Wednesday, February 10, 2010

Back!

I wanted to check back in briefly; I'm still alive, thankfully, and here's a new post. I want to tell you about Lithuanian painter Mikalojus Čiurlionis. The man was also a composer, and titled many of his paintings in rather music-flavored fashions (he's painted some Preludes, some Sonatas, some Fugues, for example). His intent was also precisely to make his art rather musical.

Anyway, one of his best-known works is Calm (or Tranquility), depending on who you ask, which is pictured above. It's inspired in large part by Isle of the Dead by Arnold Böcklin, the Swiss Symbolist painter, one of the versions of which follows. ("The time of Böcklin" is also referenced in Dramatic Symphony, a rather trippy work by Andrei Biely, which I'm currently reading. Clearly, Russians of this period had very consistent tastes, and one of those tastes was Symbolism.) (Another parenthetical: for the uninitiated, Symbolism is probably best approached through one of the French Symbolists, like Verlaine. The Russian Symbolist poets borrowed a lot from the French.)

Isn't that a glorious work? I really like how it treats light, and particularly how the interior of the isle rejects all of that light. I love objects that seem to radiate shadow as others might emit light.

That's it for a day or two, unfortunately, this is a busy week between work, fellowship applications, and quizbowl. I'm definitely going to make a point of updating frequently in the future (oh, how often that is said!) but I don't think I can maintain a daily schedule anymore.

Saturday, January 23, 2010

AFC Championship Preview

Another regular season game. That said, unlike the rematches we've already seen before, we have a perfectly predictive sample: it's not a week seventeen game where most of one team was sitting. It's a week sixteen game where the Colts stood up, then laid down. The net: three touchdowns left on the field, Colts still leading when the Curtis Painter Show began.

So, like, let's be frank. If the Colts do precisely what they did last time, and the Jets do precisely what they did last time, then the Colts will win. This is a Fact. Moreover, if we perturb the specifics of what happened last week, the Colts will win by more. A lot of the ticky-tack things where a player uncharacteristically misses his cut by an inch, slips, misjudges the ball or the receiver... those happened to the Colts. Colloquially, they got unlucky. And so if they do the exact same stuff, then we have to expect that they'll be luckier, and win by more.

Still, I'm hesitant to say that this is a done deal:
  1. The Colts will have different personnel on the field. With Mathis and Freeney playing all the time, they'll be able to maintain the same amount of pressure on the quarterback with fewer blitzes (or get more pressure with the same amount of blitzes). In the latter case, ceteris paribus Sanchez will just get sacked more. In the former case, he'll have to read a field full of more defenders, and get sacked just about as often (or perhaps more, still, because he'll see no one open and hold on to the ball).

    But this isn't a done deal! Pressure generated from linebacker and safety blitzes develops differently from pressure due to defensive line play. Perhaps Sanchez reacts better to the latter. Perhaps he perceives pressure less in the latter case, allowing him to feel more comfortable in the pocket, and read the defense, though denser, better. We don't have enough data to predict how Sanchez will react. It's safe to assume that he'll fail, but, you know, you can't be sure.

  2. A pass-rush-happy defensive line is vulnerable to the run. What would Shonn Greene like more than to run past a pass-mad Dwight Freeney? And, look, the Colts were 29th in Adjusted Line Yards. (That said, they did pretty well in power situations (third or fourth down, two or less to go), so it's possible that once the Jets fall behind and start passing, the running game won't be able to pick up short first downs with good enough regularity.)

  3. About that vaunted Colts pass rush, well, let's be honest: these aren't the 2005 Colts, which won a lot of games with defense and had the second-best adjusted sack rate in the league. That season (and 2004: 7th, 2000: 10th) has defined Freeney and Mathis as great pass rushers; the fact of the matter is that they've been positively middle of the road this year and were worse than that the past two years.

  4. The Jets defense will probably be mostly helpless against Manning, modulo the drops and mistakes that happened last time in spades. They can't consistently make him miss because his blitz identification is second to none; they can take away his number one weapon but he has four or five perfectly good weapons.
This is almost certainly a Colts win. But there's a sliver of hope, mostly featuring the Colts defense folding and Peyton Manning having another bad day. That's football.

