we’re not joining a sucker’s move by taking the over.

54 is 1.5 points higher than 52.5, so take the over. Note, however, that it’s a relatively small margin, so adjust the stakes accordingly.

Summing Up

I like the Eagles -7, and I like it A LOT.

I’ve got the Over 52.5 as the better side there, though it’s not as attractive as the spread.

Finally, for fans here, a win is much more important than whether the Eagles cover or not. So, I looked up win rates for 6-8 points road favorites since the year 2000. In those games (245 of them), the favorite has won the game almost exactly 70% of the time.

Using my projection, 12 points, I looked at the recored of teams since 2000 that were 11-13 point favorites. Road favorites won the game 75% of the time, but the sample was just 20 games. Looking at ALL games, home and away, the favorite won roughly 83% of the time.

So yeah, the odds are HEAVILY in the Eagles favor. (But that still leaves close to a 1 in 5 chance of a Dallas victory).

The Benefits of Being a High-Variance Team

Posted on December 24, 2013 by Brent

Great game yesterday. It was a nice preview of what this team COULD be if both the offense and defense play well at the same time. Can’t ask for a much better set-up for the Eagles than a win-or-go-home game next week in Dallas. The Eagles are, objectively, a much better team. The stakes should take care of the motivation aspect. Also, with Romo being out (assuming the news is accurate), if Foles shows up looking anything like “GoodFoles”, there’s very little chance of Kyle Orton keeping pace.

Now, to today’s topic.

According to Football Outsiders, the Eagles are ranked 31st in the league by DVOA Variance, at 25.4% (BEFORE the Chicago game).

Only St. Louis has been more uneven. Normally, you’d prefer your team to be both very good, and very consistent (low-variance). That’s the goal. However, there’s more to the story, and it ties in to our general underdog strategy discussion.

The Eagles are not the best team in the NFC. They might not even be in the top 5 (before yesterday, DVOA had them 7th in the NFC). That means that winning the Super Bowl will require winning multiple times against inarguably “better” teams. When I say better, I mean the expected performance of the other team is clearly higher than the expected performance for the Eagles. Of course, that’s only one part of the equation. The other, obviously, is variance.

The fact that teams don’t always perform to expectations is exactly what makes the game fun. Otherwise, there’d never be any upsets. So…taking the next step, that means if you’re a large underdog, you really want at least one of the teams involved (you or the opposition) to be a high-variance team. Remember, underdogs (both ex-ante and as a result of current conditions) want to MAXIMIZE variance.

Let’s illustrate. Below is a graphic showing the expected performance distributions for two teams. Unfortunately, the shape-options in Powerpoint are fairly limited, so the shapes are a bit crude.

Above, the width of each distribution and it’s height at each point tells you both how good the team is and how consistent it is. If this were a “to-scale” drawing, the area under the curve would add up to 1. Notice that in the above chart, there is a gap between the two teams. That means, in this case, the Red team would NEVER beat the blue team.

Let’s pretend you’re the Red team. What can you do? Obviously, you can’t do anything to Blue’s distribution. In general, the whole point of team-construction is to move the distribution to the right, so that’s option A. If you shift Red far enough to the right, you’ll catch up to Blue.

But what if it’s in-season? What if you only have one week before the game? You can’t do much to change the make-up of your team, so Option A is out. There’s still hope, though. You can WIDEN the performance distribution. This is what it means when we say teams in desperate situation must make High-Variance moves. Let’s say Red team had the same average performance expectation, but is now a High-Variance team. Then the chart might look something like this:

See the overlap? That’s the key. Although it’s still unlikely, there is now an actual possibility of Red beating Blue. Notice that there’s also a possibility of Red losing by a lot more than it would have before. However a loss counts the same whether it’s by 1 point or 40 points. (Ask any real racer….)

Back to real life: the Eagles have a very WIDE expected performance distribution; it’s reflected in their high-variance. That means that even if they’re undeniably worse, on average, than a team like Seattle, they’ve still got a decent shot at winning (compared to if they were a low-variance team.)

For example,

Prior to yesterday, the Arizona Cardinals ranked one spot ahead of the Eagles by DVOA (10.9% to 7.7%). However, the Cardinals are among the most consistent teams in the league, and rank 4th overall by Variance, ahead of the Eagles by 27 spots, and a variance margin of 18.9%.

