Spoiler alert: I just started playing daily fantasy basketball this year and I’ve never actually written a single NBA article in my life—until now—so if we were betting on the usefulness of this post, I’d take the under on “Barely Useful.”
So here’s what I know about playing daily fantasy basketball: minutes matter. You need players who are going to be on the court a lot. The more time on the court, the better. That’s not necessarily true for the players or teams in real life, but on DraftKings—a bulk scoring site on which efficiency doesn’t matter—we just want to accrue as many opportunites for stats as possible.
The fact that the best predictor of fantasy production is minutes played is the reason why you can find value on backups when the starter goes down. It doesn’t matter how large the gap in talent is between certain players if one plays substantially more than the other; no amount of NBA talent can consistently make up for a significant lack of playing time when it comes to fantasy scoring.
Of course, it isn’t like there’s a perfect correlation between minutes and fantasy points such that we only need to consider minutes when projecting players. Minutes are crucial, but perhaps certain types of players need minutes more than others.
Minutes and Fantasy Production
To find out how valuable minutes are to certain types of players, I charted the correlation between minutes played and 14 different stats for the top 200 most active players during the 2013-14 season: field goals made, field goals attempts, two-pointers, three-pointers, free throws, offensive rebounds, defensive rebounds, total rebounds, assists, steals, blocks, turnovers, points, and field goal percentage.
All of the numbers except for field goal percentage are bulk stats. Unsurprisingly, that correlation is the weakest at 0.05. That means that minutes played has very little predictive value in regards to field goal percentage; players who get more time on the court aren’t much more efficient on a per-shot basis than those who get little playing time.
The strongest correlation is with points, followed by field goal attempts, and field goals made. This makes intuitive sense since shots and points aren’t a low-frequency event. For Kobe Bryant to score 40 points, for example, he needs to be on the court; he can’t get into foul trouble, for example, and he can’t get benched in the fourth quarter of a blowout.
Now compare the strong 0.78 r-value of minutes/points to that for minutes/blocks. It’s just 0.21, meaning that although players who see the most minutes have the most blocks, the relationship isn’t nearly as strong. Since blocks are much rarer than points, they tend to come in bunches. It’s not uncommon for a shot-blocker to not block a single shot in one game, then follow it up with seven blocks in the next. Minutes matter, of course, but not nearly as much as they do for scorers.
6 Take-Home Points
Minutes are still the most important stat to consider for all players.
The point of this article isn’t to determine if minutes are important—it’s clear that time on the court is the most important stat to consider when projecting any player—but rather to see how much it matters for certain types of players.
If you’re relying on a scorer for fantasy production, he needs to see a lot of minutes.
There’s a very linear relationship between minutes played and points scored. Think about how difficult it is for a player to even double his season average in points. Now consider how challenging that is if he doesn’t see a significant amount of playing time. Whereas players regularly surpass their season average in assists and blocks by a significant percentage, it’s very difficult to do that with points unless there’s a substantial boost in minutes played.
Shot-blockers and rebounders don’t rely on pure minutes as much as scorers.
Again, because stats like blocked shots aren’t as common as points, they aren’t as reliant on minutes played. Anthony Davis is much more likely to record a high number of blocks if he’s on the court for 40 minutes instead of 30 minutes, but his upside isn’t capped in the latter scenario as much as it would be with points.
If you’re going to punt a position on DraftKings—going min-priced and hoping they return value—you’re probably best off doing that with a player who doesn’t rely primarily on points scored for production. Unless you’re confident he’s going to be on the court at an elite rate—which few min-priced players are—then you should be seeking stats that aren’t extremely dependent on minutes played.
Steals are pretty strongly correlated with minutes—more so than rebounds.
Steals are pretty low-frequency, so I’m not really sure why this is the case. It might be that the types of players capable of stealing the ball often are starters, meaning they naturally play more minutes. Either way, I don’t really have a great answer for why steals are so strongly linked to minutes played.
You can use minutes to predict turnovers better than all stats except points, field goals attempts, and field goals made.
I think this makes sense in that turnovers are fairly random. Some players are of course going to turn the ball over more than others—particularly those who touch the ball a lot—but there’s also a lot of variance here.
This same reasoning could be a potential explanation for why steals are so strongly correlated with minutes played. Perhaps the act of stealing the ball has more to do with just being around it often—which requires a lot of minutes on the court—than the actual talent level of the defender.
That’s just conjecture on my part, obviously, but it would mean that if you’re relying on a player whose fantasy numbers come primarily from points and steals—like Stephen Curry—then his play is going to be far more dependent on minutes than a rebounding shot-blocker, for example.
It might make sense to pay for points in cash games.
No matter the sport, it makes sense to narrow the range of potential outcomes by allocating the most salary cap space to the most consistent players when participating in cash games—head-to-heads and 50/50 leagues. Well, since points are extremely dependent on minutes played and minutes are fairly easy to predict, it follows that point-scorers are probably more consistent on a night-to-night basis than players who rely on other less frequent stats for their fantasy production. That idea fits with some previous NBA research I’ve done as well.
If you want consistency, spend money on scorers. If you want tournament upside, it might actually be smart to seek players who don’t rely so heavily on points for their fantasy output because those players are the most likely to exceed their average scoring by a significant margin.