This probably belongs in the IDP forum.
In a start 8 IDP redraft (2dl, 2db, 2lb, 2flex) with standard offenses and typical performance scoring, I've had excellent results with a "late IDP theory" until last year. So, I think the old idea of knowing your league is important. For the first few years these guys were letting great players slip through the cracks while drafting minimal defensive rosters of perceived studs (leaving more for me late and in free agency). Last year no such luck. They drafted very strong and loaded their benches leaving me high and dry. So, I'll have to build my D earlier than I like to this year... I think. I still think some big scorers will be around late, and I put a good bit of effort into researching who they may be.
From my experience and observing numerous other drafts, the top IDPs will start falling as early as the 3rd and always by the 5th. I miss out on those guys. Then there's ususally a bit of a run in 6 and 7 cleaning out the top LB tier and possibly nabbing a couple top DBs. I miss out on that too. But, I want to be upper tier somewhere, so I'll often start my D with a top DL or two around rounds 8-12. Sticking with the idea of being upper tier somewhere. I won't let a top tier DB fall too far.
(snipped specific info)
The benefit of this approach obviously is it allows an owner to solidify his RB and WR depth, or take Gates or Manning (both on this team with LJ, btw). Ask yourself if you'd rather have Gates and Witherspoon or Vilma and Cooley. Something that approximates ADPs anyway.
Because I target upper tier DBs and DLs this isn't "late IDP theory" but "late LB theory." It follows the rationale of late QB theory and with some geeky SoS analysis could follow QBBC too.