≡ Menu

Should Division Winners Get Home Field?

In the 2013 playoffs, the 12-4 San Francisco 49ers had to play a road game against the 8-7-1 Green Bay Packers. It’s hard to justify situations like this, especially when the NFC West was one of the best divisions in football history and the NFC North was below average. Division winners are automatically given a home game under the current NFL structure, but the NFL is reportedly open to modifying that rule. I thought this would be a good idea, with one caveat; after further investigation, I no longer even have a caveat.

Let’s start with an acknowledgment: there is no “traditional” NFL playoff format. The NFL has modified its playoff system more times than anybody other than Jason Lisk can remember. There are no automatic rules regarding home field: in 1934, West champion Chicago went 13-0 and had to play the East champion Giants in New York because the league alternated which division was the host each year, and even years were East years. Two years later, the 10-1-1 Packers had to play at 7-5 Boston for the same reason.

Even in more recent times, it was not unusual for the best team in the league to have to win a road playoff game. In the 1968 playoffs, the famed 13-1 Colts — thought to be one of the greatest teams in football history prior to Super Bowl III — had to win in Cleveland to win the NFL championship and earn the right to play in the Super Bowl.1 The 1972 Dolphins (14-0) had to play at Pittsburgh (10-4) in the AFC Championship Game, as that was the AFC Central’s year to have first priority.

Even a wildcard hosting a division winner is not unprecedented. 1982 was a very weird year because of the strike, but you’d be hard-pressed to find a weird result than this one: the NFC Central champion Packers (5-3-1) had to play at NFC East runner-up Dallas (6-3), while the NFC Central runner-up Vikings (5-4) got to host the NFC West champion Falcons (5-4). That’s because following the players’ strike, the league simply seeded teams 1-8 that year because of the abbreviated schedule.

On one hand, that just means that forcing the best teams to play on the road is not a “modern” issue. Even excluding championship round games, prior to 2013, there were 28 times where the home team in a playoff game had the worst record. But this “problem” essentially solved itself with the inclusion of a second wild card team. A team with a better record played a road game just twice between 1980 and 2001: the 11-5 Giants played at 10-6 San Francisco in the second round of the 1993 playoffs and the 11-5 Vikings also played at a 10-6 San Francisco team, this time in the second round of the 1988 playoffs.

But realignment in 2002 has turned this rare circumstance into an inevitability. With four divisions of four teams, the probability of the teams with the best four records in a conference being evenly distributed across the four divisions is pretty low. As a result, 15 times from 2002 to 2012, a wild card team went on the road against a division winner with a worse record. The Packers made it 16 times in just 13 years. The most egregious examples all resulted in losses: the 12-4 Steelers lost at the 8-8 Broncos in 2011, the 11-5 Saints lost at the 7-9 Seahawks the year before, and the 12-4 Colts lost at the 8-8 Chargers in 2008.

There is no reason why a division winner should be guaranteed a home game: that is not an immutable law of NFL history, and rewarding a team with a playoff berth is incentive enough.  Initially, I thought there might be one catch. In the vast majority of cases, when a team wins a division with a bad record, that division is a bad division. In the examples above, the 2011 AFC West went 19-21 outside the division while the AFC North went 25-15; in addition to having a better record, the Steelers posted that better record in a more competitive division. The 2010 NFC West was an abysmal 13-27 in non-division games, which makes the 7-9 Seahawks hosting the 11-5 Saints (NFC South went 24-16) even more egregious. The 2008 AFC West was 11-29, while the South that year went 26-14. Putting a 12-4 Colts team on the road at a 8-8 Chargers team is hardly defensible.

But what happens when a division is so good that it beats itself up? In other words, it would be unfair for a 10-6 team from a weak division to earn home field against a 9-7 team from the best division in football. As it turns out, to my surprise, this doesn’t really happen. The closest example I could find was in 2002, when the 9-7 Jets hosted the 10-6 Colts and the AFC East (23-17) posted a better record than the South (19-21). As it turned out, New York’s opponents had an average winning percentage of exactly .500, tied for the 15th hardest in the league, while the Colts were tied for having the 28th most difficult schedule, with an average opponent posting a 0.479 winning percentage.

