By the Numbers: Profiling Week Winners
Soon after the fantasy football season had ended last year, Rotoworld’s own Evan Silva worked to add the fresh information to his data set. With the new data, Silva looked at the past three seasons of weekly player performances and built “profiles” for the four key offensive positions in fantasy. These profiles detailed the characteristics of the Top weekly finishers at each position. For instance, in his RB profile, Silva found that of the 51 running backs that finished as the top weekly scorer from 2013 to 2015, 92% handled 18 or more touches and 71% played on a team that was favored by Vegas. These data points constructed by Silva can be used as the backbone of weekly lineup setting. There are no true guarantees in fantasy football, but if you are looking to score the top weekly RB in DFS, seven times out of 10 he is on a team that is favored.
The weekly player profiles, if incorporated into your decision-making process properly, can provide small competitive advantages. They also can become even more powerful and credible when the data set is expanded to more than just the number one finisher of each week. I took Silva’s analysis a few steps further and expanded to include the Top 6, Top 12, Top 24 etc. at RB and WR. With this new information we should be able to not only validate Silva’s findings, but also reach new conclusions about the factors that separate a good matchup from a great matchup. First I’ll jump into the RBs and then take a look at the WRs.
RB Scoring Landscape
The value this analysis can provide can be summed up with the following table:
This table highlights just how important it is to be able to make the right decisions between players when constructing a lineup. Based on the last three seasons, even identifying a Top 6 RB in a given week, leaves you missing out on eight points to the Top overall RB. Note that this point differential has been consistent over the last three seasons and is double the difference between a Top 6 and Top 12 RB. Season-long participants who make lineup decisions with fewer players at their disposal than in DFS might be more interested in just being able to start a Top-24 RB every week, which is worth 18 points on average. Season-long players can use the following RB profile to better their odds at making the right decision in this quest to start Top-24 weekly players.
The RB Profile:
In January, Silva created the following profile for the Top weekly RB:
Accounted for multiple touchdowns: 42/51 (82.4%)
Handled 18 or more touches: 47/51 (92.2%)
Played on team that won the game: 39/51 (76.5%)
Played on team that was favored: 36/51 (70.6%)
Played at home: 31/51 (60.8%)
Faced run defense that ranked 15th or worse in DVOA: 38/51 (74.5%)
Faced run defense that ranked 10th or better in DVOA: 4/51 (7.8%)
With my expanded data set (and a few added categories), the profile shook out like this:
Since this is now a lot more data to look at in the table and it might be hard to digest immediately, I find the picture helps ease you into the right mindset:
I’d like to highlight a few things before digging into the heavy analysis of the profile. First, I think it is important to keep in mind that winning in fantasy football is about gaining every little advantage you can. Some of the percentages might seem to indicate that a particular characteristic plays little significance in determining a good matchup. However, over the course of a season and over many DFS contests every small difference adds up to something significant. It is also important to focus your attention on predictive measures, i.e. the ones that are known prior to a game with near-certainty. We know who is home, we know who is favored and we know what defense a player is playing against.
“Drop off” in the chart is the percentage difference between a Top 1 RB and a Top 6 RB. A higher “drop off” indicates a category that is more important. For example, a Top RB scored 2+ TDs 82% of the time, while a Top 6 reached 2+ TDs only 47% of the time. This 35% “drop off” shows how important scoring multiple TDs is to a running back’s weekly finish. The players included in the “All” data set are all “viable” fantasy players that played each week at the position. For example, looking at the table you can see that 91% of Top 6 RBs were under the age of 30, but also 91% of all viable RBs in the NFL each week were under 30, so age does not seem to have any material impact on weekly RB scoring. I define “viable” as either a starter or a player that averaged enough touches per game to garner actual fantasy consideration.
