Possibly the closest Super Bowl matchup in history also poses the question: how much does it mean when certain aspects of an NFL team improve dramatically in the second half of the season?
21 Sep 2011
by Tom Gower and Mike Kurtz
Tom: With 11:16 to go in Sunday's game, San Francisco 49ers coach Jim Harbaugh sent David Akers out to try a 55-yard field goal. Akers made the field goal to give the 49ers a 24-14 lead. On the play, Dallas linebacker Keith Brooking was flagged for Leverage. Harbaugh elected, controversially, to take the made field goal and a kickoff from midfield rather than a first-and-10 at the 22-yard line. This decision was noted in Audibles and elsewhere, as the 49ers could have run an additional 2:00 or more and maybe even scored a touchdown had they accepted the penalty. The Cowboys eventually forced overtime with a field goal as time expired, and then went on to win the game in overtime. I also recently read this article, on what (probably minimal) influence David Romer's piece on going for two had on actual decision-making by coaches. I ended up with a very natural question: at what rate for two-point conversions would you essentially always go for two instead of kicking an extra point?
Mike: One would think that whenever the conversion rate went over 50 percent, since that would naturally mean you would get more than one point, on average, on your tries.
Tom: Yup, and not every extra point is made. So if you knew you had a 50 percent chance of converting a two-point conversion, you should go for it virtually every time. Yet there have been years where the league as a whole has converted two-point plays over half the time, yet teams didn't go for it every time. Heck, in 2006, 60 percent of called two-point conversions were converted. Mike, are NFL coaches just crazy?
Mike: No, they are terrified of looking bad.
Tom: I agree with you -- they're not crazy. I have a slightly different take though: NFL coaches, for reasons we can guess at, don't like variance.
Mike: I dunno, I like my answer better. I doubt there is a well-reasoned argument from the coaches, but rather the very real sense that anything they do that can be turned into a narrative of why the team lost, will.
Tom: True. The reason we're talking about why Harbaugh's variance-avoidance decision looks bad today is because his team lost.
Mike: Well, Harbaugh's situation is a bit atypical. He turned down a new set of downs.
Tom: The 49ers lost. He could have made a different decision that might have changed the outcome, so he looks bad. Coaches should thus be absolute win probability-maximizers then, because any decision in a game they don't win becomes part of the losing narrative. Why don't they listen more to stats people?
Mike: I disagree. Coaches are narrative-averse, which is very different. A coach that loses 21-23 could have won if he went for two after all three touchdowns. Nobody will criticize that coach, yet clearly he made decisions that harmed his team's ability to win the game. Harbaugh is subject to extra criticism because the perceived risk was so low yet he still refused the opportunity of a new set of downs, the ability to drain the clock, and the possibility of a touchdown.
Tom: Well, I can see circumstances where taking the extra point would be criticized then, such as a touchdown to go down 23-13 (or more) in the second half. I kind of get the Harbaugh criticism, and can see the case for taking the penalty. I'm just not sure it's that overwhelming or even particularly clear. If you accept the penalty, you should be in the zone where field goals are made at a high rate. Using Akers' career stats for 30-39 yard kicks, there's a roughly 88 percent chance of making a subsequent field goal even if the ensuing possession gains nothing.
Mike: 88 percent is darn good odds.
Tom: You only burn two minutes off the clock if you call three running plays or there are no incomplete passes. I feel like some of the post-game criticism is harsher than it should be precisely because Dallas made the tying field goal as time expired. That was kind of misleading, as the Cowboys had the same field position with 1:00 and two timeouts remaining.
Mike: You don't need to burn it all off the clock, though, just as much as you can. You still try to get into better field goal range or a touchdown, and even if you can't do that, your risk is 12 percent.
Tom: Maybe I'm trying to overcomplicate things, and variance-avoidance merges with risk-avoidance at some point. The old mantra "you never take points off the field" is, I would say, clearly not always the right answer, but I'm missing the overwhelming upside in this case.
Mike: I just told you the upside: lowering the likelihood of your opponent scoring by draining the clock, the possibility of touchdown, and barring a touchdown, the possibility of a shorter field goal to reduce the risk. Which is already very low at roughly 12 percent.
