It's that time of year again. It seems everybody has their season previews up, and with that comes pretty bad logic and reasoning. I could go on for days on how bad some of the mainstream analysis is, but I'll just focus on one point today.
The Raiders, as everybody knows by now, where the first team post-merger to sweep their division without winning it, and this is the basis for many, including most of the perpetually over-optimistic Raiders fans, to believe they will actually be good next year. Of course, sweeping the division is very hard to no matter who you are, but this argument is not just wrong, but is actually the complete opposite of the truth. That's right, the fact that the Raiders swept the division last year actually adds to the argument that they will be bad next year. Sounds crazy, but it's true.The main reason this is hard to believe is because many people don't realize that often times when doing previews positives turn into negatives and negatives turn into positives. When you're at the top, you can only go down, and when you're at the bottom you can only go up.
As an example, consider turnover ratio. A higher turnover ratio is obviously better, but when previewing the next season it's a reason to expect a drop-off, as shown in this graph I made plotting gains or losses in wins from 2009 to 2010 against turnover differential in 2009 (i.e. "previewing" 2010 based solely on turnover differential in 2009):
As you can see, the better the turnover differential in 2009, the worse the team did in 2010. With a stat like turnover differential that is largely due to luck, a team who relied on that stat for success one season cannot reasonably expect to be able to lean on it the next. (This is one of the main reasons I think the Falcons are way overvalued by almost everybody. Their 13-3 season depended almost entirely on these lucky numbers, which is why they got embarrassed in the playoffs by the Packers.)
The main difference is that winning division games has much more to do with skill than with luck. If you sweep your division it is usually because you're a good team. But they, like every other game, are effected to some degree by luck. The better team does not always win, as evidenced by the lack of 16-0 and 0-16 teams every year. And since division games are against division opponents, getting lucky in division games has a very large effect. Not only does it add one win to your win total, but subtracts one win from the division rival.
When you sweep the division, this effect becomes huge. Suddenly you have 6 wins while you're division rivals have 2 losses each. This gives the sweeper an enormous advantage that is almost impossible to overcome. Even if a team does sweep the division largely by luck, that team will usually win it because that 8-game deficit is so hard to overcome.
This is all fine for that season, but presents a major problem the next year. That team will probably not sweep the division again, and if they got lucky in their sweep (i.e. they were actually worse than some of their division rivals) it becomes even harder. Suddenly that 6-0 in the division goes to 3-3 or even worse, which likely gives another win or two to the best team in the division in addition to taking away three wins from the sweeper. With their advantage gone, the sweeper will struggle.
A perfect example of this is the 2009 AFC North division champion Bengals. The swept, and eventually won, their division that year. But, as we all know, they didn't do as well this year. Was this fall predictable by the dynamic explained above? Well, consider the division standings from that year if you excluded division games:
1. Pittsburgh 7-3
2. Baltimore 6-4
3. Cincinnati 4-6
3. Cleveland 4-6
Without their division sweep, they would have been tied for last with the Browns. (Also note that those standings are virtually identical to the 2010 standings). They did three games worse than the Steelers outside the division, but because they managed a sweep, they won the division. So when the Bengals went 2-4 in their division in 2010, they could not recover and fell into last place.
Now, the Raiders swept the division last year without winning it. In other words, even with the huge advantage a sweep gave them, they were still beat by two teams in the division. If that season sweep was not due to luck, then we witnessed one of the most statistically improbable occurrence in NFL history.
In short, the 2010 Raiders were a worse version of the 2009 Bengals, so I see no reason to believe they won't be at least as bad as the Bengals were in 2010.
(On a related note, the Chiefs were the only team in the last 3 years to win the division without a winning record within the division (I only checked the last 3 years because my internet at the moment is painfully slow). So we had to overcome a division deficit to win the title, which is one of the positives going into next season.)