FOR THOSE WHO "TIER" THEIR RANKINGS (1 Viewer)

[Mik789fl]

What do you use as a "break point" to determine your next tier? Is there a mathematical number? Or do you just use a "gut" feeling?

 

[shadyridr]

What do you use as a "break point" to determine your next tier? Is there a mathematical number? Or do you just use a "gut" feeling?

A tier for me is who I would be comfortable taking if I wait to pick from that position if worse case scenario I get the last guy in that tier. Basically I do it on gut. Basically my top tiers of RBs looked like this:Tier 1:Alexander, Johnson, TomlinsonTier 2:Brown, Jackson, Barber, RJohnson, Jordan, Williams, JamesTier 3ortis, Westbrook, McGahee, Dillon, Jones, Jones, Taylor, Parker, Bush, DunnBasically Tier 1 guys are studs, Tier 2 guys Id be happy as my RB1, Tier 3 guys Id be happy as my RB2, and then Tier 4 guys would be guys Id love as my RB3.

 

[imeimex]

Depends on position but I typically go 2 points per game which leaves me seperating tiers every 32 points. I start from the my top projection and subtract 32 for each tier, except when I have players that I rank nearly identical but have slightly different projections than I slide them in the tier as well.

 

[Seehawk]

What do you use as a "break point" to determine your next tier? Is there a mathematical number? Or do you just use a "gut" feeling?

I use a 5% dropExample using current WRsSteve Smith projected as #1 WR at 192 pts, a 5% drop would be approx 182 pts, so anybody projected at 182+ is in the same tier with Smith (that's Holt, CJ, Fitz & Moss).The next tier starts with Boldin projected at 177pts. 5% drop would be 168pts. Anybody projected at 168+ is in this tier.Next guy after that tier is Roy Williams at 163, 5% drop means anybody at 155+ is in the same tier.I don't claim that this 5% was arrived at by any super-secret method but I figure 5% is an acceptale margin of error and probably better then you'll actually find in most projections. My thinking is that if my projections can be off by 5% in either direction then, logically, any WR with projected points 5% above or below the guy I'm looking at is "just as good as" this guy.One thing I like about running a "hard number" is that it will bring some stand-out guys NOT in the top 10 or 15 to your attention. You'll find very small tiers (perhaps only 1 or 2 guys) that you'd be tempted to lump in with the tier above or below if you strictly did it by your gut but by running the numbers they'll jump off the page at you. If you can isolate 1 or 2 guys that are clearly (at least 5%) better then the tier below them you can really maximize your picks in the middle and late rounds.One last thing; after I tier my players I erase (or at least don't print) their projected stats/points AND shuffle the names up within the tier. If you're gonna trust your tiers, you gotta trust them all the way. If all the WRs in your 5th tier are truly "equal" then treat them equally. By shuffling them up and erasing the crutch # I find I'm more apt to sift through a tier for the guy that best fits my roster as it's being assembled rather then on the fact that one guy was projected with 878 yards and the other with 868. Watching bye weeks, or avoiding the WR2 on a team that you already have the WR1 on, etc, etc. I've found that I assemble more balanced and better constructed teams this way.

 

[Avery]

What do you use as a "break point" to determine your next tier? Is there a mathematical number? Or do you just use a "gut" feeling?

I use a 5% dropExample using current WRs

Steve Smith projected as #1 WR at 192 pts, a 5% drop would be approx 182 pts, so anybody projected at 182+ is in the same tier with Smith (that's Holt, CJ, Fitz & Moss).

The next tier starts with Boldin projected at 177pts. 5% drop would be 168pts. Anybody projected at 168+ is in this tier.

Next guy after that tier is Roy Williams at 163, 5% drop means anybody at 155+ is in the same tier.

I don't claim that this 5% was arrived at by any super-secret method but I figure 5% is an acceptable margin of error and probably better then you'll actually find in most projections. My thinking is that if my projections can be off by 5% in either direction then, logically, any WR with projected points 5% above or below the guy I'm looking at is "just as good as" this guy.

One thing I like about running a "hard number" is that it will bring some stand-out guys NOT in the top 10 or 15 to your attention. You'll find very small tiers (perhaps only 1 or 2 guys) that you'd be tempted to lump in with the tier above or below if you strictly did it by your gut but by running the numbers they'll jump off the page at you. If you can isolate 1 or 2 guys that are clearly (at least 5%) better then the tier below them you can really maximize your picks in the middle and late rounds.

One last thing; after I tier my players I erase (or at least don't print) their projected stats/points AND shuffle the names up within the tier. If you're gonna trust your tiers, you gotta trust them all the way. If all the WRs in your 5th tier are truly "equal" then treat them equally. By shuffling them up and erasing the crutch # I find I'm more apt to sift through a tier for the guy that best fits my roster as it's being assembled rather then on the fact that one guy was projected with 878 yards and the other with 868. Watching bye weeks, or avoiding the WR2 on a team that you already have the WR1 on, etc, etc. I've found that I assemble more balanced and better constructed teams this way.

Interesting strategy.I like the general thinking and just checked out a few of your "5% tiers" with my cheetsheet. As an observation it creates an awful lot of tiers. I admittedly didn't do the whole sheet, but many tiers seems to be 2-4 players. Not that this is necessarily wrong, but it doesn't seem to successfully break up my cheetsheet a whole lot more then the current rankings already do. It successfully points out where a number of players are ranked about equally but I'm not sure that the 5% drop is big enough to be very illustrative.

Perhaps I'll take a look at what a "7% tier" or a "10% tier " would look like.

Edit to add: Just took a look at some samples with a 10% tier drop and 7% tier drop. At 10% the tiers look a bit large, many at about 10 players. 7% looked pretty good to me. Many groups of 5 or 6 players but a number of very illustrative tiers with only a couple of players.

The % drop seems to be a very solid foundation. Good work!

 

[cormanman]

I use .5 tiers, for example, Tier 1 - Tier 1.5 - Tier 2. Guys in 1.5 I can't nail down my projections on. They have the upside of a 1 and the downside of a 2 but I can't make up my mind.

Cheesy perhaps, but it works for me.