Let me begin with the ratings; then we’ll get to the explanation.

Let’s begin with a look at the rushing data. The Chiefs grade out as the number one offense in rushing efficiency in 2015. Kansas City rushed for 2,044 yards and 116 first downs; giving 9 yards for each first down gives the Chiefs 3,088 adjusted rushing yards. Kansas City also had 19 rushing touchdowns, and since a touchdown is always a first down, we add another 11 yards per rushing touchdown to bring the total to 3,297. The Chiefs also had 14 kneels that cost the team 14 rushing yards last year; add those back in, and Kansas City is now up to 3,311 yards. On rushing plays, Kansas City fumbled 5 times; subtract 25 yards for each fumble (these are fumbles, not fumbles lost), and the Chiefs finished with 3,186 yards.

Kansas City recorded 436 rushing attempts last year, but we need to discard those 14 kneels. Take 3,186, divide it by 422, and you get an average of 7.55 adjusted yards per play.

Now, let’s do the same with passing. We will be essentially using ANY/A, but with a 9-yard bonus for first downs as well. To no surprise, the Cardinals take the top spot by a decent margin, while the Rams bring up the rear. Note that sack yards are already “included” in this table, as the passing yards displayed shows team passing yards, or net passing yards.

Next up, rush defense. The Jets allowed 1,335 rushing yards, 62 first downs (bringing the total to 1,893 yards), 4 touchdowns (1,937). The Jets also faced 11 kneels for -12 yards (1,949) and caused 8 fumbles (1,749). The Jets faced 373 carries, or 362 net carries, for an average of 4.83 adjusted yards allowed per run. That was the best in the NFL in 2015.

Of note: the San Francisco 49ers ranked 12th in yards per carry allowed, which might make you think the rush defense was pretty good. But that is really misleading. San Francisco allowed a first down on 27% of rushing attempts faced, the worst rate in the NFL. And the team allowed touchdowns on over 4% of all carries, the 4th worst rate. And the defense only forced a fumble on 0.8% of rushing attempts, the 27th-best rate. ^[1]I’m not sure how predictive or indicative of talent this stat is — the Saints were #1 at 3.6% — but in the interest of mirroring the offensive numbers, I think it makes sense to … Continue reading Add it all up, and the 49ers actually had one of the worst run defenses in the league, that got a huge boost in raw numbers because San Francisco faced a league-high 26 kneels. ^[2]Note that those kneels were excluded in calculating the stats explained in this paragraph.

Finally, let’s look at pass defense. Again, these mirror ANY/A, with a first down bonus. And, as you would expect, Carolina and Denver top the ratings, with the Saints at the very bottom.

One of the problems with yards per play is that it is subject to Simpson’s paradox: teams that pass often (or face teams that pass often) will generally allow more yards per play just because passing gains more yards per play than rushing. So when I came up with my offensive and defensive ratings, I took 65 parts of the team’s passing rating and 35 parts of the team’s rushing rating, which eliminates the problem entirely. You may disagree with those weights — I spent about 10 seconds deciding on them, but they felt right to me — and you have all the information you need to re-run the numbers with your own weights.

Anyway, as always, please leave your thoughts in the comments. What do you think of this method of rating teams?