Friday, January 22, 2010

NFC Championship Preview

Let's look at the past three years of the NFC.

Going into the playoffs after the 2008 season, the Panthers and Giants were the two teams to beat. The Falcons, with a dominant ground game and a hot rookie quarterback, were a popular sleeper pick. The Panthers went 4-1 in their last five, losing only to the Giants, while the Giants were 2-3, having lost three games by nineteen points total to two playoff teams and the Cowboys. The Panthers, 12-4 on the season, lost to the Vikings by ten, the Falcons by seventeen, and the Bucs by twenty-four, besides that Giants game; those latter three losses had featured three return touchdowns (fumble, punt, blocked punt). The Panthers seemed strong; the top-seeded Giants were wobblier. They lost in the divisional round too: to the Eagles, who topped them by twelve. In the divisional round, the number six seed, which had scored one point more than its opponents and lucked into an NFC West championship in large part due to the utter weakness of that division, beats the Panthers by twenty points. No return touchdowns, no nothin'. It was the worst loss the Panthers had suffered all year. Arizona races ahead of the NFC. Season type one: seeds one and two finish the season weak and strong, and a low seed beats them both.

2007: The Packers and the Cowboys are the class of the NFC. The 'boys beat the Packers, then barely beat Detroit (1) and Carolina (7), losing in the process to Washington and Philadelphia. The Packers lose to the Cowboys, then smash Oakland and St. Louis. Chicago crushes them, then they crush Detroit. The Giants wobble into the playoffs having gone 3-3 in their last six, notably losing to the Vikings (these are the Vikings with neither Favre nor Peterson, mind) 41-17. Then they win the Super Bowl. Season type one as well.

2006: New Orleans, the two seed, finishes 3-2 with a ten point and a six point loss to Carolina and Washington, respectively. (The former game sees Drew Brees not starting.) The Bears, the one seed, finish 4-1. They squeak by the Seahawks, smash the Saints, and take the NFC title. Neither wild card makes a peep. Season type three: two strong finishes by the one and two, and the one takes the conference.

This season doesn't fit the mold. Both the Saints and Vikings looked downright vulnerable towards the end of the season, according to real, quantifiable metrics. Shouldn't it be more likely that an upset would occur? Well, apparently not. Both teams entered the divisional round ready to prove that they deserved their seeds. Here are some factors that will make tomorrow's game a classic:
  1. Balance in the Saints' offense. The Vikings have a defense that is utterly impregnable against the run; the Saints rely on variation to mount an effective attack. Either Reggie Bush will really show his mettle by giving the Williamses something they haven't seen before, or Drew Brees will have to pass, pass, pass. Granted, there's more diversity in the Saints' passing offense than any contemporary pro offense. So if any offense could compensate for a shut-down running game, it's the Saints'. But that doesn't mean they wouldn't have to compensate, and I expect the manner in which they do will define the game.
  2. Vikings turnovers. The Saints, as we saw last week, are turnover-happy on defense. The Vikings have the potential to be a turnover machine: we have a quarterback who "just has fun out there" and a fumbling halfback. The Saints will try to force as many turnovers as possible in whatever way they can: last week, they didn't stop Arizona from moving the ball, in the final analysis, merely from scoring enough.
  3. Saints blitzes. Gregg Williams calls a six-blitz twenty-one percent of the time. This is an absolutely insane statistic, and it's crazy that Philadelphia blitzes more. The essential part is this: Favre does terribly against the blitz. Then again, he did great against the aggressive Dallas defense last week. It'll be interesting to look at who Favre's hot reads are, particularly on third down (principally Harvin and running backs) and whether the veteran has (finally) lost a step.
  4. Vikings pass defense versus Saints offense. The Vikings have a pass defense that sees the non-Cedric Griffin side get targeted all the time. The Saints spread the ball more than any team in the league. The Vikings are vulnerable to deep passes; the Saints throw deep more than any team in the league. The Vikings can only hope to get an effective pass rush to limit the Saints to shallower options. That said, they'll almost certainly do just that. And since some of the Saints' diversity in receiver targets is tied to their ability to throw deep, the strongest anti-Vikings aspect of their pass offense might collapse a little.
I won't pick a team for this; as the structure of my writeup shows, the way this game swings is contingent on too many factors we simply can't predict.

