Like many of my peers who had to transfer into business school, I chose to start my college career taking a number of economics courses. Those introductory courses all began with the basic assumption that everyone has complete information. When prices go up for specific goods, people want less of them. And because those classroom examples were not the real world, the ‘people’ lowered their demand by the exact right amount.
If this were true for daily fantasy sports, we would react to changes in price or performance expectations in the precisely right way.
We would all go home.
Stop writing DFS content.
Stuff our money into a mattress.
And the world would be worse.
Fortunately for all of us this is not how things work. Our Player Models and Trends tool constantly update with new data and up-to-date player projections before lineups lock, and then we “let Jesus take the wheel.”
This piece is about how to bet against those projections.
Don’t Try This At Home
Projected Plus/Minus (Proj Plus/Minus) measures “projected points minus salary-based expected points.” Higher is better, and for cash games you should look to target players with large Proj Plus/Minus marks.
As shown below, players with low Proj Plus/Minus values historically underperform. However, looking at the green bar you can see that sometimes these players hit and exceed value.
Three Little Pigs
Not all players with poor projections are created equally. Starting with the worst, I looked at three distinct groups of players that each account for about 15 percent of the total WR group (a count of about 650 within each group). Here are the groups and their Proj Plus/Minus ranges:
• ‘Worst’: -3.85 to -9.9
• ‘Pretty Bad’: -2.1 to -3.84
• ‘Not That Bad’: -0.3 to -2.0
Below are the percentages of players within each group that scored more than 2x (our Upside metric) the group’s average expected points:
‘Worst’: Average Proj Plus/Minus of -3.8
While contrarian, the ‘Worst’ group doesn’t seem to justify the risk. The top-10 performers scored an average of 18.6 DK points and only 4.4 percent of the entire group exceeded the group’s Upside. The juice from this fantasy dream team is simply not worth the squeeze.
‘Not That Bad’: Average Proj Plus/Minus of -0.7
As we might expect, this group maintained the highest success rate. The top performers from this group were elite fantasy options with difficult matchups. These players potentially make great plays, especially since they often come with reduced ownership rates. My advice is to monitor ownership rates and target elite WRs in tough matchups in GPPs.
Finding the Sweet Spot With ‘Pretty Bad’
• These contrarian WRs had an average Proj Plus/Minus of -2.6.
• The 8.7 percent success rate suggests a relatively wide range of outcomes from this set.
• An average score of 36 DK points from the top performers highlights this group’s ceiling.
This is good. THIS is what we want to find. When we search out contrarian plays we are looking for low-owned players with a high enough ceiling to compensate for the risk.
Looking to Next Year
Predicting when players with ‘Pretty Bad’ projections will peak means pulling apart a few common narratives to find potential outliers.
“…has no upside.”
A lot has been written on the importance of fragility in daily fantasy sports and the value of using players who are capable of producing exceptional weeks. When looking at players with ‘Pretty Bad’ projections, we should start our search with only those players with projected ceilings high enough to offset the relative risk within this group. Take a look at how Plus/Minus distribution changes when we consider the full group versus the players in the top 15 percent in terms of projected ceiling.
I like the chart on the right. Our sample is getting pretty small but let’s stick with it for this example.
“…looked awful the last few weeks.”
Great players have bad games but rarely have bad months or bad seasons. When we look at high-ceiling players with a string of recent poor performances, we actually see more consistent success in those who have a ‘Monthly Duds’ rate of more than 33 percent.
Again, the sample is small in this example but I would consider playing any high-ceiling player with a ‘Monthly Duds’ rate of 30-plus percent given the historical data.
“…is way overpriced.”
The final variable to consider is pricing. Proj Plus/Minus is ultimately a measure of projected results relative to pricing. Occasionally we see large pricing corrections of $500-plus that cause spikes in salary-based expected points, meaning that some high-ceiling players may end up with ‘Pretty Bad’ Proj Plus/Minus numbers.
The Cliff Notes
• Players with largely negative Proj Plus/Minus marks can be intriguing contrarian plays.
• To find which players to target, sometimes we can use narratives to figure out why a player ‘deserves’ his poor Proj Plus/Minus and whether those narratives and resulting low projections are justified.
• To do this, we want to focus on players with high ceilings. Why risk it for no biscuit?
• We also want to consider players who have had either 1) a string of bad performances using the ‘Monthly Duds’ metric or 2) a large recent salary increase.
