We have an awesome new tool on the site that shows the ownership percentages right after PGA lock for all players throughout different levels of buy-ins: It’s our DFS Ownership Dashboard.
The DFS Ownership Dashboard
Because the players who in high stakes contests tend to be more incentivized because of their larger investment, their plays tend to be sharper. Also, it’s really hard to play at high stakes continuously without sharp plays. My point is this: Even if you plan on never jumping up into high stakes PGA guaranteed prize pools, you should absolutely look at the ownership levels in those tournaments. Looking at total ownership is better than looking just at winning lineups: We want to avoid a results-oriented process with probabilistic blind spots. Our DFS Ownership Dashboard is the best resource in the industry to do just that.
If you haven’t seen it yet, here’s what it looks like:
The Volatility and GPP Grade columns measure the difference between a player’s ownership in low- versus high-stakes tournaments. If you are consistently high on players who have high Volatility and GPP Grades, you’re likely aligning yourself with the sharp DFS players.
For this article, I want to look at the sharp plays from last week’s Genesis Open and, with data from our Player Models, hopefully start to become a sharper player. To do this, I’m looking at every player’s Volatility Rating and GPP Grade from last week as well as the following data points going into the tournament:
- Round 1 Tee Time
- Odds to Win
- Long-Term & Recent Adjusted Round Score
- Long-Term & Recent Greens in Regulation
- Long-Term & Recent Driving Distance
- Long-Term & Recent Driving Accuracy
- Long-Term & Recent Putts Per Round
- Long-Term & Recent Scrambling Percentage
- Long-Term & Recent Field Percentage
- Long-Term & Recent Birdies Per Tournament
- Long-Term & Recent Missed Cut Percentage
- Course Adjusted Round Score
- Course Count
In order to analyze this data properly, I’ve adjusted for player salaries. It’s not useful to consider the correlation between a player’s Adjusted Round Score and his GPP Grade, because DFS is played within a salary cap. This is a dope lineup . . .
. . . but it’s not even close to being under the salary cap. Ownership is useful because of the salary cap restraints, so we must adapt our data.
To do this, I’ve used a process similar to the one I used in my Vegas Bargain Rating articles last year. Basically, I’m looking for players who excel at particular statistics relative to their salaries. For example, Brooks Koepka had a LT Adj Rd Score of 68.7 going into the week, yet he was priced at only $7,100. Using a 1-100 scale, we can judge players against each other for each statistic relative to their salaries. Koepka had a 100 with his LT Adj Rd Score, while Charlie Beljan (with a field-worst LT Adj Rd Score of 74.1) had a 0. In terms of LT Adj Rd Score, Koepka and Beljan were the two most mispriced golfers in the field.
The Correlations
Once we have the scaled values for each statistic, we can correlate them with our ‘sharp play evaluator’ statistics. If LT Adj Rd Score and GPP Grade have a high correlation, it’s likely that sharper players are better at identifying discrepancies in that particular statistic. Anyway, here are the scores for all of the statistics mentioned above.
And here are the correlation values. I’ve also added in correlation with DK points since I was curious:
In terms of actual DK points, it’s intriguing that the highest correlation is with Round 1 Tee Time. The weather was projected to be very bad, especially on Friday, so the late Thursday/early Friday golfers were expected to receive a significant bump in value. That overwhelmingly proved to be true, highlighting a point made on our most recent PGA Daily Fantasy Flex podcast: When weather is a factor, it is the factor.
In terms of finding sharp plays, it seems that LT Adj Rd Score is the best bet, which is good for our users, as it’s a proprietary metric developed by PGA Director Colin Davy and available only at FantasyLabs. If you did nothing else but build a model and weight LT Adj Rd Score at 100 percent, you’d immediately have an edge. It’s that useful of a statistic. Of course, there are many more factors in play other than Adj Rd Score, but it’s a fantastic starting point for identifying value.
After LT Adj RD Score, the statistics that correlate the most to GPP Grade are LT Birdies Per Tournament and LT Missed Cut Percentage. The first makes sense, as DK really values birdies in its scoring system.
However, there is certainly an edge to be found in looking at players’ ability to create birdies and scoring opportunities. Labs Co-founder Peter Jennings (CSURAM88) played young golfer Thomas Pieters in high-stakes contests last week, perhaps because Pieters was an incredible value given his salary and ability to score birdies. If you look at the table above, Pieters had a 100 rating in LT Birdies Per Tournament: He ranked 10th among all golfers with 15.5 birdies per tournament, yet he cost near the minimum at $6,700.
Along those same lines, several of the ‘sharpest’ plays last week were the value guys who have consistently made cuts over the past year: Koepka, J.B. Holmes, and even a guy like Francesco Molinari all stand out at their respective price points.
Process Over Results
We harp on process over results here at Labs, and using the PGA DFS Ownership Dashboard is an incredible way to improve your process of researching and thinking. We’ll periodically dive back into this sort of study over the next several months and consider the types of golfers sharp DFS players roster.
