This is the first hole at Davenport Golf Club, where I play the vast majority of my golf. The hole itself is about a 400 yard Par 4 that swings right at about 160 yards off the tee. This post is the start of the series, looking at how data has changed how I play holes, and changed the decision I make, positively!
In terms of my stats in the 2022/2023 season here’s the report card… In general I improved this year from previous attempts, with an average of 4.77 versus 4.92 in previous years, that’s a nice little chip off the score for the hole and adds up to a decent amount! But the question is where it comes from, and why…
Putting
Overall my putting has slipped back a little on the hole, averaging 2.1 rather than 1.9 - although when I hit the green in regulation it’s about 2.15 versus 2.3 so that’s something! The big downside is that often in previous year’s I’ve just missed the green so it’s been an easy up and down - this year not quite as much… with 1.6 putts versus 2.0!
Overall it’s the first green and it’ll always be a test, it’s hard to tell whether it’s quicker or slower than the practice green out of the gate, and I’d always expect there to be a little more variance there.
Approach
However, that’s the crux of it, when putting doesn’t quite tell the full story - there’s two key stats to track (and they’re not putts): Greens in Regulation and Greens in Regulation plus One. GIR has improved from 0.35 to 0.48 and GIRP from 0.85 to 0.87 (and yes that’s in the noise!). The question is what’s brought about that increase in GIR - even with the poor putting it’s brought the average score down, and nothing worse than a double being observed (compared with an uncomfortable level of triples or worse in previous years), albeit an increase in doubles.
The Reason
The question is why? The answer is somewhat complicated and needs a little explanation. But can be summarised by a diagram!
So… I don’t do Decade or any of the shot planning processes, but I have read Mark Broadie’s book. As a project for some students last year we developed a website to look at optimal tee shot planning for any course, with any player’s dispersion and distance parameters. It works well, and we can sketch the layout of the hole over a satellite image and change parameters and let the optimisation algorithm go to work. The algorithm picks out the optimal strokes gained landing position for every club based on easy to obtain data (taken from measurements on the range: distance, short/long and left/right dispersion plus rotation of the eclipse). It optimises over complete dispersion patter so that the outcome is as good as it can be.
The standard advice on the shot is hit down the middle of the fairway opening with a power fade. Ultimately I think that’s always been a bad take, especially as:
It’s the first hole of the day, and that’s not an easy shot to hit
If it starts at all left (and it should for a fade…) it’ll hit the trees by the front tees just short of the faraway and end up with a recovery shot from 330 yards.
The landing zone for the fade (around 230-240 yards) is strongly uphill, no runout and an awkward uphill lie into a green guarded by trees, short, long, left and right.
I’ve never had a better plan, and often ended up well left, aiming at the centre of the chute (which isn’t the centre of the fairway…!). But last year I followed the optimal line and took it down the right hand side. At worst I ended up in the left rough, but at best centre fairway landing on a flat surface (and the slope is more severe on the left hand side). The strangest thing though was playing partners always shouted “Fore” when I hit this line… It says more about the thought processes of people on the tee - that right is dead, it really isn’t and there’s so much more room than there seems off the tee.
How have I improved GIR and score? Hitting this line off the tee. This is the data-centric solution, looking at applying optimisation in a realistic representation of the hole. I’m going to do this across the course, albeit with some differences - looking at manipulating the holes with realistic information, e.g. where lost balls or danger lies - if there’s an area where a ball will go OB I’ll mark it, even if it looks like fairway!
Refining the Model
At the moment this is a very basic model - but going forwards I want to improve it, ground conditions and slopes need to be part of this in order to make the planning more realistic, and outcomes more match reality! Hole 1 works well, as there’s not too many gotchas!