Problem
In August 2021 I began to observe some interesting trends among my fellow soccer players. The group I played soccer with always had a tendency to stand and talk about what had just happened in the soccer game after we had played. They kind of tried to recreate the moments in their heads and it was clear that they thought it was fun and interesting to talk about the experience they had just had in the match. I started to direct my focus towards this target group of non-professional sportsmen, where I met a soccer team that, among other things, entered their own statistics from each game in an excel sheet. I got connected to other sportsmen who played racquet sports who were excited about the idea of being able to record their game, where they could easily find the best moments in the video and get statistics on how many times they hit the ball. My conclusion was that there was a large group of potentially non-professional sportsmen, who had an interest in using data and technology in connection with their sports which could enhance the fun aspects of using data combined with their sport.
The entire foundation for how the project was approached was based on Design Thinking.
I began to look into research that deals with the use of data in combination with football and consulted with professionals who work with machine learning on a daily basis. What I learned is that we currently see a development within deep learning and video recognition, which shows a clear technological movement that can open up completely new user experiences that we have not been able to work with before. At the same time, the smartphones that have been developed within the last 1-2 years have become so powerful and with cameras of such high quality that combined, it can be used as hardware and thus provides a low entry barrier to the product for the customer. Instead of developing your own camera, a smartphone can simply be used. With the development of 5G it becomes possible to work with live data more efficiently due to the increase in speed.
Since it is incredibly important that the user feels they are provided with high quality video of themselves and the team to look at, I experimented with finding the right angles to record from. I therefore chose to invite friends out of 2-3 rounds and play football, where I placed a smartphone on a tripod in each corner and experimented with angles and height. I found that 2 smartphones in each corner can record a total 7-man court (68 x 52.5 m) if they are at a height of 2.40 meters and point 15 degrees down towards the soccer field which fits well with the majority of grownup non-professional soccer players, as they play on 7-man pitches. Throughout the research, people liked the idea of the entertainment part of the platform, with the parts they showed interest in, was possibility to find the best moments of their just played match and to get basic statistics about the match they just played.
The app solution was therefore designed based on being able to:
1. By using a smartphone, let non-professional soccer players easily record their matches and have the video automatically segmented so that the user gets the best sequences of the match based on their own choice easily and quickly.
2. To be able to collect statistics on the non-professional soccer player in combination with video of themselves, intertwined with gamification elements.
3. To create a product where potential non-professional soccer players are motivated to become part of the sport through new technology and a new community.
The project provided fantastic insights into the work with non-professional soccer players, especially into how professional and non-professional soccer athletics differ in terms of needs from data combined with their sports. With the non-professionals, it is much more of how the data can enhance the social and fun aspects of their sport. If you look back on the last 100 years ago, sports were played without any kind of technological intervention. With time, the technology developed and was able to help develop players physically, better preparations for matches and making scouting more efficient. This made sense to a professional soccer athlete, as a very large part of their time was centered around their sport and making their everyday proffessional-life more efficient. However, from the non-professional players, most commonly, the time used on the sport does not exceed 10 hours weekly, meaning the sport is much more about the experience and the social elements with their sport. This means that the product must direct some other aspects, as they do not necessarily have any problems that needs to be solved. From a business perspective, the group of potential customers is very high compared to the professional target group. But the target group is also more different to build a product for, where a much higher focus should be to enhance their experience with their sports instead of trying to solve problems, they do not necessarily have.
Throughout the project, which included testing a prototype in figma, I gathered some key insights, which can be found below.
1. Gamification was incorporated to increase retention and to create greater motivation for using the product. However, it became clear that gamification elements such as levels do not work optimally the older the person gets. Gamification elements like levels are more suitable for children and if gamification is to be incorporated to create retention, then it will be important to try other forms of gamification. An example could be to make partner agreements with businesses, where you can obtain points based on your play, that can give a discount to buy better and new equipment.
2. A product where users can get video and statistics of themselves when they play has a very high potential. But technologically, it is still very difficult to work with event detection and player statistics, in a sport where each team has between 7 and 11 men on the pitch. The technology is still very novel, but using a smartphone as hardware to record video and using deep learning for video recognition for event detection and collecting statistics has a very large potential in sports for non-professional soccer players.
3. A product of this type must be considered a digital product as much as a physical product. In order for the players to use the product, it requires that they have the desire to bring a tripod and set them up in the two corners. It takes time to get the product ready to give them the output they need. This means you have to work with people's habits, and it can often be difficult to change people's habits or add a new habit to their existing habits.
4. By having a functionality where you record a wide video format, where the players themselves can use the zoom functionality to zoom in on themselves as they watch the video, means that the players themselves can follow themselves and determine the perspective of the video. It is about having 2 angles where each player can see the video from their own self-determined perspective at any giving time.
5. For non-professional soccer players, it's about creating a platform based on fun and social values. These are two basic values, which are the main reason non-professional soccer players meet up to play soccer, where they get the opportunity to clear their heads of thoughts and just play soccer. If you are to use data and technology in connection with their sport, then it must be based on the two values and how the two values can become stronger in their experience with the sport.
In the end, the technological challenges were too high and the technology too novel, which made me decide to shelve the project. But working with the product was an incredibly exciting journey, which gave a high learning and understanding of sports products between non-professional athletes and professional athletes, and the potential that lies in bringing digital sport products into non-professional football. With time, the technology will be further developed and products of this kinds can be incorportated into the life
Below you can find screens of the product which has been used in my testing phase.
.
.
.
Videos
The videos are recordings of non-professional football matches, with experiments in AI detecting whether the players are running or not. The actions were generated using a pre-trained model.