Smarter volleyball? AI latest way to enhance training, recruiting

By Gerry Boehme for VolleyballMag.com

Terminator, meet volleyball.  

In 1984’s celebrated movie “The Terminator,” Arnold Schwarzenegger stars as a fearsome Cyborg killer, transported back in time from a post-apocalyptic future controlled by an evil artificial intelligence (AI) called Skynet. 

While we haven’t quite reached Skynet stage as of yet, it’s clear that AI has risen to become one of today’s hottest topics. Can intelligent software really “think?” Will smart machines take away our jobs, rob us of our creativity, and destroy life as we know it?

Moreover, how will AI affect volleyball? After all, the goal on offense is to “terminate.”

In truth, programs featuring machine learning and AI have helped shape volleyball for years. Beach coaches like Stein Metzger and Brooke Niles already have tons of experience using these systems to help analyze and coach their players, and they’re excited to see what comes next.

Young volleyball entrepreneurs like former NCAA star Zoe Fleck are also jumping on the AI bandwagon, envisioning even more ways for the technology to help improve coaching and player development, especially at the high school level.

So, where did it all start?

Giuseppe Vinci began coaching volleyball in Italy when he was 16 years old. As his career advanced in Italy and the U.S., he focused more and more on technology, analyzing video and building models to help improve player performance.

In 2013, Vinci started a business called VolleyMetrics. He described his goal as “revolutionizing the statistical study of volleyball through advanced data gathering and analysis.” As CEO, Vinci expanded VolleyMetrics to serve hundreds of colleges and professional clients.

In 2017, VolleyMetrics was acquired by Hudl, well known in sports circles as a leader in performance analytics technology. Vinci now serves as Hudl’s Market Director for Volleyball, concentrating on strategy, product, and partnerships. 

According to Vinci, much has changed over the past 10 years. In the old days, Vinci said, “We had computers look up patterns of what is good and what is bad in order to determine [whether] this is a good reception, this is a better reception, this is a kill, this is an error.” The program could then “learn those patterns” to give a consistent evaluation.

Stein Metzger joined Texas this past May as beach volleyball head coach after 11 seasons at UCLA. He’s used Hudl and VolleyMetrics for years. 

“Hudl’s been at the game for a while, so they have some things figured out,” Metzger said. They send volleyball game video from Hudl’s cameras “to an actual person, then they sit there and code it, tagging all of the actions and creating statistics, which is great.” Metzger also likes Hudl’s ability to share between different organizations, and that Hudl users can upload their own videos directly into VolleyMetrics.

This past year a new upstart, Balltime, entered the volleyball AI fray.

Tom Raz is an Israeli software engineer who’s always been fascinated by AI’s potential to “help solve real problems.” He’s also an avid beach volleyball player. 

“I always wanted to combine sports and technology,” Raz said, “and I used a lot of video to analyze my own game.” 

Raz would record his matches on his iPhone, but most of the footage wasn’t actual game action. He set out to build a simple program that could “recognize the playtime versus the downtime, and just trim out all the uninteresting parts.”

Raz built what he calls “a really ugly website” and circulated it to the beach volleyball community in Israel, as well as to the Israeli national team. 

“I got really good feedback,” he says. “They shared the kind of analysis that they were doing, which was a lot more sophisticated.” Raz understood that it would take much more work to develop all the features his users wanted. But Raz also felt that, “There was bigger potential here.”

Dan Banon started playing volleyball as an early teenager and has worked for multiple startups on the sales and marketing side. 

“A little over a year ago,” he said, “I started thinking about opportunities in sports tech,” particularly volleyball. 

Banon noticed more people recording their games and posting highlights on social media. 

“I just started thinking about it, asking people around me a few questions about how they consume their volleyball footage.” 

In spring 2022, Banon saw an online post in a beach volleyball Facebook group where Raz described his technology. 

“I reached out to him, and we started chatting. I think the interesting part of this is that we’re both volleyball players, incredibly passionate about the sport. We really started looking at it, not only from the passion perspective, but how to make it into a business opportunity, because we saw a giant market that we could bring this technology to.”

