49ers 2011 Strength of Schedule: An Early Look

A couple weeks ago, Mike Sando put together a ranking of the strength of schedule for all 32 NFL teams heading into the 2011 season. The order was based on 2010 record, given that we have nothing else to go on. Based on that category, the 49ers have the third easiest schedule next year. Their opponents had a combined record of 119-137, which comes out to a .465 winning percentage. The only teams with a weaker schedule are the Baltimore Ravens (.457) and the Arizona Cardinals (.441).

Sando included a disclaimer about past records not necessarily reflecting future prospects. As we all know, a 2011 NFL team is generally going to be different from its 2010 counterpart, whether it be a few additional players or a complete overhaul. In April 2009, immediately after the draft Florida Danny put together a post discussing the 49ers trade with Carolina that netted a 2010 first round pick. At the bottom of that post (section titled Buyer Beware) he wrote a few paragraphs about the issue of strength of schedule. Given that we're now sitting here two years later discussing the issue, I asked him to provide an update on the strength of schedule issue. I've posted that info after the jump.

When it comes to strength of schedule, I do think there is some subjective value in looking at the individual records of each opponent as part of a greater analysis of each team. For example, the 49ers host the 12-4 Pittsburgh Steelers in 2011. We know the Steelers are a good football team and the record is only one indicator of how tough they'll be next year. On the other hand, the 49ers travel to the 6-10 Detroit Lions next year. The Lions have a variety of holes to fill this offseason, but if they can make the right moves I think they'll be a tougher opponent in 2011. As a part of the bigger picture analysis, I believe the previous year's record can at least be a part of the discussion. We just can't take it as gospel.

At the bottom of the 2009 post, there's a table showing the year-to-year correlations between a team's SOS based on last year's records, its actual SOS, and its Win%. The table clearly shows that although a team's Win% in a given year is highly dependent on its SOS, it has basically no relationship with what their SOS was predicted to be based on last season's records; and that's because predicted SOS itself doesn't have any relationship whatsoever with actual SOS.

I did that post after the 2008 season, so it only included 2004-2008. Here's a little updated version of the table that runs from 2003-2010, which is all seasons under the current divisional alignment:

Year

SOSPre -> SOS

SOSPre -> WPct

SOS -> WPct

2003

0.217

0.034

-0.760****

2004

0.104

-0.062

-0.361**

2005

0.125

0.045

-0.659****

2006

0.193

-0.148

-0.590****

2007

0.161

0.150

-0.345**

2008

0.408***

0.079

-0.561****

2009

0.429***

0.042

-0.606****

2010

0.312*

0.029

-0.574****

Average

0.244

0.021

-0.557

-> means "predicts"
* means prediction is statistically significant with 90% confidence
** means prediction is statistically significant with 95% confidence
*** means prediction is statistically significant with 99% confidence
**** means prediction is statistically significant with 99.9% confidence

Three other things I'd add about this table are that, even if you were to conclude from the past 3 seasons that SOSPre might be getting more predictive of SOS lately,

1.       Remember that, when people bring up SOSPre in conversation, they're implicitly saying, "well, Team X is likely to get better next year because they're going to face an easier schedule." That's what the SOSPre -> WPct correlations are answering, and - based on those correlations -WPct cannot be predicted whatsoever from SOSPre. The statement that is perhaps getting more and more accurate with time is, "Team X is likely to have an easier schedule next year because their opponents on this upcoming year's schedule were worse last year." But who cares about that if that easier schedule doesn't translate to winning more? It's one of those situations where something might be statistically significant, but totally meaningless when it comes to real-world application.

2.       Notice that all of the correlations in the table jump around from year to year. Sometimes they get stronger, sometimes they get weaker. The relationship between SOSPre and SOS became unprecedentedly strong in 2008 and 2009, but what about this past season, where it decreased in strength? What about the 5-year period between 2003 and 2007 where there was no relationship statistically speaking? It's likely that each NFL season seems to be idiosyncratic in terms of how much SOSPre is a reliable indicator of SOS, and the same can even be said for the much stronger relationship between SOS and WPct. So the question should really be, "Will 2011 be more like 2008 and 2009, more like 2010, or more like 2003-2007? My answer is, "Who knows? The stats sure can't tell us at this point."

3.       Related to the last idea, also remember that, although correlations are tidy, little statistics, I'd be more interested in R-squared, which in the current application tells you how much variation in the outcome (e.g., the "SOS" in SOSPre -> SOS) is explained by variation in the predictor (e.g., the "SOSPre" in SOSPre -> SOS). In layman's terms, R-squared says, "from 0% to 100%, how accurately can you forecast the outcome from the predictor?" In the case of SOSPre seemingly becoming a better predictor over the past few seasons, squaring the highest correlation of the bunch, .429 in 2009, gets you to only 18.4% of SOS being explained by SOSPre. Stated differently, over 80% of the NFL phenomenon called "SOS" has nothing to do with the much-discussed "SOSPre;" and that's in it's the most predictive season!

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Niners Nation

You must be a member of Niners Nation to participate.

We have our own Community Guidelines at Niners Nation. You should read them.

Join Niners Nation

You must be a member of Niners Nation to participate.

We have our own Community Guidelines at Niners Nation. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker