02 December 2011

Inside the Mind of Mike Leach

The first of two excerpts The Quad will run from a new book, Sports for Dorks College Football, co-edited by Ferhat Guven and Mike Leach, the former Texas Tech coach hired this week at Washington State. The book is an anthology of ideas by various experts on innovations taking place in college football.

Chapter 1:  Inside the Mind of Mike Leach
Thus far, Leach has espoused things that are supposed to be simple and that offensive football execution relies on the interaction between the mental side, numbers and leverage, blocking and tackling, and sound fundamentals. But what metrics are important to Leach? What matters to him?

“The most important statistics in football are wins and losses and whether or not a team can outscore his opponent.”

Groan.  Another maxim. “Isn’t that akin to saying the key to being wealthy is having with the most money?”
Mike cracks a smile and concedes the point.

“The three most important factors that I look at on offense are first downs, number of plays, and converting third downs, because those statistics tell you how well you are controlling the football.

“If you have the most first downs, you’re controlling the football. You want to be able to run more plays than your opponents, because more plays generally mean that you have more opportunities to score.

“Third-down conversion rates tie in with first downs and number of plays. Those plays extend drives. While those statistics are important indicators, a coach has to be able to interpret the results.

“Here’s the catch. While those statistics are important to us, every single one of them might be meaningless depending on the type of game we happen to be playing. If you’re in a game where you can score on three-play drives all the time, those statistics don’t really mean very much. As a coach, you have to be flexible about how you interpret them.


“Statistics and analytics have their place in sports. They help to organize data and crystallize tendencies. However, statistics also have their limitations. Football is made up of numerous variables. People who try to define the game through one or two variables underestimate the game’s complexity.

On the other hand, there is a risk that too much emphasis on statistics can overwhelm a coach’s or player’s thought process. Statistics inform judgments, but human beings ultimately make decisions. There are countless errors that can occur in a physical contest like football.”

Mike argues that advocates of quantitative analysis in sports have to be careful about overstating their case. These proponents look at historical data, and make judgments about coaching decisions, “Coach so-and-so should have done this, because the odds suggest that his probability of success would have been better if he had done x instead of y.

“For some reason, many members of the media have developed a sort of reverence for the power of statistics in football. Few genuinely understand it, and even fewer still have actually applied them to coaching in a football game.” Mike raises an amusing observation. It is doubtful that many in the media have ever run a regression analysis.

“I am open minded about the potential benefits that applied statistics might one day provide for the sport. I am someone who appreciates innovation and am always looking for advantages. Right now, it is probably fair to say that some of those innovations are not yet compelling enough for coaches to embrace completely. While there are a number of analytical tools that help to better analyze information, these tools so far are replicating the sorts of things that most coaches have been doing manually for a long time. There is no question that this sort of software has made the process of game analysis more effective. We employ these tools to help us break down an enormous amount of information. For example, we chart plays by down, by distance, by formation, by the receiver’s routes, and by the opposing defense’s formations.

“We perform a comprehensive evaluation of our play-calling tendencies using these statistics about four times a year. The statistical outputs can tell us which plays are more effective than others are and whether a certain play works better in certain formations. For example, our analysis will show us that our offense is able to gain more yards on a particular play when we send a receiver on a wheel route. We might not complete a single pass to that receiver for an entire game or series of games. However, our analysis can show that our offense is able to generate more yards on those plays, because the wheel route opens up more space for the receiver who catches the ball.

“What the stats proponents tend to forget is that while things like historical data analysis are interesting, most of those analyses are not situational. Going for it on fourth-and-three on the opponent’s forty-three-yard line might statistically be a better decision over the course of a thousand games than punting. Keep in mind, I generally try to go for it on fourth down more than most coaches, but we have yet to see data refined to a situational basis, which incorporates time of the game, any injuries, fatigue, weather conditions, personal issues (like did a player break up with his girlfriend before the game), and a host of other factors. Coaches also have to rely on their experience, the team’s psychology, and game momentum, when deciding to make a certain call.

“I coached a game two years ago in which my team was leading by five points on the road with about eleven minutes left to go in the game. We were in a fourth-and-one situation from the goal line. We ran a quarterback sneak and the other team managed to prevent our team from scoring. I was roundly criticized for not kicking a field goal and increasing the lead to eight points. I don’t exactly recall the statistics crowd jumping to my defense at the time.

“Nevertheless, the decision to ‘go for it’ was well supported by the statistics.  We even had two opportunities to cross the goal line.

“The calculus I employed was straightforward. If we score a touchdown, the game is more than likely over. If we score a field goal, the opposing team still has about eleven minutes to overcome what would be just an eight-point deficit. The opponent has a great quarterback and our defense was struggling in that game. In the worst case, if we fail to score a touchdown, the opponent’s offense will get the ball on their one-yard line. In other words, the opposing offense will have to defy the steepest odds in the game to convert the change of possession into a touchdown.”

“’Going for it’ was statistically the best choice. There were other qualitative factors which went into the decision making process.

“In terms of talent, we had a superior offensive line. Our offensive line had two All-American caliber players. Our offensive line outweighed our opponent’s defensive line by fifty pounds per player.

“As it turns out, we run the ball behind our best lineman, who accidently slips as he is blocking the defender. The opponents stop the quarterback and we fail to convert a touchdown. If we ran that play ninety-nine times, we score.”

Despite being stopped, the game is not over yet. The opponent’s offense now has the unenviable task of marching ninety-nine yards down the field to score the go-ahead touchdown.

“As the statistics would suggest, we managed to stop the ball, in this instance, by intercepting the opposing quarterback’s pass. We’re still up by five points, there is over seven minutes left on the clock, and we have the advantage of operating with good field position. Our offense, which is ranked among the top five at the time, is facing a defense that is ranked in the bottom quintile of college football. However, on our first set of downs, we punt the ball after just four plays. Still, our punter gets off a great kick, and pins the opponent at their own five-yard line with under six minutes left to play. The opponent’s offense takes over, marches ninety-five yards down the field, and scores the go-ahead touchdown.”

Murphy’s Law reveals itself. Texas Tech loses.

The worst outcome has now happened, defying all likely probabilities. “While we may not have been successful on that occasion, when I review my decision making in that scenario, I have a hard time second-guessing myself. The statistics and the qualitative assessments were overwhelmingly in my team’s favor at the time.”

About the Author: Ferhat Guven is the co-founder and managing editor of Sports for Dorks College Football. A native Texan of Turkish origin, he is an executive at a global real estate investment firm and has lived in London, England, for over eight years. Guven is a graduate of Texas Tech University and holds a master’s degree in management and finance from the London Business School.

No comments:

Post a Comment