This article kicks off Andrew Luck Week here on Grantland, which aims to provide a deeper look into one of football’s most compelling players. Luck is very obviously a great quarterback, but when you get past the broader descriptor, he’s a fascinating jumble of traits. Luck is the most pro-ready quarterback to come out of college in a generation, but he derives a considerable amount of his value from his ability to scramble and run for first downs. And despite those preternatural instincts for escaping pressure both inside and outside the pocket, the seemingly indestructible Luck has been knocked down far more frequently than any other quarterback during his two-year career. He is markedly inconsistent from game to game or even half to half, as evidenced by his stunning performance against the Chiefs in the playoffs last season.
Of course, what makes all of that even more interesting is that Andrew Luck wins. With one of the worst rosters in the NFL around him and a coaching staff and front office that appear to actively put obstacles in his way, Luck has compiled back-to-back 11-5 seasons at the helm of a team that was 2-14 the year before he arrived. Many of those wins — 14, in fact — have been by one touchdown or less. Today’s look at Luck focuses on those narrow victories.
I’ve often written in the past that teams can’t control their ability to win close games from season to season, suggesting that a team’s record in games decided by one score or less is mostly random. Andrew Luck and his Colts were 9-1 in one-touchdown games during his rookie campaign before following it with a 5-1 record the next year. 14-2 sure doesn’t feel random. With that in mind, I wanted to take a closer look at Luck in his close games and the great quarterbacks of the past to see if there’s something suggesting that Luck might be an exception to the rule. Instead of using history as a way to point out Luck’s season as an outlier, I wanted to see if he wasn’t an outlier after all. I started by finding the best two-year records from any quarterback in close games since the merger. As it turns out, Luck’s 14-2 record (.875 winning percentage) is the second-best performance1 over two years since 1970. When I saw the rest of the top five, I gasped:
You may notice some small gaps between a player’s official win-loss record and the numbers in this piece. For this article, I assigned the “win” or the “loss” for a given game to the quarterback on each team who threw the most passes. Also, all numbers go back only through the AFL-NFL merger in 1970, so quarterbacks who began their careers before 1970 have partial records for their career totals.
Maybe there’s something in the air in Indianapolis …
The Best of the Best
Anecdotally, I’ve noted in the past that some great quarterbacks have a better win-loss record in games decided by one touchdown or less than you might expect — not to the extent of Luck’s win-loss record over the past couple of years, of course, but far from the 50-50 split that history would suggest. The first time I really noticed that was in researching one of the millions of stories I’ve had to write on the Brady-Manning rivalry over the past decade, and, well, it turns out that was no accident. Here are the 20 best records produced by quarterbacks in games decided by one touchdown or less (with a minimum of 50 such games over their career):
Not a bad group. By my count, there are six Hall of Famers and another five guys who will make it when they’re eligible. But for all the hardware, there are certainly a few players in the top 20 who don’t exactly qualify as immortals; I bet you didn’t know until five seconds ago that Jay Schroeder was the second-clutchest quarterback in league history. And it really stands out just how far Brady is ahead of the pack. He could lose his next six games decided by one score or less and still have the best post-merger record in one-score games.
Besides Brady, the records for these players don’t seem to be obvious proof that they could have a meaningful impact upon their teams’ abilities to win one-score games. Take Joe Montana. His 40-29 record is impressive, but turn just five of his 40 wins into losses and he’s right back at even with a .507 winning percentage. If Montana — by all accounts the coolest customer in league history — could really manufacture wins in close games, wouldn’t the difference amount to something more than one win every three seasons? In contrast, Montana’s record in games that weren’t decided by eight points or more was 75-17, for a gaudy winning percentage of .815.
