East Coast Offense: So Far So Good!

So Far So Good!

I often cite the example offered by philosopher *Bertrand Russell to demonstrate the weakness of inductive reasoning:

It's like a man who jumps off the Empire State Building and counts the windows as he falls. When he gets to 80, he says, "So far, so good!"

It's a great way to illustrate that because something has happened many consecutive times in the past (falling past a window without any negative consequences), does not mean it will continue to happen in the future. While it's self-evident in Russell's example, it's easy to make this mistake when trying to handicap NFL games or deciding who to start for your fantasy team.

A good example was this week's Chiefs-Giants game. Andy Reid was 16-2 coming off the bye week and facing a team that had reportedly quit. Not only was the quitting narrative plainly false, but maybe Reid's prowess as a game-planner was exaggerated too. If we consider there are 32 teams, each of whom has a coach, it's likely some coach over the last 10 or 15 years has done better than expected off a bye week. Maybe not 16-2 (or 13-5 against the spread), but if you consider Reid's teams have been well over .500 for that stretch, he'd be expected to win 10 or 11 of those games, simply by keeping his usual pace. That he'd win a few more over a small sample isn't all that surprising, and this is before even looking into his opponents' collective strength.

Similarly, the Giants finally prevented a tight end from scoring a touchdown, and it's odd they stopped Travis Kelce, the best one they faced. (Kelce did have 109 yards, but on 14 targets, so he got a modest 7.8 YPT, well below his season average of 9.0.) The odds of a particular team allowing TDs to TEs for 10 straight games (dating back to Week 17 of last year) are very low, but the odds of one team of 32 doing that are a little less so. And the odds of one team of 32 having a 10-game TD-allowed streak to some particular position (TE, WR1s or slot receivers, e.g.) are better still. And the odds of some team having some quirky pattern of allowing something, e.g., the Patriots six-straight 300-yard passing games allowed to start the year, are even greater.

The point is among 32 teams and the hundreds of stats we track, there will be patterns, and we will identify a lot of them. It's tempting to conclude there's meaning in the pattern, i.e., predictive power, and in some cases there might well be. But like Russell's jumper, we need to be wary of relying on them. Last week's instance might be the window on the first floor.

This also ties into the betting ideas from last week's columns. I ran them by Rufus Peabody (who happens to be in Lisbon this month) of Massey-Peabody, and his advice for the second one (figuring out what conditions portend upsets) was to have a theory as to their causes before going into the research. His concern was if I picked out the 100 or 150 games with the most unexpected outcomes and mined them for the most common conditions I'd find something, but even if it were very unlikely by chance for that particular factor to be present across a large percentage of those games, it might be highly likely that *some* non-predictive factor would be there. Put differently, it's easy to find a false positive when you're screening for any possible factor. But if you were testing for only one or two factors you've hypothesized ahead of time, and one of them turns out to be present in a large number of upsets, it's much more likely to be predictive.

If and when I get around to the research, I actually think I'd do it both ways: (1) Go in with a theory about which factors are likely to portend upsets; and (2) Screen for anything that shows up in a disproportionate number of major upsets, pretend it's predictive and track it going forward. If it proves predictive for a season - or whatever we deem a sufficient sample, it could as a parachute for my jump, so to speak.

*I actually can't find anything online attributing this example to Russell, but one of my college professors did when he told the story - or at least that's how I remember it. Unfortunately, I can't check with him because he's dead.

DFS Automatic

We had Davis Mattek on our SXM show today talking about the Week 12 slate, and I asked him if he liked Sammy Watkins with Robert Woods out. He replied he did not because he felt it was presumptuous to think we can know what teams will do after a starter goes down. He preferred instead to target players with firmly established roles. He gave the example of DeAndre Hopkins, who gets 40 percent of the Texans targets, last week despite drawing Patrick Peterson and catching passes from Tom Savage. In Davis' view, opportunity trumps efficiency.