Charting each team against Seattle (very rough approximations here), we’d get something that looks like the following:

Notice that while the Eagles’ average is worse than the Cardinals, their overlap with Seattle is actually greater than the Cardinals. We’d have to do some calculus (and put a lot more effort into an accurate chart) in order to calculate the difference, but the overall idea is sound.

While it’s incredibly unlikely, the Eagles do stand a greater chance than similarly skilled teams to actually win the Super Bowl if they get to the playoffs by virtue of their high-variance nature.

Lastly, let’s look at the regular season variance of recent Super Bowl winners (this is going to warrant a dedicated post, but let’s just take a peak for now):

Baltimore Ravens – 15.6%, 24th in the league

New York Giants – 15%, 20th in the league

Green Bay Packers – 14.8%, 15th in the league

New Orleans Saints – 15.8%, 17th in the league

Pittsburgh Steelers – 10.8%, 8th in the league

SI Curse: Contrary Indicators 101

Posted on December 20, 2013 by Brent

My last final of the semester was yesterday, so I should be able to post much more frequently over the next few weeks, just in time for the Eagles stretch run. There’s a lot to be said there, and a few things I started and have to finish, but today I want to illustrate the “SI Cover Curse”.

Obviously, being on the cover of Sports Illustrated, as Nick Foles was last week, does not actually “curse” its subject. So why does it seem to work so often? Basically, it’s because it HAS to work; that’s how the system is designed.

Items with covers that change by installment use those covers to drive sales. Think of the gossip magazines paying millions of dollars to celebrities for exclusive pictures. They do that because people decide which ones to buy based on what the see on the cover. Beyond the brand, it’s the only real advertising available. As a result, you only get “attention-grabbing” cover subjects.

Now let’s look at a normal career arc for an NFL player (or any athlete). I apologize for the crudeness of the drawing.

Generally, players start slowly as they adjust to the NFL game (and get bigger/stronger). After an initial period of development, they plateau, and remain there until they get old and decline. Pretty simple. Let’s take that graphic and break it into those sections.

Note that this tracks closely with relative fame. Very few players are really famous upon entering the league. Similarly, very few players maintain their level of fame after they decline and retire. That has obvious implications for magazine covers. They don’t put nobody’s on the cover of Sports Illustrated (well, rarely).

Of course, that curve is far too smooth. So let’s add a layer to show the general oscillation of a player’s career skill/fame. Every player has peaks and valleys, which oscillate around the longer term average.

These are rough approximations, obviously some players have drastically different career arcs (injuries, in particular, can throw things off). However, looking at the larger point: when, in the graph above, do you think a player is most likely to be on the cover of Sports Illustrated, or any magazine for that matter?

How about at these points:

At the highlighted points above, players are at their short-term peak in fame/significance. That’s when they’re most likely to be put on the cover of Sports Illustrated. And in order for it to be a “peak”, it MUST be followed by at least a short-term decline thereafter (otherwise it wouldn’t be a peak). That’s the essence of the “Sports Illustrated Curse”. Generally, players only get on there if they’ve been playing extremely well. For example, looks at these headlines and decide which ones are actually interesting:

– No-name player plays badly!

– Famous player plays well!

– No-name player plays well!

– Famous player sucks!

You don’t have to be a marketing major to realize that the bottom two are the better “stories”. Moreover, for a “no-name” player to actually get on the cover, he has to more than just “play well”, he has to be lights out, or be involved in a singular moment of some large significance. This is why Nick Foles was featured. Outside of Philadelphia, nobody knew who he was, yet he’s putting up a historically good streak of play. That’s a story.

However, it’s only a story because he was playing SO well. If he had been merely “good”, he wouldn’t have made it. That leaves us with just two possible consequences of the following statement:

Nick Foles, relative unknown, played historically well for a short period of time. Then…

– Nick Foles was just that good, and continued to play at that level until he had broken nearly every NFL passing record.

– Nick Foles was not actually the greatest QB ever, and his subsequent performance declined soon after.

Now, if the second option happens (as the odds suggest is an almost certainty), it won’t be because he was “cursed” by Sports Illustrated. Encouragingly, Foles has been playing SO well that a somewhat significant decline in performance will still leave him as a very good quarterback.

The system within which the magazine business operates is built upon capitalizing on short-term over-performance. It’s no surprise that cover subjects experience a decline in performance/fame soon after. The level of play required to get Nick Foles on the cover of SI is almost certainly unsustainable; consequently, it won’t be sustained!