They decided to team up and start a company, which became Balltime. From May through August 2022, Banon and Raz conducted what they described as deep market research to help prepare them for Balltime’s launch.

“I think the important thing,” said Banon, “is that we didn’t draw any conclusions as to what the business should look like on our own, especially because we were outsiders looking in.” 

He and Raz also wanted to allow players to review their own games to improve. 

“The first product we ever launched was individual subscriptions for families on the recruiting journey, to help them create their highlight videos to get seen by college coaches,” Banon said. 

Banon described the enthusiastic response they received from the parents of youth athletes as their first “ah-ha moment.”

As they pursued their plan, Banon and Raz relied more and more on feedback from several key advisors, including Tony Ker (now Balltime’s head of sales and growth as well as a founding member) and Metzger.

College beach coaches Metzger, Niles on board

Already a Hudl veteran, beach Olympian Metzger, who coached UCLA to back-to-back NCAA titles,  now uses Balltime as well, specifically for practice and recruiting purposes. Stein works directly with Balltime, repping them on the beach side, and advising on software development.

Metzger likes Balltime’s ability to let users film their own videos, upload them, then have Balltime’s AI program analyze the data and provide stats with fast turnaround. Balltime also doesn’t require players to wear numbered jerseys, a big plus when it comes to getting stats for practices.

“Initially, [Balltime] was just an expansion of what we wanted to do, but couldn’t do, because of the scale of what we have to cover,” Metzger said. “There’s a lot of practice that can be captured, and we can learn from that.” 

But, Metzger says, using other programs for that purpose comes at a cost that many teams can’t afford.

“Indoor programs that have a big budget, they can hire someone to do data. When we talk about smaller indoor programs or beach programs, we don’t have the resources, or the manpower, to have someone sit there and code all the action.”

“Plus,” he continued, “when you think about beach compared to indoor, we’ve got multiple courts going at once. So, it’s just not feasible to have four or five coders sitting there coding all these matches. So, it really comes down to [personnel], and when AI can do the job that five people can do, and do it quicker, then it starts making a ton of sense.”

“Now we can expand it and do more practices, and cover more courts, and get more done, at a faster time. We can get the edited video back quicker.”

Eighth-year Florida State beach coach Brooke Niles, who has excelled as a player and coach at all levels of the sport, is always looking for the next competitive edge.

Late last summer, her husband, Nick Lucena, a two-time Olympian and now her assistant coach, was talking with his former playing partner, Phil Dalhausser. Dalhausser had spoken to Balltime’s Dan Banon and started playing around with Balltime’s software. He was impressed. 

A few months later, Banon reached out to Niles to arrange an informational call. Like many others, Niles’s Seminoles already used other programs, including Hudl. Niles herself had been using Hudl for “five or six years,” she said, “and they’ve been great. They do an amazing job breaking down our games and [providing] stats. We’re very comfortable with Hudl.”

But, Niles added, “We didn’t use Hudl as much for practices, because it took longer to upload. So, when I heard about Balltime, and listened to their pitch, I was thinking that we need something for practice” in addition to using Hudl for matches. So, she decided to give Balltime a try.

Niles saw immediate benefits.

“It cuts our practice video within an hour, so our players are able to see it almost immediately,” she said. Niles especially likes the fact that dead times are cut from the video “so that you won’t have these long pauses, and you can watch a two-hour practice in like 15 minutes or so.”

Niles finds Balltime especially helpful given the unique challenges posed by beach volleyball. “We have five courts while indoor has one, so the ease of this system has helped us, being able to go through multiple courts and making things that we were already doing a lot easier.” It also “gives me more ideas on what we can utilize to make us better, or [how to get] an advantage over somebody else.”

While coaches like Metzger and Niles are finding ways to take advantage of Hudl and Balltime to expand their effectiveness with college athletes, others have a different, yet complimentary vision.

Entrepreneur Zoe Fleck embraces the technology

Zoe Fleck is an award-winning volleyball player and entrepreneur who began her collegiate career as a walk-on at UC Santa Barbara. She later moved to UCLA before finishing her college career as a national champion and first-team All American at Texas. Fleck now plays professionally.