Key Takeaways from the RB Profile:
Touchdowns
From 2013 to 2015, 82% of Top 1 RBs scored multiple TDs but only 47% of Top 6 RBs accomplished that feat. Scoring multiple TDs greatly increases a running back’s chances of being a Top RB, however it is very challenging to predict. It’s not often we have any clue going into a game that a player is going to score multiple TDs, but looking into red zone carries per week or goal line touches can be a good proxy for trying to use this data point in weekly decisions. Maybe this could even lead you to picking up LeGarrette Blount when it’s going to be “one of those Blount games”. Few backs actually score multiple touchdowns each week, as evidenced by only 15% of the Top 24 RBs each week scoring more than once. Hence, TD scoring is ultra-important in large pool DFS contests and should be a major focus of lineup construction.
Workload
A running back handling a large workload (of at least 18 touches) tends to score more than other RBs. More than 90% of top RBs handled the rock more than 18 times. The spread is quite large between Top 6, 12 and 24, indicating that a RBs workload is very much a determinant of being a dominant weekly performer. Still, two-thirds of Top 12 RBs were given these large workloads compared to only about one-fifth of all viable fantasy RBs getting 18+ touches a week. This tells you that these high-touch guys are scattered sparsely across the league and if you have one in your lineup he has a good shot at putting up a top performance for you.
From 2013 to 2015, there were 616 games played by RBs who averaged 18 or more touches per game over the season. When these RBs saw at least 18 touches in a game, they scored an average of 18.8 fantasy points per week. I see this as more important to season-long fantasy owners who should focus on volume if feeling obliged to draft the RB position early.
Passing Game Involvement
Top weekly running backs in PPR are catching a good amount of passes (much to no one’s surprise), with over 70% of Top 6 scorers getting at least three catches. From 2013 to 2015 there were 32 RBs who averaged at least three catches per game, and these are often considered the high-floor RBs in PPR. So while a RB doesn’t have to necessarily be a pass-catching specialist, the elite scoring backs tend to be involved in the passing game to some extent. Two-thirds of top 12 RBs caught 3+ passes, so the first RB you draft this year should be a guy that’s going to catch at least 50 passes this season.
Vegas
We already know Vegas is helpful in determining viable fantasy options, but this analysis shows there is a good chance the top running backs in a given week are on home teams and/or favored teams. This is especially helpful because these factors are determined before the game is started. Furthermore, with the expanded dataset we can see that while just over half of “startable” (Top 24) RBs are on favored teams, the Top RB each week was favored over 70% of the time. Simply starting a player because he is at home seems to be grasping at straws strategically, but a majority of top performers across the board are playing at home. On the other hand, only 44% of Top 6 RBs played in games with the over/under set at 47 or higher. These 47 points are certainly a high threshold (66th percentile) but it begs the question: do lower scoring games actually lend themselves to better RB fantasy production? This could have a lot to do with game flow (as Silva discussed) which is always tricky to predict.
Opposing Defense
About two-thirds of RBs that finish Top 6 and Top 12 are competing against a run defense outside of the top-15 according to Fantasy Outsider’s Run Defense DVOA. This is another predictive measure and is one of the most consistent predictors of RB success. You should not expect top RB performances from RBs facing teams with a top-10 ranked run defense according to DVOA. In Silva’s weekly matchups column during the season, he focuses heavily on opposing defense ranks and my analysis supports the theory that sometimes player performance is less about the actual player and more about the opposition. (Note that due to the timing of my analysis, I only had access to year-end DVOA and not the weekly ranks which would be more ideal for in-season decisions.)
Combining the Factors
Combining all the takeaways from a study like this is powerful in identifying undervalued RBs in the draft, free agency and DFS contests. DFS players can apply the conclusions from this analysis in a contest with many contestants where it pays to go against the grain. Great things can happen when all the predictive measures come together at once: a RB who is home, favored, in a game with an over/under of at least 47, averages 18 or more touches and averages at least three receptions. There have been 38 RBs facing these conditions in the past three seasons and they averaged 21 points in those games with 14 of them scoring at least 25 points.