So the “Sports Illustrated Cover Curse”, while not an actual “curse”, is not complete bullshit. It’s a valuable contrary indicator for those who know how to properly evaluate it within the larger context of the player/team/league/etc…

Lastly, this isn’t just a sports-related phenomenon. It’s relatively well-known in the investment industry, though it’s tough to follow due to the long-term nature of the trends (2 year lag for an inflection point after a 30 year trend is on point, but tough to make money from). For example…

From 2005 (the market peaked in 2007):

From 2008 (the market bottomed in 2009):

And the classic (some say original), from 1979:

Mean Reversion and the 2012-2013 Eagles Improvement

Posted on December 12, 2013 by Brent

Preseason, I did a number of posts that focused on the reasons why the Eagles finished with such a poor record last year. The general thesis was that the team was bad, but it was also very unlucky. Therefore, we could expect a better record this year purely as a result of reverting to the mean in several meaningful statistics. Today, let’s take a look at a couple of them and see how they look.

First, here’s the 2012 performance dashboard I put together.

Remember that I scaled everything by historical standard deviation (last 10 years of data) so that it could all be viewed in one chart. For our purposes today, the most important terms above are Fumble Recovery %, Fumbles Lost, and Net Field Position.

Note that, for now at least, I’m going to avoid the whole luck-vs.-skill angle. I’ve explored that before and I’m sure I’ll revisit it again. Regardless of which side you believe in, the fact is that regardless of the role of luck, those statistics show NO PERSISTENCE from year to year. Note also that the three stats I’m highlighting are obviously interrelated, so it’s no surprise that terrible performance in one is correlated with terrible performance in the others.

Fumble Recovery %

In general, teams should expect to recover around 50% of all fumbles. There’s been some additional research done about varying rates for different TYPES of fumbles (Downfield WR vs QB for example), but after including all types, the overall rates converge to 50%.

Last year, the Eagles recovered just 35.09% (Teamrankings.com), which is 1.99 standard deviations below the mean. That’s really bad, and extremely unlikely to happen again. So how is the team doing this year?

46.34%

Not great, but a much more reasonable rate of recovery.

Fumbles Lost

Relatedly, the Eagles problem last year wasn’t just the rate of recovery, it was an overwhelming number of fumbles. Combined, that meant the 2012 Eagles lost a historically large number of fumbles to the other team. Looking at the chart above, we see that the team lost 22 fumbles last year, which is nearly 3 standard deviations from the mean. Like I said, historic, and a big reason why last year’s team struggled so much.

So how do things look now?

Well so far, the team has lost just 8 fumbles, or .615 per game, meaning it’s on pace for just under 10 fumbles lost, less than half of last year’s measure.

Net Field Position

Finally, for today at least, there’s Net Field Position. As a result of both special teams and the historic turnover rates, the 2012 Eagles had TERRIBLE net starting field position. Looking at the chart above, we see the team’s average drive started 6.67 yards behind the other team’s average starting position. That’s a very big difference, and it’s more than 2 standard deviations from the mean. The offense last year was actually middle-of-the-pack by yards-per-drive. The problem was that they had farther to go than everyone else.

This year? +1.4 yards, good enough for 11th overall (Football Outsiders).

Having trouble conceptualizing the significance of the shift? Well consider this:

This year, the team is averaging 33.06 yards per drive. It’s scoring 25.7 points per game.

Last year, the team averaged 31.51 yards per drive. It scored just 17.5 points per game.

Put differently, this year’s team is gaining an average of just 1.5 yards per drive more than last year’s team did.

The real difference? Mostly turnovers and field position, both of which we’re primed for mean reversion.

Lastly, the really good news

Did you notice anything else about the stats I just discussed? Let’s look at them again:

Fumble Recovery %: 46.34%

Fumbles Lost: On pace for 10

Net Field Position: +1.4 yards (11th overall)

Now? While last year’s numbers were EXTREMELY bad, and thus carried a very high probability for improvement, this year’s numbers are squarely in the middle of the expected range. That means, while last year’s team was both bad AND unlucky, this year’s team is just good, no luck caveat needed, at least as it pertains to these stats.

That means what we’re seeing isn’t likely to be a fluke. Once the season is finished I’ll look at a larger number of statistics and see where we can expect improvement or decline, but for now, it looks like the team is just good.

P.S. I’m in the middle of the law school exam period, hence the low volume of posts. Good news is I’m finished next week, meaning my break coincides with the home stretch of the season, and I’ll be able to post a lot more frequently, at least until late January.

