/cdn.vox-cdn.com/uploads/chorus_image/image/65346526/1170850770.jpg.0.jpg)
One of the discussions in which Kansas City Chiefs fans have been involved this week is whether this Sunday’s opponent — the Detroit Lions — are an undefeated team; the Lions’ record is 2-0-1 because their Week 1 game against the Arizona Cardinals ended in a tie.
“Nobody has defeated them,” say some fans, “so technically they’re undefeated.”
“Nonsense!” say other fans. “That tie will matter when the playoffs arrive!”
Both sides have a point.
But as I listened to fans go back and forth about it this week, I realized that a lot of people don’t really understand how it works — and it’s kind of interesting how things came to be the way they are.
Through 1972 — when tie games in the NFL regular season were much more common — ties were simply ignored when calculating a team’s winning percentage; it was as if the tie game had ever even happened.
But this created some problems.
Suppose two teams both lost only only twice in a 14-game regular season — but one of those teams also had a tie on their record. The record of the team without a tie would be 0.857. (12 divided by 14 is 0.857). The winning percentage of the team with a tie would be 0.846 (11 divided by 13 is 0.846).
So far, so good... right? The team without a tie has a better record.
But starting in 1973, the NFL decided to treat a tie as half a win and half a loss. So the winning percentage of the 12-2-0 team would remain unchanged at 0.857, but the record of the 11-2-1 team would fall to 0.821 — which is the result of 11.5 divided by 14.
As Andy Piascik wrote for ProFootballResearchers.org in 2005, it was sound reasoning.
A tie is really more half a win and half a loss than a game not played. To cite one admittedly extreme hypothetical example that was possible under the old system, a team with a 1-0-13 record would have finished ahead of a team with a 13-1 record. Under the new system, such a team would finish not with a 1.000 winning percentage, but one of 0.536 that far more accurately reflects its season.
Of course, the likelihood that a team would finish a season with 13 ties is ridiculously small.
But it’s not hard to imagine two teams with records of 13-1-0 and 11-0-3. Under the old system, the 13-1 team would have a winning percentage of 0.929, but the team that had three ties would have a winning percentage of 1.000. Would it be fair for the 11-0-3 team to win a division title against a 13-1-0 team? Under the new system, it wouldn’t. Its winning percentage would be 12.5 divided by 14 — or 0.893.
So in the way that we think of the word defeated, the Lions are indeed undefeated; no team has beaten them. If they had held the same 2-0-1 record in going into Week 4 of 1972, no one could challenge a claim they were undefeated, because the tie just wouldn’t have counted. But given the way the NFL now calculates winning percentages, the Lions currently have half a loss on their record; the word undefeated really doesn’t (and shouldn’t) apply.
It’s really too bad that back in 1973, the NFL didn’t abandon the traditional W-L-T method of expressing a team’s record in favor of a W-L format that included decimals. Nobody would be arguing the Lions are currently undefeated if their record was 2.5-0.5 — but in any way that matters, that’s exactly what it is.