Thursday, January 21, 2010

Théodore Géricault

Many apologies that I've missed several days worth of updates; I've been moving back to Harvard and spending time at Brandeis and, on the whole, being too busy to blog. But I promise not to abandon you!

Théodore Géricault was a really important early French Romantic painter. He's best known for a painting called The Raft of the Medusa, which depicts a real-life event: a bunch of shipwreck survivors, who had been abandoned by their captain, drift on the title raft. Around the bottom left, many weakened survivors lie piled on top of each other; at the upper right, men twirl red and white cloth, trying to attract ships.

To the left is an earlier work, The Charging Chasseur, showing a Napoleonic cavalry officer. Compare the diagonal features of both works; it's a motif that shows up often in Gericault's work.

Monday, January 18, 2010

John Berryman

John Berryman's best-known work is his collection The Dream Songs, in which personas such as "Henry" and "Mr. Bones" address issues like his father's suicide. Here's the first:
Huffy Henry hid the day,
unappeasable Henry sulked.
I see his point,--a trying to put things over.
It was the thought that they thought
they could do it made Henry wicked & away.
But he should have come out and talked.

All the world like a woolen lover
once did seem on Henry's side.
Then came a departure.
Thereafter nothing fell out as it might or ought.
I don't see how Henry, pried
open for all the world to see, survived.

What he has now to say is a long
wonder the world can bear & be.
Once in a sycamore I was glad
all at the top, and I sang.
Hard on the land wears the strong sea
and empty grows every bed.

Friday, January 15, 2010

NFL Divisional Weekend: Sunday Previews

As I write this, I've sat down to start watching the Cardinals-Saints game. I can't wait to see what unfolds. But for now:

DAL @ MN 1:00pm ET

Vikings fans should be worried.

Brett Favre struggles against big blitzes. Dallas loves to send big blitzes. Brett Favre can't rely on Adrian Peterson to run consistently (if he plays like the second half of the season, anyway): Dallas will be able to send big blitzes without worrying about draws. Percy Harvin's game has been down ever since his injury, so a great passing option is downgraded and the Vikings will less frequently have the luxury of a short field.

On the other hand, Dallas is playing hot. Tony Romo took down the Saints, right? And they've nailed the Eagles pretty hard three times now, and the Vikings couldn't beat the Cardinals during the regular season, which makes sense as the Cardinals barely beat the Giants... Stop. Judging a team's trends through a web of head-to-head matchups is pointless. What's important is that Romo is playing efficiently and comfortably, in no small part because he has a well-developed running game to rely on and a suite of draws, screens, and play-action passes to fend the defense off--

And that's where this matchup is equalized again. It is difficult to run against the Vikings; that's indisputable. It's so difficult that the choice Dallas faces is to ensure that multiple drives die or to let the run become a secondary, supplementary option. Once Dallas becomes one-dimensional, Dallas starts losing games. Dallas starts playing the way it did against the Giants. In one game, Tony Romo throws fifty-five passes, has a reasonably efficient day, and loses: an incomplete pass on third and two on their first drive to Bennett is pretty representative. In the other Giants loss, the first drive ended... on an incomplete pass on third and two to Bennett.

Even though these are two top-flight offenses, I'd expect them to be moderately unproductive on Sunday for the above reasons: Favre won't have enough time to throw, and Romo won't have enough alternatives for the Vikings not to go all out after the pass. I'll still take the Cowboys, in large part because their side has more creative solutions available, by seven.

