Heat Mapping

[Ryan P.]

Does anyone have the capabilities to heat map a hole based on how pros played it at a tourney? In my mind, it'd require the following:

1. Having someone sit on a hole and record where all shots land/scores
2. Ability to host that data
3. Ability to graph that data via data visualization software
4. Ability to graph it as a heat map (see below, Kirk Goldsberry is doing the same thing for the NBA)

UDisc could probably do this (Matt Kruger, you listening?), they'd just need to record approximate angle from the basket or location (person recording it would need to be in the right location if they used the GPS feature).

In my mind, the heat map would shot what areas of each hole lead to birdies after each shot, what area leads to pars, and what areas lead to bogeys. So, on a par 3, after 1 shot, what parts of the hole lead to birdies, what pars lead to pars, and what parts lead to bogeys? on a par 4, I'd like to see it after the first and second shots, etc.

It's obvious that it's good to be parked and bad to be 30' deep in the rough. But I'm wondering if there are other areas of holes that are better or worse to be in. Is it better to be 100' short in the fairway or 50' long in the rough? do certain holes favor the right side rough or the left side rough more?

 

[fish58320]

Dion Arlyn and I were actually riffing on this on Twitter this morning.

 

[Ryan P.]

Cool. A single tournament should be enough to get started. 500ish records of how a hole is played would start showing any areas where it's obviously better and hint at some places where it might be. If UDisc could record the location of each shot in addition to distance it'd gather a lot more data, but i don't think it'd be able to show any better nuances due to the lack of precision in GPSs.

NBA heat maps show clear patterns without having thousands upon thousands of records. Attached is Carmelo Anthony's heat map from 2012-13, when he took less than 1,500 shots. It's pretty clear he's better from the one side of the basket and terrible up close (compared to league average).

 

[Ryan P.]

Has anyone seen this for ball golf holes? Here's a link that has heat mappings for green slopes, but I don't care about slopes, I care about scores: https://ncga.org/2014/04/04/heat-maps-of-every-green-at-poppy-hills/

 

[elmexdela]

nerdy and i love it

would love to see how guardian trees/upshot placement impact the scores

 

[Cgkdisc]

Steve West, Don Ticknor and I mapped the drives and distances on the BRP driving range hole at the MN Majestic maybe 7-8 years ago mapping it to different divisions and ratings ranges. I think Steve posted the grid but might take some Perubing to find it.

 

[ru4por]

I don't use technology, certainly not on a disc golf course....but, I believe UDisc has the data. Each card at a DGPT event has a UDisc scorekeeper. I am under the impression that entering disc location (right fairway, circle 2....) can be entered.

 

[R-Ogre]

I don't use technology, certainly not on a disc golf course....but, I believe UDisc has the data. Each card at a DGPT event has a UDisc scorekeeper. I am under the impression that entering disc location (right fairway, circle 2....) can be entered.

I think these guys are looking for significantly higher-def data than just FW, C2, C1 etc.

 

[dehaas]

Definitely a cool idea, but I think has a couple of limitations. First was already mentioned, collecting enough data points to actually have some useful information. Park jobs will result in birdies obviously, but after that you lose context real quick. Is the player who landed in c2 a strong player who just barely missed their line but still canned the putt or was it an ok tee shot by a mediocre player who hit a lucky putt? Putts from further out aren't as high of a percentage so it's gonna take that many more data points to actually find the "hot spots" on a hole. Putting style comes into play too...do you push or spin? Prefer uphill to downhill? Was it windy that day?

Neat idea but would need thousands of people playing a hole to be useful. And then there's also the issue of gps margin of error...at best 15 feet in either direction.

 

[Ryan P.]

nerdy and i love it

would love to see how guardian trees/upshot placement impact the scores

exactly

 

[Ryan P.]

Definitely a cool idea, but I think has a couple of limitations. First was already mentioned, collecting enough data points to actually have some useful information. Park jobs will result in birdies obviously, but after that you lose context real quick. Is the player who landed in c2 a strong player who just barely missed their line but still canned the putt or was it an ok tee shot by a mediocre player who hit a lucky putt? Putts from further out aren't as high of a percentage so it's gonna take that many more data points to actually find the "hot spots" on a hole. Putting style comes into play too...do you push or spin? Prefer uphill to downhill? Was it windy that day?

Neat idea but would need thousands of people playing a hole to be useful. And then there's also the issue of gps margin of error...at best 15 feet in either direction.

Thanks! I'm not seeing how any of those are limitations of the idea. Yes, more data points are better. the individual situations you mentioned don't mess with the data enough, especially as there are more and more data points. And yes, grabbing even more data such as shot style, weather, etc. would be even better to allow for filters.

What I'm saying is that, even though more detail would make it even better, it'd be great to start with something. Even 100 people playing the hole should give some general indications. For instance, there may be some holes that it's better to aim 15' left of the hole than right on it, because missing left ensures a better putt than missing right. Is that obvious just from looking at the hole? In some cases, it definitely is. But just like how heat maps have redefined how basketball is played, I think it could really help DGers rethink how to play some holes.

I'm especially thinking of par 4s and 5s. When I watch videos I hear commentators say "After your drive, you want to be in this gap because it's better". Yes, it seems to be better. But do the numbers actually bear that out? I'd love to see a heat map indicating this, and it wouldn't take more than a tourney to have some good data to discuss.

Maybe the best way to say what I'm thinking is this: I totally agree that more data = better. But some data > none.

 

[dehaas]

Thanks! I'm not seeing how any of those are limitations of the idea. Yes, more data points are better. the individual situations you mentioned don't mess with the data enough, especially as there are more and more data points. And yes, grabbing even more data such as shot style, weather, etc. would be even better to allow for filters.

What I'm saying is that, even though more detail would make it even better, it'd be great to start with something. Even 100 people playing the hole should give some general indications. For instance, there may be some holes that it's better to aim 15' left of the hole than right on it, because missing left ensures a better putt than missing right. Is that obvious just from looking at the hole? In some cases, it definitely is. But just like how heat maps have redefined how basketball is played, I think it could really help DGers rethink how to play some holes.

I'm especially thinking of par 4s and 5s. When I watch videos I hear commentators say "After your drive, you want to be in this gap because it's better". Yes, it seems to be better. But do the numbers actually bear that out? I'd love to see a heat map indicating this, and it wouldn't take more than a tourney to have some good data to discuss.

Maybe the best way to say what I'm thinking is this: I totally agree that more data = better. But some data > none.

My point is the style of your individual game varies greatly from everybody else. The intended line off the tee in relation to where the disc actually landed and putting ability is lost in translation. Yes an infinite amount of data will help smooth things out, but solely looking for a hot spot and trying to land there isn't a solid game plan if you're an out of towner showing up for a tournament. Better than nothing if the data is available yes, but are you willing to put your own fate in the hands of others? Know your own game and play it.

The data portion is definitely interesting, but I still stand by too many variables out of context to dictate how I play my own game. The NBA analogy doesn't translate, yes players have preferred spots on the court they prefer to shoot from...top of key, wing, baseline, etc but that's a completely different animal than bending certain shot shapes initially and then being able to follow through with executing the hole.

Not saying it's a bad idea, just that there's a lot of untold story behind the scenes. And realistically if adequate data was to be collected, more than likely it would be performed during a pro event. That info simply doesn't translate to the mere mortal skill set.