Eagles v. Lions: Pre-Game Notes

Posted on December 8, 2013 by Brent

A few thoughts on today’s game, after which the Eagles could be in sole position of 1st place (I think Chicago’s got a good shot to knock of Dallas).

– Depending on the weather, this is a game where the general strategy should be more “aggressive” than usual. We’ve got a great offense against a mediocre defense (Eagles v. Detroit) and a mediocre offense against a bad defense. In both cases, we expect the offense to have a significant advantage, that’s why most people are projecting a shootout. That means field position has less relative value than normal. For example, giving Detroit the ball at the 20 yard line isn’t worth that much more than giving them the ball at the 40 yard line (their’s).

In other words, the bar for going for it on 4th down should be lower, and TDs should be prized more heavily than usual over field goals. Also, it’s a great game for a surprise onside kick.

If the weather is awful, it changes things a bit, but that depends on what type of bad weather there is. If the offenses are unimpaired, then the “right” play is to be aggressive, because you have to assume Detroit is going to move the ball well against the defense.

– Tough test for the O-Line. Nothing groundbreaking here, but the Lions have perhaps the best DT combo in the league (Suh and Fairley), with rookie Ziggy Ansah at DE (remember him? he was the #1 prospect on my TPR rankings). Together, those three have 16 sacks, with 9 coming from the DTs. They’re going to get pressure today, the key is how difficult it is for them. If they can consistently give Foles problems with just 4 d-linemen, the Eagles are in trouble.

The plus side is that the Lions d-line is very aggressive (wide-9 anyone?), meaning it can be pulled out of position with misdirection, which just happens to be the Eagles’ specialty. I’m expecting a lot of PA, back-side screens, and maybe an end-around. Ansah and Suh are going to get upfield regardless, might as well take advantage of it.

– Can the Eagles get pressure? Perhaps the most surprising Lions stat of all is Matt Stafford’s sack percentage this year. He leads the league at just 2.9%. Foles, by comparison, is at 8%. He also throws the ball more than anyone else in the league. Today is the toughest test for the Eagles defense since it played Denver..and we all know how that turned out. By DVOA, Dallas ranks higher than Detroit, but based on the match ups, I think Detroit poses more problems for the Eagles. Patrick Chung needs to have a much better game than last week, and the LBs have to be very alert for Reggie Bush and Joique Bell as receivers out of the backfield (they have a combined 79 receptions).

I expect the Lions to score a lot. The key for the Eagles is to create a turnover or two so the offense has extra possessions to keep up/go ahead. Any drive that ends in a Lions field goal is a success.

– The Eagles play the Vikings next week….That means a win today likely gets them to 9-5 going into that last two games of the year (Chicago and Dallas). Basically, a win today up the expected win total to 10 wins, and 10 wins probably gets you the division title. See what I’m saying? If Aaron Rodgers come back next week to play the Cowboys, a win today might make the final game of the year a moot point. So yeah, it’s pretty important.

I’ll leave it there for now. I think the Eagles are SLIGHTLY more likely to win (if you saw my odds column on BGN, I had the Eagles -2 in my projection), but it’s close enough that a single turnover or STs return could make the difference (or onside kick!). Good luck to anyone at the game…it’s days like this I’m happy I don’t have tickets.

Notes from Yesterday

Posted on December 2, 2013 by Brent

Yesterday’s game went down just about exactly as expected. I’ve got a few notes, and I’m going to try to steer clear of the obvious ones.

– The Eagles average starting field position was the 36 yard line. Arizona’s was the 19 yard line. In a game between two evenly matched teams, that’s a HUGE difference. Some of that was the turnovers, some of that was Donnie Jones. I get the sense that he’s flown under most fans’ radars, but Jones has been a very important piece this year (relatively speaking). Yesterday, he had 7 punts inside the Arizona 20 yard line. Against a poor offense (which the Cardinals are), that’s a very big deal. Remember pre-season, one of the things I highlighted was how bad the Eagles average field position differential was last year (nearly -7 yards) and that it was very unlikely to be that bad again, or even close to that bad. I haven’t seen the season numbers for this year, but its safe to say there’s been some mean reversion there, regardless of why it’s occurring (Jones? TOs? Luck? Likely all 3).

– The Eagles passing game was incredibly balanced yesterday, and illustrates a big point in Nick Foles’ favor. He really doesn’t seem to have “favorite” receivers. Yesterday, Jackson, Cooper, Ertz, Celek, and McCoy all had 6 targets. Avant had 4. That’s the definition of balance, and it shows Foles was “taking what was there”.

– Foles did NOT play that well. This definitely qualifies as a “lucky” game. He had the INT called back, but more worrisome were the handful of open passes that he missed badly on. That’s what we saw in the Dallas debacle, and it’s still without explanation. Fortunately, it was less pronounced yesterday. Perhaps I’m digging too deep for criticism, but I don’t think it’s nit-picking to say QBs should hit almost every pass when he has time to deliver and sees a wide-open receiver well within his range. Can’t expect him to be perfect, but missing more than 1-2 of those per game is too many.

– Billy Davis hurts my brain. Some questionable scheming by Davis yesterday. Most notably, and I’ve made this point before, he continues to send CB or S blitzes and “disguising” them at the snap by having the players stay in base position. Note that this means the CB is often around 20 yards from the QB at the snap. See the problem?

The fastest players in the league run 4.3 40 yard dashes (roughly). That’s with perfect conditions and no pads. So even in that situation the player will take at least 2.15 seconds to get to the QB. Moreover, when you account for the pads, sub-optimal alignment (not in a sprinter’s start), and the obvious potential for blockers to be in the way, you have to figure it’s going to take more than 3 seconds for the CB to get there. Meanwhile, while he’s en route, you’ve essentially taken a defender off the field. He’s in no man’s land. It’s quickly becoming one of my least favorite plays in football, but Davis continues to call it.

– Eagles were just 5 of 16 on 3rd down. Nothing to add here. It was a good defense, but you’d still like to see that conversion rate much higher. It’s a big reason the game was close.

– Brandon Graham had 2 sacks. That’s not really news. However, he did it with just 13 snaps. That’s a pretty good impact rate.

– The Eagles are lucky they played the Cardinals, the defensive backfield looked very vulnerable for much of the game. Have to wait for All-22 to confirm, but it looked like the Safeties especially had a tough game. Didn’t matter because Carson Palmer is not a good QB anymore (at least he wasn’t yesterday), but that will hurt a lot against a good passing offense (say Detroit?). Can’t believe I’m saying this, but they really need Earl Wolff to come back…

Big win regardless of caveats. .500 record from here on (with one game against Minnesota) gets them to the 9 wins we thought they’d get to. Whether that equals a playoff game depends almost entirely on the last game of the season.

QB Performance Frequency Distributions

Posted on November 29, 2013 by Brent

Ok, so a couple of weeks ago I posted about how often great QBs have bad performances. Today, let’s take a more detailed look at things. The overall question is, are quarterbacks equally likely to outperform or underperform their long-term average QB rating in any individual game? Stated differently, are the individual performance distributions symmetrical? Are they Normal?

Let’s start at the top. Here’s the graph for Peyton Manning. Note that the X-Axis labels show the UPPER Bound of each bar. So the “60” label means that bar corresponds to games where the player’s rating was between 50 and 60. Also, this is all games with at least 10 pass attempts. Remember that these are NOT weighted numbers, so they’ll be different from the career measures for each player. This helps to minimize the skew effects of games with a lot of pass attempts (garbage time yards in a blowout loss for example) as well as increase the weight of great games with relatively few attempts (when a team has a big lead early perhaps). Realistically, we just want to know what level of performance we’re likely to get in the NEXT SINGLE game.

That looks pretty Normal. The Mean is 97.45 (note that this is NOT his long-term average, since it’s not weighted by attempts). The Median is 95.6. Obviously, those are crazy-good numbers. That’s why he’s a HOFer.

Big-picture, if players generally follow a Normal distribution, then we can tell a lot about Nick Foles from relatively fewer games. So Peyton’s chart is really encouraging.

But here’s Drew Brees:

Not nearly as neat as Peyton’s. The Mean is 95.17, the median is 92.4, and there’s some clear skew to the distribution. Going back to the last post (linked above), notice that Brees has had more games with ratings between 60-80 than he has games with ratings between 80-100. Overall, his performance, though still amazing, is less predictable than Manning’s. The standard deviation of Brees’ game log is 29 (rounded) while Manning’s is 27. Illustrated differently, we can look at the range covering the middle 50% of performances:

Manning: 79.9-112.2

Brees: 72.8-116.7

Again, both great, but Brees is less predictable. We could raise some interesting strategic questions here as far as which one you want in which situation, but I’ll save that for another day. For now, just imagine tying it back to our “David Equation” (that’s what I’m calling it now). Brees might not be as good, but his higher-variance play might be preferable for an underdog team, while you’d rather have Manning if you’re the favorite.