Indeed, you’ll often find that the quarterbacks with the best records in games that aren’t close also have the best records in the tight ones. Among those aforementioned passers with 50 close games or more, Brady has the best win-loss percentage in close games (.711) and contests that aren’t close (.823). The same quarterback also has the worst win-loss percentage in both categories, too; coincidentally enough, it’s Archie Manning. Peyton and Eli’s dad was just 17-61 (.218) in games decided by more than seven points, a figure that rose to 16-34-2 (.327) in one-score contests. It all makes sense; it’s far easier to imagine an awful team like Archie’s Saints winning a squeaker than blowing out the competition, while a dominant organization like Brady’s Patriots is more likely to lose narrowly than to win by a comfortable margin.2 With all of that being said, for those passers in the 50-game sample, just more than 18 percent of a player’s winning percentage in close games is accounted for by their performance in games decided by eight or more.
If you use a player’s win-loss record in games decided by more than one touchdown to inform how many wins he “should” have recorded in close games, the best clutch quarterback of all time is … Jeff George? No, really: George’s teams were 17-47 (.266) in eight-plus point games and 25-27 (.481) in one-score contests.
As the samples get smaller, naturally, the results get crazier. If you expand the sample out to 20 close games over a career, the best record in one-score contests belongs to, of all people, Vince Young (20-6, .769). Among the sub-.500 crowd: Troy Aikman (31-32, .492), Warren Moon (42-46, .477), Kurt Warner (22-24, .478), Aaron Rodgers (17-20, .459), and, yes, even Doug Flutie (18-19, .486). And all of those players have close-game samples significantly larger than that of Luck.
Given the numbers above, it’s safe to say that Luck won’t win 88 percent of the close ones over the remainder of what’s sure to be a long career. Even finishing with a win percentage above 60 percent might prove to be difficult. But there is one example of a player who managed to pull off something resembling Luck’s magical spell over a much longer time frame who doesn’t appear in the chart above. Daryle Lamonica did his best work as a member of the Oakland Raiders just before the AFL-NFL merger, sneaking into the NFL for only the final three seasons of his career. If we include his AFL performance, Lamonica was the primary quarterback in 32 games that ended up decided by one score. In those games, Lamonica’s team was 24-2-6. 24-2-6! With the ties counting as half wins, that’s a win percentage of .844. So the closest passer amounting to Luck over a fuller career came and went 50 years ago, played in a different league, and still managed to come up short of Luck’s prodigious win rate in close games. That’s how incredible Luck has been these past two years.
King of the South
Colts fans probably aren’t concerned about what might happen to Luck in close games over the rest of his career, though; they’re likely more interested in what will happen with Luck and his team in 2014. Last year, Luck’s 9-1 record in close games seemed unlikely to recur. 5-1 isn’t quite as extreme as 9-1, but it’s hardly regression back to the mean. Is two years enough to suggest that Luck and his Colts will be better in close games than we might expect?
In an attempt to figure it out, I formed a roughly representative sample of quarterbacks with two-year runs in close games similar to Luck’s. It’s impossible to come up with 30 other Lucks, because his performance has been so incredibly rare, but I was able to find 33 pairs of quarterback seasons from 1970 to 2013 from passers who played in 15 or more close games over a two-year stretch and won at least two-thirds of them. Those passers were 399-154 (72.2 percent) in close games during their impressive two-year stretches. The following year, those passers went a combined 106-88, winning just 54.6 percent of their one-touchdown games. That’s not an even 50 percent, but it’s close enough to suggest that the quarterbacks — a group that includes the likes of Peyton Manning, Dan Marino, John Elway, and even Vinny Testaverde — are unable to keep up winning close games as a skill, even if they’ve done it two years in a row. It seems unlikely Luck will be able to keep it up in 2014. Unless he’s Daryle Lamonica.
Do You Ever Feel That Way?
It seems fair to give most of the credit for these close wins to Luck, right? It obviously hasn’t been the running game, and while the defense has several talented contributors, every single one of Luck’s clutch performances has seen the offense produce at least 17 points. On five occasions, Indy’s defense has allowed 27 points or more in one-touchdown games. They’re 5-0 in those games. And naturally, all close games aren’t created equal; some are close contests that are literally decided during the final snap, while others qualify as a “close” contest only late in the fourth quarter, after a mostly meaningless score. It seems worth asking those questions about these 16 close Colts games. In other words, how much of Indianapolis’s incredible run actually comes down to some magic from Luck late in the contest?
