One of the most talked-about components of daily fantasy sports — outside of fantasy point projections — is the understanding of ownership at both the individual and overall roster level.
Many use ownership reports to unearth overlooked players or fade popular plays due to a discrepancy between projections and reality. While this is an important process in some instances, it may not be as repeatable as developing a system on how to view ownership in the correct context.
Let’s start at the beginning and work our way through the stats that drive ownership, and investigate how we can utilize this information to make better roster decisions in daily fantasy contests.
Subscribe to PFF's EDGE or ELITE subscriptions get access to the following tools, charts, etc.
– All Premium Article Content
– Fantasy Football Rankings & Projections
– Fantasy Football Draft Guide, Cheat Sheets & Advanced Stats
– Fantasy Football Mock Draft Tool (DraftMaster)
– Fantasy Football Strength of Schedule Matrix
– NFL Player Grades
– NFL Premium Stats
What is ownership, and why does it matter?
Ownership is the percentage of rosters that utilize a particular player. For example, Patrick Mahomes was the quarterback on 23.5% of rosters that entered the Draftkings Millionaire Maker in Week 4 of the 2019 season.
This is actionable information because Mahomes had a mundane performance in Week 4, putting up 21 DraftKings points with a salary of $7,500. You were toast if you played with a roster that featured Mahomes that week, as eight quarterbacks with a lower salary outscored him, but you were cashing checks if you rostered Jameis Winston (at less than 1% ownership for $5,700) after he put up 33.3 fantasy points.
At the individual level, ownership is a way to differentiate your lineup from the majority of rosters in your contest to hopefully maximize profits if things break your way. By doing this, you can leverage the difficulty in projecting individual player performances every week, allowing you to gain an edge on the field that you are playing against.
The salary component of daily fantasy sports means that, theoretically, every player on a given slate has at least some small chance of ending up on the roster that wins the top prize in that contest. The relationship between the player’s probability of ending up on the winning roster and his actual ownership is the exploitable difference that results from ownership percentages.
How this plays out at the quarterback position
The quarterback position in daily fantasy sports is unique because it only allows one individual in that position group to be used. It also offers a significant portion of starting quarterbacks as worthwhile roster options.
In contrast, the tight end position can also be used in the flex position, but it offers fewer worthwhile plays on a weekly basis compared to quarterbacks simply because of the volume component.
This plays out in ownership, as we see rosters less concentrated on certain individual quarterbacks, which broadens the options owned at quarterback on a given week.
We never see quarterback ownership percentages in the 30%-or-higher range, making a fade on a particular player less appealing than at other positions. Site pricing also has an impact on quarterback ownership being dense and highly concentrated in the 5-15% range.
Typically, players experiencing the highest ownership are value plays inserted due to injuries after initial salaries are released. DFS sites price up backup quarterbacks to eliminate this opportunity for a free fill-in play. As you can see in the chart below, quarterbacks have the narrowest range of ownership compared to the other skill positions.
Distribution of ownership by position
What drives ownership at the quarterback position?
If you are to have any chance at success when playing daily fantasy sports, you need to implement a worthwhile, repeatable process when it comes to looking at ownership and how that affects your particular roster construction.
One way we can begin to research is by identifying the particular stats that drive ownership and looking at how they relate to projecting fantasy points.
Since site operators have spent years perfecting salary pricing, it is no surprise to see salary perform exceptionally well when predicting ownership and fantasy points. So, after adjusting to account for salary, we can look at the previous n-game average for a variety of popular metrics and see how worthwhile they are in predicting ownership and fantasy points in the subsequent week.
This information can not only help explain what drives ownership and fantasy points, but it can also help guide the question of which timeframe to use with certain historical stats. Note that we are using DraftKings’ salary and ownership data from Weeks 1-17 of the 2017-2019 seasons.
Fantasy points typically come in chunks, with passing touchdowns making up a significant portion of quarterback output. Taking a moving average of the previous n-game ranges, it seems as though daily fantasy players overvalue recent passing touchdowns at a substantial clip. This is most apparent in the previous game’s passing touchdowns, which explains a significant portion of quarterback ownership percentage but almost nothing related to quarterback fantasy output in the subsequent game.
When it comes to large-field guaranteed prize pools (GPPs), fading a quarterback who is coming off a monster passing touchdown performance is one of the safest bets to make. Even if the quarterback has two great games in succession, this comes with increased ownership and thus makes it harder to roster this scenario and come out ahead.
