NHL Best Bets for Tuesday (11/19)
Kenneth Le
Over the course of the PGA season, I’d like to do a few deep dives into the world of statistics in golf. There are plenty of stats to choose from, and quite frankly, it’s hard to wade through it all.
In the end, people focus on the granular data because it’s available but I’m not sure how useful it is across all applications. I’ve found the most success over the past three years, using what I call my Power Ranking (PRK) Score. It’s a single number that attempts to predict how a golfer’s skill is relative to an average PGA Tour golfer on an average course.
How PRK is Calculated
PRK is a weighted average of a golfer’s scoring across different timeframes. It’s great to be able to look through different components of scoring separately with counting stats like Greens in Regulation or with strokes gained data, but at the end of the day, we’re not trying to predict that. We’re trying to predict how a golfer will score. There’s a lot of noise that comes along with all of those other stats and sometimes the simplest answer is the correct one.
Raw PRK is one way that I measure this which is basically an expected SG: Total number. An elite number would be somewhere in the -2 range. For example, Jon Rahm was right at -1.95 heading into this week’s tournament. An average golfer would be right at 0 and anything above 0 means that I’d expect that golfer to lose strokes to the field with their average play.
PRK Score is what goes into the Volatility Report data table. It converts that raw number into a field relative scale from 0-100 for a particular tournament.
Winners
When it comes to betting or DFS, we want to find winners on the PGA Tour. On average, there are only a few golfers each week in the highest group with a PRK Score of 90-100. Golfers in that group account for just under 14% of the winners on Tour since the start of 2018. We can see that the PRK scores are directionally correct by looking at the win rates of golfers in each range.
|
PRK Score Group |
Win Rate |
|
90 – 100 |
3.7% |
|
80 – 90 |
2.5% |
|
70 – 80 |
1.3% |
|
60 – 70 |
0.8% |
|
50 – 60 |
0.3% |
|
0 – 50 |
0.3% |
These numbers seem low, but there’s obviously only one golfer who can win each tournament. As you go down by each level in the PRK Score, you significantly chop your win equity.
Other Finish Positions
Picking winners is tough, and it’s not the only thing we should be trying to do. We can have profitable seasons without hitting any winners if we’re careful with our risk and how we select golfers.
|
PRK Score Group |
Top 5 |
Top 10 |
Top 20 |
|
90 – 100 |
18.9% |
31.1% |
49.7% |
|
80 – 90 |
12% |
21% |
36.3% |
|
70 – 80 |
8.1% |
15% |
26.4% |
|
60 – 70 |
4.8% |
9.2% |
19.4% |
|
50 – 60 |
4.1% |
7.7% |
15.6% |
|
0 – 50 |
2.2% |
5.4% |
12% |
We see a similar trend here. Every single cohort level that we go up, improves our odds of hitting all of the finishing positions. This is important because it shows directional accuracy with the model. This isn’t to say that there’s a guarantee here, but we can use these baserates in betting and compared to DFS salaries to improve our chances of being profitable week in and week out on the PGA Tour.
