Forecasting Adjusted Defenses (NFC Edition)
Heading into the 2017 NFL season, we are presented with an opportunity to measure what teams accomplished the prior season. Many projection systems (including pricing algorithms on major DFS sites) simply stop here. This is an area we can attack in the early weeks of the season.
How do we attack it? We need to measure how defenses performed on an adjusted basis. We can do this relatively quickly by applying a Simple Rating System (SRS) to the late weeks of the 2016 season (I used Weeks 12 -16 for this study). This will take the average output allowed by a defense (or generated by an offense) and adjust it to schedule strength.
It appears that DFS pricing algorithms stop at this point. We can leverage this information and find weak points, but first we’ll have to assess offseason changes to each team’s coaching staff, offensive/defensive philosophy, schedule difficulty, and personnel. Once this “adjusted” SRS is in place, we can quickly judge a matchup and more accurately assess the differential between an offensive and defensive unit on a level playing field.
Simply? We’re taking a known quantity, using our best judgment to make it more accurate, and then targeting mispriced players.
Let’s look at the big picture. First, here are adjusted DraftKings Point projections, by position, for every offense and defense heading into Week 6. You’ll notice Floor projections as well. Floor projections attempt to provide the baseline production earned or allowed. This, in turn, should provide context to the Matchup Projections that you’ll see below.
The next two tables transform the information above into Projected Matchups for Week 6.
The first table (Week 6 Implied Points) shows the offense, their opposing defense, and the projected Mean DraftKings Points implied at each position. For each matchup, this is the average DraftKings points we can expect each respective position to score. Keep in mind that this is a mean expected value within a range of outcomes. Expected DraftKings Points are calculated using a “secret recipe” that includes each team’s adjusted points coming into this week, home / away considerations, and weighting of each team’s offensive/defensive strengths, among other factors.
The second table (Week 6 Floor Points) is very similar to Implied Points, except that I am attempting to show the baseline performance for each positional group in each matchup. The intent in this case is to provide the “worst case scenario,” useful for cash game lineups or seasonal lineups where a high ceiling may exist at other roster spots.
Before diving into the detailed adjusted preseason projections, I should explain the layout of the data and define the variables. Each team is divided into Offense and Defense table sets. The first table shows the adjusted point value for each position as well as the projection adjustments for each.
Would you like to create customized implied projections of your own? Check out my “Adjusted Matchups”:http://FantasyADHD.com/matchups app!
The columns are defined as:
L5 aFP – Adjusted DraftKings Points over the last five games (excluding Week 17). Raw fantasy points scored at each position are adjusted for opponent strength per SRS methodology.
Rank – Positional Rank per L5 aFP.
SOS – Strength of Schedule. This schedule strength is calculated using SRS methodology.
SRS – Simple Rating System score. This is a measure of a team’s “power” at the given position, where zero is considered league average. Positive numbers indicate more power.
Coach – Subjective adjustment for coaching changes. This can encompass new personnel, philosophy changes, and/or general improvement.
Script – Subjective adjustment for game script. This can encompass projected schedule strength, changes in offensive philosophy, or pace of play.
Roster – Subjective adjustment for roster changes. This can encompass new personnel acquired via Free Agency or Draft, departed personnel, players returning from injury, or players lost to injury during the preseason. Also included is improvement or decay of skill level.
1st 4 – This is the new Adjusted DraftKings Points that we can use to assess the first four games of the 2017 season. After four weeks, we will have enough data (63 games!) that we can confidently use SRS going forward. Blue indicates positive adjustment, and red indicates negative adjustment
The second and third table show the matchups for the first four weeks of the 2017 season, with adjusted projections (and plus/minus against projected league average) for each opponent at each position.
Now, without further ado, let’s see what the adjusted projections look like in detail!
It will be nice watching the Arizona offense with its full complement of weapons again. I’ve adjusted their game script toward a more pass-heavy approach to reflect the offense being at full strength.
Defensively, the Cardinals should be in much better situations than late last season, when teams passed against them at will. Tyrann Mathieu’s return should considerably improve their allowance to QB and WR. Opposing offenses (particularly IND) will likely target RB and TE more frequently, boosting their output.
Atlanta enjoyed other-worldly offensive output in 2016. I’ve adjusted them down sharply at each position to reflect competitive game scripts and regressed efficiency. We’re not sure what to expect from the changeover from Kyle Shanahan to Steve Sarkisian, but Sark will likely not push the buttons as well as Shanahan did last season.
On the flipside, opposing offenses that play in more competitive games should run more productively and need to pass less often.
