Writers of Pro Football Prospectus 2008

12 Jul 2012

Chicago Bears: I'm Just Sayin'

Easiest Projected Schedule in Football Outsiders Books, 2005-2012
Book/Year Team W-L
Prev Yr
That Year
Pro Football Prospectus 2005 Bears 5-11 11-5
Pro Football Prospectus 2006 Bears 11-5 13-3
Pro Football Prospectus 2007 49ers 7-9 5-11
Pro Football Prospectus 2008 Patriots 16-0 11-5
Football Outsiders Almanac 2009 Patriots 11-5 10-6
Football Outsiders Almanac 2010 Chiefs 4-12 10-6
Football Outsiders Almanac 2011 49ers 6-10 13-3
Football Outsiders Almanac 2012 Bears 8-8 ???
Average Wins   8.5 10.4

No guarantee, Bears fans, but I'm just sayin'.

That effect would be even stronger if you didn't have a 16-0 team that lost the league MVP to injury in Week 1.

This other one is not quite as strong, especially because of the 2008 Steelers, and of course because of Peyton Manning.

Toughest Projected Schedule in Football Outsiders Books, 2005-2012
Book/Year Team W-L
Prev Yr
That Year
Pro Football Prospectus 2005 Chargers 12-4 9-7
Pro Football Prospectus 2006 Broncos 13-3 9-7
Pro Football Prospectus 2007 Bills 7-9 7-9
Pro Football Prospectus 2008 Steelers 10-6 12-4
Football Outsiders Almanac 2009 Broncos 8-8 8-8
Football Outsiders Almanac 2010 Bengals 10-6 4-12
Football Outsiders Almanac 2011 Chiefs 10-6 7-9
Football Outsiders Almanac 2012 Broncos 8-8 ???
Average Wins   9.8 8.0

How odd is it that the Bears have been on the "easiest projected schedule" list three times, and the Broncos have been on the "hardest projected schedule" list three times?

Football Outsiders Almanac 2012 is available here.

Posted by: Aaron Schatz on 12 Jul 2012

29 comments, Last at 16 Jul 2012, 7:56pm by BroncosGuyAgain


by tuluse :: Thu, 07/12/2012 - 2:18pm

I like hearing it

by Steve in WI :: Thu, 07/12/2012 - 5:20pm

Yeah, especially since the Bears have made so many moves that make me think they'll be much better than they were in 2011 regardless of the schedule. If the important guys stay healthy, I'm feeling really good about this season.

by snoopy369 :: Thu, 07/12/2012 - 2:36pm

I'd assume the Bears are on there because they frequently a) have relatively weak schedule in the NFC North and b) some years have matchups against other large weak divisions (NFC West, AFC South, etc.)

On the other hand, the Broncos I can't explain. The Chiefs and Raiders have been reliably bad lately...

by tuluse :: Thu, 07/12/2012 - 2:38pm

And why don't the Lions or Packers have an easier schedule. Surely you would think 2 games against the Bears vs the Packers would affect things.

by justanothersteve :: Thu, 07/12/2012 - 4:10pm

Because the Packers and Lions have to play the #1 and #2 teams respectively of the other NFC division teams. The Packers get the Saints and Giants. The Lions get the Falcons and Eagles. They all get the sad AFC South and NFC Worst (West). The Vikes may get the #4 teams, but also get the Packers, Lions, and Bears twice each. I'm sure all four teams are amongst the easiest schedules.

by JasonG (not verified) :: Thu, 07/12/2012 - 2:57pm

Interesting info and all, but when I clicked the headline I thought this was going to be in depth about the Bears, how you expected Marshall/Jeffrey/Tice (and don't forget about Bennett's return, not to mention Campbell/Bell depth) infusions would lead to vast improvement on offense and perhaps also how McClellin might unleash Peppers, thus projecting a savage D once again. O-line and Forte unknowns make it much harder I know, but I've heard every other site's blah blah about the new Emery Bears, but really I want FO's take. Got to buy the book I guess.

by chemical burn :: Thu, 07/12/2012 - 3:40pm

Isn't this sorta weird though in the sense that it's the PROJECTED easiest schedule and not the actual easiest schedule? How often do those end up being one and the same? I can't imagine it's all that frequent. I guess maybe having an easy if not the easiest schedule is enough to account for it, but I'd be curious for the same kind of record comparisons for the Top 5 or so ACTUAL easiest schedules end up being.

The kind of "this actually isn't meaningful" noise that I would expect does show up in the "Projected Toughest" table, so there's a reason to believe that this is just coincidental. I mean, as it stands, the above table only shows 3 instances of a team dramatically improving with the easier schedule - and 2 of them are associated with coaching changes.