Tom: I don't mean to suggest 12 percent is absolutely the right chance Akers had at making a subsequent field goal. That's just intended to be a ballpark estimate, and I'm sure Harbaugh had a better sense at Akers' range and accuracy than I did.
Mike: I doubt he really considered it, honestly. After all, he had his points. On the other hand, I'm not going to rip Harbaugh personally too much for this. He had a few seconds to make an extremely important decision.
Tom: In the post-game press conference, Harbaugh apparently expressed that he had confidence in his defense to get a stop, which obviously didn't happen.
Mike: Any coach will say that. Norv has said that.
Tom: I didn't mean to make this column about Harbaugh per se. I just wanted to talk about how coaches evaluated risk, and how I think they’re variance-averse as much as anything else.
Mike: While I think you may have something to your anti-variance theory, I think it's more applicable to strategic decisions rather than spot decisions like this one.
Tom: I think we have seen coaches respond to the Romer paper. There is an increasing awareness of the mathematical odds of going for it on fourth down -- more than the article indicates, anyway. It's just an incomplete evolution, and one that's likely limited because coaches aren't absolute pure win-probability-maximizers. I'm also not sure I get your distinction between strategic decisions and spot decisions -- every spot decision is a strategic decision, it seems. It's just most of them don't matter that much.
Mike: No, spot decisions are not part of any coherent or long-term strategy. They are simply a coach weighing what option he thinks is more likely to result in a win.
Tom: So you're talking more about tactical decision-making, which I see as strategic in-game decision-making?
Mike: Honestly, I don't think there's much time for strategy in the game itself. You just don't have enough time to weigh all the options fully.
Tom: Sure, and with the NFL's restrictions on things like iPads and laptops and calculators on the sideline, the coach has to make the decision on the fly with only his built-in heuristics to guide him.
Mike: You are the only person I know who would describe "gut feelings" as "built-in heuristics." And again, I studied engineering.
Tom: I have "gut feelings," too. They're just things like "I'm thirsty, I think I'll have another Diet Coke."
Tom: That's not something that, even for me, is particularly amenable and responsive to reasoned decision-making. This may seem odd for somebody who keeps bringing up decision-making, but I'm sort of off the idea that coaches should be absolute win probability-maximizers. Like I thought of the Belichick decision against the Colts, it wasn't clearly wrong, and there are a bunch of squishy factors any reasonably competent coach has a better handle on that could skew a close call one way or the other. It's one thing for me on the couch to yell and scream about things, but the coach is making many, many decisions in the lead-up to the game and during the game itself. While a guy like Andy Reid could benefit at times from a clock management guru, he's still doing a much better job than you or I or most people not employed by an NFL team would at the whole spectrum of his job tasks. So maybe we all just need to relax a little.
Mike: While I agree in general, to use the Reid example, he should simply hire someone to handle that. So while it is nitpicking, it's something eminently solvable.
Tom: I think the Herm Edwards-Dick Curl experiment says that's actually harder to implement in practice than you think, and Reid would be better off reading this book in the offseason. (No, I haven't read it and don't own it yet myself. I have too much to read already and am not actually a coach.)
Mike: It never ceases to amaze me that these books are actually published.
Tom: I knew chess wasn't the game for me when I saw a book titled something like "128 Essential Openings." That's a little far afield, but I want to hear from you, reader, at what percentage point would you as a coach be willing to go for the two-point conversion virtually every time?
Tom: Another week, another week where I had the league's highest score. This time, it was the other non-Staff League team.
Mike: And so modest, too.
Tom: Josh Freeman had a very average 14-point performance, but all three of my wide receivers were better than him. I also benched Roddy White for Stevie Johnson in a decision that paid off, and started Ryan Mathews over Ben Tate and Joe Addai. The biggest performer was Miles Austin and his 143 yards for three touchdowns.
Mike: Fitzpatrick is a strange situation. I think I might actually start suggesting people start him.
Tom: I have Stevie Johnson in a couple leagues. I'm not sure Fitzpatrick is a regular starter, but in the right matchup, sure.
Mike: The only downside is that most people with iffy quarterback situations have already nabbed him up, so the teams that could use a player like him are probably out of luck.
Mike: Your other team?
Tom: My other team skated out a victory by the slimmest of margins, 286.70 to 286.20.
Mike: To be fair, that is not actually the slimmest of margins.
Tom: Well, fine. I was just glad the Rams didn't run a play on fourth-and-28, because if Kenny Phillips had made a tackle, I would have lost. I was projected to win the game by 37 points, but my team underperformed and my opponent's team outperformed expectations. And it was easy to single out particular players for blame and praise: Reggie Bush for me and Denarius Moore for him.
Mike: Still you persist with Bush. How long are you going to keep starting him?
Tom: Well, this might be the last week in non-PPR leagues. He had a great performance in Week 1. The game was a bit closer than it should have been. I left a questionable Champ Bailey in the lineup instead of benching him.
Mike: The Broncos stick to their injury report pretty well. I'd never start a questionable Bronco.
Tom: I shouldn't have started him, I just didn't pay enough attention to his status.
Mike: Ah. One of the things I like about Yahoo! is that it has a little red letter next to the player for injury status. That makes it very easy to swoop in right before gametime and shuffle things around.
Tom: That plus the little box with an asterisk for recent news made it easy. I just missed him in my run-through.
Mike: That's an asterisk? I thought it was just the card being shiny. Weird.
Tom: Eh, I still skated my way through and am now 2-0 in that league.
Mike: I had somewhat commanding victories in both of my leagues.
Mike: Vincent Jackson's career day ended up being a double-edged sword, actually.
Tom: How so?
Mike: On one hand, he gave me 31.2 points in one of my leagues, but I was up against him in the other league, where my opponent received 34.2 points. In fact, Jackson was the main reason my second game was close. He accounted for nearly a third of my opponent's total points.
Tom: That kind of thing does happen sometimes.
Mike: That said, Philly Almost-Hero Jeremy Maclin did even better, with a 35.7. Darren McFadden's 27.8, Drew Brees' 23.0 and Mike Wallace's 22.6 all made up for the 0.00 I received from Antonio Gates and the 2.00 from Chicago DST. I think McFadden's going to have a very good year.
Tom: Oh, the league where I had the week's highest score? 2.00 from Bears DST prevented me from an even better performance.
Mike: Chicago was a disaster this week.
Tom: I'm still a little bit skeptical McFadden will have that good a year, but he has outperformed my expectations to date. Granted, two games and whatnot.
Mike: McFadden has big draw on PPR leagues, since he is a legitimate receiver out of the backfield. In non-PPR leagues I'm slightly less optimistic, but still see him as a solid RB2 if not RB1.
Tom: I guess I agree with that. I'm still not entirely comfortable thinking of players in terms of their fantasy production as opposed to how good I think they are.
Mike: As far as my first league, it was a pretty workmanlike effort. Once again, Brees and one other player (this week Jackson) led the way, supported by good but not fantastic days from the rest of the peanut gallery. Adjusting your thinking takes some getting used to. The end result is that I'm in second place in both leagues, although I'm not second in points in my family league and next week I have to actually deal with Robert Meachem and Lance Moore. I’m probably going to go with a healthy Moore, but I'm still nervous.
Tom: I'd go with Moore for sure, especially in a PPR league.
Mike: Although ... I have Julio Jones, who is clearly a target.
Tom: At least he was this past week.
Mike: In basically any coverage, as this week showed. He was the week prior, too, with five catches for 71 yards. That’s not bad for a WR3.
Tom: I mostly managed to restrain myself in Audibles, so I'll repeat the feat here. It's getting tough, though.
Mike: You see, dear readers? We can show restraint! Sometimes. When we feel like it.
Wagstaff's Ringers (Tom, 1-1) 78 def. Intentional Rounding (Danny, 0-2) 75
I'm going to have to check on those other league victories about which Tom keeps bragging, because Wagstaff's Ringers is a two-year-long train wreck. On the bright side, the Ringers didn't actually post the lowest total this week; merely the second-lowest. Still, a win is a win. Of course, Rounding isn't much better, despite having talent like Michael Vick (13 points), Hakeem Nicks (7) and Chris Johnson (6) on the roster. I'd say Bills DST (-4 points) was Danny's downfall, but Tom received a mighty -7 points from Chargers DST, which I think is actually a record for defenses that were actually started. Seppuku for everyone.
Reverse Jinxes (Elias, 1-1) 132 def. Known Chumpsky (Rivers, 1-1) 92
We go from the closest match to the biggest blowout. Everything went right for Elias this week, from Tom Brady's monster game (28 points) to Fred Jackson's monster game (25) to Miles Austin's super-ultra-godzilla game (32). Chumpsky simply could not keep up, much like the Ravens DST, which posted a mind-boggling -1 points.
Dyscalculia Plus Ones (Will, 2-0) 112 def. Edmonton Eulers (Tanier, 1-1) 83
The Dyscalculias continue to ride the Cam Newton train to fantasy victory, although there are some dark clouds overhead if the second half of Week 2 is an indicator. For now, Will's team will keep chugging along hoping to catch lightning from a cast of decent but inconsistent stars (Tony Gonzalez, Jets DST, Wes Welker). The Eulers, meanwhile, somehow did not predict that Rob Gronkowski was going to have a career day! I know, impossible, right?
That's Great Hustle! (Sean, 2-0) 145 def. Los Pollos Hermanos (Rob, 0-1) 123
Poor Rob drew the best performance of the week as his opponent, instead of, say, Tom. Both of these teams has great weeks, actually, Sean pulling double-digits from 6 of his slots and Rob from all but one. The real difference was the ceiling: Los Hermanos' top score was 21 points (Aaron Rodgers and Peyton Hillis), while Hustle had three players that topped that total (Matthew Stafford with 25, Jahvid Best with 23 and Vincent Jackson with 29). Rob really should look into unloading Sam Bradford and Ryan Fitzpatrick, both riding the pine, while their values are high.
Equippo del Jefe (Aaron, 1-1) 102 def. Parts Unknown Mufflers (Ben, 0-1) 92
The difference-maker here is the Detroit Lions' mascot, who magically caused Jamaal Charles to land awkwardly and tear his ACL. Charles is now on IR, and Ben is in some serious trouble. Aaron received great production from running backs Darren McFadden and LeSean McCoy (24 and 23, respectively), but missed out on the Jeremy Maclin experience in Atlanta. The Mufflers have no one to blame but themselves, however, as the only surprise on the bench was Ben Tate, and honestly, who is going to start Ben Tate against Miami?
Tom: By request of Mr. Tanier, and please note that Scramble is not the Brady-Manning Irrational Thread.
Mike: Although, as the featured comment notes, this means Peyton Manning will dress up in a pickle costume. This pickle commercial will, of course, be awesome. Unlike this ... thing.
Tom: I don’t know. I thought Peyton's latest commercial was distressingly average.
Mike: To which one are you referring?
Tom: Reebok, I believe.
Mike: Oh, yeah. Pretty lame, sadly does not play to Manning's talents. In any case, still probably better than this Ugg commercial.
Mike: Who thought stutter was a good visual effect? Especially for a graceful, powerful athlete?
Tom: One thing I noticed: the only time we see both the legs and Tom Brady in the same shot is at the end when he's sitting down in the locker room. For most of this commercial, the shoes don't have to be on Tom Brady's feet at all. It could be a body double.
Mike: In fact, it probably is, walking in rather weird ways at certain points. Of course, they then have Brady making weird faces while waving a stick in the air, so maybe he's going through some sort of magical incantation.
Tom: I blame the Friskies CGI cat. And in fact he could be using the Friskies CGI cat's ability to open wormholes into other dimension to change both the shoes on his feet as he goes through different settings. If you've read Dan Simmons' Hyperion Cantos, think of the trip down the River Tethys through farcasters.
Mike: It would be much easier for you to just make a Harry Potter joke, but OK.
