SITE NEWS: We are moving all of our site and company news into a single blog for We'll tag all PFR content, so you can quickly and easily find the content you want.

Also, our existing PFR blog rss feed will be redirected to the new site's feed. ยป Sports Reference

For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.

Playoff Proposals: No Automatic Home Game for Division Winners, and “Making the Cut”

Posted by Jason Lisk on December 31, 2007

[NOTE from Doug: this post was written prior to yesterday's games. It's my fault, not JKL's, that it didn't get posted at the end of last week. As you'll see, it doesn't affect the content of the post, but you may have to mentally adjust some tenses.]

This year, seven of the eight divisions were decided prior to week 16. The only one that was not, the AFC Central, was determined when Cleveland lost at Cincinnati on Sunday. All four of the teams that would have bye weeks in the playoffs were decided prior to week 16 as well. Only six of the sixteen games in the final week of the 2007 season have any potential impact on which teams make the playoffs, or specific playoff seeding.

This year was not particularly different from other seasons since 2002, when the league expanded to 32 teams and went to an 8 division, 4 team per division format. From 2002-2006, only 30% of division runner ups ended the season either tied on record with the division champ, or within one game in the standings. In contrast, from 1990-2001, when the league featured a six division format, 50% of division runner ups ended the season either tied on record with the division champ, or within one game in the standings.

In the previous six division format, the top wildcard team would earn the #4 seed and a first round home game. Now, a wildcard team is basically prohibited from earning a home game--the only potential exception being the unlikely scenario where the #5 and #6 seed both reach the conference championship game. Most of the time, the wildcard team that earned the home game was not decided until the final weeks of the season, as 83% of the #5 seeds were within one game of the #4 seeds. The net effect of all of this is that far fewer games played in the last two weeks of the season are deciding who earns home games in the playoffs.

Of course, we are only dealing with five previous seasons with the same structure, plus this year. To truly answer whether the current schedule format and playoff structure lends itself to fewer games, we would need to build a model, like Doug did here with the current 32 team format. Then, we could compare the results of the two to see if there were substantial differences.

So maybe Doug will add this to his list of constantly expanding things to do, to see if I am wrong. My belief, despite the small sample size of years, is that the current format does systematically decrease the number of meaningful games late in the year.

Is this a problem or not? Personally, I would rather see more incentive for teams to play the full schedule, than less. Competition is a good thing, and I would like to see more of it. So, with that in mind, here are a couple of ideas that should increase the number of meaningful games late in the year. Further, at least as to the first, it is an idea that I think makes sense in the current format. Feel free to comment on my ideas, or also treat this as an open discussion on other playoff ideas, realistic or fanciful, that you have. If you think things should stay absolutely the same for all time, you can say that too.

Before I get to the ideas though, let me point out that the NFL has been a league of constant change when it comes to the number of teams, schedule, structure, and playoff format. Do you know the longest continuous stretch of years since the AFL-NFL merger where the league had the same number of teams, schedule length, divisional structure, and playoff format? You're in it right now--six years. If you discount the strike year of 1987, where the league played just one fewer game (and three with replacement players), but otherwise had the same structure and playoffs, then the answer is still only seven seasons (1983-1989). So, the NFL has shown it is willing to change and adapt.


If you want to automatically guarantee division winners a playoff berth, fine. But I fail to see how simply being the best of a historically grouped, geographically grouped, or randomly generated group of four teams automatically merits a home game, without any consideration for how that team did relative to the rest of the conference. I don't think there is any stone tablet that decrees that division winners must play at home, even if they have a worse record than a wild card qualifier.

Take Jacksonville this year. Why shouldn't they have an opportunity to be the #3 seed, when they are probably the third best team in the AFC? They beat both San Diego and Pittsburgh this season, and will likely finish with a better record despite playing in a tougher division, and yet get rewarded by having to go back on the road against one of those teams.

If wild card teams could earn higher seeds than #5, then the increased competition for home games would increase the number of meaningful games late in the year. Do you think Tampa Bay would have been so laissez faire in the game at San Fransisco if they had to make sure they finished ahead of New York in order to get a home game in the first round, rather than travel to New York in January? Think back to the 2005 season. Would everyone in the AFC have played hot potato with the #3 and #4 seeds in the final weeks if losing a home game was an option for the division winners? New England and Cincinnati had clinched their respective divisions, and the "winner" of the #3 seed would likely get Pittsburgh. New England, 10-5 entering the game, played an exhibition game against Miami at home and lost in overtime with Matt Cassell throwing passes, and Cincinnati went to Arrowhead and got thumped while resting starters. Under my proposal, Jacksonville would have been playing for a home game as well that weekend, and Cincinnati and New England would have both had incentive to compete and earn a home game (against each other) in the first round.

