Fantasy Golf – The Players Championship Projections
[Editor’s Note: We’re continuing to expand our growing coverage of DFS sports, which now includes PGA. We will be offering these first two weeks of DFS PGA projections free of charge. After that, we expect to move PGA coverage to a paid subscription, in line with all our other current DFS offerings.]
Welcome to the first of what I hope will be a regular column in which I share my fantasy golf projections. My projections for last week’s Wells Fargo Open did very well. I’ll talk about how I made the projections, then get into who I like this week. Here are The Players Championship Projections.
I forecast the following for each player: final score to par, number of DraftKings points, percent chance to make the cut, and percent chance to win the tournament. The ‘Value’ column is based on the player’s expected points per dollar of salary.
Each forecast number is an expected value. If you’re not familiar with the concept, it’s just what will happen on average. For instance, it’s very likely that the winner of the tournament will have a score much lower than Jason Day’s projected -5.2 below par. Almost by definition the winner of the tournament must perform better than his projected average. My simulations put the average winning score around -12, very similar to the winning score the last few years. The flip side also holds true, the players that finish near the bottom will under-perform their expected totals.
I used the following procedure to obtain the spreadsheet you see below:
- Using every complete round played on the PGA tour from 2010-2016, I calculated a z score for each golfer. I assumed that golf scores are normally distributed, and then standardized them. I took each player’s score, subtracted the mean of all players for that round, and divided by the standard deviation of scores for that round. If you’ve taken stats you might recognize: A negative z score is good – it means your score is below average, which is good in golf.
- I corrected each z score according to the strength of each field (not all fields are created equal).
- I calculated a moving average of adjusted z scores and the standard deviation of those scores.
- I found the average score and standard deviation for The Players Championship over the last five years. I took the two parameters for each player, adjusted z score and standard deviation, from step 3 and scaled them to fit the current tournament.
- Then via Monte Carlo simulation, I played the tournament 20,000 times using the tournament specific parameters for each player to randomly determine their scores.
- Finally, I matched each player’s simulated score to his past DraftKings scores when he’s shot the same score under par.
The Players Championship Projections
The projections are in the table below. If you’d rather work directly with a Google spreadsheet click here.
The main focus with these projections is that golf is played against a field. A golfer’s scores are adjusted for the difficulty of each field he’s played against. More credit is given for a low score relative to a strong field than versus a weak field. The scores are agnostic to which course was played. The most important factor for real world and fantasy success is how a golfer performs relative to the field.
There is no opinion in these numbers. I didn’t cherry pick data points to support a narrative or carefully construct a reasoned argument based on premises that have not been proven to work. Instead, I painstakingly constructed an unbiased mathematical model based on solid statistical principles. I back tested it for accuracy and made numerous tweaks and improvements before I felt it was worth sharing. I will continue to analyze its performance in future articles. Still, golf is hard to predict. I will make bad calls, but hopefully more good than bad.
Players I like
There is an extremely strong field this week, typical for The Players. That leaves some very good players with low salaries. Henrik Stenson appears a little under-priced this week at $9500, likely in part to missing the cut last week for the first time since March 2014. I also like Charles Howell III, much better than a typical $6700 golfer, to bounce back this week after missing only his second cut this year.
Anyone with a dark green cell in the ‘Value’ column is a good value – you probably didn’t need me to tell you that. Freddie Jacobson is the best value out there, but he comes with some risk. He’s a volatile player with a wide range, large standard deviation, of outcomes. Jamie Lovemark and Chez Reavie are two solid golfers that always seem a little too cheap on DraftKings.
Two players with good upside and reasonable prices this week are Jimmy Walker and Louis Oosthuizen. Both players have the ability to shoot really low scores. They just lack the consistency of the very top players.
Below are two sample lineups. The first is constructed to have the best chance of all players making the cut. The second is more of a tournament lineup. Please use the projections to construct your own as well.
|Charles Howell III||55.44||0.738||6700|
Please don’t rely on me to update the projections for injuries or withdrawals. I will make an effort to update the spreadsheet, but no guarantees. Also, be cautious with Jim Furyk. He’s only played once, last week where he missed the cut, since coming back from a long layoff due to injury. I did manually adjust his score downward but it was just an educated guess.
Stay tuned for more content and projections in the weeks ahead. I’ve put a lot of work into what you see here and I’m always trying to improve it. If you have questions or comments, please feel free to email me: firstname.lastname@example.org. Good luck this weekend.