I was watching the Scottish Championship on the European Tour on Thursday, and they were interviewing eventual winner Adrian Otaegui after he opened with a tournament-leading 10-under 62 at Fairmont St. Andrews. Otaegui was telling the interviewer how he felt a good round and a great performance had been coming, even though his finishes and scores didn’t necessarily indicate it was imminent.
Always looking for an edge and a better way to look at players, I wondered what kind of markers we might be able to find for players who are on the come — even if their results aren’t showing it.
The best single statistic we have available is something we already track. We compare a player’s average strokes gained tee-to-green per round in the prior 50 PGA Tour events to the same measure in the prior 10 PGA Tour events. This gives us an indication if a player is trending above or below their long-term baseline performance in the one stat that can really tell us if a player is ready to breakout.
However, averages are problematic from a statistical point of view, particularly with a relatively small sample size and the possibility of erratic changes in values in the data set. We typically see some pretty dramatic ranges in strokes gained tee-to-green from event to event, for a variety of reasons. Averaging out that data creates a comparison point between players, but it doesn’t tell us the players who are more consistent from event to event, course to course, grass to grass, field to field.
That’s why we should be using long-term variance to give us a better clue as to the best players on Tour in terms of consistency.
To see our hidden picks and weekly model, join Forebucks!
For just $50 for 12 months or for free with a qualifying sportsbook deposit, Forebucks members get access to our winning weekly PGA Tour event model, in-depth DraftKings picks, searchable database of PGA Tour results from 2011-present and top-15 PGA Tour trends, as well exclusive members-only content.
You’ll also have access to our course fit modeling, Quality Stokes Gained data, course demand insights, as well individual access to fantasy expert Ryan Ballengee for your unique questions.
Variance takes averages and helps us understand how much deviation there is from the average in each part of the data set. It’s calculated by finding the difference between a data point and the average of the data set, then squaring each of those differences (deltas), adding them and dividing them by the total number of data points.
Unto itself, population variance doesn’t necessarily tell us much. However, when used to compare a wide variety of similar data sets, we can get a measure of consistency in values. That’s especially useful with strokes gained tee-to-green. We not only want to know which players are gaining the most strokes tee-to-green but also how wildly they deviate from their average. The less variance, the more consistent a player is.
Let’s look at this using the CJ Cup field and data leading into last week for the last 30 PGA Tour events.
|An, Byeong Hun||3.031||0.725|
This chart features the top 25 players in that field based on average strokes gained tee-to-green by round. The number before is is the variance for each player. We can use that as a comparison point to help us identify consistent players.
Corey Conners and Harris English jump off the page as solid performers who turn in similar SGT2G efforts week to week, across events, designers, grasses and fields. Shane Lowry is similar. Keegan Bradley is consistently great from tee to green (and also consistently bad with a putter). Brooks Koepka hasn’t been the same powerhouse he was since his injury, but even still, he’s pretty consistently solid.
Even the best players are solidly consistent, with Justin Thomas and Collin Morikawa doing pretty on a regular basis. Jon Rahm and Tony Finau don’t show a lot of variability.
Dustin Johnson on the other hand? Whoa. Of course, turning in 80-80-78-WD will do that to a guy.
Viktor Hovland plays in a pretty wide range with a solid SGT2G game. He can stink it up one week and be top three the next. It averages out similarly to Keegan, who plays in a much tighter range.
Our next goal will be to look at the whole population — a bigger sample — and work with standard deviations to help give our numbers even more clarity.
For now, though, this is helpful in finding those foundational players we want to roster each week.