This week’s analysis focuses on DFS values, with clear implications for owners in typical points leagues. The goal is to answer critical questions, such as: Which positions are most plentiful among the top-200 DFS players? Which are most scarce? How many DFS points per game do shooting guards average, compared to every other position? Which statistical categories account for the largest and smallest percentages of DFS points? How strongly does a player’s usage, scoring, rebounding or defensive stats correlate to their overall value?
Answering such questions allows even casual DFS owners to make strategic decisions about their lineups. We will see, for example, that point guards are under-represented in the top-200 overall players, yet as a group they average the highest DFS points per game (28.67). As such, you might think about paying up to secure one of those relatively scarce but valuable PGs to anchor your lineup. This is just one example of how to apply the data, which we should examine before going any further.
Editor’s Note: The NBA Season Pass is up and running. Get weekly projections, rankings, the popular Pickups of the Day column, exclusive columns and customizable stat options. It’s the extra tool that can take your fantasy basketball teams to the next level.
Scoring Systems
Most DFS scoring systems are very similar, as you’ll notice in this comparison of three prominent sites (subsequent analysis refers to FanDuel for simplicity’s sake):
| Yahoo | FanDuel | DraftKings | |
|---|---|---|---|
| Points: | 1 | 1 | 1 |
| 3-pointers: | 0.5 | n/a | 0.5 |
| Rebounds: | 1.2 | 1.2 | 1.25 |
| Assists: | 1.5 | 1.5 | 1.5 |
| Steals: | 2 | 2 | 2 |
| Blocks: | 2 | 2 | 2 |
| Turnovers: | -1 | -1 | -0.5 |
| Double-Double: | n/a | n/a | 1.5 |
| Triple-Double: | n/a | n/a | 3 |
| Rosters: | 8: PG, SG, G, SF, PF, F, C, UTIL | 9: PG, PG, SG, SG, SF, SF, PF, PF, C | 8: PG, SG, G, SF, PF, F, C, UTIL |
Scoring: Yahoo and FanDuel have nearly identical scoring systems, with Yahoo adding 0.5 points extra for made 3-pointers, while Draft Kings includes a few minor tweaks and the major addition of bonus points for double-doubles and triple-doubles.
Rosters: Yahoo and Draft Kings are identical, while FanDuel requires an extra player and greater specificity in terms of positions. These differences are crucial to keep in mind, depending upon which league you are in, but they’re not so large as to preclude an over-arching analysis. Let’s begin.
Category Impact
A glance at the scoring systems above makes it clear that ‘points’ as a category are going to be extremely valuable. In fact, I’ve done previous analyses in which points account for roughly half of all DFS value. Let’s see if that’s changed at all this season, using FanDuel as our reference.
Here is another way to visualize that data:
So far this season, the top-200 players have relied on points for more than half of all DFS scoring in this system (the inclusion of turnovers skew the numbers slightly in the pie chart). Let that sink in. Rebounds are second at 25.3%. Assists, at a relatively meager 16.4%, still account for more overall value than both defensive categories combined. Turnovers have a marginal impact.
The supremacy of points-scored cannot be overstated when gauging DFS values. Part of their power is that they’ve been uncoupled from the potentially damaging impact of percentages – when a poor FT shooter like Dwight Howard gets hacked, his DFS owners get points even if he goes 6-of-13 at the line. If Devin Booker shoots 7-of-22 from the field, his owners aren’t penalized for his high-volume inefficiency. Linger over these numbers as long as you’d like before pressing on with the analysis.
You can follow me on Twitter @Knaus_RW!
Value by Position
There are multiple methods to parse DFS value by positions. A simple approach is to determine how many qualifying players at a given position crack the top-200 for overall DFS points per game (ppg). The top-200 equates to 6.7 players per team, on average, if every team were active...that sounds about right. Once this top-200 group is aggregated, I’ll break them out to show how positional distribution changes when you only look at the top-150, top-100 and top-50.
