Download Week 2 Percent Rostered File - Steve Buzzard
Estimating the percent rostered in FanDuel mid-sized GPP contests
Download Week 2 DraftKings Percent Rostered File - Steve Buzzard
Estimating the percent rostered in DraftKings mid-sized GPP contests
Why is Percent Rostered Important?
Each week this year I will be projecting the percent rostered rates for the medium priced contests on both DraftKings and FanDuel. Percent rostered rates are arguably the most important aspect of DFS, especially for tournaments. As players continue to get better every week thanks to great coverage like we have here at Footballguys it becomes more important to play against your opponents than just picking the “best” value plays. Consider the following example:
You are playing a simple game where you and 100 of your friends have to pick who will score more points between Drew Brees and Tom Brady. All the losers contribute $10 to the pot for the winners to split. Let’s say your prediction is about the same for both of them, 20 points. You might think it doesn’t really matter who you pick. You will win half the time and lose half the time. But let’s say you also know that a lot of people really like Brady and you find out about 70% of the people are going to pick him. Now the obvious choice is picking Brees.
Let’s see what happens to the people who pick Brees when he wins. 70 people that picked Brady must contribute $10 to the pot to be split between the 30 Brees pickers. So each player who picked Brees receives $23.33 calculated as ($10*70)/30 = $23.33.
Following the same process for the players that picked Brady they will receive $4.28 every time he wins.
If your projection is correct and each has about a 50% chance of winning half the time you would receive $23.33 from Brees winning and half the time you would lose your $10 making. This means picking Brees would have the expected value of $6.65 calculated as $23.33*50% - $10*50% = $6.65. Alternatively, Brady would have an expected value of -$2.86.
It is a big difference between the two despite their projections being nearly identical. In this example, you could be way off on the projection for both Brady and Brees but still easily pick Brees as the correct play by seeing how much more often Brady was being picked. This is basically how I have won countless tournaments in all sports over the last few years and you can do the same with these percent rostered projections.