Tuesday, October 22, 2013

Week 7 - Retrospective

Tier 1 median: 20.1
Tier 2 median: 29.9
Tier 3 median: 22.8
Tier 4 median: 20.8
Tier 5 median: 16.6
Tier 6 median: 5.7

QB tier predictions were mostly accurate last week.  Tier 2 was the noisiest, due to Culter and Foles both leaving early due to injuries.  Injuries are impossible to predict, so if you add their substitutes points to their totals, they would have 18.9 and 13.0, which are tier 2 worthy numbers. Aside from that, outstanding weeks from Luck, Stafford, and RG3 brought the median to 29.9.

Tier 1 median: 13.2
Tier 2 median: 9.4
Tier 3 median: 5.9
Tier 4 median: 9.9
Tier 5 median: 12
Tier 6 median: 2.4
Tier 7 median: 4

In tier 4, FJax, Mathews, Woodhead, Ridley and Bell all had good fantasy days (9.9 points or higher) bringing up tier 4.  

In tier 5, without Chris Johnson's 66 yard touchdown play, he would have scored 4.4 points. But that's fantasy football.

Tier 1 median: 7.5
Tier 2 median: 7.7
Tier 3 median: 7.8
Tier 4 median: 4.4
Tier 5 median: 4.1
Tier 6 median: 4.8
Tier 7 median: 6.6
Tier 8 median: 2.4

Similar to last week, tiers 1, 2 and 3 have a large separation from tiers 4, 5 and 6.  I only have two weeks of retrospective, but I have a feeling this won't be the only two weeks we'll see this trend.

Wright (9.8), Boykin (16.3), and Hopkins (13.6) really inflated tier 7 this week. Had I updated the charts after knowing the James Jones was out, Boykin probably would have saw a rise.

Tier 1 median: 5.5
Tier 2 median: 7.7
Tier 3 median: 6.7
Tier 4 median: 2.5
Tier 5 median: 5.3

Tight end tiers come in as one of the nosier positions again. I suspect that a possible explanation is that a relatively high percentage of TE points come from touchdowns, not yards.

Tier 1 median: 7
Tier 2 median: 7
Tier 3 median: 6
Tier 4 median: 11

Kicker tiers looked meaningless last week (though this is expected to be the noisiest position).  

Note: Hauschka has a name mismatch the two data sources I'm pulling, so he appears as 0.  I'll fix next week.

Tier 1 median: 10
Tier 2 median: 7
Tier 3 median: 5
Tier 4 median: 2
Tier 5 median: 4

DST predictions looked good last week.


  1. Boris,

    The image for the WR tiers is not linked properly. Here is a link to the full sized one:


  2. I'm new to following this blog, so if you've addressed this already I apologize, but which tier tends to be the noisiest in terms of the predictions matched to actual values? It seems to me like it should be DT just based on my own anecdotal evidence but I'm curious.

    1. it really too early to say since i've only done this 2 weeks, but early indications point to TE/K as the noisiest.