In 2021, she and a partner founded Zoe Fleck Volleyball Camps, which runs camps and clinics for girls of all levels. She recently expanded her services to include online mentorships to provide one-to-one guidance, coaching, and support to help aspiring female athletes pursue their dream of playing in college.

Fleck did not watch much volleyball when she was growing up. 

“My mom played, and I played a bunch of different sports as a kid, and I just fell in love with volleyball,” she said. However, “I wasn’t watching any volleyball on TV. Before I got to college, I watched no volleyball at all.” 

At college, Zoe said, “I suddenly had VolleyMetrics, because it was provided to me by my university. You get access to watching all of the college volleyball teams [and] a bunch of professional teams.” 

When she started watching film of herself, as well as more advanced players, something clicked. “That really changed how I saw the game.”

About four months ago, Balltime’s Banon reached out to Fleck after noticing the sizable following she attracts on Instagram. As they talked about how Fleck’s video study helped her rise from what she describes as “a nobody in the sport to a professional athlete,” Fleck got the feeling that their values were very much aligned.

Fleck already ran in-person volleyball camps, but she now finds herself spending eight months a year playing professionally in Europe. She was looking for a way to continue and even expand her personal coaching, and she thought Balltime might help.

A couple of weeks ago, Fleck partnered with Balltime to offer an online mentorship program that is scheduled to run for the next six months. Fleck plans to use Balltime’s AI software to teach young players how to watch volleyball and learn from it.

When she first started watching athletes playing at a higher level, Fleck says, “I was able to figure out solutions to problems that I was having that I wasn’t able to come up with on my own. I call it ‘standing on the shoulders of giants.’

“If I wasn’t watching professional liberos, I’d have to solve all the same problems for myself. But because I was able to watch players who have been playing volleyball for ten years, I could take the solutions that they came up with and implement them in my game ten years before they were able to.

“One of the cool things about Balltime is that you can take a YouTube link and just paste it in, and then [Balltime] will code the whole match for you. So, if there’s one specific player you want to watch serving, then you can just watch all of that person’s serves. You can watch them serve 20 balls rather than having to watch an entire match to see those 20 serves.” 

Fleck plans to use the program to teach players “what to look for, [the factors] that allow that serve to be at such a high level, so people can then take that technique and use it in their youth age groups.” By watching players who are playing at a higher level, Fleck believes, “They can learn more about how to upgrade their techniques, how to get better at the sport they play, instead of just going into practice every day, having no idea what to work on.”

She also is quick to mention that this idea isn’t new. 

“This is the same thing, technique-wise, that youth athletes are doing in other sports,” Fleck said. They learn by “just watching players at a higher level. All the high school football players are watching college and NFL games. We don’t have that in volleyball.”

Accordingly, she believes new and cheaper AI programs finally make that possible and also thinks these new resources will be especially valuable for female athletes. 

“Boys will get together, and watch sports together, and talk about what their favorite players are doing, and then practice it the next week. But among girls,” Fleck noted, “it’s just not really a cultural thing to watch sports. We don’t know how; we don’t know what we can possibly get out of it.” 

And, Fleck adds, “Even though volleyball is the number one female participation sport in the U.S., it just doesn’t get very much film coverage.” 

While professional and college volleyball matches can be watched on streams today, Fleck says, they’re available in “more obscure ways” than other more popular sports like basketball and soccer. 

“One of the things that I think is really interesting about AI, VolleyMetrics and Balltime, is that, if you ask any young basketball player if they knew who Michael Jordan was, they’d say, of course. [But], if you ask any youth volleyball players about (USA Olympic gold-medalist) Jordan Larson, most of them will have no idea who that is. There’s just not really the culture of female athletes watching female athletes.”

Fleck hopes to change that by using Balltime to mentor girls to “not only watch high-level volleyball but also see what’s possible with the game in a way that you just don’t get if you’re not watching. We’re trying to take players of the game and turn them into students of the game, as well as fans of the game.”

Metzger sees similar benefits.

“I think in the beach volleyball world, athletes, especially at the junior level, have not spent that much time analyzing their own video and their own play. And I think this is going to make it accessible to them to watch that. It’s not part of the culture yet, but this is giving the culture an opportunity to expand in this part of the sport.”

