Analyzing Trends in FanDuel's Perfect Daily Fantasy NBA Lineups
In daily fantasy sports, you often get out what you put in.
To have extended success, you need to work hard and learn -- especially learn from your mistakes. You'll never find perfection and win every game, but perfection isn't a bad teaching tool.
Now, you can always take the Charlie Brown walk of shame after you decide to fade Russell Westbrook when Kevin Durant is out and leave it at that, and you can try to convince yourself that playing J.R. Smith in cash games was a great idea.
You can see what types of trends can be found when studying optimal lineups (i.e. the best lineup you could have possibly constructed based on the actual outcomes).
So, what do you need to know from the 64 NBA slates played in the 2015-16 season from tip off through the end of the calendar year?
FanDuel Perfect Lineup Analysis
If you don't already know, in FanDuel's NBA format, you fill out a nine-player roster with one center and two of every other position: point guard, shooting guard, small forward, and power forward. You have a $60,000 salary cap to construct the best lineup possible. That's enough background for now.
Of the 64 main slates (we're not looking at early-only or late-only contests) since the season tipped off, only 10 optimal lineups actually used all $60,000 of the salary cap. Five of them came during the 37 instances of a slate featuring seven or fewer games (13.5%); the other five came in the 27 slates with eight or more games (18.5%). That's something to keep in mind if you feel obliged to zero out the salary cap with every lineup.
Stacking -- or playing multiple players from the same team -- is a fairly unpopular strategy in daily fantasy basketball. After all, if one player scores, another player didn't. Aside from an assist to a score, basketball stats are pretty much mutually exclusive. On FanDuel, you can play up to four players from the same team, though. Should you? Nope.
But, here's something interesting: of the 48 slates since Opening Day that featured at least six NBA games, 46 optimal lineups featured at least one pair of teammates. I tried to factor out super short slates, when overlapping teammates is more understandable, but even on medium-sized slates, you can find teammates in the optimal lineup.
Four of the lineups featured at least a trio of teammates. Further, 30 of the 48 lineups featured at least two -- if not three -- pairs of teammates. More than half of the optimals in six-plus game slates featured at least two pairs of teammates. No lineup featured a full stack (i.e. a four-player stack).
In the right conditions, stacking can make a lot of sense in daily fantasy basketball.
For our purposes, I'm referring to short slates as slates with seven or fewer games. I know. I know. That's not really a "short slate," but I didn't want to segment the data too many times and reduce the sample sizes. We have 37 such slates from tip-off of the season through New Year's Eve. What should we know from these?
Well, of the 333 players in these 37 lineups (9 per lineup), 64 had a salary of $4,500 or less (19% of the subset). 49 players (15%) had salaries of $9,000 or more.
Only 6 of the 64 "value" players (i.e. those costing $4,500 or less) were point guards (9.4%); 14 of the "stud" players (i.e. players costing $9,000 or more) were point guards (28%).
Speaking of Stars and Scrubs...
Here's how the 37 optimal lineups fared in terms of combinations of sub-$4,500 players and plus-$9,000 players.
The "stars and scrubs" philosophy of loading up on extremes -- whether high-dollar or low-dollar -- has been key to the optimal scores in these short-to-medium slates so far this year. Of course, finding the values that actually explode is hard, but it looks like an important factor for big fantasy point totals.
Still True for Big Slates
A similar trend exists for bigger slates (i.e. those with eight or more games). 51 of the 243 optimal players were values (21%), and 36 (15%) were studs. At the small forward position, 13 of the 51 values were small forwards (anyone who plays on FanDuel knows how crucial -- yet hard -- it can be to pin down a cheap, high-scoring small forward, let alone two), accounting for 25.5% of the values.
Only six (16.7%) of the 31 studs in the optimals were small forwards.
Every optimal lineup with at least eight games on the docket featured a player priced $4,500 or less.
With so many games, you need to be willing to find and roster cheap players to help improve your chance at a monster fantasy sore.
Something to Note About the Point Guard Discrepancy
A big reason why that might be the case is that only six of the values were point guards (11.8%) and 11 of the 36 studs were point guards (30.6%). A cheap point guard may not always be the best ingredient for a high score when players such as Russell Westbrook and Stephen Curry are playing on the bigger slates.
The average score of the 37 optimal lineups on our "short slates" was 371.81. Of the 27 bigger slates: 407.89.
None of this information is going to tell you who to pick on a given night. That's not what this is about. Every slate is fluid, and the fact of the matter is that not every slate features multiple viable options who are at the bottom of the salary barrel.
Also, value picks are -- naturally -- a big part of high scores when you're dealing with a tight salary cap, but that doesn't mean punting just to punt is how to approach this. Pinpointing the low-dollar options primed for a big night isn't easy, but if you're chasing those big scores, you probably don't want to roll out a super balanced lineup and be unable to fit in superstars.
One last thing: it doesn't take an optimal lineup to win, especially in a cash game. Balance can be a viable option when there aren't value picks.
As the data continues to accumulate, I'll revisit these trends and some others. If you have anything you'd like to see discussed in future articles centering on optimal lineups, drop a comment below.