The 9th Commandment
Turnovers are player play-making failures, not play design failures.
No coach has ever thrown an interception. No coach has ever fumbled the football.
Returning to the kitchen analogy, a chef might have the freshest and tastiest ingredients and, using a brilliant recipe, combine those ingredients into a delicious soup. But if the waiter spills the soup between the kitchen and the customer's table, both the recipe and the ingredients will be wasted. Likewise turnovers (TOs) can waste productivity in football. Just as it is impossible for the chef to know when and under what circumstances a waiter will spill the soup, it is impossible for a play designer to know when and under what circumstances his players will fumble the ball or throw an interception.
Yet, "if you asked football coaches what was the key to winning games, many of them would say turnovers," wrote Eddie Esptein in Dominance. "Sometimes it almost seems that coaches are obsessed with turnovers as if nothing else really matters. I doubt that most coaches really believe that--I hope they don't--but the subject seems to come up quite a bit. A team's giveaway/takeaway ratio does correlate well with winning games although other numbers correlate better."
Epstein's concern that coaches are overly obsessed with turnovers or, more likely, their powerlessness to prevent or create turnovers, is based on his conclusion that fumbles and interceptions essentially are random. "NFL coaches won't believe this, but a team's turnover margin is due in no small part to luck," Epstein wrote. "How do I know this? Forty percent of the time a team's turnover margin flip-flops from one season to the next; that is if they were positive one season they're negative the next season or vice versa. That's kind of a high percentage if turnover ratio is primarily a team skill. The correlation between turnover margins in successive seasons is rather small, not zero, but small enough to strongly suggest luck plays a large role in turnovers."
Likewise, in a 2006 New York Times op-ed, economist David Berri, co-author of The Wages of Wins, wrote that "a look at turnover numbers over the years suggests that there is little in a team's style of play or emphasis on avoiding mistakes that affects the number of times it turns the ball over."
However, since Bill Walsh's West Coast System became widespread, the league-wide interception rate has fallen from 6% to 3%. Likewise, the turnover rate on fumbles is 3%. Thus, it appears that turnovers are "negative black swans." That is, turnovers are a "large-impact, hard-to-predict, and rare event beyond the realm of normal expectations."
There can be little dispute that turnovers can have a large impact on which team wins and which team loses a football game. In 43 Super Bowls, the more productive team is 38-5 (.884 wining percentage). However, in 4 of the 5 games where the more productive team lost the Super Bowl, the more productive team had a negative turnover margin. Some giveaways have been fatal. For example, in Super Bowl 36, New England's Ty Law returned an interception of a Kurt Warner pass for a TD to provide the difference in the Patriots 20-17 upset of the St. Louis Rams. In Super Bowl 37, Oakland was nearly as productive as Tampa Bay, but 3 times the Buccaneers returned interceptions for touchdowns to turn the game into a blow-out, 48-21.
Turnovers are compound events in that not only does the defense fulfill its primary mission by taking a down from the offense, but the defense also creates a first down for its own offense or, even better, scores points on its own. Such points are costless. The compound flavor of a turnover was vividly demontrated in Super Bowl 43 when Pittsburgh's James Harrison returned an interception of a Warner pass 100 yards for a TD to provide the difference in the Steelers 27-23 win over the Arizona Cardinals. As advancednflstats.com pointed out, Harrison's return came on first down and goal from the 1-yard line. "There, the exected points for the Cardinals offense was +6.1 points," advancednflstats.com wrote. "The return was effectively a swing of 7 points for a touchdown, plus 6 expected points denied the Cardinals. It was the Super Bowl's first pick-thirteen."
To determine how effectively past interceptions predict future interceptions, advancednflstats.com examined the correlation between various past statistics and future statistics. "The defensive interception rate stands out as the least enduring, least consistent team stat," advancednflstats.com found. "This indicates there is a lot of randomness in interceptions, which is no surprise. Producing defensive interceptions does not appear to be an enduring, repeatable ability of a team. Instead, it appears that defensive interceptions are more of a function of 1) randomness, and 2) their opponents' tendency to throw interceptions. In other words, interceptions are very random, and they are 'thrown' by an offense much more than they are 'taken' by a defense."
