What’s On My Mind This Week
I am always surprised at what I learn each and every week as I progress through this data project. … This week, in particular, I visualized a few things that ran counter to my perception regarding maximization of fantasy match-ups.
- Against what my normal decision-making process would tell me, I started Case Keenum in a 2QB ahead of Joe Flacco. Why did I do this? It’s simple, honestly. I trusted my data, the manipulations I’d performed, and did what it advised me to do. A year ago, I would’ve never made that decision.
- Also, I sat Jordy Nelson (vs. DAL) for Davante Adams because the data I had compiled told me that Rod Marinelli’s defense was subtracting WR1’s from opposing offenses with masterful efficiency. This flew in the face of a long-held belief in “starting your studs.” While this decision ultimately didn’t work out (Adams bumped his head and left the game early), Nelson still underperformed in line with what my data was telling me. Process: 1, Bankroll: 0.
A large part of evolving as a decision-maker and evaluator stems from challenging your own beliefs (however constructed) and performing quality control on your process. NFL front offices (the good ones, at least) frequently undergo self-evaluation to reduce subconscious tendencies (like calling a run behind the left guard on 1st down) and leverage them against competitors who are tracking those tendencies. The gold standard remains the Patriots, who are an amoeba on offense from week to week. Their tendency is that they have no tendency. That is the paradigm we should aim for, I believe.
Scattered Thinking About Week 6
Lies, DAAM Lies, & Statistics
DAAM (Defensive Adjusted Aggregate Metric) has grown up a little each week, and the past week is no exception. I beamed about migrating it to an automated process in R last week (2 minutes vs. 2 hours), and I leveraged it last weekend to make the lineup decisions above.
What I’ve tacked on this week are head-to-head metrics for weekly match-ups, which you can see below. In this form, I’ve written an Excel document (for improved visualization) that draws on in-season back data of a match-up for instant analysis.
Note: the “hm” column (home = 1; away = 0) refers to the offense, always.
So, in short, the top half of the table draws in the last six weeks of positional data from the offense, and the bottom half draws on the last six weeks of positional data from the defense. In the middle, the green row represents the imminent matchup. The “projection” in the green row is just a spitball right now… a rough average of the home/road split from respective offenses and defenses depending on where the game is being played.
Not shocking, in this matchup at least, is that both teams perform better on their home fields. But is this really the case, or is it just circumstantial based on “luck of the draw” in their opponents?
Game Script Splits (GSS)
I’ve uploaded GSS plots representing the first six weeks of the 2016 season. You can find them in the links below.
Passing Game Breakdowns (PGB)
I’ve also uploaded PGB plots representing the first six weeks of the 2016 season. You can find them in the links below. These are broken down into Offensive and Defensive PGBs, with slice plots for each position.
Hyperfocus On Week 6
Jacquizz Rodgers @ SF (FA $4000, FD $5600, DK $4300)
When it boils down to it, Tampa Bay has been pretty bad this season. The flash of light in Week 1 against Atlanta turned out to be an exploding bulb, given that Atlanta is giving up all kinds of fantasy scoring regardless of opponent. Since then, the Bucs have disappointed on the scoreboard outside of Mike Evans. Here’s what their GSS (Game Script Split) plots look like (full game, 1st half, 2nd half):
Last week, though, Jacquizz Rodgers stepped in and shouldered 36 touch opportunities, converting those into 129 total yards. With the bye week to rest up, and the fast-paced, ever-punting 49ers on tap, ‘Quizz might be in line for another 30+ opportunity game.
What jumps out to me is how San Francisco’s last four running back match-ups have been brutal for their defense. DAAM has these teams rated 2nd (ARI), 4th (DAL), 5th (BUF), and 9th (SEA)… pretty much a Murderer’s Row of a month for the Niners. Add in the loss of Navorro Bowman, and we see is a mushroom effect on their run defense.
Mike Davis vs TB (FA $3000, FD $4500, DK $3000)
On the flipside, San Francisco hasn’t done much but play from behind. Their defensive GSS plots tell the story of how teams are happy to run against them, and frequently.
Let’s see what DAAM has to say…
The 49ers, notwithstanding a biorhythm game in Buffalo last week, have slightly underperformed expectation in the run game (-3.25 FPPOF), but a quick peek at the defenses they’ve faced (-4.14 FFPDF) shows that maybe things aren’t so bad after all. They’ve yet to run into a “soft” run defense.
I’m interested to watch what San Francisco does with a lead should they acquire one. I expect they’ll run, and run, and run… but will Mike Davis be efficient? At his price, it might not matter.
Marcus Mariota & Demarco Murray vs IND
I think I see a ripe opportunity in the passing game for the Titans this weekend, but I’m fearful it may be too ripe.
Above are Passing Game Breakdowns of the Tennessee offense and Indianapolis defense. From a touchdown perspective (green dots), I see a lot of overlay, which might indicate these teams are a good “match” for TD production. Additionally, the lack of red dots on Indy’s defensive PGB are a good sign for Marcus Mariota in the pass game, and we shouldn’t see much ball-hawking of his usual throws.
Let’s dig a little deeper and look at air yard distributions to the running backs.
Now, this is kinda eerie, don’t you think? Despite having not played one another yet, the overlay on these PGB plots indicates DeMarco Murray is a prime candidate for a receiving score this weekend. As you’ll see below on the DAAM tables, Murray is also in line for very healthy fantasy scoring against the Colts.
Indianapolis has been absolutely gashed the last three weeks by some very poor-scoring rush offenses in Houston (-4.00 RB DAAM), Chicago (-4.73 RB DAAM), and Jacksonville (-7.24 RB DAAM). The Titans, on the other hand, are a high-scoring rush offense, putting together a 3.34 RB DAAM through the first six weeks of the season. I will say, however, that it’s highly likely this is all touchdown-based score differential (with equivalent yardage outputs) given the sample size, so tread lightly.
All said, I wouldn’t be very surprised if Demarco Murray leads the league in RB scoring this week, and that could very well include a TD reception from one Marcus Mariota.
Asides & Errata
Thanks for spending your time on me this week. Hopefully the visualizations and ideas above provide some insight this week as you make selections.
If you have questions about what I presented, or want to discuss in more detail, please find me on Twitter @FantasyADHD.