DFS Turnaround - Week 10



It’s NFL season and that means you’ve got a chance at $100,000 with Sunday Night 7. Predict what will happen on Sunday Night Football and watch along to see if you’ll be the next big winner. Click here to get started with the NBC Sports Predictor app.

The dynamic game of Daily Fantasy Sports (DFS) requires much more than simply knowing the sport for which we're entering contests to be successful. We must be adaptable, precise, and open to learning from previous endeavors, the latter of which will be the primary focus of this weekly written piece. Game Theoretic methodologies will allow us to analyze and dissect the previous week's winner of the largest and most prestigious Guaranteed Prize Pool (GPP) tournament on DraftKings – the Millionaire Maker. These same tenets of Game Theory, which can most simply be explained as the development of decision-making processes given our own skill and knowledge, assumptions of the field based on the cumulative skill and knowledge of others playing the same game, and the rules and structure of the game itself, will allow us to further train our minds to see beyond the antiquated techniques of roster building being employed by a large portion of the field. Approaching improvement through these methods will give us insight into the anatomy of successful rosters and will help us develop repeatably profitable habit patterns for the coming weeks. We'll start by looking at the previous week's winning roster, extract any pertinent lessons for future utilization, and finish with a look ahead towards the coming main slate.

Winning Roster

Week 10 Milly Winner
Week 10 Milly Winner


Lessons Learned

The Value of Tight End Points

For the second consecutive week, the optimal roster contained Justin Fields paired with his tight end, Cole Kmet. Since we've covered why it is theoretically optimal to stack a tight end with his quarterback, let's talk quickly about the value of guaranteed points at the tight end position – and it circles back to how tight ends derive their fantasy value, broken down into tiers. If we lump tight ends into two distinct tiers, one containing the tight ends that can put the slate out of reach at the positions (Travis Kelce, Mark Andrews, and George Kittle) and one containing all the “variance tight ends” (the ones that require touchdowns as the main driving force of their fantasy value), we can then formulate a plan to attack the most variant position in the game of NFL DFS.

When we play a tight end in the latter grouping, volume is not present, meaning we're looking for bulk scoring at the position (which comes in the form of touchdowns). We can even move George Kittle into this grouping as well due to his low volume expectation this season, but his per-touch efficiency keeps him more aptly placed in the former grouping with the tight ends that can put the slate out of reach. The variance associated with a player scoring multiple touchdowns means we can capture both sides of those points by playing them paired with their quarterback; however, there is intrinsic value in capturing “guaranteed points” at the position instead of variance-hunting. Enter the former grouping of tight ends (minus George Kittle), or the only two tight ends that we can count on for consistent volume. More consistent volume leads to a range of outcomes that is shifted to the right (higher), meaning we don't need these two players to hit their 95-99% outcome in order to be optimal. Typically, Travis Kelce and Mark Andrews can provide enough volume for fantasy players to need only a 70%+ outcome to remain in the hunt for the top overall spot in GPP tournaments, meaning we are reducing the variables associated with a highly variant position by playing them. As such, those two players (Travis Kelce and Mark Andrews) are the only two at the position where theory dictates one can stray from playing their quarterback in a stack with them.

The Value of One-Offs

DraftKings user funkymonk1369 shipped $1 million with a haphazard group of secondary correlations and one-offs, but he or she wasn't winning anything without the three-touchdown game from Christian Watson, who single handedly offset the requirement for the top overall fantasy score on the slate – Justin Jefferson (38.30 fantasy points). As is often the case, the optimal roster involved very little correlation, but we need to divorce the idea of correlation pertaining to the optimal roster. We don't need the optimal roster to win. We aren't playing a computer, we are playing other flawed human beings – human beings that make mistakes. That said, we stack and correlate to remove variables, to lessen the number of things we need to get right on a slate in order to generate a repeatably profitable habit pattern. As such, the idea of the optimal roster bears little importance to us. What we're hunting for is the theoretically optimal processes that lead to winning rosters, backed up by Game Theoretic methodologies and the data from previous winners.

