Joe Posnanski and NFL Tiebreakers

Joe Posnanski is my favorite sportswriter. He's smart and funny, and kind without being saccharine. I wish I could write the way he does; the man is a tremendous storyteller. I write about sports statistics, and it's interesting if you really like stats. Joe writes about sports statistics, and it's interesting if you have a pulse. He's also got a fantastic background in baseball.

There are plenty of writers out there who don't deserve their fame. They got where they are because of connections or good fortune, or one genuinely good work that they've been milking ever since. Joe Posnanski is not one of those. Yeah, he was friends with Buck O'Neil, and he's pals with Bill James and Tom Watson and a bunch of other people. That's because he's a nice person, and he's truly passionate about sports.

All of that is why I'm always let down when Joe writes about football and basketball. For some reason, he doesn't apply the same logic to other sports that he does to baseball. Posnanski is one of those people who, when you talk about clutch hitting in baseball, will mention that it's usually random, that it's not a repeatable skill, that we have better methods to judge performance. But this spring, he bought in completely to the LeBron James narrative. When the Heat were rolling through the playoffs, going 12-3 en route to the NBA Finals, it was because LeBron was the best player in the world. When the Mavericks pulled an upset, it was because LeBron choked, because he doesn't have that Reggie Jackson/Michael Jordan/Joe Montana quality. Come on, really?

When he writes about baseball, Posnanski sets aside subjective judgment:

"What is more likely … the stats are wrong or that three people, who watch the Royals nightly and who you appear to trust, are wrong?" Well, I have absolutely no doubt in my mind: To me, it is much, much, much, much more likely that the people who watch the Royals nightly are wrong. After all, the people I know saw Rick Ankiel play just 27 games with Kansas City (and a few scattered games as a visiting player). The numbers have charted every single game that Rick Ankiel has played in the big leagues. The people have watched Ankiel in the larger context of nine players on the field. The numbers break him out and look at him individually. People, almost without fail, allow any great play or terrible misplay to become outsized. The numbers give credit for the great play, take away credit for the misplay, but don’t let it skew expectations, or blunt whatever happens next.

That's four muches in front of "more likely" there. But when he writes about basketball, Pos sets aside the numbers and gets sucked in by the narrative. Karl Malone's 1996-97 MVP award? Bogus. Sure, Malone had the stats that year, and Jordan didn't, but everyone knew Jordan was the best player. When Jordan, whose team also included Phil Jackson, Scottie Pippen, and Dennis Rodman, won the championship, well that just confirmed it. This is exactly the sort of thing Posnanski resists in baseball, but for other sports, he's in, hook, line, and sinker. How can the same logic that applies to baseball be withheld from other sports? It doesn't make sense.

When he writes about baseball, Posnanski resists conventional wisdom, as when dismissing Thomas Boswell's assertion, "When stats WILDLY contradict common sense, always doubt the stats." Posnanski wrote:

Yes, this seems a solid premise. Only, you know what? It isn't. It is, when you think about it, a horrifying premise — I cannot believe that Tom Boswell, my hero, really believes that. Common sense says that the universe revolves around the earth. Common sense says that thunder clapping means God's angry. Common sense says that when your car is sliding you want to turn your wheel away from the skid. Common sense says that a fast guy with no power who might or might not get on base is the perfect guy to put in the leadoff spot. ... Common sense says a lot of incredibly stupid things and if you are going to automatically choose common sense over, you know stats and facts and results, well, that's a good way to crash into trees and lose your shirt in a card game and get stuck with Omar Moreno.

Common sense says that in a clutch situation, you'd rather have David Eckstein than Barry Bonds. Common sense says that a huge, strong running back is ideal for short-yardage situations. Before this year, common sense said that Dirk Nowitzki couldn't win the big one. Common sense says that Michael Jordan was still better, in 1997, than Karl Malone. Apply these things to baseball, and Joe writes logical and entertaining columns dissecting the flaws in the argument, but switch to another sport, and he's the one promoting the flawed arguments in the first place. It's mind-boggling.