(Even some of the smaller improvements like the Bears going from 11-5 to 13-3 go hand-in-hand with a significant improvement like switching from a woefully raw Orton to the intended starting QB Rex Grossman.)

by chemical burn :: Thu, 07/12/2012 - 3:45pm

And now that I'm looking at it and not just accepting the commentary, the Projected Tougher list contains just as much if not more evidence that having a projected tough schedules hurts your team than a projected easy one helps.

(And I keep forgetting that I disagree with how FO calculates strength of schedule. Consider me in the "this info is pretty much useless camp.")

by some guy (not verified) :: Sat, 07/14/2012 - 12:21am

I do know that the 2009 patriots went from having the projected easiest schedule to having actually the hardest schedule. But I think that it doesnt matter, what they are trying to say is that most teams with the easiest project schedule had ~8-8 records the previous year and had ~10-6 records the next year and obviously the bears had an 8-8 season last year so I'm just saying...

by chemical burn :: Mon, 07/16/2012 - 12:47pm

Yeah, but this is exactly what I'm talking about: the Pats went from projected easiest to projected hardest and their win total went down - well, what a freakin' surprise. If a team ends up with an easier schedule, I think we can all agree that increases their chance of winning. If a team ends up with a harder schedule, that most likely decreases their chance of winning. But throwing in nonsense about "projected" schedules is just causing unintelligible noise. There's a few very good reasons to think the Bears will be better this year: most notably the Cutler from Hanie improvement. And if their schedule does in fact end up being the easiest (and not the hardest a la the 2009 Pats) then there's a great chance they will end up 10-6 or better. But that doesn't make this table useless junk.

by DenverCheeze (not verified) :: Thu, 07/12/2012 - 3:53pm

Given the Pack and Lions are both playoff teams from last year...are the Vikings THAT bad to balance out the schedule for Da Bears and make it so easy? How is this calculated? Formulas please!

by chemical burn :: Thu, 07/12/2012 - 4:00pm

Their method is pretty worthless - team A plays 3 above average teams and 1 horrendously terrible team has the same SoS as team B which plays four mediocre teams. Clearly, those schedules have 2 different strengths...

by Arkaein :: Thu, 07/12/2012 - 5:24pm

The SoS method may not be perfect, but the chart is strong evidence that it is far from worthless.

A worthless SoS would predict random or inversely correlated results.

by chemical burn :: Thu, 07/12/2012 - 9:37pm

The charts above? They're evidence of virtually nothing. The Easiest table features 2 teams that had slightly worse records (1 or 2 games), 1 team with a much worse record (5 games), 3 teams with much better records (6 or 7 games) and 1 team with a slightly better record (2 games).That's as close to random as I could come. It's literally a 3 to 4 split on improvement vs. decline - and as I point out above, 2 of the 3 with wild improvement got new coaches so that clouds the source of improvement further.

On the Toughest table, we have 2 teams that kept the same records, 1 team with a slightly better record (2 games), 3 teams with significantly worse records (3 or 4 games) and 1 team with a much worse record (6 games.) Again, that looks very close to random to me - although much better than the Easiest table (despite the author of this little commentary's analysis to the contrary.) That's 2-1-4 between same/improve/decline - hardly decisive evidence of anything.

But more importantly, these tables aren't based on ACTUAL SoS, but on preseason PROJECTED SoS. So they really truly tell us nothing about FO's ability to calculate SoS. Literally nothing. FO's SoS numbers aren't featured anywhere up there.

by Jimmy :: Fri, 07/13/2012 - 8:47am

I would take a different point from that chart (I am a Bears fan after all). The info I take is that apart from an appalling Niners team, every other team on it has won at least ten games. Thats the playoffs baby, wild card at least.

by tuluse :: Fri, 07/13/2012 - 11:48am

Agreed, people are too focused on the change in wins rather than the absolute number of wins.

by Paddy Pat (not verified) :: Sun, 07/15/2012 - 7:27pm

2003 Dolphins? 2008 Patriots? What are you talking about? Since 2003, 10-6 has only a 5 in 6 chance of making the playoffs.

by Ferguson1015 :: Thu, 07/12/2012 - 11:17pm

Huh, the entire AFC West dominates the "Hardest schedule" I'm surprised the Patriots had the easiest schedule the year after they went undefeated. It seems like something is wrong with that.

by Brent Hutto (not verified) :: Fri, 07/13/2012 - 9:25am

You really have to account for the likelihood of very good or very bad team regression toward 8-8, at least somewhat, no matter what their schedule.

So for instance the 5-11 Bears are more likely to go 6-10 than 5-11 the following season regardless of factors that represent improvement or not in the quality of their team. And the 16-0 Patriots were far more like to go 15-1 (or 14-2 or 13-3) the next season than have another 16-0.

As a quick-and-dirty way of assessing that I took your "W-L Prev Yr" column and moved it one game toward 8-8. So a 10-6 team becomes 9-7 and a 6-10 team becomes 7-9 (and 8-8 team stays 8-8). Then I calculated the total plus or minus chance in each listed team's schedule relative to that corrected expectation.