I’m hesitant to be so positive about Carolina’s offensive prospects entering this season. Why? I don’t trust Mike Shula to evolve his scheme. I’m going to give him the benefit of the doubt, however, and project adjusted outputs reflective of a modern offense that utilizes dynamic receiving talent.
If Christian McCaffrey is utilized the way most believe he will be, I expect RB output to surge significantly. Logically, we have to assume that Cam Newton will cease to be the primary red zone rushing option in this scenario. I’ve regressed QB output as a result.
Defensively, the Panthers should be in the driver’s seat relatively early. They do face four heavyweight run offenses, however, and I expect their allowance to increase as a result.
Despite a litany of changes on the offensive side of the ball, I couldn’t see much reason to modify projections much from where they were at the end of last season. Signing Mike Glennon and drafting Mitchell Trubisky is more or less a lateral move, from a talent perspective, from Brian Hoyer and Matt Barkley.
The Bears play relatively solid defense for a team that doesn’t win much. I’ve projected their adjusted allowances to increase slightly since they will face four robust offenses to open the season.
Dallas faces a very difficult defensive schedule to open 2017. We’ve known this for a few months now. Many expect that this will stifle their offensive production, but I’m skeptical. The Cowboys choose to stifle their own production through limited play volume and ultra-conservative play-calling. For this reason, I don’t foresee their output changing all that much outside of slight game script variations.
Defensively, the Cowboys will face three sluggish offenses. What they won’t enjoy, however, is low rush volume with opposing offenses forced to throw most every play.
One item to note about WR production against DAL, however. Their defensive philosophy is to erase the WR1 from the offensive gameplan, which leads to greatly increased production for other pass catchers in the offense. For Week 1, monitor O’Dell Beckham Jr. closely. If he plays, Brandon Marshall is the hammer. If he doesn’t, I believe Sterling Shepard and Evan Engram become the hammers.
Detroit, thanks to having a poor defense, will opt to play slow, conservative offense the same way they did in 2016. They do this to limit the defense’s exposure and keep the game from getting too far away from them.
The complexion of the offense changes significantly, however, thanks to the return of Ameer Abdullah. I expect Jim Bob Cooter to reallocate his gameplan more toward the RBs and opt to strike less frequently (but more effectively) with his WR and TE group.
Defensively, losses continue to mount from a roster perspective. In 2016, Detroit were a sieve against passing games, and I’m projecting that to remain the case entering this season. Larry Fitzgerald, Brandon Marshall, Julio Jones, and Stefon Diggs should all enjoy productive games.
Green Bay Packers
Offensively, it’s difficult to see how much changes for GB. They are still pass-first. I don’t expect they will be as ridiculously pass-heavy as last season, however, with the transition of Ty Montgomery to full-time RB.
I’ve greatly reduced WR production since game scripts (and personnel) should allow them to run more frequently. Facing four solid defensive units to open the season also tempers output projections.
Defensively, the Packers open against three good/great passing offenses. Teams practically abandoned the run against GB last season as the offense got hot and built large second-half leads. These four games to open the season should be much more competitive.
Los Angeles Rams
Anything has to be a vast improvement over Jeff Fisher’s failed regime, right?
I am highly bullish on the Rams to open this season. I don’t think they’ll even approach league average in any area offensively, but by virtue of not being Dead Ass Last in nearly every statistical category should make players like Jared Goff, Todd Gurley, Sammy Watkins, etc viable on a week-to-week basis depending on matchup. I’ve projected adjusted output at each position to increase due to good coaching and reasonable game scripts.
Defensively, hiring Wade Phillips was about the sharpest thing this team could’ve done. While Gregg Williams is a sharp defensive mind, his philosophy doesn’t have a long shelf life. I project solid improvement across the most lethal phases (QB, RB, WR) due to coaching and game script improvement.
Minnesota opens with three dynamic, explosive offenses that will test their very stout defense. Overall, I expect the Vikings to play the same conservative style of offense they have the past three seasons. Adding a healthy Stefon Diggs and talented Dalvin Cook into the offense should make them much more productive themselves. They’ll need to be to keep up on the scoreboard.
The first three weeks will help us determine whether Minnesota’s defense dictates production against elite offenses, or simply tries to keep the bleeding to a minimum. Are they a shut-down unit? We’ll find out quickly.
New Orleans Saints
We may wonder how this offense could possibly improve after it sent Brandin Cooks to New England, but I believe it has, and my projected adjusted output reflects this.
