HOW STATS CAN LIE

A few observations on how statistics can lie and the difficulty of handicapping as a result.

Many handicappers like to review box scores of games to analyze things like total yards, first downs, and time of possession to see if the winning team really dominated the game or maybe just cashed in on a few lucky turnover plays. Anyone who has ever watched a game or pre-game show (with the sound on) has heard the old adage that whoever wins the turnover battle, will win the game. The problem with that is trying to predict which team will turn the ball over. Well, today I am prepared to debunk that myth.

Even if you knew who would win the turnover battle, you might not know who will win the game

As an example, consider Sunday of Week 6 in 2011, on which twelve games were played. Of those 12, in three games each team had an equal number of turnovers and in the remaining 9 games, 5 times the team with the turnover advantage won the game, but 4 times they lost. This is hardly the lead pipe cinch that we have been programmed to accept as the gospel.

But let’s dig a little deeper in our examination of that Sunday’s box scores. In one game, neither team had a turnover, team A had an extra minute of possession, won the first down battle 22 to 20 and outgained team B 424 yards to 399 yards from scrimmage and had an advantage in return yards of 93 to 27. While these stats would suggest team A won a close game, you should realize that if that were the case, I would not be using this game as an illustration. So, who are team A and team B? Team A is the Rams and Team B is the Packers. Final score: Packers 24 Rams 3. Surprised?

Let’s try another one. Team A and Team B have roughly identical total yards and time of possession, but Team A has 21 first downs to Team B’s 14 first downs. Team B turned the ball over twice and Team A did not turn it over at all. Team B had 9 more penalties than Team A. If you figured Team A won fairly comfortably, you would be wrong. In fact, Team A (Detroit) lost to Team B (San Francisco). New England defeated Dallas in a similar game, except that the first downs were nearly equal and although Dallas was also plus 2 in turnovers, they did turn the ball over twice themselves.

Last one. Each team with one turnover and an equal amount of first downs, a nearly identical number of total plays from scrimmage, an equal amount of penalties and penalty yardage. Who wins that game? The team that scored one touchdown on a kickoff return and another on a fake field goal–The Oakland Raiders.

So, what’s the lesson here? Does the team that dominates the stats always lose? No, but they make the most entertaining teaching examples. Is there some way to decipher these stats and use them? Maybe, but I don’t know what it is–because the stats discussed above do not exist until after the game is over and there is no guarantee that the team will put up a similar performance in the following week. Yes there were some hidden stats, which I didn’t use in the examples, that may explain the outcome of those games, but they probably don’t help you figure who to bet on next week. For example, the Rams failed to convert 3 fourth down plays in Packer territory–yeah, that helps explain the result of the Week 6 Packer-Ram game, but what does it tell you about what either of these teams will do in Week 7?

So here is my theory………..it is not the number of first downs, total yards, penalties, penalty yards, return yards or even turnovers or big plays that determines the outcome of games…………it is the timing of these events. In the examples above, San Francisco and New England both lost the turnover battle, but won the games because they made the big plays down the stretch, when they counted the most. St. Louis had more first downs than Green Bay, but it was the 3 first downs they failed to convert on fourth down attempts that allowed Green Bay to win a statistically even game by 3 TD’s. Oakland dominated the time of possession and total yardage, but it was 2 big special teams plays that won the game for them.

Last thing. Of course, traditional thinking is not always wrong and I do not mean to imply that it is. But in my opinion, it would be far less interesting to discuss the five teams that won their games thanks in large part to at least a 2 turnover advantage over their opponent. Hopefully, this helps you realize that picking games is not so simple and that stats can be misleading and sometimes not at all helpful in picking winners, and therefore you will not try this at home on your own. Rather, you should visit this site every week and let us help you Win on Pro Football.

P.S. You should be comforted to know that none of the stats discussed above enter into the selection process of the Green Machine, whatsoever. Looking forward to next week.

Jimi Green