NYJ @ SD 4:40pm ET

This is going to be a game of matchups.
  1. Jets CB Darrelle Revis against... someone.
    Darrelle Revis has been in the league since the 2007 season; he had declared for the draft after his junior season at Pitt. By the end of the third season, he has turned himself into the best cornerback in the league. Of course, last year most people would give Nnamdi Asomugha that tag; it's probably fair to say that any judgments made on Revis now are colored by the hype that's surrounded him this season. It's unquestionable, though, that Revis is one of the elite cornerbacks in the league.

    It's hard to say what San Diego will try to do offensively to try to account for Revis. Their running game is not nearly reliable enough: ranked by DVOA they're last in the league. This comes because of poor RB play, in part. LT is the forty-first most efficient RB in the league; Darren Sproles is running poorly (and infrequently). Their offensive line is below average at run blocking, but it's not the problem.

    The Chargers line has a 4.5% adjusted sack rate, good for fifth in the league. Clearly, the Chargers are going to be passing if they're going to win. It goes without saying that they won't be throwing at Revis. Conveniently, San Diego has two or three primary options: Vincent Jackson, Antonio Gates, and Darren Sproles (who, despite his terrible running stats, is the most productive and second most efficient receiving RB in the league), which is one or two more than the number of Darrelle Revis the Jets have.

    Unfortunately, the Jets don't get the best defense in the league by shutting down a team's top receiver and letting everyone else run fancy-free. In fact, the Jets are second in the league (behind the Houston Texans?!) in shutting down passes to RBs. If these guys have a weakness, it's passes to #2 WRs. So the key to the Chargers' success is putting the game in the hands of Legedu Naanee and Malcom Floyd... or playing Gates at wide receiver. The latter seems safer.

    I wouldn't be surprised, actually, if they let Revis pick his man, play-by-play. That would allow them to disguise coverage to a certain extent, for one thing; it would also prevent the Chargers from knowing who's more likely to be open.

    That said, if Revis has historically been vulnerable to anyone, it's been tall receivers. He's the tallest Jets corner, too. I wouldn't be too surprised if Philip Rivers targets his man aggressively towards the beginning of the game, just to throw him off. If nothing else, it'll give him an idea of what he'll need to be able to do in later drives.

  2. Jets RB Shonn Greene and RB Thomas Jones against the Chargers run defense.
    The Chargers are a weak run defense, too; they're twenty-fifth in the league. Cincinnati was none too strong either, granted, so I wouldn't expect a much better result than last week's, but I think the Chargers are more affected by injuries to their defensive front (relative to the majority of the season, during which they built up that rating) than Cincinnati was last week. The Chargers are great against right tackle runs, fifth in the league (while other positions they're in the twenties). That negates the biggest Jets advantage, leaving them only middle runs and left tackle runs.

    I wouldn't be surprised if an unstable run game leads the Jets to favor short passes to their RBs, both of whom can run through or around defensive backs. That way, they build Mark Sanchez's confidence, avoid the teeth of the Chargers defense, and get the ball to their biggest offensive playmakers.

  3. Chargers OL against the Jets pass rush.
    Philip Rivers relies on the deep ball more than perhaps any quarterback but Drew Brees. As such, he needs time more than any other quarterback. The Chargers have a tremendously strong offensive line, but it's been hit by injuries and inconsistent play lately.

    The Jets won't let them have it. They're only an above average pass rush in terms of sacks, but their rush is much more than that: it's more about hurries and hits, and, in the end, preventing complete passes. They do that excellently, because they frequently don't need to worry about leaving receivers in man coverage. So they'll be able to rush more men, get more hits, and stop more passing plays.
What do I see? I see a Jets victory by three.

EDIT: Jimmy Johnson just said that in Tampa, the Saints went up 17-3 and "had the game won," thus exculpating them from the fact that they, uh, didn't win. What?!

NFL Divisional Weekend: Saturday Previews

Last week, we saw one incredibly exciting game (Packers-Cardinals), one end of an era (Eagles-Cowboys; I don't think the Eagles will be thought of as a playoff regular going forward, particularly if continued skepticism of Reid-McNabb leads to losing one of them), one defensive slog (Jets-Bengals), and one defensive whuppin' (Ravens-Patriots). This week looks like it'll be very different: the divisional round has a lot of strength vs. strength matchups, and I wouldn't be surprised if the result is a lot of one-sided games, determined from the outset by whose strength is... stronger. Anyway, on to previews.