# of players in top-200, by position | |
Position | Total # |
SF | 43 |
SG | 43 |
C | 41 |
PF | 37 |
PG | 36 |
Grand Total | 200 |
There’s nothing nuanced about this table. It shows that swingmen are relatively plentiful, while power forwards and point guards are relatively scarce. The disparity isn’t all that great, however, with only seven players separating PGs from SFs. Knowing how many players from each position make the cut is only part of the equation, though, since it’s important to know where they are distributed among the top-200.
This provides a more detailed picture. We see, for instance, that there are 14 point guards who currently crack the top-50 for overall FanDuel value. That’s more than all the top-50 power forwards (7) and shooting guards (6) combined. Centers are also well-represented at the top of the pile. Low-end shooting guards appear plentiful in this analysis, while small forwards emerge as a position that can be found in any range.
Average DFS points per game, by position
Another method I’ll consider is determining how many DFS ppg those players average, by position. To qualify, a player must have appeared in at least 10 games this season (as of Tuesday, Dec. 6)...that means no Jeremy Lin, Dirk Nowitzki, Will Barton, etc.
Top-200 per-game value, by position | |
Position | DFS ppg |
PG | 28.67 |
C | 25.43 |
SF | 25.11 |
PF | 24.83 |
SG | 22.71 |
Grand Total | 25.25 |
Here the script is flipped from the positional-eligibility table above, with PGs emerging as the most valuable on a per-game basis, with SGs bringing up the rear. As with the previous chart, however, we want to know how the picture changes when you look at the distribution of players across different groups (top-150, top-100, top-50).
I intended to look at the average values per player across these different groups, an oversight I’ll correct the next time I’m doing a deep-dive into DFS. The chart above is instead a reflection of the total per-game values for all players who qualify -- PGs lead the top-50 by a wide margin with a cumulative 538.3, SFs and Cs are close behind, etc.
If you desire, you can easily figure out how much each group averaged per player...we saw above that 14 point guards qualified in the top-50. Those players have a cumulative average of 538.3 DFS point per game, so simple division (538.3/14) yields an average of 38.45 DFS points for the top-50 group of PGs. Again, feel free to mull over these tables as long as you’d like...there’s still a lot to cover!
Correlations for DFS
This section is designed to determine one thing -- whether there is a correlation between a player’s usage, rebounds, assists, etc. and their DFS points per game (and if so, how weak or strong that correlation might be).
To clarify the term ‘usage’ for those who don’t know, here is the definition from the estimable Basketball-Reference.com: “Usage Percentage...the formula is 100 * ((FGA + 0.44 * FTA + TOV) * (Tm MP / 5)) / (MP * (Tm FGA + 0.44 * Tm FTA + Tm TOV)). Usage percentage is an estimate of the percentage of team plays used by a player while he was on the floor.”
Here is the correlation result for usage:
The key number is “0.72606,” which indicates a strong positive linear relationship between players’ overall FanDuel values and usage rates. Roughly speaking, when one of the variables (usage) increases, the other variable (DFS value) also increases. To emphasize just how strong the relationship is, here are the correlation results for every scoring category:
Unsurprisingly, given what we saw above, points have the strongest correlation with overall value. Usage is second and it’s interesting to see turnovers with the third-strongest correlation -- they are a part of usage, for one thing, and their negative impact is outweighed by the fact that players who commit the most turnovers also tend to rack up the most counting stats in other categories. My takeaway is that, for DFS purposes, you can basically ignore turnovers.
Assists also have a stronger positive correlation than rebounds, which is another interesting wrinkle. We saw earlier that boards accounted for 25.3% of all DFS value among the top-200, with assists at 16.4%. The implication is that assists are more likely to be found in players who produce more overall value (scoring, steals), while rebounds are concentrated in players who contribute less in other categories. This supports the earlier result showing point guards with the highest average DFS value as a position. It all starts to tie together when you dig deep enough.
Values by Players
I’ll conclude with a straight-forward list of the top-150 DFS players as gauged by FanDuel’s scoring system (which, again, is very similar to other major sites):
I hope this helped you think critically about new ways to gauge DFS value. If you have any questions or insights, you can always reach me on Twitter @Knaus_RW. Good luck this week.
