Research Shows Marino Is Comeback King, Not ElwayBy Mike TanierEverybody knows that John Elway holds the career N.F.L. comeback record.Everybody is wrong.Elway is credited with 47 fourth-quarter comebacks, the highest total in league history. According to the widely accepted number, Elway engineered 10 more such comebacks than his longtime rival Dan Marino.But the researcher Scott Kacsmar found major problems when he started investigating the data. “Comebacks” aren’t an official statistic. That means that each team’s public relations department gets to use its own definition of comeback. The Dolphins used a strict interpretation for Marino: he got credit for a comeback only if he took the field in the fourth quarter with the Dolphins trailing and led a game-winning drive. The Broncos used a more lenient definition: Elway got credit for leading a comeback if the Broncos were tied in the fourth quarter. There were some other variations: in one instance, Elway got credit for a comeback in a game that ended in a tie!Once Kacsmar standardized the definitions and corrected some media guide oversights, he discovered that Elway had 34 true comebacks. Marino had 36. If you prefer to count the game-winning drives that started with fourth-quarter ties, then Marino still has a 51-49 edge. Any way you slice it, Marino is the real “comeback” king.Elway may not even be second. Johnny Unitas was credited with 31 comebacks by Colts media guides, but they only counted regular-season comebacks, while Elway and Marino both got credit for playoff games. That means Unitas doesn’t get credit for the 1958 championship (the “Greatest Game Ever Played”), and Kacsmar found two other comebacks for Unitas. That means that Unitas and Elway are tied, though with scant play-by-play data available for Unitas’s career, there may be a few errors in the data.Kacsmar’s data doesn’t suggest that Marino is better than Elway, or that he was a more “clutch” quarterback, or that fourth-quarter comebacks are a fair way to evaluate a quarterback’s career. His research does show that a well-known, widely-reported N.F.L. statistic is completely wrong, and that this erroneous number colors our reputations of two of the game’s greats. Media guides still use varying definitions of “comeback” when compiling their figures; when comparing Peyton Manning’s comebacks with Tom Brady’s, or Brett Favre’s to Joe Montana’s, you may be comparing apples to oranges. It’s not a good idea to judge a quarterback by his comebacks, but since many people are going to do it anyway, the least we can do is make sure that the numbers are right.Kacsmar’s research can be found at ProFootballReference’s blog. Further analysis can be found at Football Outsiders.(Mike Tanier is a Fifth Down contributor and a co-author of Football Outsiders Almanac 2009, which is available in print and as a download at FootballOutsiders.com.