Daily Fantasy NBA: Adjusted FP Allowed Per 36
I wrote an article last week about how looking at fantasy points allowed (or DvP) in DFS NFL is not very effective at determining optimal matchups, mostly because of strength of schedule, luck, and injury impact can skew the results. However, this is likely not the case in the DFS NBA, mostly because we are able to look at much larger samples. This takes away a lot of the variance element, but strength of schedule and injury impact can still cause the data to be noisy. With strength of schedule, a team could give up a lot of fantasy points to one position or another just because they have faced a lot of very good players at that position, or vice versa. Injuries can cause a good fantasy defender to be replaced by a bad one, creating more vulnerability at that position for that team.
This inspired me to create a new tool for daily fantasy NBA called NBA Adjusted FP Allowed. Instead of looking at fantasy points allowed by position, like you can find on websites like Rotogrinders or Rotoworld, it looks at the difference between fantasy points allowed per 36 minutes and the average fantasy points per 36 of opposing players at a given position. It looks at this difference over the entire season as well as in each teams last 10 games. In essence, it’s a more advanced and accurate DvP tool.
To mitigate the impact of strength of schedule, I decided that instead of looking at fantasy points (FP) allowed per 36 by position, I would look at the difference between FP allowed per 36 and average FP per 36 of opposing players. This means that if a team faced a lot of good opposing players at a given position, they would only be considered bad at defending that position if their FP allowed per 36 was greater than the average FP per 36 of opposing players. You can look at the results in the NBA Adjusted FP Allowed tool. (Note: You may want to download the file in a csv or excel document to sort by position and then by other statistics to give the data better organization)
At this point in the season, strength of schedule has not had a large impact on FP allowed. The spread of average FP per 36 allowed is only a couple points between teams. But in some cases, those couple points make a big difference. Houston would appear to be the best fantasy matchup for PFs, but Philadelphia is actually a bit worse, they have just had a much softer schedule of opposing PFs, while Houston has faced the best opposing PFs in the league. The Lakers have been the worst team against opposing PGs this season, but look even worse when you see they haven’t faced a lot of good PGs. Utah looks like an average matchup for SFs, but Utah has faced a brutal schedule, leading average FP per 36 against by a full 1.5 FP over the field, and therefore are a very bad matchup for opposing SFs.
Using adjusted FP allowed is also beneficial when working with smaller samples. Injuries can cause teams to give worse quality of players playing time and therefore make them vulnerable. Adjusting for strength of schedule, we take away that factors impact on the FP allowed data. Sometimes, looking at last several games can help us see if injuries have impacted a teams fantasy defense. If there doesn’t appear to be a reason why a team has been worse in the last 10 games than they would be on the season, I probably would assume that last 10 games are variance.
The L10DIFF column gives us the difference between FP allowed above average in the season versus the last 10 games, and can help us identify where injuries may be impacting FP allowed. Golden State has the biggest L10DIFF at on the day I wrote this article, as they have been average against SFs for the season but by far the best matchup for opposing SFs the last 10 games. This makes a lot of sense. Harrison Barnes has been out, replaced by Brandon Rush in the starting lineup. Rush is not a good defender and is a poor offensive player, so the Warriors are much more vulnerable to SFs recently than they have been on the season, and probably are the best SF matchup until Barnes returns. Similarly, Phoenix has lost their starting C and PF the past several games, have been starting Jon Leuer at PF, and often going small late in the game. It’s hard to tell whether Leuer is a bad defender or their small ball lineup is vulnerable to big men, but either way injuries have clearly caused them to become a better matchup for opposing PFs.
Fantasy points allowed statistics are useful in the NBA because we can look at large samples. But strength of schedule can still have a substantial impact on that data. With adjusted FP allowed, we should have a better picture of which teams truly give up a lot of fantasy points to any given position. Adjusted FP allowed can also help us examine data in a smaller sample to identify injury impact. Use this tool to help you evaluate matchups in daily fantasy NBA.
View all posts by Daniel Steinberg