Turns out not to matter much. Your table shows the easy schedule teams improving by 1.9 games, with my (admittedly crude and minimal) correction that becomes +1.7 games instead. And the tabulated -1.8 games for the hard schedule teams becomes -1.4 games.

That said, even if those differences hold up when someone really accounts for regression to the mean (HINT HINT!) we're talking a game or two on average. Which sounds reasonable for "easy" or "hard" schedules, doesn't it?

by Brent Hutto (not verified) :: Fri, 07/13/2012 - 9:33am

If it were a rainy Sunday morning and I had nothing better to do, I'd go back and look up historically what the average following-year record was for 5-11 teams and so forth. Use that as my "expected regression to the mean" column then compute the plus or minus games difference for the easy and hard schedule teams each year. That should be a pretty fair estimate of the "easiest effect" and "hardest effect".

by dmstorm22 :: Fri, 07/13/2012 - 10:06am

This isn't a distribution of similar teams. In the NFL every team doesn't have an equal chance of going 8-8. I don't think there is any evidence that teams tend to go towards 8-8. Remember that NFL teams are not normal. In most season there are as many 10+ win teams as there are with 7,8,9 wins.

Also regression to the mean isn't about regressing to 8-8 but to the true level of the team. Some of these 5-11 teams could have been lucky to go 5-11 and "regress" to their natural mean level which isn't necessarily 8-8.

Also NFL season records are stochastic. The team quality may be more deterministic but in general a team that goes 5-11 given their actual level (using Pythagorean projection or dvoa or whatever) is around 5-11 is no more likely to go 6-10 than out is 4-12.

by Brent Hutto (not verified) :: Fri, 07/13/2012 - 11:02am

None the less given a 16-0 record one year, surely the best guess for the following year's record is not 16-0 is it?

by dmstorm22 :: Fri, 07/13/2012 - 11:09am

Yes, because no team is actually 16-0 good. The only way to truly have an expected record of 16-0 is to shutout all your opponents. Even getting an expected record of 15-1 is really difficult. I don't have the new pythag information in front of me, but I believe using the original method (with an exponent of 2.37) the record in a 16-game schedule was the '85 Bears, which was still under 15 wins.

A team that goes 16-0 is unlikely to do it again because they almost assuredly overachieved what would normally be expected from a team. They wouldn't not go 16-0 because teams regress towards 8-8, they would be expected to go less than 16-0 because they aren't 16-0 good.

by dbostedo :: Fri, 07/13/2012 - 9:59pm

To add to what dmstorm22 said, think of the "expected wins", the "mean", of a team as the average number they'd have if they played the same season over and over again a large amount of times. That number, depending on team quality would most like be between 4 and 12 wins, with a few outliers above and below, and with the largest group in the middle range of 6-10 wins.

Those are win totals team are said to regress toward, not 8 wins. For instance, if a good team, like the 2007 Patriots, could play their season thousands of times, they have a lot of seasons of 16-0. But they'd have more with 15, 14, or 13 wins. And they'd have a few where they wind up with 8 or 9. So their "expected wins" might average out to something like 13 for their mean number, or expected wins. That would be the most likely outcome.

But we only can see 1 season. So when we see a 16-0 season, we expect then to be worse the next year, because 16-0 is playing above what would be expected for their mean number of wins. By the same token, if they went 8-8, you might look at history and players and know they're a good team; In that case you might say they'd regress toward 13-3 the next year from 8-8.

by Revenge of the NURBS (not verified) :: Fri, 07/13/2012 - 3:28pm

I'm not sure how much stock to put into these tables as a whole, but I definitely don't put any stock into the data points for the 2006 Bears and the 2009 Patriots. Those teams supposedly had the easist projected schedule two years in a row, so a change in SOS can't possibly explain the change in record in the second year.

by The Ninjalectual :: Sat, 07/14/2012 - 3:45am

Teams with the easiest projected schedule win on average 2 more games per year than the year previous? I would say that's significant.

by mvhuber :: Sat, 07/14/2012 - 7:51am

Met Harry Carson at a charity event a couple weeks ago. When I mentioned the Giants could have a tough year due to SOS, his response was "That doesn't matter. You still have to play the game!!" Apparently, it doesn't matter if you play Florida State or FAMU. Man, I wish I was old school.

by Revenge of the NURBS (not verified) :: Mon, 07/16/2012 - 9:07am

That's not really old school. I'd bet that any current player or coach would say the same thing. They get graded by wins and losses, not by how they performed in relation to their SOS.

by BroncosGuyAgain :: Mon, 07/16/2012 - 7:56pm

This fails every possible test of statistical validity. Just sayin'.

Still, a fun note.

Good luck to my Bears-fan friends (of which I have many).