DIY Stat Analysis

[BLW15]

I've been doing some at-home analysis of my stats for the past 16 months and wanted to share in the hopes that you might be able to provide some ideas or insight into what else could be done with this information.

After every round I input date, score, round index, number of: birdies, pars, bogeys, doubles, triples, putts, GIRs and fairway percentage into a google drive sheet. I then update scatter plots that show score vs.: birdies, pars, bogeys, double, triples, GIRs, putts and FIR and a separate set that shows round index vs each of those stats. My initial goal was to see how each of these stats correlated with my score/index and to see if I could use this information to improve my game.

Here's a summary chart of my findings:

I'm not a stats wiz per se, but did take a class in college and know my way around excel/google drive. For the uninitiated, the R-squared of a regression tells you how much the two variables correlate. The highest possible R-squared value, 1.0, would indicate that two variables correlate perfectly and the lowest possible value, 0.0, would indicate the two variables do not correlate at all. For the sake of simplifying the analysis I used a second degree polynomial regression and the following correlation ranges: 1.0-0.7 Strong, 0.7-0.4 Medium, 0.4-0.2 Small, 0.2-0.0 None.

Some insights from the data:

1. First of all, I was surprised that none of the stat categories had a "strong" correlation to my score/index. This makes me wonder if there is something else that does correlate well or if the score/index is just influenced by too many different variables for one to really shine through.

2. I rarely make birdies (4 total over 30 rounds of golf) so it makes sense that they would have no correlation with my scores.

3. Because birdies are so rare, pars are essentially the lowest score I make on a hole. It therefore makes sense to me that these have the second highest correlation to my score/index. The highest correlation belongs to triples, likely because they represent the highest possible score on a give hole. Its not rocket science to know more of the lowest score and less of the highest score lowers your overall score/index but its good to see the data back it up.

4. There is hardly any correlation between score/index and bogeys or doubles. This is likely because it matters less how many of each I make as opposed to what they "would" have been. By that I mean on any given day I could score well with a lot of bogeys if they otherwise "would" have been doubles or I could score poorly with a lot of bogeys if they otherwise "would" have been pars. The same goes for doubles that "would" have been bogeys or triples.

5. All else being equal having less putts should mean a lower score but number of putts doesn't tell the story of how you got to the green. I am much more likely to three-putt after a GIR because I am typically starting farther from the hole. Likewise I am more likely to one-putt when I've chipped it close which usually means I missed the green so an extra stroke had to be taken. For this reason it makes sense to me that putts have a small correlation on score/index. However, I can't explain why the putts would better correlate to index (borderline medium correlation) compared to score. See below for more on this.

6. I would have expected GIR to correlate most with score/index and am surprised at how little it does. My only explanation here is that as with birdies, I don't have many GIRs overall so the impact they can have is limited. Similar to putts there is also a disparity between the correlation between score (borderline medium) and index (none) that I can't explain. I am also confused by why the disparity between the two stats is reversed from putts to GIR. The only semblance of explanation I can think of is that I was lucky enough to St. Andrews this past year and their oversized greens produced outlier numbers by increasing both my number of putts and GIRs. However, this was only two rounds out of 30 (both produced slightly better than average scores and indices), is the data that sensitive?

7. There is absolutely no correlation between FIR and score/index, this is the clearest answer of all the stats. This is backed up by recent Shot Scope data that farther is better than straighter (assuming both are in bounds).

 

I think I'd be most curious to see if other people find similar correlation (or lack thereof) between their stats and score/index, what might change based on handicap, and what other analysis could be done with this or similar data?

[TR1PTIK]

I've been tracking stats in Excel off and on for years. The thing about tracking your own stats is that you need to figure out what will be the most beneficial to you and that can take a little experimentation. Here's a look at what I have currently:

Typically, the middle image is hidden as that simply holds a variety of formulas used for my charts and scorecard. The "Quota Game" seen on my scorecard is something I pulled from No Laying Up's Strapped. In order to make that work, I also created a separate sheet linked to this one to calculate my handicap using the pre-2020 formula. I'm sure I'll eventually fork over the money for an official USGA handicap, but I'll also probably keep this going as well.

I was messaging back and forth to the Asst. Pro at my course back home and he shared some metrics with me that I will begin playing around with and could be beneficial for you as well. He keeps track of: Tee shots w/ Driver, 140-210yd shots, 130yds & in, Chips/Pitches, and Recovery shots. He also tracks a subset of what you have, # of putts, fairways, greens, penalties, scrambling, and 3-putts. There's one very specific stat that is clearly important to his game which is simply called: "Times I short sided myself". 

