DFS Strategy: Late Swap Game Theory

Should you keep Patrick Kane late in a GPP slate or swap him out?

Of the numerous differences between the two major daily fantasy sports sites, none is greater than the late swap feature on DraftKings. FanDuel lineups lock at the start of the first game of the night, regardless of when later games begin. On DraftKings, your handicapping is just beginning once you send your roster in and the action starts.

Many players you will be competing against in GPPs make no site-to-site adjustments, and this fact leads to inefficiencies. It’s a mistake not to factor in the scoring differences across sites, for example. A simple example from the hockey world would be players who plug their rosters into both sites on a night where one game figures to be high scoring on a short slate. The stack of players from both sides of the same game that is logical on DraftKings becomes a serious error when posted to FanDuel due to FanDuel’s accounting of plus/minus in their scoring.

Aside from the scoring differences however, there is even greater opportunity of an edge on DraftKings due to in-game play. Again, many of your competitors will not pay any attention to late news, such as in hockey where an AHL call-up gets the assignment on Line 1, or the many late NBA moves of starters getting the night off, which are often announced less than 30 minutes prior to tipoff. Some of your competition will be busy, some will say they ‘just don’t like changing what I came up with,’ some just want to watch the games at the bar with their buddies and drink.

Knowing that a portion of our competition effectively ‘checks out’ at the start of the first game of the night, it can be more than reasonably deduced that in-game play is out of the question for these entrants. But it gets better; even those who will make a last-minute roster change based on late news will either never give a thought to game theory at that point or will be those who say, ‘For better or worse, this is my squad.’

This mindset leaves a gaping hole in long-term ROI due to the structure of payouts in a typical GPP. A standard example would be in the DraftKings nightly $27 buy-in NHL Faceoff, where a recent representative tourney had $35,000 in prizes for a field of 1,445 and the top 300 got paid. The top three places paid $6,000, $3,250, and $2,000 for a total of $11,250. A little over 32% of the total prize pool went to the top three finishers, top five returned 38% of the pool, and the top 10 got 43%. So, of the 300 top finishers, 10 chopped 43% of the pot, while the other 96.6% of prize winners were playing for the other 57% of the money.

How does this little math class play into game theory? Well, that depends on someone’s personal level of risk they’re comfortable with. In an example, you’re in 35th place with one game left and 20 people in front of you have a player remaining. Some quick math and the open spot on their roster lets you know that they are all rostering Patrick Kane playing the west coast game against Vancouver. You have him as well.

He is projected to easily outscore any one player going in that game. Whether he scores four goals or gets hurt on his first shift, your ceiling is exactly the same: 21st place. You cannot win, you cannot hit the top three, and you cannot even hit the top 20. If you’re happy with a likely finish in the low 20s, then fine. It’s nice to finish as high up the board as you can, especially if you’re new to the game. I personally would never leave him in and would switch to someone from a completely different line, knowing most, if not all, of the people in front of me will not do so.

You put in Andrew Ladd, and leave a few thousand in salary unused, which is completely irrelevant at this juncture of the game. If Kane gets held down and Ladd pots a couple, the sky’s the limit. In the more likely event of Kane doing what he does, not only do you not finish in the top 20, but those close enough behind you that also have Kane will blast past you.

This is what keeps some inexperienced players from fully employing game theory; the fear of the more likely downside. This is the point in the tournament where your chances of winning are greater by using the player less likely to succeed, so once you are at peace with this regardless of outcome, you are at the point where your game will see long-term ROI benefits.

Now, take what you’ve learned and sign up for DraftKings.