No one is sure what the RB splits will look like, but this will be a very productive unit once again. I’ve projected them for slight improvement even after being ranked 2nd in Adjusted Output during the measurement period. They’ll be without LT Terron Armstead this week, however.
Defensively, I don’t forecast much change. Teams will still be forced to keep up with New Orleans on the scoreboard (whether from ahead or behind).
New York Giants
New York is very difficult to project going into this season. I find them very polar. They could be a fantastic offense, and they could be discombobulated. I lean toward them being a mix of both at this time, having Jekyll & Hyde personality on a week-to-week basis. They open the season against slow-paced teams, so they may struggle to score as frequently as they’re capable.
Defensively, I think they will struggle to play as stout as they did at the end of last season. They are a stifling unit, particularly against the passing game. I’ve projected slight increases in their allowance across the board as they face three good offenses (DAL, PHI, TB) that should get their share of lumps in.
I’ve boosted projected adjusted output across all of Philadelphia’s passing game. Why? They’re built to throw, throw, and throw. With no conventional RB on the roster, it’s difficult to see the Eagles running more than 35% of their plays.
LeGarrette Blount appears to be an ill fit in their scheme, but I can’t believe the Eagles signed him without realizing that he’s very poor from shotgun formation. It’s very possible they were only trying different ideas with him and saving conventional plays for live fire. Until I see this, however, I’m projecting RB output to suffer somewhat.
Defensively, nothing about the schedule causes concern. I’ve tweaked adjusted allowances slightly to account for improvements in their pass rush and positive game scripts.
Seattle’s offense will go as “(player-popup #russell-wilson)Russell Wilson”:/players/russell-wilson-13560’s health goes. The OL is still horrific, and sadly too reliant on Wilson’s elite skillset for the front office to put more effort into its edification. I’ve projected their adjusted outputs to increase across the board since they will either play in high-scoring affairs against equal teams (GB, TEN), or have their way against woefully overmatched opponents (SF, IND). Game scripts, regardless, will favor offensive output.
Defensively, it’s difficult to forecast anything but regression. During the measurement period, SEA simply didn’t allow much production despite playing a league average schedule. Game script should force opponents to throw more often.
San Francisco 49ers
This will be a bad team, but I think they’ll be salty enough on offense that teams will need to score in the high 20s to win.
Brian Hoyer is a significant drop-off from Colin Kaepernick at QB, but he should run the offense well enough that he is close to league-average in output. His schedule isn’t particularly scary at this point.
The 49ers will enjoy the largest bump at the WR position, where an infusion of real talent (Pierre Garcon, Marquis Goodwin) provide them with enough firepower to keep defenses honest.
Defensively, I expect a wreck. Opposing offenses should amass excellent rushing numbers against a defense that will play from behind and lacks talent.
Tampa Bay Buccaneers
Most are expecting Tampa Bay to become a heavy passing offense after adding Desean Jackson, much like they were in the first half of 2016. I’m taking the side that they won’t necessarily be more pass-heavy, but rather much more efficient due to game script and personnel.
Game script and coaching philosophy should allow
Opposing offenses will be forced to throw more against the Buccaneers by virtue of game script, so I’m projecting their adjusted allowances higher in the passing game and lower in the running game.
Washington Native Americans
Overall, I expect business as usual in Washington. I’m projecting minimal adjustments on both sides of the ball. This will still be a heavy passing offense that runs with a lead. Efficiency is the greatest question.
I am still wary of Kirk Cousins as a “good” QB. I believe he is a product of coaching and receiving talent, but many times that is enough in the NFL. For this reason I’m leaving QB adjusted output static.
Washington will miss Desean Jackson despite adding Terrelle Pryor to the offense. Jackson is a very unique player in that he boosts entire offenses on his own. I hope Pryor improves, because I like him as a player, but I cannot forecast him having the same impact on the offense that Jackson did.
Defensively, game script will dictate whether production will come from the passing game or running game. We’ll need to monitor the Native Americans closely and make quick adjustments depending on how they perform out of the gate.
By and large, teams will have difficulty shaking the adjusted fantasy point trend that they accrued late last season. In many cases, we only make slight adjustments based on offseason changes. As the season matures, we will blend calculated adjusted allowances from Weeks 1, 2, etc with these projections to update our projections as best as possible. Once we complete Week 5, we will have enough in-season data for our calculated adjusted allowances to stand on their own.
Have a question about my process? Don’t be afraid to reach out!
If you’re new to Draft, be sure to use promo code “GRINDERS” and receive a bonus on your first deposit.