May 6, 2008
Some early SOS thoughts
As soon as the draft was over, it was time to seriously look into this year’s Strength of Schedule. As you may know, we figure out our SOS a little differently than others. We project SOS; we don’t go by last year’s numbers. We will be importing this year’s schedule into our SOS tool shortly, so you can see this year’s SOS by position in your scoring system and for a range of weeks or the whole season soon, but we do view our projected SOS as more important.
The SOS tool goes completely off of last year’s numbers and this year’s schedule, whereas our projected SOS goes off the projected grades we give each defense against the run and the pass compared to each team’s schedule.
The first thing we do is analyze off-season movement in terms of additions, subtractions, coaching changes, etc. and we give each defense a number grade against the run and the pass. Anything over a 90 is an A, and an 80 would basically be considered a D. Here are the grades we’re currently giving each defense against the run and the pass:
Team RG Team PG
Ari 85 Ari 83
Atl 81 Atl 81
Bal 89 Bal 87
Buf 86 Buf 84
Car 86 Car 88
Chi 84 Chi 87
Cin 83 Cin 85
Cle 85 Cle 81
Dal 90 Dal 88
Den 83 Den 88
Det 84 Det 86
GB 87 GB 88
Hou 84 Hou 82
Ind 86 Ind 88
Jac 87 Jac 88
KC 86 KC 84
Mia 84 Mia 84
Min 91 Min 87
NE 87 NE 86
NO 85 NO 85
NYG 88 NYG 87
NYJ 85 NYJ 86
Oak 83 Oak 87
Phi 86 Phi 88
Pit 89 Pit 87
SD 90 SD 87
Sea 87 Sea 87
SF 86 SF 86
Stl 84 Stl 85
TB 87 TB 87
Ten 88 Ten 87
Was 86 Was 86
We then take the 2008 schedule and put in the grades for each team’s 16 opponents against the run and the pass. Since the lower the grade the worse the run defense, the team that has the lowest average grade for their 16 opponents has the best SOS. This year, according to our projections, the San Diego Chargers have the easiest schedule against the run. The average grade of their 16 opponents is 85.0. The team that has the hardest SOS against the run is the Cincinnati Bengals. Their average grade is 87.0, mainly because of their tough division (they’re always near the bottom).
We’ll be producing a comprehensive SOS overview with charts and rankings against the run and the pass for the season, the first half of the season, the second half of the season, with plenty of analysis next week. But for now, here’s the season chart for both the run and the pass. I’m sure you can come to a lot of the logical conclusions yourself simply by looking at how these teams are slotted.
Rnk Team Avg. Grade vs. Run Rnk Team Avg. Grade vs. Pass
1 San Diego 85.00 1 Denver 85.38
2 Kansas City 85.25 2 NY Jets 85.44
3 Carolina 85.44 3 Buffalo 85.50
4 Minnesota 85.44 4 Miami 85.50
5 Buffalo 85.50 5 Oakland 85.50
6 New Orleans 85.50 6 St. Louis 85.56
7 New England 85.56 7 Carolina 85.63
8 NY Jets 85.56 8 New England 85.63
9 Tampa Bay 85.56 9 San Fran 85.63
10 Denver 85.69 10 Baltimore 85.69
11 Oakland 85.75 11 Philadelphia 85.69
12 Miami 85.81 12 Pittsburgh 85.81
13 San Fran 85.81 13 Cincinnati 85.88
14 St. Louis 85.81 14 Jacksonville 85.88
15 Atlanta 85.88 15 San Diego 85.88
16 Baltimore 86.00 16 Tennessee 85.88
17 Seattle 86.00 17 Indy 85.94
18 Jacksonville 86.06 18 Seattle 86.00
19 Tennessee 86.06 19 Kansas City 86.06
20 Chicago 86.13 20 Tampa Bay 86.06
21 Green Bay 86.19 21 Dallas 86.13
22 Detroit 86.25 22 Minnesota 86.13
23 Houston 86.31 23 New Orleans 86.13
24 Philadelphia 86.31 24 Washington 86.13
25 Dallas 86.38 25 Arizona 86.25
26 Arizona 86.44 26 NY Giants 86.25
27 Washington 86.56 27 Chicago 86.31
28 Cleveland 86.63 28 Green Bay 86.31
29 Pittsburgh 86.63 29 Detroit 86.38
30 Indianapolis 86.69 30 Cleveland 86.50
31 NY Giants 86.81 31 Houston 86.56
32 Cincinnati 87.00 32 Atlanta 86.69
I use SOS big time. Not sure if alot of members pay attention to it, but its a great tool to strategize for mid season trades. Just wanted to give you some Dap.
Is there a formula that would include O-Line & D-Line? Or is this part of the equation. KC might have the 2nd easiest run SOS but if thier O-Line grade is a D then that should have some effect.
Jbeau,
I use SOS alot also but SOS can change over the course of a season. Tennessee became a very good run defense then what thier SOS was at the beginning of the year. They finished 5th in the league against the run giving up just 92.4 yds a game.
I don’t use it with drafting Crew. Reason being that no one can really predict how a defense is gonna be. With the exception of certain teams…Vikes and Ravens. They are usually stout against the run, but who would have thought that last year that the Bears would be bad against the run. I only use it around mid season, where we have an idea of how teams are playing. I then try to trade for certain guys who have easier schedules toward playoff time. for example, John said to target jamal Lewis last year. He wasn’t going crazy at the time, but he was solid, and showed a good burst. his SOS against the run was ridiculously easy. I traded Crayton and Donnie Edwards and obtained Lewis and Bodden. Crayton had a few good games, but I knew he wouldn’t keep it up. Rode Lewis through the rest of the season to the championship game. Thats how I use SOS. sorry to be long winded.
NYBrewCrew: That’s an interesting idea, and I might look into a way to handicap the teams and grades based on the o-lines. Good one.
We like to think that the projected SOS we do takes into account the rapid changes, as opposed to going off last year’s numbers. Last year’s preseason analysis was only ok, but the 2 years before were really spot on.
With a spread of only 2 or 3 points out of 100, I’m not sure how useful these projections are.
The point spread between the best and the worst defenses is just 2 points. Given the denominator is in the 80s, there does not seem to be statistically significant difference there. Please help me better understand the scoring system.
Also, over the past three years I have looked to see if the forecasted season strength of schedule had an impact on production. In most cases I could not find a statistically significant correlation. Great RBs, Wr, etc. seemed to do well no matter the SOS and poor players still did poorly. Some of the folks in the middle seemed to be helped or hurt as little but it was less than a 50% correlation.
Am I missing something?