Tom: This commercial is fundamentally not aimed at people like me, a universe that consists of at least me, even if nobody else. If you want me to buy footwear because Tom Brady's endorsing it, show me Tom Brady doing things like playing quarterback well, not walking from place to place. Heck, at least show me other people admiring Tom Brady's awesome footwear.
Mike: The commercial, as much as the commenters on YouTube want to pretend it isn't, is aimed at women. Sure, they make men's shoes, but men don't sit around looking at shoes. And men definitely don't care about lifestyle shots of Tom Brady, no matter how jarringly filmed. This commercial is clearly aimed at women. Even on those terms, it fails at its basic goal, which is to beefcake it up with Tom Brady.
Tom: In that case, why show this commercial during Monday Night Football, a show that I assume skews heavily male? Why not show it during, say, Two and a Half Men?
Mike: Why show Spanish-language commercials during "Please Love Us Latinos" month? Or all the NFL women's apparel commercials the league airs?
Tom: This feels like another commercial that exists because people in advertising agencies aren't like me.
Mike: There is a not-insignificant number of women watching, and they're desperate to see something they're interested in.
Tom: Sure, there are a number of women watching. I'm just skeptical that it's a cost-effective buy.
Mike: Tom Brady is very relevant to their interests.
Tom: NFL apparel for women appeals to both sexes, even.
Mike: Are the bored housewives watching Two and a Half Men going to buy Uggs? There are few places where a spot like this would work. Might as well get the exposure.
Tom: Ok, fine, buy commercials on some show the kind of people who would buy Uggs watch, like Breaking Bad or whatever.
Mike: Another male-skewed viewership.
Tom: Commercials during NFL games are expensive because you can reach a big and relatively diverse audience. Buy one NFL commercial or three during other times?
Mike: This conversation is really boring. We're debating ad slot purchasing strategy.
Tom: I prefer to think of it as plumbing the mystery of Ugg's intentions.
Mike: I doubt our readers would agree.
Tom: If you're trying to sell me a shoe, though, show me the same shoe in the entire commercial. Don't change shoes on me mid-stream.
Mike: But they have so very many shoes! Because men are famous for buying lots of shoes of varying styles.
Tom: I don't know about you, Mike, but I only wear one pair of shoes at a time. If I put on a pair of shoes, I want it to be as versatile as my life. I don't want to have to put on a different pair of shoes if I want to go walk on the grass.
Mike: That is why you are not a famous football player! Clearly we have found your problem. Oh, there's a related video: "Top 100 Players of 2011." That sounds like a reasonable topic! Let's see who it's by ... "BostonSportsFTW" ... oh.
(Tom makes a joke that is better off not being made. Bad Tom! Bad!)
Mike: Anyway, what direction do you think Brady/Not-Brady had for this? "Wander aimlessly while we film your torso/feet?" There's one bit where he runs quickly in one direction, then stops, and starts walking back from whence he came.
Tom: "Walk down this block." Mid-block: "Ok, start walking back up the block."
Mike: And then there's a dog, but not TOO MUCH dog.
Tom: I knew Air Jordans wouldn't give me Michael Jordan's ability, but at least his shoes might have been reasonably connected to what he was doing.
Mike: But while the dog is around we have to wobble the camera back and forth. This is the artsiest commercial we've ever covered, except maybe ANA.
Tom: Levi's Walt Whitman? That was really artsy, I remember thinking.
Mike: Drat. It’s true. It was also even worse than this one, although I guess that is impressive in and of itself, making Tom Brady mediocre.
Tom: I'm not so sure. I thought the Walt Whitman kind of added to the artistic ethos of that commercial. This Mos Def, though, I'm not sure what it's supposed to do, but it doesn't add anything for me.
Mike: Well, it doesn't do much, because they're scared of alientating the old white people watching the game. So while it is Mos Def, it's basically generic quiet hip-hop background #2362. That said, nothing can do more damage than the creepy stare Brady gives us at the end. It's like he knows we miss Catholic Match Girl and is trying to destroy our beautiful memories. Bad Brady! Bad!