I don't doubt he's correct - the guy is a far more serious DFS player than I am, and the difference between getting 10 targets and five is much larger than getting 9.0 and 7.0 yards per target. A pedestrian receiver getting 10 targets, should get 70 yards, while an elite receiver getting five is looking at about 45. Volume is the bigger differentiator.

But in a large-field tournament, which most people shouldn't play, given the rake and the huge percentage of the payout going to the top few spots, I like to find high-efficiency players whose opportunities might spike - if game flow and coaching whims cooperate. As such, I'm in on Watkins this week (especially if Marshon Lattimore is out for the Saints) and also Martavis Bryant who saw more work last week and could see a boost with Juju Smith-Schuster banged up. My third player in that vein is Hunter Henry, but unfortunately he plays on Thanksgiving.

Week 11 Observations

• The Eagles annihilated the Cowboys who have been exposed the last couple weeks. And it's not only missing Tyron Smith – though his absence has been huge – or Sean Lee or Ezekiel Elliott. Dez Bryant simply doesn't separate anymore, and the Cowboys don't have other playmakers. Jason Witten is old and slow, Brice Butler is 27 and rarely sees the field, and Terrance Williams and Cole Beasley are just average. There's no danger for the defense. It also doesn't help when your moronic coach punts down 14 with 18 minutes left in the game on 4th-and-2 from his own 47 yard line. Jason Garrett is underrated in his terribleness.

• It's frustrating to root against Carson Wentz, as he has a knack for escaping what seem like sure sacks and making downfield throws. He was a game manager in this one, but did what he had to.

•Jay Ajayi is like Alshon Jeffrey – big-time talent, but one of many cogs in the diverse wheel that makes up the Eagles offense. Ajayi had 101 YFS, but only seven carries, while LeGarrette Blount had 13 carries and Corey Clement had six. Clement and Kenjon Barner scored the touchdowns. Because Ajayi is guaranteed neither a big workload, nor the goal-line carries, it's hard to make a case for him. He's not a great pass catcher, either.

• Jeffery led the team in receiving and scored a TD and a two-point conversion (thanks to a Jake Elliott injury), but had only seven targets. Zach Ertz and Nelson Agholor each had five and Torrey Smith four. Game flow dictated modest passing stats, but only Ertz and Jeffery are remotely reliable,

•Dak Prescott had his worst game of the year – four sacks, three picks, one lost fumble, 4.7 YPA, no TDs. The Eagles defensive front is stout, and he was without Smith at right tackle, but he played poorly even under the circumstances. The Cowboys need to threaten down the field, and right now they're not doing it.

•Alfred Morris had a solid game – 17 for 91, but the Cowboys never made it to the goal line, and he doesn't catch passes. Rod Smith – 8-for-11 – got stuffed.

•Dez Bryant had eight catches for 63 yards, but on 14 targets. The Cowboys don't target him down the field, and defenders were on him the moment he got his hands on the ball.

• The Patriots minus 6.5 were my best bet, and in retrospect should have been the best bet of the year. Their offense is humming, they faced a banged-up, bottom-of-the-barrel defense and were in an unusual site (high-altitude Mexico City), meaning the coaching staff's ability to get its players ready would likely be an even-larger-than-normal factor.

•Tom Brady had seven incomplete passes and took one sack. Otherwise, he had 339 yards on 30 completions, three touchdowns and no picks. For some reason the Patriots still had him on the field in the fourth quarter, and he took a shot to the knees, but seemed fine. It's a hell of a risk for Bill Belichick to take, but he seems to do it every year.

•Brandin Cooks torched the Raiders for 9-6-149-1 and now has 20 targets the last two weeks. He's a top-five WR.

•Danny Amendola, now healthy, also saw nine looks for a more modest 66 yards, but he caught eight balls and scored a TD. With Julian Edelman and Chris Hogan out, someone has to play white receiver for them.

•Rob Gronkowski had only three catches for 36 yards. He wasn't needed given the game flow, but I wouldn't be concerned about him going forward.