Ironically, the Colts were in this same position last year, as 11-5 Indianapolis went on the road to face 10-6 Baltimore (to be fair, the Ravens did rest starters in a meaningless week 17 loss, so the Ravens may have finished with 11 wins if different rules were in effect). The AFC South was only negligibly worse (19-21) than the AFC North (21-19), but the Colts average opponent last season won just 44.1% of their games, the third lowest rate in the league (Baltimore’s was at 49.6%).

But this is not a significant cause for concern. In general, if the second best team in one division has a better record than the best team in a different division, the wild card team probably came from the tougher division, anyway. I could get behind a one game buffer rule, where the wild card team only gets home field if they won at least 1.5 more games than their opponent. That would prevent the most serious injustices while also (1) preventing any inequities going the other way, and (2) perhaps making week 17 more interesting.

If you’re the type of person who believes winning your division should guarantee a home game, allow me one last chance to persuade you.  And I’ll do that by explaining how we ever got to a place where a division winner was given home field.

Beginning in 1937, the NFL implemented a uniform 11-game schedule. There were five teams in each division, so each team played its division opponents twice and had one game against 3 of the 5 teams from the other division. That means 73% of all games played were division games. In that world, it’s easy to envision a scenario where one division is weaker than another, and as a result, a good team in a bad division would post a better record than a great team in a tough division. This arguably happened in 1939.

That season, 4 of the 5 teams in the NFL West went .500 or better. There were only 15 non-division games, but the West went 9-5-1 in those games. So how much did it mean that the West champion Packers, who went 9-2, had a worse record than the East champion 9-1-1 Giants? That’s probably why the league decided to just alternate the site of the championship game: after all, the NFL couldn’t confidently determine which team was better based just on record, and this was a fair approach.

1965 is another good example. That year, there were two 7-team divisions and a 14-game schedule. Each team played its six division rivals twice and had two games against teams from the other division, meaning 86% of all games were division games. Just one team in the NFL East had a winning record: Cleveland, at 11-3. The Browns faced an average team with a 0.383 winning percentage. Does having one extra-half win against that schedule make the Browns better than the 10-3-1 Packers, whose average opponent went .500? Of course not. There were only 14 inter-division games that year, but the NFL West went 13-1 in those games. That’s as stark a disparity as you can imagine, and the league’s rotating structure worked out: the Packers got to host the title game.

It is much more difficult to infer which division is best when the vast majority of the games are intra-division affairs. But now, only 37.5% of games meet that definition. As a result, concerns about a tough division producing a 9-7 champion are overblown: that team either had a losing record inside the division or went no better than 6-4 outside the division. Yes, it’s conceivable that a 10-6 wild card team could be somewhat fraudulent because of a weak division — it might run up a 5-1 record by sweeping the bottom-feeders and splitting with the champ — but that still means they at least won half of their games outside the division.

There will always be teams that have better records than teams due purely to strength of schedule. But in the modern NFL, that’s rarely because of a weak division. It’s time to eliminate automatic home playoff games for division winners.

  1. After realignment in 1967, the West division was selected to host the title game in ’67 (the Ice Bowl) and ’69 (Vikings over Browns), while the East would get home field in ’68 and … well, the East drew the short end of the stick when the AFL-NFL merger took effect in 1970. []

{ 10 comments… add one }

  • ammek February 20, 2014, 6:35 am

    Not that it changes the substance of your argument, but the “weird” example you chose from 1982 is a little flawed. The Vikings (5-4) hosted Atlanta in the first, ‘wildcard’ round, while the Packers (5-3-1) didn’t travel to Dallas until the second, ‘divisional’ round. Green Bay previously won a wildcard game at home to the St Louis Cardinals.

  • ammek February 20, 2014, 6:40 am

    Re-reading, I get your point. It’s not a “flawed” example at all, though it is a little confusing. Heck, the whole 1982 season is confusing. But the NFL did away with divisions after week two anyway, didn’t it?

    • Chase Stuart February 20, 2014, 12:08 pm

      Yeah, 1982 is always an oddball.

