Welcome back to the second edition of Finding Frankenstein. If you missed last week, then you don’t know how I burned down a cow pasture with G.I. Joe action figures.
Finding Frankenstein is where I leave all my opinions and narratives at the door and rely only on what the tools at FantasyLabs tell me. I experiment with the redonkulous amount of data by using our Trends tool to run trends for literally every possible metric specific to rounds played at that week’s course. I use those results to construct a tournament-specific model in our Player Models tool. Once the model is built, I take my experimentation even further, trying to create a monster player capable of taking down a tournament.
The Valero Texas Open Recap
Last week our Frankenstein ended up being Billy Horschel, who finished T4 at nine under par with 88 DraftKings points, the seventh-highest point total for the week. His ownership was higher than I had hoped for, at 16-19 percent in most tournaments, but anytime we get a Frankenstein who finishes inside the top five, this Mad Scientist will be smiling wide.
Three of the final four players in consideration for the Frankenstein title played well, including Brendan Steele, who led for a large portion of the tournament but shot a 75 on Sunday and finished T13 at seven under par. He ended up with 87 DK points, the eighth-highest total for the week. Luke Donald also finished T13 and was in contention all the way up until his 74 on Sunday. Bryson DeChambeau did his best actual Frankenstein impression and played like a real donkey, missing the cut at three over par.
In hindsight, the Valero model apparently should have sprinkled in some course history, which I completely ignored. But enough of last week, it’s onto the Zurich Classic of New Orleans played at the TPC Louisiana. Read another article if you want to know how many yards the course is and how many Par 3s and Par 5s there are. We don’t have time for that noise here.
The Zurich Classic of New Orleans
So when I run the trends, I run them at least twice for every filter. I look at large sample sizes first and then narrow the sample way down in an attempt to see if the golfers on the elite end of that spectrum truly are more profitable than the ones in the middle. Once I have analyzed all those, I start to decide which “Score” sliders will be weighted in my Frankenstein model, and how much each one should be weighted. This is borderline crazy-town stuff, but I actually enjoy it.
The first thing that struck me this week was how many metrics produced positive results. Last week in San Antonio, I really had to dig deep to uncover just a few positive trends. This week there’s an almost overwhelming plethora of options. Noticeably absent from the upper realm of encouraging trends is Long-Term Adjusted Round Score, although for people who like the stat Bryan Mears has written about finding value this week with LT AdjRd. This metric has been heavily weighted in all my models since FantasyLabs released its PGA tools, so this experiment is becoming more interesting and slightly scarier by the minute.
Since the options were many, the first thing I did was to isolate which trends showed up positively in both large and smaller sample sizes. Both Long-Term and Recent Driving Distance, as well as Long-Term and Recent Scrambling showed up as the most consistently profitable trends, so these have been weighted most heavily in this week’s Frankenstein model.
Recent Driving Distance was exceptionally consistent and profitable over a large sample size. Quite a few of the Par 4s at TPC Louisiana play as “less than driver,” so this is incredibly interesting to me. With this narrative out in the PGA DFS community, perhaps people won’t consider driving distance as terribly important this week. With that said, it could be a metric that differentiates Frankenstein from the other golfers people are on this week, so it will be weighted most heavily in my model.
Long-Term Scrambling showed a massive Plus/Minus over a small sample size, and Recent Scrambling was profitable very consistently over a larger sample, so scrambling is next in the weighting.
In order of weighting, Long-Term GIR, Long-Term Missed Cut, Recent Bogey, Recent PPR, Recent AdjRd Score, Long-Term Scrambling, Recent Driving Accuracy, Recent Field, and finally Long-Term AdjRd Score were used.
If Colin Davy reads this for some reason, I’m quite certain he will need to be resuscitated after seeing Putts Per Round weighted that high in comparison to Long-Term AdjRd Score. But this is what the creation of the Zurich Frankenstein calls for!
As always, the top-tier golfers must be removed from Frankenstein consideration, and this week that’s all players priced above $10,000. When you’re trying to create a hideous wretch suitable for tournament play, you must get rid of all the beautiful people. So we say goodbye to World No. 1 Jason Day, last year’s New Orleans winner Justin Rose, #SB2K16 Superstar Rickie Fowler, last week’s winner Charley Hoffman, 2013 New Orleans winner (and last week’s Frankenstein) Billy Horschel, and Daniel Berger, whom I couldn’t immediately think of a catchy lead-in for but is probably a swell guy.
It’s worth noting that Day and Rose were rated Nos. 1 and 2 and Hoffman No. 5 in this week’s Zurich model. The 25 remaining Frankenstein options this week were:
The most consistently profitable trend that couldn’t be weighted in our model was Recent Par-4 Form. This is where we will begin narrowing down the contenders for this week’s title. If you don’t show up in this 146-count sample, frankly, you don’t deserve to be Frankenstein.
The Par-4 metric eliminates only six golfers. Thankfully Keegan Bradley was one of them. I honestly would not have survived the tilt if I had to crown Bradley as Frankenstein this week, so I can exhale.
Next up in the creation of the monster player is Recent Bogey Score. Although this trend was weighted in the model, it produced such a consistently positive Plus/Minus over a large sample size that I wanted to dig deeper into the results.
Bogey Score eliminated another 10 names including Charles Howell III (apparently Frankenstein hates dudes with IIIs in their names). Unfortunately, DFS-Hero Will Wilcox was also eliminated from creation consideration at this point, but check out his Twitter feed to see if he’s feeling frisky this week. He evidently hung out on Bourbon Street last night, which any self-respecting American d*mn well should do while visiting New Orleans.
Only nine remain:
Some frightening names remain for sure, but lets see where the data leads us.
Recent Par 4 is a strong indicator of success over a large sample size, but it is even more consistent and more profitable over a narrow sample size, so I’m going back to the top-tier of recent Par-4 to trim some additional fat from this monster and get him in playing shape.
Five additional players were eliminated, including Marc Leishman, whom I thought would make a fine Frankenstein, but I’m not allowed to choose my monster! The data creates him!
We have left a Frankenstein Final Four for the Zurich Classic of New Orleans:
Recent Par 5 produces a massive Plus/Minus over a small sample size, partly due to the fact that six of the qualified golfers in 2015 and one golfer in 2014 scored over 100 DK points. One hundred DK points is exactly what I’m looking for from Frankenstein, so it seems only logical to use this metric as our next creation step.
Take a look at some of these massive scores from the past results tab:
Delicious is all I can say. Let Frankenstein eat on those Par 5s in New Orleans, baby!
Of the four remaining qualified golfers, only Jamie Lovemark, who has a Recent Par 5 score of -6.7 matched this trend. Lovemark is our Frankenstein for the Zurich Classic. He is the hideous wretch whom the data has deemed most suitable for taking down a tournament or two this weekend.
With all the course history debate going on, it’s worth noting that Lovemark has missed the cut in New Orleans in his only two attempts: 2012 and 2014. You can bet that the course history buffs will not be on him this weekend, which is all the more reason to like his “Frankenstein behind” in tournaments.
So load up a Frankenstein lineup, and see what kind of monstrous results you can achieve. Feel free to berate me on Twitter throughout the tournament if Frankenstein doesn’t play up to your standards.
Good luck this weekend.
