Welcome to the fifth chapter of Finding Frankenstein, a PGA experimental series on FantasyLabs.
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. First, I experiment with an unreal amount of data by using our Trends tool to run trends for literally every possible metric specific to rounds played at the week’s course. Next, I use those results to construct a tournament-specific model in our Player Models. Finally, once the model is built, I take my experimentation even further, trying to create a monster player capable of taking down a tournament.
The Players Championship Recap
Last week’s Frankenstein, Branden Grace, played the weekend but finished T57 at three over par with 60 DraftKings points. He was rostered in 13.05 percent of lineups across all contests, which isn’t terrible as far as ownership goes. Grace was going along just fine at four under par heading into his final hole of the second round, but he doubled that final hole and then had a disastrous third round.
If you are a glass-half-full person, you’re probably happy that Frankenstein survived the cut. If you’re a glass-half-empty person, you’re probably pissed that Frankenstein didn’t notch a top-ten finish.
The Players was a very strange tournament, with seemingly every golfer shooting a round of at least six under par on Thursday or Friday, making the cut, and then shooting somewhere between seven to 10 over par on Saturday.
A New and Improved Process
In light of back-to-back uninspiring monsters, I have decided to fine-tune the creation piece of my experiment.
The model works. Matt Kuchar, who was the highest-rated Frankenstein-eligible player in last week’s model and my PGA play of the week for The Players, finished T3 with 91 DraftKings points.
However, the way I go about fashioning my monster has undergone a significant change this week. Previously, I isolated only a few of the most profitable or consistent trends, typically five or six. Unintentionally, I was leaving a lot of predictive data out of the creation of my monster.
This week, I am using a formula that I have developed for scoring each golfer in all of the profitable trends. Whoever has the highest total score across all of these trends will be crowned Frankenstein.
Building the Frankenstein Model for the AT&T Byron Nelson 2016
We’re enthusiastically moving onto the AT&T Byron Nelson at TPC Four Seasons Resort in Irving, TX — and, finally, the trends have been kind!
Last week, the results were dreadfully cloudy. This week, we see a rainbow of clarity over the Lab. There are some very sharp historically profitable trends for rounds played at Las Colinas. Oh, how I’ve missed you, Long-Term Adjusted Round Score (LT Adj Rd Score).
Here are a couple of observations I made in the Lab:
- Long-Term Greens in Regulation (LTGIR) produced a massive Plus/Minus of +11.57 with a lovely 61.7 percent Consistency over a relatively large sample.
- Long-Term Birdie Score was nearly as productive, with a Plus/Minus of +9.88 and Consistency of 57.4 percent on the higher end of the sample.
- LT Adj Rd Score is back, providing a Plus/Minus of +5.91 with 53.7 percent Consistency over a gigantic sample size of rounds specific to Las Colinas.
If you plan to build your own Frankenstein model this week, all the other sliders included in order of weighting are as follows: Long-Term Missed Cut, Recent Field, Recent Adj Rd Score, Upside, and just a sprinkling of Long-Term and Recent Scrambling with Recent Bogey Score.
For rounds played this year, the model has produced a Plus/Minus of at least +3.39 for golfers rating in the top 20 percent:
Creation of the Monster
As always, I eliminate the top-tier golfers from Frankenstein consideration, and that’s anybody priced above $10,000 this week. Jordan Spieth and Dustin Johnson, who have never missed a cut in a combined eleven starts at the Byron Nelson and were ranked Nos. 1 and 2 in the model are gone along with five other appealing plays. Five of the top eight golfers are out of creation consideration.
Now we have reached the point where my process has changed. Rather than select which trends should be considered and which should not, I have assigned a point total to each of the twenty-nine trends that have been historically profitable and consistent for rounds played at Las Colinas and the AT&T Byron Nelson. The twenty-five Frankenstein eligible golfers are scored based on their performance in each trend.
You may be interested to know some of these creation specifics:
- Long-Term Par-5 Score was worth the most creation points, since it produced a Plus/Minus of +5.19 with 59.3 percent Consistency over a sample of 86 golfers.
- Golfers who matched my “Elite GIR” trend were awarded 4.28 creation points.
- Long-Term Putts Per Round were scored but also received the fewest points, ranking 29th out of the 29 trends.
Monster Scoring Results
Here are some intriguing outcomes from the creation scoring:
- Louis Oosthuizen, the highest-priced golfer in consideration at $9,900, scored 47.11 creation points.
- Kyle Stanley, the lowest-priced golfer in consideration at $6,100, scored 40.06 creation points.
- Thomas Aiken was the only golfer in the field of 25 not to receive any points for Long-Term Birdie Score.
- Adam Hadwin received points from only seven trends and with only 20.31 creation points was by far the lowest-scoring contender.
- Brandt Snedeker received points from thirteen trends and Russell Henley received points from eleven trends, but they both scored exactly 35.08 creation points. They were the only golfers in the 25-person field to tie.
- Everybody’s favorite DFS homeboy, Will Wilcox, scored 33.54 creation points, ranking him 20th.
The following four golfers were the top scorers not crowned as Frankenstein this week (their total creation points are in parentheses):
- Gary Woodland (54.06)
- Brooks Koepka (53.5)
- Colt Knost (49.2)
- Ryan Palmer (48.97)
Frankenstein Revealed
Lucas Glover’s 55.79 creation points represent the best total of the week. He also received points from a field-best twenty of the twenty-nine trends. Glover is my Frankenstein for the AT&T Byron Nelson! He is the unsightly beast whom the data has deemed most apt to help take down a tournament this weekend.
Glover has no course history at the AT&T Byron Nelson but enters the event with five straight made cuts, including an eighth-place finish last week.
So load up a Frankenstein lineup, and see what kind of monstrous results you can achieve. Feel free to let me know what you think of the new process on Twitter!
Finding Frankenstein will return on June 8, 2016.
