Everyone is familiar with the perils of hoping a player can repeat a big season. “Regression to the mean” and “sophomore slump” are terms that are commonly thrown around – different ways of suggesting that when you use an early pick on a player, you’re more likely to get something less than what you’re expecting. But is there a sweet spot where the math flips?
That is, if it makes sense that the very best prospects will underperform (as a whole), then shouldn’t it make sense that the lesser players should finish with better-than-expected stats?
It’s a slippery and difficult concept to get into with individual players. (It would best be addressed with ADP from Sept. 1 for the last number of years, which is not something I have.) But we can look at entire offenses. That provides a limited, zero-sum game, with exactly 32 options each year. We can look at production in any given year, then compare it to what the offenses did in the following season.
For this study, I looked at the last 20 years, looking at overall rushing and receiving production, using 6 points for touchdowns and 1 for every 10 yards. For each season since 2002, I looked at the totals for each team, giving it a rank between 1 and 32 for each of the last 20 years. That gave me 20 groups of 32 teams (the 20 first-place teams, the 20 second-place teams, etc.)
For rushing, the numbers show that the good teams will tend to get worse, while the bad teams will tend to get better. Teams ranking 27th to 32nd in overall rushing production, for example, on average all improved by at least 20 percent statistically. Only four other offenses improved by at least 13 percent, and they all ranked 22nd or lower.
These numbers suggest that it makes sense (in general terms) to pick guys from seemingly bad offenses, including Dameon Pierce, Rachaad White.
The biggest decliners, meanwhile, were the teams near the top. Teams with top-3 rushing offenses tended to decline by 16-23 percent.
If you look at all 420 teams that ranked in the top 7 in rushing, they declined (as a group) by 15 percent in their next season. All 420 teams ranked in the bottom 7, meanwhile, improved by 25 percent. Using last year’s stats, the teams that ranked high in rushing posted numbers that were 78 percent better than the 420 lesser teams in those seasons. But that gap closed to only 22 percent the following year. (A difference of 13.5 to 7.6 fantasy points per week closed to only 11.5 to 9.5.)
Teams that finished with top-5 rushing numbers last year (making them more likely to decline statistically): Eagles, Bears, Giants, Cowboys, Falcons. Teams with bottom-5 rushing numbers last year, making improvement easier: Bucs, Texans, Colts, Dolphins, Chargers.
With rushing offenses, you get down to the 14th spot before offenses tend to putting up better numbers (rather than worse).
| RUSHING OFFENSES | |||
|---|---|---|---|
| Rk | Run/G | Next | Imp |
| 1 | 15.7 | 13.2 | -16% |
| 2 | 14.4 | 11.1 | -23% |
| 3 | 13.8 | 11.6 | -16% |
| 4 | 13.2 | 12.0 | -9% |
| 5 | 12.8 | 10.9 | -15% |
| 6 | 12.4 | 11.3 | -9% |
| 7 | 12.2 | 10.2 | -16% |
| 8 | 11.8 | 11.7 | -1% |
| 9 | 11.6 | 11.2 | -3% |
| 10 | 11.3 | 10.6 | -6% |
| 11 | 11.1 | 9.6 | -13% |
| 12 | 10.9 | 9.6 | -11% |
| 13 | 10.6 | 10.1 | -5% |
| 14 | 10.5 | 10.7 | 3% |
| 15 | 10.3 | 9.8 | -6% |
| 16 | 10.1 | 10.3 | 2% |
| 17 | 9.9 | 10.1 | 2% |
| 18 | 9.7 | 10.1 | 4% |
| 19 | 9.6 | 9.9 | 4% |
| 20 | 9.3 | 9.5 | 2% |
| 21 | 9.2 | 9.8 | 6% |
| 22 | 9.0 | 10.7 | 18% |
| 23 | 8.8 | 10.1 | 14% |
| 24 | 8.6 | 9.7 | 13% |
| 25 | 8.4 | 8.8 | 5% |
| 26 | 8.2 | 9.5 | 15% |
| 27 | 8.0 | 10.0 | 25% |
| 28 | 7.9 | 9.5 | 20% |
| 29 | 7.6 | 9.5 | 24% |
| 30 | 7.4 | 9.2 | 25% |
| 31 | 7.2 | 9.3 | 30% |
| 32 | 6.7 | 9.1 | 37% |
The passing numbers are similar, but there isn’t as much decline at the top. The top 4 spots are filled by offenses that declined by only 13, 6, 13 and 10 percent. Two of the offenses in the top 8, in fact, didn’t decline at all. I take this to mean that when you have these franchise-type quarterbacks – Mahomes, Allen, Burrow, etc. – they are more reliable than the elite running backs, who are more likely to be affected by injuries and aging.
If you look at the top 25 percent of passing offenses (the top 8 of the 32 teams) they declined only half as much as the corresponding rushing games.
If you’re looking for a passing offense that will improve by a lot, best to focus on the bottom dwellers. Teams ranking last or next-to-last tended to improve by 26-plus percent. (That would be Falcons and Bears this time around). The other four slots where teams improved by at least 10 percent would be those who previous ranked 27th to 30th (again, using standard fantasy scoring). Teams from last year in this group: Steelers, Ravens, Panthers and Titans.
To me, Pittsburgh and Baltimore look like teams that will be passing it a lot better. I’m of the school of thought that Lamar Jackson is a much better passer than he has shown in recent years – that he’s been held back by the run-oriented system they were using previously. With the Steelers, they’ve got Kenny Pickett moving into his second year.
I’m also expecting a move forward for the Giants, with Daniel Jones and Brian Daboll moving into their second season together (even last year, Jones was much more prolific as a passer in the second half of the season).
| PASSING OFFENSES | |||
|---|---|---|---|
| Rk | Pass | Next | Imp |
| 1 | 28.2 | 24.4 | -13% |
| 2 | 26.6 | 24.9 | -6% |
| 3 | 25.7 | 22.5 | -13% |
| 4 | 25.1 | 22.7 | -10% |
| 5 | 24.6 | 22.9 | -7% |
| 6 | 24.0 | 23.9 | 0% |
| 7 | 23.5 | 21.5 | -8% |
| 8 | 23.0 | 23.0 | 0% |
| 9 | 22.5 | 21.5 | -5% |
| 10 | 22.2 | 20.7 | -7% |
| 11 | 21.9 | 22.2 | 1% |
| 12 | 21.6 | 20.4 | -5% |
| 13 | 21.2 | 21.1 | 0% |
| 14 | 20.8 | 19.5 | -7% |
| 15 | 20.5 | 19.4 | -5% |
| 16 | 20.2 | 20.5 | 2% |
| 17 | 19.9 | 19.9 | 0% |
| 18 | 19.6 | 19.3 | -1% |
| 19 | 19.3 | 19.7 | 2% |
| 20 | 19.0 | 19.5 | 3% |
| 21 | 18.8 | 20.0 | 6% |
| 22 | 18.6 | 18.8 | 1% |
| 23 | 18.4 | 18.6 | 1% |
| 24 | 18.0 | 18.2 | 1% |
| 25 | 17.8 | 19.0 | 7% |
| 26 | 17.4 | 18.9 | 9% |
| 27 | 17.0 | 19.1 | 13% |
| 28 | 16.5 | 18.7 | 13% |
| 29 | 16.0 | 17.8 | 11% |
| 30 | 15.5 | 18.0 | 16% |
| 31 | 14.8 | 19.0 | 28% |
| 32 | 13.9 | 17.4 | 26% |
—Ian Allan