Now let’s take a step down and look at some non-future-HOFers.

Here’s Sam Bradford:

Much uglier, as expected. The mean is 80.48, the median is 81.2. The standard deviation is actually much lower than either Manning or Brees, at just 21. His middle-50% range is:

66.3 – 91.3.

That’s what a bad QB looks like. Now we should probably caveat all of this by saying it’s a bit unfair to evaluate the QBs in a vacuum, with no regard to the talent level they’re working with. That’s the case with just about every NFL evaluation, it’s just the nature of the game. Still, Bradford hasn’t been good enough, and I’m skeptical he ever will be…

How about Eli? (You knew I couldn’t leave him out)

Mean of 82, Median of 81.65. Standard deviation of 27.3. Hmmm…those numbers look vaguely familiar. What were Sam Bradford’s again? (mean of 80.48, median of 81.2, stdev of 21).

Interesting. Future HOFer Eli Manning’s average a median performance are almost identical to secret-bust Sam Bradford (secret because nobody seems willing to say it outright. I will, Bradford is a bust, unequivocally. Closing in on 2000 pass attempts, he has a com % below 60 and a Rating below 80).

What about Eli’s middle-50 range? 63.6 – 100.7.

That’s just about the definition of mediocre (and maybe even a bit worse, we’ll see later).

Before I move on, let me repeat one thing.

Eli Manning’s MEDIAN performance is a rating of 81.65. So HALF of his starts are WORSE than that.

Going back to the initial question I posed: Are individual QB performance distributions symmetrical? Almost, though we have to grade it as inconclusive since I only looked at a few QBs. So we might be able to use that to infer some info about Foles. Clearly, though, they’re not Normal… There’s a lot more we can do with this type of data, but I’m going to have to wait for another day to start on it.

Also, as is usually the case, I think I stumbled onto something more interesting (the middle-50 ranges). So rather than go through each QB and post a chart of they’re distributions, I’m going to end this post now and start making a table of every starting QB’s Middle-50 range. Then we can start talking about which is “best” in a given situation, with some real data to go from.

Before I go, here’s Nick Foles. He has just 11 qualifying games, so small sample is an understatement, but it’s fun to look anyway.

Average is 97.2, median is 96.6. Standard deviation is 35.8.

And here’s Mike Vick.

Average is 80.9, Median is 83.6. That’s 2 points BETTER than Eli’s median performance…. Vick’s standard deviation is 28.6.

Happy Thanksgiving

Revisiting PreSeason Projections: Overrated Teams

Posted on November 25, 2013 by Brent

Sticking with the preseason review today (no better time to do it than a bye week, other than after the season ends, of course). This time lets revisit the list of overrated and underrated teams I posted before the season started and see how things stand.

Overrated

The original post is here. Below, I’ve included the summary table.

To refresh, I simply took the Football Outsiders projections and compared them to the Vegas lines to see where the biggest differences were and what they could tell us about the expected performance of teams. As you can see, pretty much nailed this one. The most overrated teams, by this measure, were Atlanta and Minnesota.

Atlanta, of course, is currently 2-9 and arguably the biggest disappointment in the league this year (though I’d obviously argue that they shouldn’t have been so highly rated to begin with.

Minnesota was another gimme. Anyone taking the Over (7.5 wins) for this team was either crazy or stupid. Specifically, I said:

I have the Vikings in the 4-5 win range, which is even lower than FO. While there are a number of explanations (including the points above), it really comes down to Christian Ponder versus a very tough schedule. I think it’s likely to be a train wreck.

To date, the Vikings are 2-8-1. Ponder hasn’t exactly been a train wreck, but with 9 interceptions (to 7 TDs) and a rating of 78.7, he hasn’t been good either.

Miami started the season strong, winning the first 3 games, but has lost 6 of 8 since then. Currently at 5 wins, the Dolphins are likely to surpass the FO projection, and may even get to the 8 wins O/U (I doubt it though).

All told, you’d have done pretty well by following this chart. Certainly regarding the teams identified as the most significant projection dislocations. Said differently, if you looked at the original post and took the under on the Vikings and Falcons, you’re pretty happy now.

The Eagles, of course, are not on this list. FO projected the team to win 7.8 games, while the O/U was set at 7.5, making them Underrated, though only marginally so. My personal projection was for 9 wins, which still looks pretty good.