A lot of it, actually. For a team that has played so many close games over the past couple of seasons, they seem almost dependent upon a formula: Get a score from Luck about halfway through the fourth quarter, come up with a stop on defense, and then finish off the game with a first down shortly thereafter. Of Luck’s 16 close contests as a pro, just five are of the backed-in variety, like the Broncos game in which Matt Prater kicked a field goal with 17 seconds left to turn a 39-30 scoreline into a 39-33 contest.
More frequently, the Colts needed more than that simple formula. They needed Luck to make things happen. The other 11 wins all featured Luck leading a scoring drive in the fourth quarter that either gave his team the lead for good or extended the game into overtime. One of the more impressive performances actually came in Luck’s second loss as a pro, the Week 3 game against the Jaguars in 2012. In that game, Luck took over on his own 33-yard line down two points with 1:33 left and no timeouts and drove the length of the field for a lead-taking field goal with 56 seconds left … only for Blaine Gabbert to throw a game-winning 80-yard touchdown pass to Cecil Shorts on Jacksonville’s next pass from scrimmage. And even that paled in comparison to Luck’s run versus the Lions in 2012, against whom the Colts scored two touchdowns in just more than four minutes to claim an unlikely victory.3
I’ve mentioned that Luck’s record in close games over the past two seasons has been extraordinary. The worst record an NFL quarterback (minimum: 15 games) has produced in close contests over a two-year stretch since the merger: the 2012-13 seasons of Matthew Stafford, who’s gone 6-14 (.300) over the past two years.
It’s possible the Colts could have won some of these games even without Luck around, but it certainly seems like they’ve gotten more out of having him in the lineup than they would have from a typically mediocre quarterback.
The Wyatt Earp Effect
For all the research and comparisons above, I can hear my friend Chase Stuart screaming from his laptop about Luck’s performance in close games being explained by the “Wyatt Earp effect,” which he does a superb job of explaining here. The Wyatt Earp effect holds that while something might be statistically unlikely, it can often be explained away by the fact that a large sample will inevitably reveal some statistical quirk over a small period of time. The Earp reference implies that the famous gunfighter was simply the product of a large sample of men fighting in duels; inevitably, by chance, somebody would arise who had won a large number of duels, even if he had no specific skill in doing so.
Likewise, it’s fair to wonder whether Luck (and perhaps even Lamonica) are subject to the Wyatt Earp effect. If you think a passer really has exactly a 50-50 shot of winning a game decided by one score or less, his chances of playing 16 games and winning 14 or more times are astronomically low — about 468-to-1, an event that would happen 0.2 percent of the time. That’s exceedingly unlikely to occur, but it’s also still feasible that it will eventually happen randomly. Luck could very well be that eventual randomness sprung to life.
So, Luck won’t be able to keep up his performance in close games, which means the Colts will fail to live up to their 11-5 record in 2014, right? It’s not quite that clear. For one, it’s easy to envision a scenario in which Luck isn’t anything special in close games and it’s of no consequence going forward, because the Colts win by eight or more points on a more frequent basis. Luck is just 8-8 in two-plus-score games during his professional career, but the safest way for a quarterback to consistently win is to blow out the competition. Given the strides Luck made between his freshman and sophomore pro seasons, it wouldn’t be a surprise to see Indy take a step forward in terms of raw level of performance in 2014.
If the Colts are dependent on close games for half of their wins again in 2014, though, they’ll be engaged in a very dangerous exercise. Luck proved the numbers wrong in 2013, but history tells us that time is on the numbers’ side. Whether his initial run was due to randomness or a stroke of genius, Luck simply can’t expect to win more than 85 percent of his close contests as the years go on. His team may feel that pinch as early as this season.