ARI @ NO 4:30pm ET

Week 12 was the high-water mark for the Saints. They had come off a 38-17 win against the Patriots. In that game, Tom Brady threw two interceptions and no touchdowns. The Patriots never converted good field position into points (Welker had a forty-one yard punt return; zero points) and their two touchdowns came at the end of eighty and eighty-one yard drives. Drew Brees threw as many incompletions as touchdowns (five). I can't imagine there were many people the following Tuesday who really expected a non-Saints team to represent the NFC in February.

That would have been pretty reasonable, on the whole. The New Orleans story up until that point was an explosive, unpredictable offense based on the zany principles Sean Payton cooked up. That offense was complemented by an unforgiving, turnover-forcing pass defense, led by Darren Sharper's remarkable ability to return interceptions for touchdowns. The challenge for opponents, essentially, had been being content to be run-heavy even while the Saints offense could respond, seemingly automatically, with a touchdown drive where Drew Brees (with the assistance of a two yard run by Pierre Thomas) goes 4/4 for 74 yards.

Then the Washington game happened: a game the Saints could have lost if Washington had had a kicker worth anything. Let's put this in perspective: the Saints never sacked Jason Campbell. Our man JC threw only one interception. Drew Brees was forced to throw the ball forty-nine times. Early on, the Saints tried running on third and short and failed; then they switched to passes that nfl.com lists as "short middle," and Washington figured that out. This was the game that the Saints' weakness was revealed: drop enough men into coverage, and they can't do a thing. (Here's another wrinkle: this game could well have been lopsided in favor of Washington if not for the Brees interception that Meachem stripped to return for a touchdown and if not for a bizarre muffed punt by Washington.)

So the Saints had a weakness, but the news didn't spread quickly enough. They barely beat Atlanta in a game that saw Brees have a fantastically efficient day (to the tune of 31/40, 3TD). Would this game have gone differently had Atlanta played more like Washington defensively? Maybe. Then the loss to Dallas, and the Saints looked just mortal.

I think this result was prefigured by the Saints-Eagles game in Week 2, which featured Kevin Kolb at quarterback racking up 391 yards. Why this wasn't a cause for concern: the Saints still won by twenty-six. Also, Kolb threw three interceptions, one of which was returned for a touchdown. So were the Saints in this game bad at giving up yardage compared to how good they were at keeping it from it becoming points? Yes, essentially: both touchdowns were due to big plays (DeSean Jackson catch, Ellis Hobbs punt return), though there were two field goal drives of over sixty yards. Forcing turnovers help that.

But this is Kevin Kolb. And, indeed, this defensive hiccup came back later in the season. Over the whole season, the Saints gave up 31.01 yards per drive, good for twenty-second in the league (Kansas City: better. Seattle: better). Their defense was, indeed, buoyed by their ability to produce turnovers (third in turnovers per drive, with 13.9%). (Both of those statistics, by the way, are thanks to http://footballoutsiders.com.) That fell apart in their Dallas loss (no turnovers, and 439 net yards, 43.9 yards/drive) and their Tampa Bay loss (two interceptions for a swing of at least ten points, but 439 net yards, 48.8 yards/drive, and Tampa Bay, everyone: Tampa Bay and a rookic quarterback in one of his first starts). That's what I see here: New Orleans can't force turnovers as once they could, particularly not against playoff teams. And that will come back to bite them against Kurt Warner.

I come to the Cardinals, and I have little to say that hasn't already been said: we saw what they could do against a thought-fantastic pass defense last week, and I shudder to think what they might do against a once-fantastic one. Now, of course, last week was a statistical anomaly: the Cardinals caught their fair share of breaks, in which a defensive back chose to play inside instead of outside technique, or shaded his coverage to the left instead of to the right, or whatever. So I don't think that they'll put up another forty-plus offensive points, even against a defense that I think is worse than Green Bay's. (It certainly doesn't have the linebackers.) The last thing you want to give Kurt Warner is time to throw, and the Saints will have little choice but to do just that. The Cardinals don't operate well when they're just dropping men in coverage, as we saw last week; if Aaron Rodgers can read you, I bet Kurt Warner can too. So it'll be an exciting, high-scoring game, but I'll pick the upset here. Arizona by ten.