Can it help in recruiting?

In addition to coaching current players, Metzger sees huge value in using more accessible AI analysis to help recruit future stars at Texas.

“The growth of the sport at juniors’ is exploding, and [so is] the skill level,” Metzger claimed. “The difference between the top players and the mean or the middle of the group was pretty significant. It’s picking up now, so it’s hard to discern who the better players are, because they’re a lot closer in ability and skill sets.”

“The training, and the growth of the sport in general, is making everyone better. So, it’s getting harder to recruit in those senses. Being able to edit video so that we can watch a 40-minute [clip] in 10 minutes is going to actually make this whole thing manageable.”

Metzger went on to say that “Either the players upload their own video to YouTube, and I can grab the link and throw it into Balltime, or I can take the video myself when I go to tournaments. In fact, I can upload it right there at the courts. By the time I get back to the hotel, I have edited video.”

Niles also appreciates Balltime’s value in the recruitment process. She says it allows her to “take player videos from their YouTube channel and put them into Balltime, then Balltime can send back their stats in an hour or two. That’s really helpful for us on the recruiting front.”

While Niles and Metzger both believe that the influence of AI driven programs like Hudl and Balltime will continue to grow, they also think that improvements need to happen, especially in terms of reliability.

Niles says that the stats provided by Balltime “aren’t quite there yet, but they “are getting better and better. I wouldn’t say that I trust all the stats yet, so sometimes I’ll go through the video and just make sure that, ‘Hey, this is what I see.’”

Metzger concurred. 

“Accuracy is important. Statistics are important. I think that’s the future. How accurate can AI get over time? Eventually, I think they’re going to be extremely accurate.” But, he added, “If the statistics aren’t accurate, if there’s a margin of error that’s too big, [we] can’t really use it.”

But, while Metzger and Niles see the need for future development, neither buys into the doomsayers who warn that AI will replace coaches and players while ruining civilization as we know it.

“I certainly think about AI lately and the stories I’ve been reading,” Metzger says. “But everyone seems to be excited about this. I don’t see this form of AI replacing anybody.”

In fact, Metzger thinks it’s only helped. 

“If you look at our sport, the top teams are now carrying 20-25 athletes, and they only have three coaches, so the ratio is difficult, and there is only so much time in a day. This is going to be a real asset for sports that have those tougher numbers when it comes to player-coach ratio.”

AI instead of coaches? Not so fast … 

What about jobs like video operators? Might they be threatened by this new technology?

Brooke Niles doesn’t think so.

“Yeah, I listen to all those AI podcasts about how it could get out of hand. I don’t necessarily understand that aspect as much, but I think there’s always room for coaches and their experience. For instance, my husband and I have played professionally. He’s been to two Olympics. I think you can’t replace that experience factor in coaching athletes.”

When she or her husband connects a game situation to their own experiences, Niles says, “All of the girls’ ears perk up,  so I don’t know if you get that from an AI. So, hopefully, there’s still a job for us in the future.”

Instead of replacing coaches, Niles thinks that AI will “probably just make our job a lot easier.” She believes that people like video operators could even expand their roles, giving them time to “get more hands-on with players on the court. I see it as just being a tool to help with the student athlete experience.”

Tom Raz views it the same way. 

“I think it’s really just freeing up their time, removing the mundane task, to give [coaches] more time to do the qualitative work of analyzing the patterns,” he said. “Finding the key thing that they’re looking for in the data, rather than inputting it, Raz said. “It enables them to do more qualitative coaching.”

Raz adds that, before using Balltime, coaches said they “would use practice time to go over video with the players. Now, players can watch those clips, and the feedback of the coach, while they’re on the ride back home, and they can use their time more efficiently to actually practice at the gym.”

Zoe Fleck also doesn’t believe that AI will replace coaches. But it does raise the bar. “It forces coaches to coach at a higher level,” she says. “This kind of software has been available in football for years, and the coaching strategies of football get better every year, because of this kind of software in coordination with the human element.”