If you’re looking to use previous passing touchdowns as a guide, looking at the previous five games is best when it comes to predicting fantasy points.
Similar to passing touchdowns, previous-game passing yards are overvalued by the average DFS player in the weeks immediately after a big performance. However, it offers almost no predictive value on fantasy points in the subsequent week. The term often used to describe this in the DFS industry is “point-chasing,” and it is one of the biggest errors new players tend to make.
Yards per attempt
Once thought to be the marquee passing stat, yards per attempt has since been lapped by more worthwhile efficiency metrics. It does a reasonably good job of predicting quarterback ownership over the previous two- or three-game samples but is unreliable when used to predict fantasy points. It should be left out of any decision-making process with the hope that others you are competing against still place some type of value on it.
Adjusted yards per pass attempt (AYA)
AYA is a metric that is similar to fantasy points, except it attempts to convert touchdowns and interceptions into a yards metric as opposed to a points metric. It rewards a value to touchdowns and interceptions, among other stats, and is combined into a value on a per-pass attempt basis. As is the case with the other metrics so far, it is overvalued by DFS players in the short term but does move the needle from a fantasy points prediction standpoint over the longer term.
PFF’s overall passing grade is a useful predictor of fantasy points, especially over the five-game range. DFS players rush to play a quarterback coming off a monster game or two, which doesn’t predict fantasy points as well as taking a further look back over the course of five games. Utilizing a longer timeframe without overreacting to a recent performance is key when integrating passer grade into your process of selecting the appropriate quarterback in DFS.
Average depth of target
Despite aDot being among the stickiest of stats for predicting wide receiver performance, it receives no such distinction when evaluating the quarterback position. It is essentially meaningless when used to predict fantasy performance for quarterbacks — it barely moves the needle when predicting ownership over the most recent timeframes.
Passing Air Conversion Ratio (PACR)
PACR is an efficiency metric made popular by Josh Hermsmeyer and his development of the airyards website. This metric can be expressed simply by dividing yards per attempt by average depth of target. The exciting part about PACR is that it is most worthwhile when predicting fantasy performance using the previous three-game range. It doesn’t predict ownership nearly as well as fantasy points, which makes it the first worthwhile metric we have encountered during this process. Utilizing PACR in your research will highlight plays at the quarterback position that will fly under the radar of the average DFS player.
Yards after the catch
A highly unstable stat, yards after the catch struggles in any sort of prediction facet. It may be quick and easy to judge wide receiver units that have been exceeding expectations, but this isn’t sustainable going forward. It is not worthwhile to include in your DFS process.
DraftKings fantasy points
The average number of fantasy points is typically the first research point by most DFS players, as it is essentially forced viewing on any DFS site. Thankfully, it is at least somewhat worthwhile from a fantasy point perspective, but it does significantly better when predicting ownership.
It seems that the high view count affects DFS players, especially since it can be within eye’s view as roster decisions are made. Similar to passing touchdowns and yards, the most recent performances are overvalued by DFS players, which is an exploitable edge that should continue since gamblers are rarely known for taking a cautious and patient approach to predicting future outcomes.
DraftKings fantasy points from rushing
The Konami code of fantasy football has always been discussed as the mobile quarterback who offers a reliable floor thanks to his legs. We don’t have to look far for a recent example, as Lamar Jackson and his fantasy MVP 2019 season was in large part due to his rushing ability. We see this played out in our linear regression model as previous DraftKings fantasy points from rushing plays is the most worthwhile predictor of fantasy points in subsequent weeks.
The ability to generate production with your legs appears to be undervalued in recent seasons. This is a situation that could see an overreaction from either DFS players or site-pricers who could continue to move up the price of the subset of mobile quarterbacks. This is actionable information but is on the verge of becoming too apparent, as we’re seeing that it’s more of a driver of ownership than actual fantasy results.
An initial understanding of what drives ownership and how these metrics interact with fantasy points is worthwhile when trying to perform better in DFS contests.
Most of the stats are prone to recency bias from DFS players, but that provides information as to what direction your opponents will lean toward. The most worthwhile stats for quarterbacks are PACR and passing grade over an appropriate time frame where the DFS players who chase points have forgotten about the performance, and the fantasy player can be had at lower ownership.
Generating fantasy points from rushes is the most worthwhile when predicting fantasy points. Still, the rise in ownership means this is most likely an edge that is diminishing before our eyes.
Quarterback ownership in general is less significant than at other positions, so we will continue to dig into each position as it relates to ownership and fantasy points.