Busy week (exams approaching), but I’m hoping to post performance distributions for prominent QBs to continue our discussion (in more detail) about reasonable expectations, so look for that sometime in the next couple of days.

Checking in with Preseason projections

Posted on November 20, 2013 by Brent

The bye week is a good time to revisit our preseason projections, so today, let’s look back at how I though the season would play out. If you remember, I took two different approaches. First, I used a basic, back-of-the-envelope points for and against projection matrix. The base-case assumption there was 9 wins, though the average outcome of the matrix was a bit higher (9.1 wins). Here is the matrix:

After I did that, I went through the schedule and tried to come up with benchmarks, the clearing of which were (a) reasonable, and (b) would lead to 9 wins.

Let’s start with the win projection.

The base case projection (9 wins) was built from an assumption that the Eagles would produce points at a level 15% better than the league average and allow them at a level 5% worse than league average. Over the course of the season, I projected that to equate to an overall point differential of +37, which I plugged into the Pythagorean projection model to get to 9 wins.

Through 11 games, the Eagles’ point differential is +16. They’ve scored 276 points and allowed 260. Based on that, my projection looks pretty good. If the Eagles kept up at that exact pace, they’d end up with a point differential of about +23, just two TDs from the +37 projection.

However, I think it’s instructive to dig a bit deeper and take a look at how each side (point production and points allow) has compared to our expectations.

As I said above, I expected the Eagles to score 15% more points than the league average. To date, the team as averaged just over 25 points per game. The league average (including the Eagles), is currently 23.4 points per game, meaning the Eagles have scored 6.8% more points than the league average. That means the Eagles offense (technically “point production” to include defense and STs) has been worse than I expected, by approximately 1.8 points per game.

However, some of that can be attributed to the Matt Barkley game, as I obviously didn’t plan on him getting a start this year, and we can also assume that the offense would have scored at least a few more points if it hadn’t been playing with big leads (especially against the Redskins).

On defense, I expected the team to be 5% worse than league average. Thus far, they’ve allowed roughly 23.6 points per game. The NFL average, as we saw above, is 23.4 points per game, meaning the Eagles have been worse than average, as expected, but by a very small margin (less than 1%).

Stepping back, the offense has been slightly worse than expected, though I think we know why, and the defense has been slightly better than expected. Overall, though, the base-case projection looks to be pretty damn close.

The Benchmarks

I then tried to game out the season, and assign benchmarks for various portions. So how do the Eagles look when compared to the roadmap I set out? Well it just so happens that I set one of the benchmarks to the Bye week (naturally). If you check the link (up top), you’ll see that, for the team to get to 9 wins, I felt it had to have a record of at least 5-6 at the Bye week.

Of course, the Eagles have exceeded that by one win, and currently stand at 6-5.

Rather than re-hash how we got here, I’m just going to look forward. Things didn’t go exactly to plan, but the general path is far from what I expected. The key, of course, is what happens now.

In my pre-season roadmap, I called the section after the Bye week “The Dessert”. Based on the team projections, it looked to be the easiest section of the schedule, and I somewhat aggressively said that 4-1 would be a reasonable expectation of performance for the Eagles through this stretch. Has anything changed? Let’s go through them:

Cardinals – Their 6-4, but have a point differential of just +2. They’ve won their past 3 games, but have faced Atlanta, Houston, and Jacksonville over that span. They feature a good defense and a mediocre offense.

Lions – A bit schizophrenic, as usual. Hard to peg how they matchup against the Eagles. The downfield passing game (Stafford to Johnson) looks to be a terrible matchup for the Eagles, but it’s not as if the Lions’ defense has impressed (ranked 22nd by DVOA).

@ Vikings – Not a good team and it’s offense revolves around the rushing game, which he Eagles have had success against.

Bears – By DVOA, this is the toughest game remaining on the schedule. The Bears rank 5th overall by Football Outsiders. Similar to the Lions, the primary matchup concern for the Eagles will be a great, big, WR (Marshall).

@ Dallas – Already looks to be the key game of the year and has a good chance of deciding the division (and probably the only shot at a playoff spot). Lost to Dallas once already, so have to account for that. However, Dallas’ other wins have come against the Giants, Rams, Redskins, and Vikings. Not one of them ranks higher than 20th by DVOA. Call me unconvinced…(not that the Eagles’ win resume is that impressive either).