“So,” Fleck adds, “I don’t see AI taking over for coaches. I think that it will raise the standard of how coaches coach, and how well they coach. It will just raise the standard of volleyball in general.”

Giuseppe Vinci agrees that AI’s potential to raise coaching standards will make a big difference in helping athletes improve their skills. “We believe technology can do a lot to level the playing field with regard to making athletes stand out and making athletes know what they have to work on,” he says.

Moreover, Vinci thinks that many athletes “don’t get much coaching, maybe twice a week,” and that there’s often just one coach in the gym with 10 or 14 players. Vinci believes that “technology can help athletes, or help coaches, be prompted with things to watch, things to learn, things to notice.”

“That’s a way that AI can help,” said Vinci, “In giving every athlete a chance to get the shot they deserve in the sport they love.”

If there is any downside, Zoe Fleck believes, it’s that this new tool might lead young athletes to pay perhaps too much attention to statistics.

“The only thing I can really think of is, if players get too obsessed with the numbers rather than actually just learning from watching the film,” Fleck says. She believes that an excessive focus on statistics is “not always healthy for an athlete’s development.” But she continues, “I think the benefits clearly, and by miles, outweigh the potential downfalls of having this kind of information and video available to all athletes.”

Where does volleyball AI go from here?

While Hudl and Balltime have certainly garnered a great deal of attention among the volleyball community, other companies also serve the sport. 

Pixellot uses an array of cameras and AI-automated video production to follow the action on the court and produce a live or recorded video feed. AI algorithms analyze the video and identify the location of the ball, players, and other objects on the field. Pixellot also offers data and analytics capabilities.

Another venture, Veo, uses its own camera to record and analyze games and practices automatically.

As the current market leader, so to speak, Giuseppe Vinci knows that Hudl faces many challenges, and challengers. However, “if you look at Hudl as a whole, there aren’t really businesses that are doing the breadth of things that we’re doing, from video editing to recruiting, to tagging, to stats, all of these pieces.”

Regarding other AI providers, Vinci said, “We probably don’t really know who all of our competitors are.” But, he added, “We know what our objective is, and since we’re volleyball people, I would say we very much focus on our side of the net, and on our game.”

Vinci’s goals include “integrating our camera with sensors that you wear, and have them talk to each other.” Vinci also sees advanced artificial intelligence as supplying much more than automated tracking and tagging of a game. 

“There is a lot that AI [can do] to help support athletes in [things like] anticipating injuries and recognizing trends that might push the coach to say ‘oh, I have to act on this.”

Vinci believes that future AI will help coaches and athletes focus on areas where “I as a human can make the difference because I have certain skills that a machine is not going to achieve.”

As far as Balltime is concerned, Tom Raz and Dan Banon intend to remain locked-in on their mission. “In a start-up,” Raz explained, “You always try to find the place where people have the biggest pain and try to solve that.”

Banon described how Balltime’s development has always been driven by the consumer. “We’ve been fortunate enough to be able to talk to so many people that are involved with volleyball at different levels, different types of people, club directors, recruiting coordinators, college coaches, student athletes, parents, recruiting consultants, and it’s just been wonderful to build out features based on what people want.” 

Banon continued, “It’s just been, you know, a community effort essentially centered around volleyball.” While they anticipate expanding into other sports, Banon says, “Our first love is, and remains, volleyball, and we’ve been able to achieve all of this in talking to the larger community about what we should be building.”

Niles thinks that compiling stats for practices rather than games is “a lot more challenging, but that’s where it’s headed. It’s more about showing our players what you want to show them, especially based on what the focus was at practice that day.”

So far, she added, “It’s been a great learning tool for our players, and they love using it, so we’re excited to see where it goes.”

Metzger sees the future of AI in volleyball as delivering “what I didn’t think could be done, that I thought had to be done by the human eye, by coaches.”

For example, “If I serve to a certain location, what are the statistics in terms of their ability to put the ball away or not put the ball away?” Those are the types of data that Metzger sees “coming down the pipe that I think are really exciting. Eventually, it’s gonna make for better volleyball players.”

Maybe even Terminators.

The post Smarter volleyball? AI latest way to enhance training, recruiting appeared first on Volleyballmag.com.