How will Artificial Intelligence Change the World of Sports?

In short, AI will change the world of sports drastically and it will change it for the better.

First, I have to explain what Artificial Intelligence is. Artificial Intelligence is defined as software that is capable of making its own conclusions on the way of reaching a goal. In that process, AI software has to have the ability to learn, self-optimize and, ideally, return data output. In order to write an article about artificial intelligence on this blog, we have to look at the subject from two corners: the marketing corner and the sport performance corner.

Marketing Corner

In marketing, just like in any other field, there is always a need to improve and to outshine the competition. Of course, every marketer wants to succeed using different roads. On another hand, everybody, even slightly familiar with how the IT operates and advances, is well aware that every human task that can be replaced with technology will eventually be replaced with technology.

Enter Artificial Intelligence. Many brands already use AI as a tool to reach their marketing goals.

In marketing, Artificial intelligence can autonomously:

• study customer data and generate customer specific context
• operate as chatbot that carry on any type of communication with the customers
• write basic articles on various topics
• track content performance
• track athlete performance, improve health and prevent fatigue or injury

It is old news that AI is replacing simple tasks, but now the growth and presence of AI is bigger than it ever was. As you can see, AI already eases everyday activities for marketers, but the best is yet to come. In the world of sports, 2016 was the year of data analysis and 2017 will be the year of AI.

Pandora is already using AI. It “studies” your music taste and delivers to you the music it „thinks“ you might like.

Another AI success story was created Down Under by Mars Chocolate Australia and Clemenger BBDO Melbourne. They used AI to analyze social media „anger rates“. When they discovered anger spikes in a certain area, Snickers bar prices would drop in that area. Well, chocolate does have the feel good factor (although for a short time).

A while ago I read that Amazon decided to send items to their customers before they even ordered them. But at that time, this information was slightly overblown. They wouldn’t check your Amazon history and, based on that data, decide you will surely order a certain item. What they would do is, for example, send lots of Patriots merchandise to Boston area when the team won the Super Bowl. To expect that Patriots merchandise would be in high demand in Boston area after the Super Bowl triumph was not such an incredible prediction.

But now, with AI, this all leads to a market where marketers will know what customers want before they know it themselves. Amazon’s Alexa, Echo and Echo Dot also already use AI (mostly successfully).

“I think the biggest opportunities for advertisers are those that allow for conversational experiences like Alexa skills and messenger bots, because they allow you to give the brand a voice that isn’t possible through more traditional ad formats”  – Laurel Boyd, MullenLowe Mediahub’ SVP Director of Media Content and Innovation

I say “mostly sucessfully” because earlier this year, a little girl in Texas managed to place a $170 order through Amazon Echo for a dollhouse and cookies. Now this is a security mistake, but a fixable security mistake. The problem grew bigger when TV stations reported on this story and said “little girl told Alexa – “order me a doll house”“ on-air, causing other Echos to interprete this as a command and to execute it.

This was a bigger mistake, since it made the news worldwide. It only shows that not even the biggest brands are immune to hiccups, and that AI has great power and is helpful, but that it still needs improvement.

One of the most known AI solutions is IBM’s “Watson”. At the moment, The North Face and Under Armour are testing Watson to figure out the way to implement AI in attempt to incorporate and personalize real time data.

“So Watson might know that product A works better in poor weather than another, or than product B is very popular with customers who are very active on Facebook. Watson is able to process large volumes of data and is thus simply capable of creating new correlations and generating new ideas that even an experienced marketer wouldn’t think of that quickly.” – Marilies Rumpold-Preining, IBM Watson Customer Engagement Executive

The biggest question marketers have for the AI specialist is „Can it be creative?” The answer is simple: „It can!“ For example, a recently launched service Logojoy creates logos using AI. Its clients testify that relationship with Logojoy “feels like you’re working with a real designer“. Logojoy is now a machine that „earns“ $70,000/month in revenue. But this is just the beginning. With time AI will learn and specialize itself to fulfill various needs different brands will have. It will be able to create, run and monitor marketing campaigns on its own.

Creative AI operates better when it has access to more data. Therefore the challenge is to steer AI to collect significant amount of correct data, and use that same data to create. In 2017 this technology will only grow and it will be directed to personalization.

“There’s a great likelihood that by 2022 artificial intelligence will be able to create brand identities and experiences. It’s definitely top of mind for anyone in the marketing industry, mainstream consumer adoption is making it a more urgent consideration for all marketers”  – James Trump, creative director with agency Moving Brands San Francisco

In a world where AI will become creative, today’s creatives have nothing to be affraid of. They will be mentors to the computers and their creativity will be on a higher level. Their role will change, of course, since they will have to be creative about their use of creative machines.

As you can see, the world of sports joined the advanced analytics party and mastered it in less than a year. Does anybody really wants to be late to the AI party?

Sport Performance Corner

As I mentioned earlier, Under Armour is using AI to improve their relationship with the customer. Furthermore, Under Armour is a fast growing sports gear manufacturer, and as such they started to rely on AI to improve their products. Their new Health Box is an AI powered personalized fitness system. And this is just a start since Under Armour plans to turn their sports gear into gadgets using AI.

Researchers from Disney Research, California Institute of Technology and STATS have introduced an AI approach to sport preparation that can help teams and athletes better prepare for specific opponents by using both advanced analytics to study the opponents and AI to calculate the predictions of opponents’ most likely reactions to certain on-field actions.

“Given that this is the location and speed and acceleration or whatever of the current players, can we predict where all these players will go?” – Yisong Yue, assistant professor at CalTech

To make it reality, this approch takes individual behavioral data and from it builds individual models using AI’s learning and AI’s prediction abilities. Athlete’s predicted moves are then visually presented as a „ghost“, which explains the title of their research paper: “Data Driven Ghosting using Deep Imitation Learning”. Ghost is a transparent image that shows predicted movement of a certain athlete. Everybody who used to play car games and tried to improve lap records knows what this is all about.

You simply know this will make its way to game preparation routines, if not even games themselves. Especially with improvement of VR. There will no longer be a need to film the opponents plays in VR. AI will gather data on certain teams and players, use it to predict their reactions and, with the help of VR, visually present it to the players.

“At a high level, AI is simply about designing computer programs that can behave in increasingly complex ways. In this case, we want the trained AI agent to mimic the decision making of professional soccer players as represented by trajectories. Coaches do this all the time when they draw up plays on a chalkboard, and ask their players to execute certain trajectories in different situations.” – Yisong Yue, assistant professor at CalTech

STATS provided the data for this research. A lot of it. About 10 frames per second per player. Meaning that a single play during the game could alone provide 2,200 data inputs for AI to study and use.

Future Corner

Well, creativity is not the last thing AI managed to master. A software called DeepStack managed to beat 10 out of 11 world poker experts using AI’s new human-like ability – intuition. Each of the poker players played 3,000 hands against DeepStack and the only player who managed not to lose was the Australian poker star Martin Sturc.

I specifically say „not to lose“, since Sturc didn’t beat DeepStack. He was just the only player who lost to DeepStack by such a small margin that it was statistically marginal and it was not possible to decide the overall winner.

So, AI is growing into a creative intuitive smart machine that can improve sports from both the marketing corner and the on-field corner.

Everything in sports and technology is improving so fast that there is no scenario in which ignoring it would end well.

Article sources:

ispo.com

campaignlive.co.uk

shutterstock.com

sporttechie.com

thedrum.com

theoutsidegame.com

cnn.com

yahoo.com