The Thanksgiving Curse: a look at the data

The Den
1

Sup ninjas. Today we’re going to examine the data behind the Thanksgiving Curse of the Detroit Lions. We’re going to answer key questions: Is the curse real? Yes! Why does it exist? Mostly the NFL’s scheduling decisions. Are there any other reasons for the curse? God has abandoned us, or God is discipling us, or God is testing us. Insert your own theological explanation. Here we go…

First, the obvious: the Lions lose a lot on Thanksgiving. Since realignment in 2002 the Lions are 6-18. That’s a .250 winning percentage. In non-Thanksgiving home games the Lions are 84-85, the 24th best home record over that period. Not good, but either low-mediocre or high-bad depending on where you make the cuts. During that same period, the Lions road winning percentage was .328. The Lions have played better on the road than they play on Thanksgiving.

What’s going on here? Primarily it’s opponent strength. The average Thanksgiving opponent had a winning percentage in the prior year (when the league puts the schedule out) of .591, basically a 10-win team. The non-Thanksgiving home opponents had a winning percentage of .497. So the league is almost always giving the Lions a tough opponent on a Thanksgiving. The probability of this happening by random chance is 1.07% To put that in perspective: if you simulated the last 24 years of NFL scheduling 100 times, the Lions would only get a Thanksgiving schedule this difficult once.

Something similar happens when you look at AFC opponents. For most of this the Lions played 2 AFC teams at home; with the 17 game schedule they host a third every other year. That means that there has always been a “better” AFC home opponent and a “worse” AFC home opponent. Of 8 AFC Thanksgiving games, the Lions played the better team 7 times. The probability of that occuring by random chance is 3%.

Okay, so it’s scheduling. The league wants a good matchup for Thanksgiving so it has a good team play the Lions. That strategy is at least questionable. I’m not so sure it’s anyone’s idea of a good game when Peyton Manning throws six touchdowns and absolutely clobbers a hapless opponent. I can’t imagine the second half television sponsors are too pleased with blowouts. But if that’s what they’re doing, I get it: put the flashy teams on display…

…except they don’t do that with the Cowboys. Until the Cowboys hosted the Chiefs this year, the team the Cowboys played on average had a losing record. (Hosting the Chiefs this year brought their opponents record to exactly .500). So the NFL gives the Lions a playoff team, but gives the Cowboys a mediocre team.

Okay, so it’s not a curse, the league just screws the Lions.

No, there’s something else going on here. The Lions Thanksgiving opponents have a .591 winning percentage, but the Lions play good teams at home in other weeks. In non-Thanksgiving games, the Lions have a home winning percentage of .346 against teams with at least 10 wins. On Thanksgiving their record is .250. This means that the Lions are 9% less likely to win on Thanksgiving compared to other home games, even controlling for opponent strength.To put it another way, if the Lions host a playoff team on a Sunday, they’ll win about a third of the time. If they host a playoff team on Thanksgiving, they’ll win only a quarter of the time. That’s the curse. That 9% difference is the curse.

But the inequitably hard schedule - compared to the cupcakes the Cowboys get - isn’t helping.

1 Like
2

But that is the point for the NFL. In years past, they wanted to make sure at least one team will be there to keep eyes on the game, regardless of the outcome. They didn’t ever have to worry about that with the Cowboys, so their opponent mattered very little. With the Lions however, until about three years ago, it very much did matter. That is why you got Tom Brady and Peyton Manning on Thanksgiving.

Also, you have to remember that until about 5 years ago, it would always rotate between playing an NFC/AFC team because CBS and FOX only would be able to show the opposing conferences game if it was an away game. So of course CBS is going to want to showcase the best of the NFC against the Lions in that scenario.

The only way for the Lions to get out of this perpetual cycle is to become a team that people will watch, no matter what.

3

Well posted. I was thinking (I’m 62), the Lions have been AWFUL on T-day in most of my lifetime.

The 21st century has been really rough. Nice explanations / theories.

4

Yes, that’s generally understood to be the idea. If the Lions are bad, get a good team in so people will watch. But blowouts aren’t good television, either. You want people watching commercials in the fourth quarter, not doing other Thanksgiving stuff.

5

I thought there was a time the league talked about taking the game away from us because we were terrible and boring, and ford threatened to pull advertising. IIRC.

Honestly it might be time to take it away anyways. We invariably lose and then have to wait even longer for another game.

what kills me is it SHOULD be a home field advantage. we don’t travel on a short week, etc. But nope, no matter how good the team has been lately, we suck on thanksgiving.

1 Like
6

Yep–year before the Dan Campbell hire.

7

Further exploring the idea that the NFL wants put a good team on television so that more people watch. The hypothesis is that because the Lions are generally bad, they need a good opponent to get people to watch the game. So I ran a regression with the Lions’ winning percentage as the independent variable and their opponents winning percentage as the dependent variable, going back to 1980:

1000004688
1000004688512×307 22.1 KB

The P-value is around 40% so it’s not significant. What this means is that the team that the Lions play on Thanksgiving doesn’t get better or worse depending on how good the Lions are. In other words, when the Lions are good, the league does not say “enough people will tune in to see Barry Sanders run for 200 yards or Calvin Johnson catch three touchdowns, we can give them the Browns.”

Now let’s look at the Cowboys again. Hypothetically, because the Cowboys are America’s Team, people will be watching regardless of whether the team is good or bad. They ratings should be pretty consistent, regardless of whether there’s a good team playing or they both stink. But that’s not what happens:

1000004689
1000004689512×307 30.7 KB

As the Cowboys get better, their opponents get worse (p < 0.10).

So basically the league is saying that the Cowboys game only needs one good team. For the Lions game, the opponent always needs to be pretty good, and how good the Lions are doesn’t matter.

8

That’s the mysterious 9% that I mentioned. Yeah the league gives us tough opponents. But we should have won three more Thanksgiving games since 2002 because we have a decent chance of beating playoff teams at home when it’s not Thanksgiving.