•Dion Lewis led the backs with 10 carries (for 60 yards.) Rex Burkhead and James White had five each. Lewis also led the backs in receiving with four catches for 28 yards and a TD. Burkhead had four catches of his own for 21 yards, and White failed to catch his only target. From Super Bowl hero to third fiddle.

•Stephen Gostkowski had a monster game including a 62-yard field goal in the altitude. That was worth 6.2 points in my NFFC league, i.e., he kicked a touchdown. But the Pats should have committed a delay of game penalty on purpose to give him a 67-yard try and the all-time record as the 62-yarder looked good by 10 yards.

• Stick a fork in Derek Carr. The Pats defense is playing better, but 4.8 YPA against them in good weather on a neutral site is terrible. His fantasy line was mildly redeemed by a garbage-time TD to Amari Cooper, but Carr's average at best with an average-at-best supporting cast.

•Michael Crabtree saw 11 targets, but had only six catches for 51 yards. Cooper saw seven targets but had little to show for them except the late TD. Seth Roberts had nine targets for 36 yards.

•Marshawn Lynch ran well (11 carries for 67 yards), but as usual saw a modest workload. He also caught a pass for 10 yards.

•Andy Dalton eked out 6.2 YPA and three TD passes (with no picks) in Denver, an upgrade from what he's done for most of the year. A.J. Green salvaged an otherwise bad game (nine targets, 50 yards) with a late TD.

•Joe Mixon has been a paragon of inefficiency this year, and that didn't change against a stout run defense – 20 carries for 49 yards, two catches for zero yards. It's his job at least.

•Brock Osweiler didn't lose the game, but he did nothing to win it, either. Expect Paxton Lynch to see an audition as soon as he's deemed ready.

•C.J. Anderson and Devontae Booker split carries, with Anderson scoring – and also fumbling in a key spot. Booker had five catches for 54 yards, though, to Anderson's two and 19. Jamaal Charles was a non-factor.

•Emmanuel Sanders had eight targets for 15 yards. Demaryius Thomas caught five of nine for 64 yards and a TD.

• I had a good fantasy day, in large part because I picked up the Chargers defense everywhere. I knew they were good, and they'd be a tough matchup for a rookie QB on the road, but five picks in the first half and two defensive TDs were a nice bonus.

• It's inconceivable Sean McDermott would start Nathan Peterman next week. If he did, I'd expect the team to mutiny, and he knows that. Benching Tyrod Taylor was obviously a mistake, but sticking with the decision for Week 12 would be coaching suicide.

• Somehow LeSean McCoy had a big game in the blowout, 126 YFS and two TDs. Kelvin Benjamin got hurt again, and the Bills just don't have many good alternatives. Zay Jones led the team with four catches for 68 yards.

•Philip Rivers had a solid game, thanks mostly to Keenan Allen who went 13-12-159-2. Allen won't beat many teams deep (his long was 29-yards), but he's the AFC's Michael Thomas – a player who will catch most of what's thrown his way.

• Top-10 draft pick Mike Williams had eight targets, though only five catches for 38 yards. His involvement in the offense is growing, however. Hunter Henry caught only two of five targets for 25 yards.

•Melvin Gordon saw 20 carries, but again was less efficient than Austin Ekeler. I doubt their roles will change much by season's end, but maybe Ekeler goes from 30/70 to 35/65. Ekeler is also the better receiver.

• I have to admit 7.2 YPA, two TDs, one interception and one sack, Tom Savage's numbers Sunday, were more civilized than I expected.

• If we're grading on a curve DeAndre Hopkins gets an A for his 9-4-76-1 line with Patrick Peterson defending and Savage throwing the ball. Bruce Ellington was useful in PPR with six catches on seven targets for 63 yards.

•D'Onta Foreman was unstartable, put up a monster game and now is likely out for the year with a torn Achilles. Lamar Miller, who struggled (22-for-61) has even less competition for the job.

