The build up for the season can be excruciatingly long for a football fan. The preseason games are a poor substitute for the real thing and only give a small fix. We analyze the opening game and talk about it for weeks, and then it comes and we are given our first bit of information and excitement for the season. I want you to take a quick test.
Pick your favorite team and try to identify who they played in week 1 last year. Then keep going and see how long it takes you for each week, and how long until you get a game out of order. If you are like me, you probably got the opening game opponent (Baltimore against Kansas City for me) fairly quickly, maybe you got the second one (Oakland?), and then it started to blur together (Giants or Redskins or Eagles?).
What you are experiencing is the primacy effect, which is the tendency to better remember the earliest items in a series, and place greater emphasis or importance on those earlier items. Every play seems a little bigger, and first impressions can be difficult to break. [Ed. Could this be why I think Mark Sanchez will be fine? He dominated the Texans in his first career start. Damn primacy effect!] The primacy effect is reduced when a sequence happens rapidly (think quickly showing someone ten consecutive flash cards) but is increased when the sequence is spaced out. The NFL season, then, should be a strong example of the primacy effect, because the games are spaced out a week apart and we get to spend that week analyzing every aspect of the games.
To test this, I wanted to take a look at cases where the initial information conflicted with what happened thereafter. The gambling market is more sophisticated than the average fan, and those that invest heavily tend to put a lot of time into studying just about every angle. So, it shouldn’t matter when a good team has a bad game, right? That’s just one piece of information to consider. I looked at the teams that ended up winning 11 or more games, but lost the season opener by 15 or more points, since 1995. I compared them to 11+ win teams who had a 20 point loss or worse in games 5 to 8. So, I’m comparing good teams that had a bad loss (there were only nine in that group, and I had to lower the point limit for week 1 to get that many) to other good teams who had an even worse loss in the middle of the season. I then looked at the “against the spread” records over the second half of the season for both types of teams.
The week 1 losers covered a whopping 73.6% of the time over the second half of the season. The week 5 to 8 losers: 57.9%. These teams won roughly the same number of games over the second half of the season and the season overall, so the main difference between the groups was when they happened to have their bad game. The week 1 losers simply had far fewer wins that were non-covers because the line was just a little too high, and I would hypothesize that the primacy effect had something to do with that. The 2003 New England Patriots had the largest opening day blowout (31-0 to Buffalo), and were famously called out by Tom Jackson, who said that the players hated their coach and were done (according to Patriot Reign, Belichick later got personal revenge on Jackson.) All they did was go 14-1, and 14-0-1 against the spread the rest of the season, and the line never caught up to them despite the fact they immediately turned things around.
If gamblers can give undue influence to early results that conflict with later information, then I think we all are susceptible. I know that I will try to be more consciously aware of whether my opinions on a team or player are being too heavily influenced by a first impression. [Photos via Getty]