DailyOverlay Grading Process
Below is the process that we use to record, grade and score DFS experts across all sports.
We collect all of the expert’s picks from the various sites via a manual process. Prior to each sports season, we accumulate a list of at least 15 different experts/sites that we will pull picks from for the upcoming season. We try to pull a wide variety of selections including free articles, paid articles, optimizers and projections sets. We try to focus on the pulling cash game selections, if noted, from articles while only pulling the primary picks that have been written up or emphasized in the write-up. For projections sets, we typically sort by points per dollar and use the top few plays at each position. The selections pulled from optimizers are typically done as close to when we produce our Weighted Consensus article for the slate as possible. If given a choice, we will typically pull picks from DraftKings focused articles, optimizers and projections since DraftKings scoring it the basis of our grading process.
Recording Actual Scores
After all the games of a given slate are completed, we pull the actual DraftKings points that were scored by every player on the slate. We then assign those points to each player that was recommended by a DFS expert for the slate. Any players that were recommended by an expert but did not play in the game for the slate are removed from the data set for that slate. For example, if a hitter is recommended for MLB and he is not in the starting lineup that night, he will be excluded from that expert’s list of recommended players for that slate. This also includes all players from any game that is postponed or canceled. Once this step is completed we then have a clean set of experts picks that can be put through our grading process.
Player Grading Process
Now that we have a clean set of data we are ready to produce a grade for each player. Each player is graded on three different criteria.
- Value Grade – Each player’s value grade is determined by their relative value score (points per dollar) compared to the top value score (points per dollar) at their position for the slate. Players receive a letter grade (A+, A, B, and so on) depending on the percentage rate that their value score generates compared to the top value score at their position. (Such as greater than 93% = A+, greater than 85% = A, and so on).
- Ranking Grade – Each player’s ranking grade is determined by how they rank compared to all of the other players for the slate at their position in points scored. The cutoffs for how many players receive an A+, A, B, and so on are determined based on a sliding scale depending on the number of games that are played on the slate and their position.
- Gross Score Grade – Each player’s gross score grade is determined by their relative points scored compared to the top point scorer at their position for the slate. Players receive a letter grade (A+, A, B, and so on) depending on the percentage rate that their value score generates compared to the top point scorer at their position. (Such as greater than 93% = A+, greater than 85% = A, and so on).
Each player’s Final Grade is determined by taking the highest individual grade from the three grading metrics above.
Note that what constitutes a “position” varies between sports. For football, we separate positions by QB, RB, WR, TE and D/ST. For baseball, we separate positions by hitter versus pitcher. For basketball, we treat all players as the same position.
Expert Scoring Process
After all individual players have been given their Final Grade for the slate, then each expert gets their DailyOverlay Score generated based on the player’s grades that they recommended.
Each expert’s DailyOverlay Score is calculated as an average of all of the individual player’s grades using the following weighting.
A+ = 10
A = 5
B = 3
C = 2
D = 1
F = 0
Experts scores are then tabulated for each slate and overall to produce the final DailyOverlay Scores and rankings.
NOTE: DailyOverlay Scores are also used to help weight the Consensus Rankings for future slates, so the top expert’s recommendations receive more weight than the bottom ranking experts.
If you have any questions or feedback on the grading/scoring process email us at firstname.lastname@example.org