• Bocce Labs
  • Posts
  • Is Computer Vision for Bocce Ball even Practical?

Is Computer Vision for Bocce Ball even Practical?

Probably not for an average club but I won't let that stop me

Howdy šŸ‘‹

Last week, I introduced Bocce Ball Computer Vision Technology, my pipe dream for integrating technology into a sport that is many thousands of years old.

This week we're going to explore whether the idea is even practical (spoiler: it is not in most situations).

Welcome to the 8 new subscribers this week and a special thanks to šŸ”ŗSergio for spreading the word! (šŸ”Œ Sergio is doing a live show on 12/8/2022 in Chicago and he has a link in his bio if you want to grab a ticket like I did).

Is it practical to bring technology to a sport 1000s of years old?

I tend to be a pessimist. So let's start with an optimistic thought courtesy of bocce baller Andy Z. Back in 2020 Andy, Alex, Matt, and I were shooting the shit about bocce tech ideas. Andy said something along the lines of:

Imagine when someone back in the day pitched an idea for broadcasting the sport of golf on TV. Do you really think someone was initially open to putting 100s of TV cameras and cameramen on an 18 hole course as well as a blimp in the sky?

Andy (not a direct quote)

Golf has grown and grown since the major tournaments started broadcasting consistently on TV. There's even a whole TV channel dedicated to golf now. Some fans who watch golf all weekend get inspired and go out on the course and pretend they can drive and putt like Tiger Woods and Phil Michelson. That's not a bad thing.

Imagine if bocce ball had a TV fan base that drooled over how good the pros are. That could happen one day. But bocce would be competing against mainstream sports backed by big sponsors and heavy technology investment.

As an example, the behemoth of Disney pours money into the MLB's artificial intelligence, computer vision, and statistical technology. Read about why Disney bought MLB's technology unit for 1Billion dollars. Also check out how the NFL has teamed up with Amazon's cloud services to provide cutting edge technology for fans and viewers. It certainly doesn't stop there.

Okay - those are inspiring reasons why I should be optimistic about technology for bocce ball. Now, onto all the reasons why it is longshot.

Bocce technology is a longshot

First off, technology can be super distracting to the experience. For professional sports, these are things like jumbotron closeups, loud speakers, buzzers, shot clocks, a sideline full of reporters tapping away at their laptops, screens showing rotating sponsor logos, and more. Do social bocce sports leagues need to add technology distractions like cameras, digital scoreboard displays, automated stats, and a website/app? Probably not, but maybe for the pros.

Part of the "American Bocce experience" is having a ref make calls with a good old fashioned measuring tape. Take away the suspense as the ref bends down to measure the balls (uh oh! is a double measure?!) and the game is forever changed. In the tennis world, they still have line judges that yell when a ball is out ā€” if a player challenges the call, then computerized animation is displayed on the jumbotron. The rest of the time it is just collecting data quietly during the game for stats. A similar experience could be designed for bocce to minimize the tech distractions.

For me, going to a bocce league night is my time away from technology. Bocce and Petanque probably originated in Neanderthal days as they were throwing rocks. Maybe bocce without tech is just how things are supposed to be.

And finally, bocce clubs likely can't afford to invest in this tech. The project in my living room alone has cost more dollars than I'm willing to admit. But if I want something broadcast-ready, it would cost even more for the higher-resolution and higher-speed camera setup. That's a tough sell to a club that is focused on ensuring their people have a great time. The pessimism in me thinks that most players wouldn't care about half of the stats that become possible (please reply to this email if you disagree).

Well... What's next then?

So there you have it ā€” a number of reasons why my bocce software project is likely to remain a pipe dream hobby for the time being.

But what about the future?

Will this fizzle out? Will it become a reality?

If you can dream it, you can do it.

Walt Disney

As I continue to play bocce, I build more and more faith that this sport has potential to grow and turn my pipe dreams into reality. Now I just need to start showing up (even as a spectator) at some of the larger regional tournaments and meeting folks beyond the Chicago city limits.

Here are a few things that keep me going:

The potential for landing sponsorships (big ones) are possible at any time. This may sound like a hail mary, but it truly is how the sporting world works. As American Bocce Co, The Bocce Bros, and the United States Bocce Federation grow the sport in the U.S. (larger tournaments, more players, youth programming), we're sure to build some relationships with prominent names for advertising and sponsorship. Maybe one day bocce could be a college sport...

Camera costs have come way down. As an example, Apple is putting a 48MP camera in the iPhone 14 Pro for $1,200 without a cell plan. In contrast, my Ethernet machine vision cameras from Allied Vision are only 5.1MP and they cost $1,400 each (and that's without the separate $500 lens). Perhaps I could build this entire system with several iPhones, an iPad, and an iOS app? That would certainly make the system more user-friendly, and potentially cheaper.

I'm bound to find the right collaborators if I continue to write and share progress. I'll be getting more specific about my progress, challenges, and roadmap for this bocce software project in upcoming blog posts. In the meantime, if you see my short demo videos on šŸ“ø Instagram and šŸ¦ twitter, be sure to amplify them by liking and resharing. That will help me connect with folks. And if you found this newsletter interesting to read, click the fwd button in your email system and send it to someone.

And finally, this might be a longshot, but if one bocce club were to have some cool stating/training technology on a court, other clubs might become interested too. My friends at AccuTennis are growing slowly, but so far player and coach feedback has been extremely positive.

Or maybe we just need bocce at Dave 'n Busters with those goofy animations that 90s->modern bowling alleys have on their TV scoreboards?

A few updates from the week:

Dinner with bocce ballers Jamie and Chris

Bocce Team at Chicago Distilling Company

ā’ø Jessie

My new team went 1-1 for the night at Chicago Distilling Company.

A visit from my cousin, Molly, (who was in town all the way from Vermont) for dinner and drinks

Updates from the lab:

Dave H on Instagram: "Nerd alert! Calculating Euclidean distances between ball centroids (and drawing measurement lines) is easy. Calculating accurate distances to balls is tough due to shadows and light flickers. I have a SMA (simple moving average) activated on each bocce ball to eliminate jitter. Ultimately I need faster cameras and a faster computer to run this pipeline. These Ethernet cameras can do 24 FPS on my network, but this is running at about 1.3 FPS due to the computational pipeline. Of course Iā€™ve sped it up for illustration purposes so it doesnā€™t look as laggy. Iā€™ll be discussing camera trade offs on my blog on an upcoming Tuesday publication. Link in bio to subscribe if you are interested. @americanbocceco @oddballchicago @sticky.bandits.worldwide #bocce #bocceball #computervision #python #opencv #tensorflow"

39 Likes, 6 Comments - Dave H (@drhoffma) on Instagram: "Nerd alert! Calculating Euclidean distances between ball centroids (and drawing measurement lines) is easy. Calculating accurate distances to balls is tough due to shadows and light flickers. I have a SMA (simple moving average) activated on each bocce ball to eliminate jitter. Ultimately I need faster cameras and a faster computer to run this pipeline. These Ethernet cameras can do 24 FPS on my network, but this is running at about 1.3 FPS due to the computational pipeline. Of course Iā€™ve sped it up for illustration purposes so it doesnā€™t look as laggy. Iā€™ll be discussing camera trade offs on my blog on an upcoming Tuesday publication. Link in bio to subscribe if you are interested. @americanbocceco @oddballchicago @sticky.bandits.worldwide #bocce #bocceball #computervision #python #opencv #tensorflow"

Thanks for reading and I'll see you next week!

~ Digital

p.s. if you found any of this interesting, please fwd it to someone via email.

Reply

or to participate.