Typically, those winning rosters contain the top overall raw point total on the slate, which for this week was Justin Jefferson. Most times are able to capture exposure to the top overall score on the slate through the game environments we choose to attack, but that isn't necessarily always the case. This idea should dictate the types of players we choose as one-offs for GPP rosters, with the primary focus placed on top-end ceiling. It's not enough to be wasting a roster spot on safety through our one-off selection, eliminating the vast majority of the player pool in the process.

The Value of Touchdowns

This might seem like an obvious point to make – hey, play the players that score touchdowns because that's how you capture bulk points! Profound! While it seems superficial, there is value in first identifying how the pieces in the game we're playing score their points. Let's relate this discussion on variance to baseball and homeruns. What do optimal rosters look like in MLB DFS? It is typically littered with a bunch of guys who score bulk points on a slate, or the ones that hit home runs. The problem with playing the game of MLB DFS that way is that the percentage chance of even the best players hitting a home run in the isolated sample size of one game is about 6%. Considering there are eight hitters on a DFS roster, that would mean the chance of capturing eight different players to hit a home run is about 1.68 per 10 billion from a pure mathematical sense. Thusly, we don't need optimal in order to win – we need to develop repeatably profitable habit patterns utilizing optimal theory!

In our example of MLB DFS, typically we leverage the data on opposing pitchers and bullpens in an attempt to identify spots where the team will score runs, which gives the players in a primary stack everything from bulk scoring opportunities to additional at bats to the chance to piece together similar scoring to a home run through multiple different acts. In NFL DFS, that process looks like primary correlations, stacking, secondary correlations, and attacking game environments – all of which increase the chances of capturing bulk scoring while simultaneously reducing the number of variables in order to succeed. All of that said, touchdowns reign supreme when it comes to bulk scoring, it is simply difficult to predict their outcome as one of the highest variant acts in the sport. This leads us back to the process of attacking game environments, teams, and upside situations in our hunt for overall upside from individual offenses, as opposed to trying to guess nine roster spots right.

Looking Ahead

Justin Fields + (The Value of Touchdowns +)

There is a psychological aspect of DFS that comes with player pricing, in that just two weeks ago Justin Fields was priced 50% less than he is for Week 11. This makes people feel like they “missed out” on the opportunity to play Fields profitably, particularly considering the presence of Josh Allen, Lamar Jackson, and Jalen Hurts on the slate. The reality is that Fields is very likely to still be underpriced relative to his upside (duh, he's scored 43.4 fantasy points or more in consecutive weeks!). This presents an interesting opportunity to continue playing Fields at ownership that is highly likely to be lower than it should be in a game against the sieve-like Atlanta defense. The value of touchdowns comes through the high touchdown share that Fields carries in his growing role on the Chicago offense. The “plus” part of the equation is found through always stacking a mobile quarterback with exactly one of his pass-catchers, as we've covered before in this space.

Brian Robinson (The Value of Touchdowns)

The Texans have allowed 14 total touchdowns to opposing backfields this season and the Commanders' three-headed backfield is highly likely to remain a two-headed monster into Week 11 with the injury to J.D. McKissic. The path to a 100-yard game with multiple touchdowns is there for Robinson to absolutely smash his modest price of just $5,600. The same can be said for backfield-mate Antonio Gibson, albeit coming with a wider range of potential outcomes due to his lower snap rate. Embracing touchdown variance in ways the field are unlikely to do is one way to generate leverage smartly, without introducing suboptimal plays.

Garrett Wilson (The Value of One-Offs)

Wilson provides the rare opportunity to access a highly concentrated offense without the need to attack the expected game environment, which for a game against the plodding Patriots is a plus. Wilson has become the focal point of the Jets offense with a 27.5% targets per route run rate (ranks 16th in the league). DFS managers can safely insert Wilson into lineups as an upside one-off with such a massive share of the offense, even in a game environment expected to be rather poor.

Mark Andrews (The Value of Tight End Points)

Andrews has seen his price dip below $7,000 for the first time since Week 5 after two lost games due to injury and the team's Week 10 bye. When on the field and healthy, Andrews makes up a massive portion of the Ravens offense, responsible for an unreal 87.2% route participation rate, a massive 31.2% targets per route run rate, and the top team target market share at the tight end position (29.9%). Locking in points at the highly variant tight end position is one of the ways to tackle the position, as we discussed above.