I plan on breaking up the distances a bit more than he does, but that's probably the biggest change I'll make, and all of these are fairly easy to track on a scorecard or notepad.

EDIT: BTW, if you're on Instagram, you should check out the Asst. Pro's account @tgriffgolf. He's been posting a lot of images and video lately with COVID keeping everyone inside.

[BLW15]

@TR1PTIK thanks for the input. What do you find most beneficial from the stats you keep?

I've seen stat tracking sheets with different distances broken out and can see the benefit. It seems like tracking outcomes from each shot like that would give you more information about why you made a particular score on each given hole as opposed to the whole round in general. That is, more of a micro view compared to the macro view my stats currently provide. This would be more similar to a strokes gained type calculation because you could see how you perform in different facets of the game (driving, approach, greenside, recovery, putting, etc.) or from particular yardages or with particular clubs.  Overall that would probably be more beneficial to improving your game because you could more readily identify strengths and weaknesses.

I get the impression this is what tech like Game Golf and Arccos are trying to provide but I have not tried either of them. I wear a Garmin S60 Approach GPS watch while playing and it keeps score and some stats (putts, drive accuracy, penalties) as well as club tracking if I input which club I used after each shot. The best part about the club tracking is that it gives me my true average distance for each club. The downside is that it doesn't provide much in the way of analysis of the effectiveness of each club.

[TR1PTIK]

@BLW15, I'm happy to help. I've become a bit of a spreadsheet nerd since originally developing the first version of my stat tracking sheets nearly 6 years ago. It's how I taught myself the various functions of Excel and it's an aspect of golf that I really enjoy talking about.

There are three stats (IMO) which are the most beneficial to shooting better scores: GIR, scrambling, and 3-putts - or 3-putt avoidance if you'd prefer to look at it that way. There's plenty of arguments to be made otherwise, and it will vary depending on the individual, but if you play regularly and can shoot in the low 90's - upper 80's that's it. You have to hit greens, you have to get up & down when you miss, and you need to avoid 3-putts.

Breaking down your approach shots is definitely beneficial to identify weaknesses in your approach game and how you break it down is completely up to you. You can do the same thing with the short game and putting via proximity to the hole, but that can be a touch more difficult to manage on the course unless you're confident in your ability to eyeball distances.

I went ahead and updated my spreadsheet last night then populated with an old round (I chose a good one so I wouldn't have to input a ton of chips or penalties) to test.

As you can see, I broke up the yardages into 25-yard increments (24 technically to prevent overlapping of data) moving from distances of 200+ yards down to 125 and in. I broke it up this way based on how far I hit my clubs and my perceived ability to find target from those yardages. From 125 yards, I should have a wedge in my hand every time. In fact, 125 is about the crossover point between my pitching wedge and my gap wedge. So even on a windy day, I'll know when I look at the stat I was hitting wedge without having to mark down the club. Next up is 125-149 which covers PW, 9i, & 8i. Again, there's a crossover point here to make sure I'm only capturing the clubs I want to. From there it goes, 8i, 7i, 6i then, 6i, 5i, 4i, and finally 4i, hybrid, fairway wood.

Similarly, I went ahead and broke up my driving stats based on club using the driver, fairway, and hybrid. It's pretty rare that I hit hybrid off the tee on a par 4 or 5, but there have been occasions where it's the play due to trouble further up. 

On the last note you made about Arccos and Game Golf, that is essentially the information they provide - though less customizable, but also considerably more detailed. I've used Game Golf in the past and have Arccos currently. I love all the detail, but it's starting to seem almost too granular. Also, until they release the new Link device, I just hate messing with my phone on the course.

I'm playing Sunday which will be my first round to experiment with capturing this new data for my spreadsheet and I think I'll try and do it 100% Arccos free. I purchased a golf log book after Christmas that came pre-formatted to capture most of the data I need (sadly, I've barely used it). All I had to do was add 3 rows by hand to capture tee club, recovery shots, and unforced errors. I'll let you know how it goes and share my post-round stats. It will certainly make for an interesting study (at least for me) as I'm finally starting to turn the corner again with my game after several setbacks and I'm playing a course that's completely new to me. Additionally, the course has removed all the flags from the course during COVID which means I will only be firing at the center of a green unless I have a good indication of where the hole is.

