Leroy Hoard said:
Good points.
Personally I have had it with going after preseason "top" defenses. I'd rather draft one late and play matchups all year than give up a pick that could have turned into a decent RB or WR.
Oh absolutely. I almost always draft my defense in the next-to-last round. But my goal is always not to stream. Ideally, I'll either draft a D late that turns out to be set-it-and-forget-it, or else I'll pick one up early on in the season and ride them the rest of the way. If none of that pans out, then I'll fall back on streaming.
My point was that, even in cases where you pick right and get that SIAFI defense, you're still going to have some ups and downs. Let's look at the top three defenses through Week 14 (NE, Dallas and Buffalo). I'm guessing that in most leagues, they were either drafted relatively late (NE/Buff) or not at all (Dallas). But all three have become relatively reliable weekly options.
So let's look at how they've performed using the somewhat arbitrary criteria of "bad" games (less than 7pts in standard scoring), "average" (7-9) or "good" (10+):
- New England: 5-1-7
- Dallas: 3-4-6
- Buffalo: 5-2-6
That means that if you had stumbled into one of those defenses in the draft and started them every week, on average more than half of the time you would have ended up with a single digit score, and a third of the time you would have ended up with a result that hurt your team's chances. True, some of them were landmines you might have seen coming and known to swap in a replacement (like Buffalo vs TB this past Sunday). But other stinkers came in games that looked like smash plays, like Dallas vs Denver or NE vs Houston.
I still think this approach is preferable if you can pull it off, in part because you don't have to pay the mental tax of choosing streamers each week. But it's hardly foolproof. As with just about every other aspect of fantasy, DST scoring is always going to have a large element of randomness to it