Remember that the Eagles outperformed my projection before the bye week, so the team only needs to win 3 games to get to 9 wins. Can it? Absolutely. Will it? I think so. Beating the Vikings means you need to got 2-2 against the Cardinals, Lions, Bears, and Cowboys. As far as I’m concerned, those teams all count as “mediocre”, though the Bears are on the upper edge.

If we break that up even further, the immediate question becomes:

Can the Eagles get one win from their next two games? (Cardinals and Lions).

If the answer is yes, then 9 wins still looks like a likely outcome. (and the answer IS yes)

Notes from Yesterday

Posted on November 19, 2013 by Brent

Just a few notes from yesterday’s game:

– I don’t understand Chip’s decision to dial down the offense in the second half. It makes complete sense to become more conservative and to take fewer risks when you have a lead (think equation). HOWEVER, when the opposing team is basically begging for you to take a shot, you should take it.

Early in the game, it was clear the Eagles were going to take shots downfield when they had Cooper matched up one-on-one with a DB with no safety over top. It almost led to an early TD (Cooper lost sight of the ball). Anyway, late in the game the Redskins were packing 8 in the box and playing a single deep safety. That means you’ve got both D-Jax and Cooper against a CB, and the safety can only help on one of them.

Somehow, a situation Chip was hoping for and targeting early in the game lost its appeal. Keep in mind that this is not a high-risk play. Throwing it deep to Cooper when he’s in single coverage is very unlikely to produce an outcome worse than an incomplete pass.

Given that we saw this exact same situation play out last time the team played the Redskins, after which Chip claimed he learned his lesson, I’m worried this will be a recurring issue. Obviously, that would necessitate having big leads, which would be awesome, but it’s still a bad habit. My only guess as to the reasoning is that Chip still doesn’t fully trust Foles.

– Overall a good win, but let’s remember that the Redskins aren’t a good team. We’ll learn a LOT more about the team when it faces Arizona and Detroit after the bye week. To date, the Eagles “best” win came against a Green Bay team playing with its 3rd string QB. It remains to be seen whether the Eagles rank within the “mediocre” division of the NFL. They’ve lost against Dallas and San Diego…which would suggest they’re at the bottom of that subset of teams. If so, they’ll have trouble against the Cardinals.

– Still researching the topic, but safe to say that Nick Foles is at least close to doing something unprecedented.

He now has a career rating of 97.6, with 22 TDs and just 5 interceptions. He’s also rushed for 3 TDs.

His rating this season is currently 128. The single-season record is 122.5 (Aaron Rodgers).

Under Chip Kelly, he’s seen significant playing time in 6 games…he’s won 5 of them.

His career interception rate is now 1.2%. The NFL Record for a career rate is 1.7% (Aaron Rodgers…yeah, he’s really good).

As I showed at the end of last week, few QBs have, at ANY point in their careers, had a career rating that as high as Foles does now. The fact that Foles has it 15 appearances and 11 starts into his career is a very good sign.

Naturally, it’s a safe bet that Foles won’t maintain this level of play. The next question, though, is:

What are the odds a “bad” QB could have a stretch of games like this?

How about a “mediocre” QB?

We could probably turn to Bayesian analysis to help out, but for now, it’s enough to know that the odds of either situation aren’t very good. When you then consider that fact that he’s doing it to start his career, I think it’s safe to say Foles’ odds are now pointing heavily in favor of at least “solid NFL starter” and potentially much higher.

– Last point. Chip was correct in going for it on 4th and 1. There’s just not much to gain from punting the ball there, especially in comparison to the relatively high likelihood of maintaining possession. I was much more concerned about the play-call. It looked like a delayed handoff, which would be an inexplicable call (especially to Bryce Brown). However, it may also have just been a miscommunication. Unfortunately, the announcers had already stopped calling the game and were too busy to bother talking about it. I don’t think we even got a replay. I’ll have to review the film, but my first impression was: right strategy, wrong play.

P.S. It’s week 12 (practically) and the Eagles are entering their bye week in first place.

Eagles Rewind

Objective and Analytical Analysis of The Philadelphia Eagles and the NFL

Eagles vs. Cowboys: Breaking down this week’s betting lines.

The Benefits of Being a High-Variance Team

SI Curse: Contrary Indicators 101

Mean Reversion and the 2012-2013 Eagles Improvement

Eagles v. Lions: Pre-Game Notes

Notes from Yesterday

QB Performance Frequency Distributions

Revisiting PreSeason Projections: Overrated Teams

Checking in with Preseason projections

Notes from Yesterday