[cnosil]

Reading through yours post you don’t seem to understand the reason for tracking individual distances. These distances are ihow strokes gained numbers are generated. It also helps you assess how well you hit each club. If you divide strokes gained into tee shots, approach shots, short game, and putting you can start to see the good and bad parts of your overall game. Individual rounds are interesting, but the real power comes when you measure over periods of time.

The problem with metrics is that they are hard to find a correlation like you are looking for. looking at FIR or GIR if you just miss a fairway or green it is the same a missing a fairway and being in the woods or missing a green by 30 yards. The other thing is that playing well in one aspect may counter a poor performance in another.

Just looking at score is tough because you might make a 40 foot putt for par after hitting 3 bad strokes or hit 2 great shots and lip out a 3 footer and make par. Two completely different issues with those two scenarios.

I’d look at things like missed 3 footers, first putts that end up outside of 3 feet and first putt distance, greens hit from various distances, shits that are short of the pin on approach shots, shots that require a chip out or resulted in a penalty stroke.

If you do a search on the forum for EdTeergolf and read some of his posts, he provides some good information on playing and assessing the game. He does have solid credentials and knows what he is talking about...when I take lessons, he is my goto instructor; additionally he coaches a women’s college golf team.

[Bucky CC]

Watch out for penalty strokes as well. Nothing ruins a good round faster than hitting two shots in the water or OB.


Sent from my iPhone using MyGolfSpy

[TR1PTIK]

1 hour ago, cnosil said:

Reading through yours post you don’t seem to understand the reason for tracking individual distances. These distances are ihow strokes gained numbers are generated. It also helps you assess how well you hit each club. If you divide strokes gained into tee shots, approach shots, short game, and putting you can start to see the good and bad parts of your overall game. Individual rounds are interesting, but the real power comes when you measure over periods of time.

@BLW15: This is a good point that I hadn't addressed in my previous posts since I was mostly sharing how I track in Excel. An individual round won't tell you much on it's own. You need a database of rounds to compare to and a set of cumulative data to extrapolate from in order to make the most of your analysis.

[Camille Bissarro]

I've tried to reinvent the wheel a few times on this (even having my own note books printed at a local shop). When I've created my own spreadsheets, only the obvious stuff seems to pop out, ditto when I've used various apps (GolfShot, 18Birdies, some other one...). Plus, on that latter point like @TR1PTIK I really don't like having my phone out while I play. Basically, like @cnosil says, when things go pear shaped, they tend to do it in a multitude of compounding ways.

Back to reinventing the wheel, it turns out strokes gained is a pretty good set of metrics for helping sort out the recursive shambles that are many of our golf games (each one unique, though, like so many snowflakes). Recently I've been using this method via the Shot by Shot website. As it is a paid service ($24 a year for the LITE version, which is a fine place to start), I know this is not a DiY solution. Intellectual property rights not withstanding, there is enough info on the site (or on the web more generally) to try to reverse engineer you own system, albeit without the benefit of a few hundred thousand other rounds at various handicap indicies against which to compare your results.

[KristianC]

Hello...new here and I LOVE looking at numbers from my golf game to make it better so this is right up my alley! 

Haven't fully read through this entire thread but from what I've read so far very interesting!  I had a basic spreadsheet but never thought of making it as detailed as your guys....great stuff! 

[Sixzero]

Hi I think that stats like fairways hit, GIR's and number of putts don't tell you alot more than the score card, and don't tell the true story.

For a start how wide are the fairways you play, how big are the greens. When I use to go on holiday to Florida to play golf (when my handicap was 18 20 years ago), I could hit 100% of fairways and about 80% GIR's, where as where I play in the UK with narrow fairways, and greens a quarter the size of the ones in Florida the numbers were no where near those %s, I got great at getting out of trouble though!! LOL

The other thing is the score that you record on the card, dose not reflect how well you may have played a hole. Here is an example that I always remeber each time I play this hole. 308 yard dogleg right par 4 off are yellow tees. I scored par, hit fairway. GIR 2 putt. Still does not tell the true story. The truth is I hit a bad drive, got lucky ball ended up on the left edge of the fairway just short of the ruff and 150 back from center of the green and I can seee the flag, I hit the best 8 Iron I have ever hit, it finished 12 inches from the hole for birdie, and I pushed the putt right and made par. So in truth I only played 1 good shot on that hole.

So what I record is every shot I take, what club and did I hit it thin, fat, off the toe etc and did it finish were I was hopeing for, for driver a quick spray of foot powder, and I can see why the shot ended up where it did, with a geat big ball mark low on the toe!