Two issues this week:

1. When a team's defense improves, does their offense typically improve also?

2. How should cheatsheets be altered for non-dynasty keeper leagues?

Teams like Seattle and Cincinnati figure to be better defensively this year (can't get much worse). Does a big jump in the defensive rankings lead to an appropriate improvement in offensive numbers, since the O figures to have better field position/more possessions?

Keith

Good question. The answer, in general, is maybe.

Here's the plan: we'll take a look at all teams from 1989-2002. We'll measure how much their defense improved, in terms of total yards allowed, between the previous year and that year (technical note: my database does not, at present, include sack yards, so what I'm calling "total yards" is really rushing yards plus passing yards which, for fantasy purposes, is probably a more important indicator anyway). For example, the Colts' defense allowed 820 yards fewer in 2002 than it did in 2001, so the 2002 Colts go down as a +820. That's a pretty big improvement.

Now we list all 409 teams from most improved to most imploded. The most improved was the 1998 Raiders (+1547 yards). The biggest collapse by a defense during that time period was perpetrated by the 2001 Titans (-1539). Now we'll cut the list into equal quarters, called the "Vastly improved defenses," the "Mildly improved defenses," the "Mildly declined defenses," and the "Vastly declined defenses." Here is how each group's offenses did, on average:

 

Group	Avg Off. Change
Vastly improved defense	-64 yards
Mildly improved defense	-16 yards
Mildly declined defense	+46 yards
Vastly declined defense	+91 yards

The relationship between defensive improvement (or decline) and offensive improvement (or decline) is extremely weak, but what there is of it seems to indicate that defensive improvement coincides with offensive decline and vice-versa.

Insterestingly, if you look at points instead of yards, you get the opposite result, although it's still very weak. Same chart, but with "improvement" and "decline" measured by points instead of yards:

 

Group	Avg Off. Change
Vastly improved defense	+19 points
Mildly improved defense	-1 point
Mildly declined defense	+1 point
Vastly declined defense	-10 points

What's going on is that improvement on the defensive side tends to coincide with a shift from passing to running on the offensive side.

 

Group	Change in Rush Atts
Vastly improved defense	+32
Mildly improved defense	+8
Mildly declined defense	-13
Vastly declined defense	-38

With issues like this, you've got to be careful with cause and effect. All we can say is that defensive improve coincides with more running and less passing. We cannot say that defensive improvement causes more running and less passing. It may (or may not) be more accurate to say that more running and less passing causes (statistical) defensive improvement.

Also keep in mind that, unless there is major change on the defensive side, the differences on the offensive side tend, in general, to be pretty small.

I'm starting a 4-year keeper league this year. We will be retaining 4 players, so this draft is crucial. There are volumes of rankings on the net for redraft leagues and traditional keeper or dynasty leagues, but nothing for a 4-year league such as mine. Who would make your top 15 RB list in a standard basic performance scoring system(1pt per 10 yards rush/rec, 6pt per TD)?

Players 30 something like Marshall & Holmes have to be marked down...but how much? Is Alexander a better pick than Faulk for a 4-year keeper league? These are my quandaries. Your feedback would be heaven sent.

Britt M.

I can't speak for everyone, but I'd be comfortable betting that most sharks would tell you that age should only be a consideration when two players are very close. That is, use age to move guys around within a tier, but don't make a habit of vaulting young mediocrities past proven producers. For example, and to answer your specific question, I would put Shaun Alexander ahead of Faulk (I'd take Deuce before Marshall as well), but I would not put Travis Henry ahead of him. Your mileage may vary if your tiers look different than mine do.

Let me give you a couple of other bits of advice:

• Even though you can keep players for four years, never attempt to look ahead more than a year or two at the most. Had you started your league before the 1999 season, the four-year terms would just be expiring. Take a look at the top 15 rated RBs going into 1999 (list taken from a popular fantasy mag that I just happened to still have on the shelf

Think about all the ups and downs those 15 guys have been through in the last four years. Nearly every player on that list has been rendered nearly worthless by injury, retirement, or plain old ineffeciveness in at least one subsequent season. No one has any idea -- none at all -- who the top backs will be in four years. Clinton Portis is not an even money bet to be a top 10 back in the 2006 season. Neither is Tomlinson. Neither is Alexander or Ricky or Deuce. No one is. Concentrate on trying to figure out 2003, where you have some prayer of being right.

• Keep-4 leagues are very different from dynasty leagues. In dynasty leagues (especially those with deep rosters), it makes sense to take some young deep sleepers and stash them for a few years. But in a keep-4 league, who's going to be kept? Assuming it's a 12-team league, 48 players. Assuming your league is as running back hungry as most, you're probably looking at about 24 RBs, 10-15 WRs, 10 QBs, and a handful of TEs.

Think about that for a minute. When you draft your WRs this year, remember that you're not going to end up keeping them unless they are at least top-15 material going into next year. In a dynasty league it might make sense to draft, say, Antonio Bryant ahead of Rod Smith. In a keep-4 league, I don't think it does. Most would agree that Smith is likely to be more productive that Bryant this year. And it's extremely unlikely that you'll be keeping either one of them, so why not take the production now? Don't even consider drafting someone "for the future" unless they've got a real shot at being a keeper. Surprisingly few WRs, QBs, or TEs fit that description.

• Be aware of the value of future draft picks. When trade offers involving your next year's first round pick start to materialize, you might be tempted to reason that, since each team has four keepers, the first round of the draft is really like the 5th round of a normal draft. That's not quite true. There are typically (not this year, but typically) at least a couple of big-time blue chip rookie running back rookies in the draft -- guys that would be drafted in the 2nd round of a re-draft league. Getting those guys in the draft is really huge. But once they're gone, the rest of the first round really does look like the fifth round of a re-draft. Moral of the story: there is usually a huge difference between a high first-round pick and a low first-round pick. When you're negotiating deals involving future first-rounders, make sure you're taking into account where the first-rounders you're talking about will be.

I hope this helps you some. Good luck with the new league