As a reminder, you can access the full results here.
Kicker: No offense for Seattle, 0 points for Steven Hauschka. Like the sun in the East.
Wide Receiver: Remember how Harry Douglas was supposed to have a big role in Atlanta this year? Ok, so Chansi Stuckey’s 0 was low man this week, but Douglas joined Andre Roberts, Eddie Royal, Anthony Armstrong, Victor Cruz, Jordan Shipley, Legedu Naanee, and Golden Tate with 1 point each.
Running Back: You might have started Felix Jones on your real fantasy team and gotten 2 points, or even Mark Ingram, Cadillac Williams, or Arian Foster and gotten 3.
Quarterback: Luke McCown couldn’t even get his name spelled right on a Jaguars preseason telecast last year. This week, he wishes it was some guy named “Luke McGowen” that had put up his -6.
KEEP CHOPPING WOOD: So, two years ago, the Jaguars went into New York and knocked off the Jets on the road. David Garrard was their quarterback that day. Quarterback Luke McCown’s penchant for throwing the ball to the Jets was one of the reasons they were not able to repeat that feat on Sunday.
MIKE MARTZ AWARD: The rare dual award this week. After stopping the Eagles on a fourth down late in Sunday night’s game, the Atlanta Falcons and Mike Smith elected to kneel the ball immediately three straight times and then punt. While the Falcons still won, they faced a DeSean Jackson punt return and a play from scrimmage -- if they’d tried to run time off the clock they might have been able to end the game by punting out of bounds. Not impressed with this display, Rams coach Steve Spagnuolo simply demanded co-winner status by conceding Monday night’s contest against the Giants by punting. Yes, down two scores with less than two minutes to play, on fourth-and-23, the Rams’ chance of winning was minimal, but it wasn’t exactly zero.
COLBERT AWARD: Facing fourth-and-1 from the 10 in a tie game in the third quarter, Titans coach Mike Munchak elected to eschew the safe field goal to take the lead and instead went for it. Javon Ringer took the pitch for a score and a lead the Titans would not relinquish in the week’s biggest upset.
Nickorton86: I play in a pretty crazy 12-team league with 3 WR, flex, 1 QB/Flex (so essentially 2 QBs), 1 PPR. I have Andre Johnson and Peyton Hillis starting, and will probably start Darren McFadden (NYJ) and Wes Welker (@BUF) despite not great matchups. So I have one WR and one flex left for Shonn Greene (@OAK), Santana Moss (ARI), and Reggie Wayne (PIT). My gut is to start Moss and Shonn Greene and leave Wayne on the bench based on matchups, but Greene has not exactly lit it up this year. Thoughts?
Tom: At the start of the year, even with Collins, I was slightly optimstic Wayne would still be a decent fantasy play. I haven't seen much of the Colts-Browns game yet, but I've changed my mind. I thought Wayne was declining last year, and Kerry Collins plus a leaky offensive line makes him a deeply marginal fantasy play. The Colts basically can't throw the ball downfield at all, which hurts wide receivers' fantasy value.
Mike: The Colts are a disaster in general. If you've talked yourself into starting a Colt, you should swing by because I have a fantastic bridge you're sure to be interested in. Wayne is worth holding on to until Manning returns, but in the meantime starting him is basically insane.
Tom: I wouldn't go that far. I think you just have to think of him as more like a WR3 and play him based on matchups. The Steelers are not a good matchup, so you don't play him.
Mike: Fair enough. In any case, I agree. You have to go with Greene and Moss. And pray. Good luck with your really rough week.
Tom: Shonn Greene has underperformed this year, but I liked him coming into the year for a reason and I think those reasons are still valid. Even with the right tackle and Mangold's injury situation, play him. But yeah, what Mike said, good luck.
Bask in the comic! Bask I say! We'll see how long we can keep this up. In the meantime, send your questions to scramble-at-footballoutsiders.com or drop by the forum topic, which Tom assures me will actually exist this week.
55 comments, Last at
28 Sep 2011, 8:38am by
55 comments, Last at 28 Sep 2011, 8:38am by Mr Shush