On Friday, Posnanski posted something on his blog titled, Why Hitting Streaks Matter (Sort Of). The premise was that our interest in sports statistics is sort of random, that something like a hitting streak or a 3,000-hits kind of magic number is pretty arbitrary, that these things matter because people are interested in them, but don't necessarily mean much otherwise. And sure, I'll go with that. Look at Brett Favre's consecutive-games record, or Cal Ripken's. By themselves, I don't think those records mean much of anything, but for many people, that's Favre's defining accomplishment. I'm much grumpier and more taciturn than Joe, so this actually sort of bugs me, because I don't think the consecutive-games record makes Favre a better QB than Fran Tarkenton or Bart Starr or Roger Staubach. Maybe he was, but let's argue that on its own merits.

Anyway, one of the examples Posnanski used in his column was the NFL's playoff tie-breaking system. But I think what he wrote about it was totally off-base. First, Joe cited the system as an example of something that has legitimacy because fans say so, when I think the reality is that fans complain every year about the system. And while some of these complaints are perfectly legitimate — conference record is not a sound tie-breaker — others are not. Strength of schedule and point differential (both of which Joe cited as nonsensical) are great tie-breakers, and what's ridiculous is that they're so far down the list.

"We break playoff ties based on puzzling mathematical breakdowns that don't really tell us which team is better. Why is there not intense outrage and mutiny in great American cities when their teams get knocked out of the playoffs even though they have the exact same record as teams in the playoffs? Why should my team's 10-win season be worth less than theirs because of our conference record (come on) or strength of schedule (which we don't control) or point differential (are we supposed to run up scores now?)"

It's true that teams don't control their strength of schedule, but they do control how they perform against the opposition they get, and to suggest that a team that went 10-6 against a .475 schedule is as good as a 10-6 team that played a .525 schedule is absurd. It's probably not as good as a team that went 9-7 against a .525 schedule. This is precisely the kind of tie-breaker we should be using. Just last season, consider the Kansas City Chiefs (who got annihilated in the playoffs), or the Tampa Bay Buccaneers, who went 10-6 without beating anyone who had a winning record (until a meaningless Week 17 game against the Saints, who had already clinched playoff position).

Point differential is one of the best predictive tools we have for evaluating team strength. Pythagorean winning percentage works in baseball, and it works in the NFL. Hey, no one wants to encourage running up the score. But point differential is the sort of stat that helps you separate the 10-6 Bucs (+23) from the 10-6 Packers (+148). Green Bay actually had the best point differential in the NFC last season, even better than the 13-3 Falcons (+126) or the division champion Bears (+48). This is often an incredibly valuable and telling statistic. Or hey, want to guess who had the best point differential the year before? The New Orleans Saints.

Aren't the tie-breakers designed to help us get the best teams into the playoffs? Strength of schedule and point differential probably are the most meaningful team statistics outside of record itself. If baseball somehow considered run differential (runs scored minus runs allowed) as the new criteria for homefield advantage in the World Series, I imagine Joe would throw a party to celebrate. But the NFL uses the same thing as its fifth tie-breaker (which almost never comes into play, maybe once every four years or something), and it's grounds for criticism. If Joe wants to complain about the NFL's tie-breakers, he should be grumping about how seldom these meaningful figures actually come into play in the tie-break formula.

I'm not trying to complain or criticize so much as to point out inconsistencies in the work of someone who smart enough to understand why he's wrong, humble enough to admit it, and honest enough to correct himself. A throw-away paragraph about NFL tie-breakers isn't life and death. But to people like Posnanski and myself, little things like this matter enough to attract our attention. I've written whole columns on reforming the NFL's overtime rules, pages on how we should evaluate punters, a detailed breakdown of how AFL passing stats support the league's exciting reputation. Joe wrote an imaginary interview with an esoteric fielding statistic. If anyone can appreciate why these things matter to a sports fan like me, it should be Posnanski. The strange things that we decide matter in sports: that's what Posnanski was really writing about on Friday.