  • Justin February 20, 2014, 12:03 pm

    I feel like this would take away the advantage of a team that wins a division over its division opponents. Let’s say we have a team that finishes in the third seed with a record of 11-5 and the next best team in their division finishes 10-6 and goes to play a 4th seed who ended up 8-8. This means that the division winner gets to host a playoff game and the runner up has to go on the road. Under your system there would essentially be no difference between the team that won the division and the team that finished runner up. This problem will only be exacerbated if the NFL expands the playoffs. This past year it wouldn’t have mattered who won the NFC South. Both teams would have ended up hosting a playoff game in the first round. The Week 16 game would have been mostly irrelevant.

    • Chase Stuart February 20, 2014, 12:07 pm

      That’s a reasonable point, but it seems minor in the grand scheme of things. Sure, the battle for the 3/4 seed between two division winners might be of slightly less import, but that seems like a worthy trade to avoid an 11-5 team going on the road to face a 7-9 team in the playoffs.

      And there would still be some importance to that game. The CAR/NO battle was as much for the first round bye as for anything else. Now your example is different (a 3/4 seed), but still, there is some value to being the 3 over the 4.

    • Ian February 20, 2014, 4:37 pm

      You don’t think it’s important to avoid a team like Seattle until the conference championship? If you want to go by merit, your problem is not a problem at all, but a feature. With everyone agreeing that San Francisco being the second-best NFC team, and probably NFL team, the lack of fortune of being in the same division as the number one team shouldn’t be as punishing as it was. In the NCAA basketball tournament, for example, you’ve had up to 3 teams from the same conference get #1 seeds, and the conference championship is only a small prize.

  • George February 20, 2014, 4:15 pm

    Just a quick thought on this; I think it would be worth trying to determine exactly what home field advantage was worth? In my personal opinion I feel it has determined who has represented the NFC to certain extent in the last two years (this year I’m not 100% sure – but if San Francisco has Green Bay come to them does it make a difference when they end up playing Seattle?, but last year I feel if Seattle hadn’t had to travel to Washington for a Wild Card and then go back across half the country to Atlanta (or around 6000-7000 miles travel) and say at least the Washington games goes to Seattle – I don’t know if this quite works out, Seattle goes to San Francisco with a lot less travel behind it and something different might have happened I think) – you know where I am going with this.

    I definitely feel that there is merit in looking at the value of home advantage (and have found a different way of doing this on an individual basis, by combining it as part of my solver model but it is important to look at it as a long term trend e.g. since say 2002) and I think the approach suggested whilst it would change the relevance of certain situations, I guess it would be a reward for being the best team that you can be.

    Just thinking about it, you can have ratings for teams at the end of a season, I have home field advantage values I feel okay with – you can stick it in a play-off model and see what happens in terms of percentages? I’ve worked a model up for the existing set-up of how the play-offs work, but I’m sure it could be tweaked.

  • eagle97a February 21, 2014, 12:05 am

    A bit off topic but we’re forgetting one thing, the owners will always think about the potential profits of a home playoff game and that’s a roadblock to any change in the playoff format.

  • EranUngar February 21, 2014, 10:16 am

    I tend to agree with you that if a team finishes 2nd in it’s division and is still more then 1 game ahead of a division champion it should host the wildcard game.

    However, as for your logic you missed a strong argument against it. It’s not only the inter division games that effects the strength of schedule it’s also which other conference division your division faces. Next year the best division of the AFC plays the best of the NFC west against – AFC west. You’ll find 5 playoffs teams plus another team with 10W that didn’t make the playoffs. After such a season winning your division even at 9-7 is a big thing and should award you home field advantage.

    Also, it would make winning your division practically meaningless.

  • Justin February 21, 2014, 11:19 am

    Another suggestion that is worth looking into is trying to delay intra-division playoff games as long as possible. In the pre-1994 expansion of the playoffs, a wild card team could only play its division champion in the Conference Championship game. For example, if an NFC East team won the wild card game and the NFC East champion was the No. 1 seed, the NFL simply sent the NFC East wild card team to play at the No. 2 seed and sent the No. 3 seed team to the NFC East champion / No. 1 Seed. This format resulted in some classic title games between heated rivals. Unfortunately, this would be impossible to guarantee under the current format, but there are several times when a slight tweak here and there would delay intra-division games. In my mind, this is the best possible change to the current playoff rules.

Leave a Comment