To seed the teams, I would just use the current tiebreaker rules, break ties within divisions first, then across divisions, but allow a 2nd place (or 3rd place finisher) to finish ahead of another division winner, once their own division winner was seeded, if they had a better record or won the tiebreaker. These are the seeds, heading into week 17, under my proposal:

1. NEW ENGLAND (15-0)
4. SAN DIEGO (10-5)
5. PITTSBURGH (10-5) (would move to #4 with win and SD loss)
6. TENNESSEE (9-6)

1. DALLAS (13-2)
2. GREEN BAY (12-3)
3. SEATTLE (10-5)
4. NY GIANTS (10-5)
5. TAMPA BAY (9-6) (would move to #4 with a win and NY Giants loss)


I'll admit that this next one is the more fanciful of the two proposals. There has been some discussion in the recent past of adding a seventh (or eighth) playoff team in each conference, though I am not sure how serious these discussions were. The problem, as I see it, is that in some seasons, there are not seven legitimate playoff teams that would be reasonably competitive, while in some, there are.

I will borrow an idea from golf. In golf, a minimum number of golfers are guaranteed to make the cut after the first two rounds. Let's say there are 144 golfers in the field. 72 golfers (plus ties) may be guaranteed to make the cut, regardless of score. In addition, more golfers can make the cut if they are within 10 strokes of the leader. So, the leader can influence how many more players make the cut.

Applying this concept to the NFL, my proposal would be as follows:

1. Six teams are guaranteed to make the playoffs each year, just like now.
2. The seventh-ranked conference opponent makes the playoffs only if they are 8-8 or better, and are within 3 games of the #2 seed in the standings;
3. The eighth-ranked conference opponent makes the playoffs only if they are 8-8 or better, and are within 3 games of the #1 seed in the standings;
4. The league could offer these potential additional games to NBC, or, if they really wanted to start a riot, put them on the NFL Network.

Thus, the top seeds may have incentives to keep winning in the final two weeks, even if they have already clinched a top two seed. If you are getting a #2 seed at 11-5 thanks to tiebreakers, well, you likely wouldn't get a first round bye under this system. Applying this rule to this season, Cleveland would clinch a playoff spot with a win over San Fransisco by getting to 10-6. Indianapolis could eliminate Tennessee by winning at home, and earning a first round bye, but a loss would put a 13-3 Indianapolis team at risk of having to play 10-6 Cleveland in the first round. In the NFC, Green Bay would be playing a must win game to guarantee a first round bye. A loss, coupled with a Minnesota and Washington win, would mean Minnesota gets in and gets to play in Green Bay.

Here are the additional playoff games that would have resulted from this proposal, since 2002. I've added the team's regular season conference SRS ranking in parentheses:

2002	AFC #7 Denver (4), 9-7 at #2 Tennessee, 11-5
2002	AFC #8 New England (5), 9-7 at #1 Oakland, 11-5
2002	NFC #7 New Orleans (5), 9-7 at #2 Tampa Bay, 12-4
2003	AFC #7 Miami (7), 10-6 at #2 Kansas City, 13-3
2003	NFC #7 Minnesota (6), 9-7 at #2 St Louis, 12-4
2004	NFC #7 New Orleans (12), 8-8 at #2 Atlanta, 11-5
2005	AFC #7 Kansas City (5), 10-6 at #2 Denver, 13-3
2005	NFC #7 Dallas (5), 9-7 at #2 Chicago, 11-5
2006	NFC #7 Carolina (6), 8-8 at #2 New Orleans, 10-6

I included the SRS conference ranking to see what types of teams would be getting in with this change. For the most part, they are teams that had pretty good SRS ratings that may have played tough schedules or lost out on tiebreakers. Seven of the Nine were in the top 6 in the conference in SRS ranking (which is another post for another day). Only 2004 New Orleans was not seen as at least average for the conference by the simple rating system.

While nine more teams would have made the playoffs under this system in retrospect, not all nine of those teams would have made it if the system was actually in place. Atlanta basically shut it down at 11-3 in 2004, having clinched the #2 seed. Saint Louis 2003 was 12-3 going into week 17, and lost at 4-11 Detroit. Last year, New Orleans lost at home to Carolina in week 17 when they rested their starters. The nine higher seeded teams above went a combined 11-7 over the last two weeks of the regular season.

Combine these two proposals, and ten of the sixteen games would have some meaning this weekend, which means that fewer NFL season ticket holders (who already have to pay for two preseason games as part of the ticket package) are subjected to a third. You are still going to have some dog games like my Chiefs at Chase's Jets, but there will be fewer of them. Jacksonville would be the #3 seed after a very successful season in which they have gone 11-4 in the toughest division, rather than San Diego, who has somehow managed to finish ahead of Denver, Kansas City and Oakland. Indy and Green Bay would have to win to insure a first round bye and potentially eliminate a division rival. Seattle, New York and Tampa in the NFC would be fighting for the #3 and #4 seeds and a home game in the final week. San Diego and Pittsburgh would be in must win situations to earn the #4 seed and a home game. Cleveland would earn a playoff spot if they get to 10 wins, and Washington, Minnesota, New Orleans, and Tennessee would be fighting for slots as they are now.

So, that's it. Throw out your crazy or not so crazy playoff ideas, or tweak and critique mine. Happy New Year, and bring on the playoffs.

This entry was posted on Monday, December 31st, 2007 at 7:38 am and is filed under General, Insane ideas, Rule Change Proposals. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.