BAL @ IND 8:15pm ET

This game happened during the regular season, much like the whole damn wildcard round. That game was all manners of anomaly: Peyton Manning made bad reads and forced throws, the Ravens tried to run Ray Rice out of passing sets, nfl.com tries to get me to click "WATCH HIGHLIGHT" on each of the game's six field goals.

It's tough to draw inferences directly from that game, and it's probably not a great idea. The Ravens offense knows what it is now, and what it is is running Ray Rice. The Colts won't be able to stop him any better than the Patriots did; indeed, they'll do a worse job: the Patriots have about a 0.2 yard edge over the Colts in adjusted line yards, per FO. (Also, Rice's style doesn't match up well against the Colts: they can stop smaller backs much more effectively than larger backs.)

Strangely, the Colts would be predicted to do better defending passes to Rice; they're eleventh, versus the Patriots' twenty-sixth, against passes to running backs, though they give up comparable yardage. This, even though the Pats didn't give up a Rice catch last week. So I guess it's possible that Flacco loses this game, too, by throwing an interception into triple-coverage again.

Flacco's mechanics are not going to be the best this game, due to injuries. The Baltimore offensive line prevented Mathis and Freeney from getting a sack last time; they may well do it again, but giving Flacco time might simply not be worth it: the secret to this game for the Ravens is the ground game.

Let's be honest about what will happen when the Colts have the ball: Peyton Manning will complete passes frequently and drive down the field easily. The Ravens don't have the pass rush they did in the past, despite what happened last week. And the only way for a pass rush to be particularly effective against Manning is when it is very, very fast: the '05 Steelers beat the '05 Colts in the playoffs because Manning didn't have time to get to a hot read with any kind of accuracy (and because that rush could come quickly without sacrificing much in terms of coverage; the Steelers six-blitzed three or four times, but other than that got fast pressure with four man zone blitzes). So Manning will, in all likelihood, have almost as much time as he wants to scan the field. And the Ravens have such a depleted secondary (Frank Walker and Domonique Foxworth are starting) that receivers will be open.

Really, this reminds me of the Indianapolis-Miami game. Indy moves the ball at will, but only holds it for fifteen minutes. Miami gains about four yards a minute, takes eighteen plays to move fifty yards, whatever. Forty-five minutes of ball control, and a loss because Indianapolis can match Miami's pace. That's the thing with time-of-possession: if your offense is just as skilled as your opponent's, then it usually means diddly to get a high ToP. It doesn't mean diddly if:
  1. The reason you have a high time of possession is because you have more possessions, via turnovers. So your equally talented offense turns more possessions into more points.
  2. You have the offensive flexibility to stop running the ball all the time and start passing once the defense gets winded
  3. Your ball-control offense doesn't get a high time of possession by running a bunch of eleven play drives that end in punts because they go thirty-nine yards. Drive conversion into points is still the name of the game, and if your offense doesn't have a good selection of plays that can work as a guaranteed one or two yards, then it cannot run the ball more than half the time or it will fail to score points on that drive. If you fail to score points and end up behind in the score, then every second you run off is also one you needed to come back.
The Ravens obviously lack the second criterion. They might have the first. I think they have diverse Ray Rice plays to satisfy the third criterion, but the Colts are good against running backs. I really don't see this working out any better than it did for Miami.

I wish the Ravens could win this one, just because a Ravens-Jets AFC Championship would be literally unwatchable by anyone but me (and I like watching football alone or with Hannah or my parents, not with crowds). And I think there's a fine chance that they pull it off. I just think that's only a bigger chance than the no chance at all most people give them. I'll pick Colts by fourteen.