•Blaine Gabbert put up good fantasy numbers – three TDs, 257 yards, 13 rushing yards – and his 7.6 YPA was solid. But he threw two picks and took two sacks.

•Adrian Peterson was throttled – 14 carries for 26 yards, and he offers little as a receiver. His upside the rest of the way is modest even if David Johnson doesn't return.

•Larry Fitzgerald had a strong game, catching nine of 10 targets for 91 yards and a score. His long catch for the day was only 22 yards, though.

• Someone named Ricky Seals-Jones, a backup TE, caught two TDs. He probably played with Gabbert during the scrubs portion of practice.

• After a monstrous start to the year, Kareem Hunt has crashed to earth. He's probably still a top-12 player the rest of the way, but the Chiefs look like the team we thought they were before they beat the Patriots in Week 1.

•Tyreek Hill caught all seven of his targets, but for only 68 yards. I still doubt Hill's ability to be effective running intermediate routes.

•Sterling Shepard was scratched with an illness, Evan Engram caught only one of six targets, and Eli Manning inefficiently spread the ball around to 10 different receivers.

•Orleans Darkwa had 20 carries and saw four targets. It wasn't pretty – only 90 YFS, but he scored the game's lone TD. He's the unquestioned full-time starter.

•Drew Brees finally had to make some plays, and he did to the tune of 385 yards (9.4 YPA) and two TDs. He did throw a pick and took two sacks, but led the Saints to an overtime win.

• The Saints receiving tree is still fairly narrow even with Coby Fleener getting involved this week. Michael Thomas, Ted Ginn and Alvin Kamara saw targets on 26 of Brees' 41 attempts.

•Mark Ingram (155 YFS and a TD) and Kamara (122 YFS and a TD) are both top-10 PPR backs right now. Ingram has a case for No. 1 overall.

•Kirk Cousins had a monster game (10.2 YPA, three TDs, zero picks, two sacks) on the road against an improved Saints defense. He also finally got something from an outside receiver as Josh Doctson had four catches for 81 yards.

•Samaje Perine broke through with 23 carries for 117 yards and a TD and one catch for nine yards. With both Rob Kelley and now Chris Thompson (broken fibula) likely out for the year, Perine should see a heavy workload.

• Funny that there was even talk of replacing Case Keenum (7.4 YPA, one TD, zero picks/sacks) with former game-manager Teddy Bridgewater.

•Adam Thielen broke out last year with a whopping 10.5 YPT, but he's kept up his elite efficiency with 9.5 YPT this year.

• After it looked a few weeks ago like Jerick McKinnon was clearly the back to own in Minnesota, Latavius Murray has crushed it the last couple weeks, scoring two more TDs and running more efficiently. McKinnon is still the pass-catching option, however.

•Jared Goff simply couldn't get it going in Minnesota. He didn't turn the ball over, but he didn't complete a single pass of 25 yards.

•Todd Gurley had a modest game, though he had 15 carries, four targets and scored a TD.

• A healthy Ryan Fitzpatrick is far better than an injured Jameis Winston, and honestly I don't think there's much different between the two healthy versions. The same can be said for Matt Moore and Jay Cutler, incidentally.

• Congratulations for those who waited on Doug Martin. He's an average back running behind a weak offensive line.

• For God knows what reason Adam Gase started Damian Williams and barely used Kenyan Drake.

• Moore's favorite target is Kenny Stills who had 180 yards, including a long TD. Jarvis Landry also had 95 yards and a score, while DeVante Parker, Jay Cutler's guy, caught four of nine targets for 26 yards. Parker was shoved to the ground in the end zone on Cutler's first pick – it should have been an obvious PI – but it was uncalled.

•Mike Evans had 10 targets, while no one else had more than five.

• The Packers are terrible, but the Ravens defense is legitimately good – right there with the Jaguars.

