One of the most fundamental aspects of DFS Strategy is understanding the “Range of Outcomes”. Many newer DFS players make the mistake of treating the projection like a prediction. The projection of a player does not tell us what is going to happen. No matter how good the projection source, it will not tell us the outcome of a single player, game, or stack. It is impossible to predict the future. So what good are projections if they don’t predict what’s going to happen? They tell us the average outcome in the distribution or range of outcomes.
Every outcome has a corresponding probability. In DFS, we plot these probabilities on a Normal Distribution curve. What is a Normal Distribution curve you might ask?
The above image is an illustration of Normal Distribution. The midpoint of the curve is where the projection falls in the range of outcomes. That is to say, the average outcome will fall at the midpoint of the curve. The distribution is segmented by Standard Deviation (more on that later), and each segment has a corresponding probability that the event falls within that range in the distribution. The graph reads as follows:
The same can be said for the inverse:
How does this apply to DFS Strategy?
Every player has a “Range of Outcomes” of their fantasy point output ranging from 0 (starts the game, and rolls his ankle in the first minute, and leaves with injury) to 100+. It’s very important to note that any outcome is possible within the range of outcomes. You’ll see from the above graph that most outcomes (64%) fall within 1 standard deviation from the mean/projection (either positive or negative).
Okay, so what the heck is Standard Deviation?
Let’s say Player A has a data set of 3 games and has scored: 5, 10, 15 fantasy points. The average/mean would be 10.
Let’s say Player B has a data set of 3 games and has scored: 9, 10, 11 fantasy points. The average/mean would be 10.
So in this very simplistic model that didn’t factor in any other variable, both players would have their projection (average/mean) at 10 fantasy points.
Standard Deviation is a measurement of the variation between the numbers. We won’t dive deep into how to calculate the standard deviation in this article by hand as there is a very simple formula in Excel or Google Sheets that you can use to calculate it (stay tuned for future excel based content). In our situation, the Standard Deviations of the two players are:
Standard Deviation for Player A = 4.1
Standard Deviation for Player B = 0.8
Remember, most outcomes (68%) fall within 1 Standard Deviation so we can likely expect:
Player A to score between 5.9 and 14.1 (mean – standard deviation and mean + standard deviation)
Player B to score between 9.2 and 10.8 (mean – standard deviation and mean + standard deviation)
What does that tell us about these two players?
As we can see, the top end of Player A’s range of outcomes is 22.3+ fantasy points whereas the top end of Player B’s range of outcomes is 11.6+ fantasy points. Knowing how the Standard Deviation impacts the ceiling and floor of both players enables us to use them more efficiently. Player B may make more sense in a lineup that wants a higher floor (Cash), while Player A may be better suited for a lineup that wants to increase the variance of the score (GPP).
The key takeaway with all of this is to look at projections as a range of possibilities. Some outcomes are more likely than others, but nothing is off the table. If a player scores one, two, or even three standard deviations above or below their projection, it doesn’t mean that the projection was incorrect. Every player is going to hit every part of their range of outcomes with some degree of frequency. Our job isn’t to try to predict the outcome of the slate. All we can do is figure out what the most likely range of outcomes are and leverage them to our advantage.
You must log in to post a comment.