Charts:
Charts are on the menu on the right.
Week 8 - QB | WR | RB | TE | Flex | K | DST
In this fantasy football model, I take aggregated expert ranking data from Fantasypros.com and pass it into a clustering algorithm called the Gaussian mixture model. The algorithm finds players who are ranked similarly and discovers natural tiers within the data. The results is a series of simple charts that can help you decide your fantasy football line up each week. Match ups, injuries, strength of offense and defense are all accounted for. For more intuition behind the model, you can read this introduction.
Tips:
(1) NYTimes link: Week 8 in-depth commentary
(2) Favor higher tier players.
(3) When choosing between players of the same tier, flip a coin or go with your gut. If experts are 50/50, there's no telling.
Rest of Season:
I'm also happy to announce the return of ROS tiers. View them here:
Select Week 8 Questions and Answers
Would you roster Allen Robinson or Andre Holmes?
It's very close, but I think I would take Robinson over Holmes (same tier ROS, see ROS WR chart), Robinson has shown more consistency and but Holmes has more blow up potential, so the answer should depend on your needs at WR. However, if you have other Jags on your team, I'd take Holmes. This highlights a concept I've gotten more fond of over time, which is team diversity. Game flow is a huge dictator of how fantasy points end up distributed, and having players on different teams is often advantageous. I'd much rather have Marshall/Cobb than Marshall/Jeffery, because if the Bears have a bad day, only one of my receivers is affected. Similarly, I'd rather have Bortles/Holmes than Bortles/Robinson.
What do you think of McKinnon and Asiata going forward?
I think McKinnon is here to stay. He's already a weekly RB2 and he's more likely to move up than down. There's a small chance the pendulum of touches and snaps swings back closer to 60-40 for McKinnon-Asiata, but I doubt it. Even in that case, Asiata won't get getting much yardage on 10 carries a game, and that doesn't translate to much on 3.6 yards per carry. Asiata has some value in the passing game (20.3 receiving yards per game), but there's just no upside appeal. Would I drop? it depends who else is on the wire and who your weakest link is, but I would have dropped him for any of Tre Mason, Anthony Dixon, Denard Robinson, and Bryce Brown. Check the ROS RB chart.
Why is Bradshaw so low? Why is Richardson so low?
TRich just hasn't looked like a good runner this year. He's not particularly explosive and he's lacked vision. He's averaging 51.1 yards per game and 3.4 yards per carry. He's only been able to put up numbers becuase he's getting 14.4 carries per game. It's the volume that's keeping him relevant. I think we're seeing TRich's ceiling right now. It's hard to put Bradshaw any higher because Trent is getting 14-15 carries a game. Bradshaw is only averging 10.0 carries per game and 4.3 receptions. No one will argue that Bradshaw is not the better runner here, but as we've seen, even Jamaal Charles won't get much done with that kind of usage. This is similar to the Spiller/FJax scenario in the last year or two, and neither of those guys ever got rated above the middle tiers.
Thoughts on Ronnie Hillman and Montee Ball?
If Hillman continues to play like he does, I'd be surprised if Ball gets the starting job back. Hillman tallied over 100 yards from scrimmage in his last two games, against the 49ers and Jets.
Thoughts on Julio Jones ROS?
The Atlanta offense line is in shambles. Until they figure something out, Matt Ryan is just going to get crushed against good and even mediocre defenses. Julio is to good to trade for 80 cents on the dollar. He's still a good bet for 80 yards in any matchup, and if you're lucky, a garbage time TD bomb. I'm only dealing him if I can get someone like Brown, Dez, Lynch or a good package deal like Olsen + Cobb.
I'm a worried Foles owner. Sell or Hold?
ROS QB chart shows he's tier 2, and I agree with that. Philly has a top 5 offense in the league and arguably the fastest. Also their defense isn't great which helps Foles in fantasy.
Can I bench Keenan Allen this week?
I wouldn't strongly advocate it, but I wouldn't argue against sitting Allen for guys 1 tier below him this week. Benching him for Torrey Smith, Colston, or Cooks seems acceptable. Andre Holmes probably not, and definitely not Allen Robinson or Brian Quick. I'd give Holmes a 20% chance of outscoring Allen, and Robinson/Quick a 10% chance.
Is it time to bail on Seattle DST?
I think their low scoring acceptable within the range of random variation. I'd be surprised if they still keep coming in with bottom shelf numbers. They're still one of the top defenses in real football, especially at home. That's pretty evident when you watch their games. In the same vein, I'd be surprised if the Eagles continue to score like they do.
Dallas or Denver DST this week?
I like Dallas this week against Colt McCoy. They're not a great defense, but they're doing an amazing with time of possession control via DeMarco Murray. As of a week or two ago, the Cowboys were averaging a league low 56.3 opposing snaps per game. As for Denver this week, Rivers is even better than his stat sheets say and Oliver is as good as his stats, so I'm not sure I'd be confident starting any defense against SD, and Denver is no exception.
FAQ:
(1) Why are the flex charts inconsistent with the RB and WR charts?
This happens because flex is an after thought for many experts. Less experts submit rankings to Fantasypros, and they're not graded on them. Flex tiers are always one of the more debatable charts.
(2) How accurate are these tiers?
Here is a link to last year's accuracy.
(3) When do you update the charts?
Every day. Check the last updated timestamp.
(1) Why are the flex charts inconsistent with the RB and WR charts?
This happens because flex is an after thought for many experts. Less experts submit rankings to Fantasypros, and they're not graded on them. Flex tiers are always one of the more debatable charts.
(2) How accurate are these tiers?
Here is a link to last year's accuracy.
(3) When do you update the charts?
Every day. Check the last updated timestamp.