• There are only three players really worth mentioning in this game: Alex Collins (20 carries, seven targets, four catches), Jamaal Williams (18 carries, six targets, four catches) and Davante Adams (10-8-126.) That Devante Mays lost a fumble is good for two reasons: (1) It solidifies Williams (whom I have in two places) as the team's feature back; and (2) it makes it less likely we'll have to spell check yet another version of D(a)(V)ant(e) any time soon.

•Mike Wallace, and Jeremy Maclin will have roles, but there's little upside given the way the team plays, and Danny Woodhead (six targets, five catches) should be useful in PPR.

• I keep taking the Browns ATS, and they keep finding Chargers-esque ways to blow the cover. This time, it was a fumble-six with a minute left. Still wasn't as bad as having CLE +5.5 in this game.

•Corey Coleman had a nice game in his return, especially given the opponent. He's clearly the team's No. 1 receiver, though it'll be interesting to see what happens if/when Josh Gordon comes back. DeShone Kizer, who also get points with his legs, could be an option in the fantasy playoffs.

•Duke Johnson had a nice game, but hurt his shoulder. Isaiah Crowell got stuffed.

•Leonard Fournette was a game-time decision but managed 128 yards (4.0 YPC) against the league's toughest run defense. He said his ankle will be an issue the rest of the year, however.

•Marqise Lee had nine targets, Dede Westbrook six. Neither did much with them, but Westbrook seems like the team's No. 2 WR.

•Jordan Howard has quietly lived up to his early-second round draft slot. Tarik Cohen, who had nine carries and four catches, also scored. Adam Shaheen caught all four of his targets, including a TD.

•Matthew Stafford had 299 yards and 9.6 YPA but spread the ball around, with no receiver having more than seven targets. Marvin Jones led the team and caught a TD, but Eric Ebron (four catches, 49 yards) was also involved and is in danger of becoming just relevant enough to let you down again.

• Did Pete Carroll feel bad for DeShone Kizer? Hard to explain faking the field goal at the end of the half otherwise. Especially the play design which would have got the ball into Luke Willson's hands with 30 yards to run. How slow does he think Atlanta's special teams are? In Carroll's defense, Tyler Lockett was gashing them for 40-yard returns with ease every time they kicked off.

• Wilson had 258 yards passing, two TDs and 86 yards on the ground and another score. That equates to 430 passing yards (remember rush yards are worth double) and 3.5 passing TDs. It's why Wilson should be ranked ahead of Tom Brady going forward.

• The Seahawks still haven't found a solution at running back. J.D. McKissic and Mike Davis both did more in the passing game than on the ground. If I had to pick, I'd take McKissic who got six targets.

•Jimmy Graham caught 7-of-11 targets for 58 yards, a TD and a two-point conversion, but he saw at least three looks from inside the five and a deep shot into the end zone on a Wilson scramble that he should have caught. The way they post Graham up from in close, don't have red-zone wideouts and don't run much at the goal line means Graham is the favorite to lead pass catchers in TDs the rest of the way. Even Wilson, who scored a short TD, doesn't do a lot of designed runs from in close a la Cam Newton.

• The wideouts were quiet except for a 29-yard Doug Baldwin TD on a free play after the Falcons had jumped offsides.

•Matt Ryan had only 195 yards, but on the road in Seattle, albeit against a defense at far less than full strength, he played well, getting 7.2 YPA, two TDs, one sack and no picks. Ryan made a few key third-down conversion throws under pressure too.

•Tevin Coleman is explosive, but doesn't look like the natural between-the-tackles runner that Devonta Freeman is. Terron Ward ran harder and with more wiggle.

•Julio Jones had 10 targets, a couple great catches, one bad drop and even one failed fade in the end zone, but wound up with a modest five catches for 71 yards. He might have a monster game in him at some point, but timing it has proved to be awfully hard.

• What a weak kick by Blair Walsh. A 52-yard attempt is like a 45-yarder 20 years ago – you can miss wide, but you can't come up short.