Advancednflstats.com's conclusions are consistent with history. In Dominance, Eddie Epstein highlighted the incredible ability of the 1958 and 1959 Baltimore Colts to intercept passes. In those years, Baltimore intercepted a whopping 75 passes. More importantly, the Colts percentage of passes intercepted was far greater than the league average in both years (9.6% vs. 6.2% in 1958 and 11.4% vs. 6% in 1959). In both years, Balitmore defeated the New York Giants in the NFL Championship Game.
A team's ability to repeat dominance of a single statistical category may indicate a systematic design edge in favor of that team (e.g., Green Bay's YPA superiority in 1966 and 1967). But QC's review of the history books failed to uncover any such design advantage in favor of the '58-59 Colts. Clearly, Baltimore had some players with a nose for the football when it went in the air. Don Shinnick, who played middle linebacker in '58 and outside linebacker in '59, holds the NFL record for career interceptions by a linebacker with 37. Cornerback Milt Davis is one of only 8 players to twice lead the NFL in interceptions in a single season. (Former Dallas Cowboys cornerback Everson Walls is the only player ever to lead the league 3 times.) But despite its best efforts, QC could not uncover any innovative coverage or pass rush schemes pioneered by the '58-59 Colts that would explain their success.
However, in his book Always A Winner, Don Shinnick described a defensive design that Baltimore used against San Francisco's Y.A. Tittle in one game in 1959 that featured Baltimore's 6-6, 288-pound defensive tackle, Gene "Big Daddy" Lipscomb, that sounded like a prehistoric version of Dick LeBeau's zone blitz designs. "We mixed up the defensive patterns quite a bit to keep Tittle guessing," Shinnick wrote. "Five or six times Big Daddy Lipscomb played back from his line position while we linebackers blitzed. In effect he acted as a linebacker and laid back for the trap and draw plays and the screen passes."
Colts' defensive end Gino Marchetti told Sports Illustrated's William Nack that he and his teammates called Lipscomb "our fourth linebacker." According to Nack, Baltimore coach Charley Winner considered changing Lipscomb's position. "I remember one game against Green Bay," Winner said. "They had a fast back named Tom Moore, and Big Daddy dropped off the line to cover a pass. He chased Moore for 40 yards and then knocked down the pass in the end zone. I wanted to make him a linebacker, but I couldn't replace him on the line."
Still, there is no record of Baltimore regularly employing this design as a pass defense. More likely, Lipscomb caused more interceptions with his disruptive height than he did with his pass coverage. Bill Walsh said that a quarterback's ability "to see downfield is the very basis of professional quarterbacking. If a quaterback begins looking at pass rushers instead of receivers, he is doomed." In order to be successful, the quarterback "must be able to avoid the pass rush and find a throwing lane between defenders," Walsh said. Thus, it is not surprising that tall defensive tackles like Lipscomb and San Diego's Ernie Ladd (6-9, 315) played on teams that intercepted a lot of passes. (In Ladd's rookie year, 1961, the Chargers set a pro football record that still stands with 49 interceptions and then picked off 6 more passes in a 10-3 loss to Houston in the 1961 AFL Championship Game.)
Before the 1964 season opener, the Baltimore News-Post asked Shinnick to divulge his secret recipe for intercepting passes. Shinnick wrote that he had no secret recipe. "I try to make myself scarce when the opposing quarterback is setting up for the snap," Shinnick wrote. "If he doesn't know where I am going to be, I've got an important advantage. Whenever possible I keep moving up and down the line behind those big guys up front in hopes that I'll get lost in the crowd. Naturally it also helps to know the passer and receivers. This comes from actual playing experience and watching game films."
Finally, no historical source even bothers to so much as identify the name of Baltimore's defensive coordinator in 1958-59. It is doubtful that head coach Weeb Ewbank came up with a design that produced the Colts' prolific interception totals. "Several players told me that Ewbank's forte was assessing talent," Epstein wrote. Thus, it appears that, with the possible exception of the 1959 game against San Francisco and Tittle, Baltimore's interceptions resulted from superior player play-making, not from superior defensive play design.
Because turnovers are like clumsy waiters and other black swans, such events are beyond the control of the designer. Thus, QC treats turnovers as player play-making failures, not design failures.