Gaming and esports – why you should care

One of the biggest beneficiaries of the COVID-19 pandemic was without doubt the gaming and esports sector. But with many countries returning to some sort of "normality", the question is: Will the sector lose prominence in the public eye? The signs point to "highly unlikely".

The sector has been through rebirths and surges since the 1970s but, with it being worth an estimated EUR 23.3 billion in 2021 in Europe alone, it won't be going away or needing a rebirth any time soon. Gaming and esports does not need to be limited to the "traditional" players (e.g. publishers, or developers) - there are a growing number of opportunities in this space for "non-traditional" players.

Gaming and esports are digital by nature and therefore the possibilities arising from data use, data collection and interpretation of that data (including the use of AI) are growing. In addition, various learning points also arise which can be used by other industries.

And this is why you should care.

We explore the opportunities below and with a particular focus on the opportunities and use cases resulting from data and AI within gaming and esports.

Sector overview

Gaming and esports is a booming business and widely reported to be largest entertainment sector in the world, larger than the traditional powerhouses in this space being box office and music. Gaming includes console gaming, video game streaming, mobile/tablet gaming and PC gaming. The global video game market was estimated to be worth USD 178.37 billion in 2021. To put that into context,  in 2021 the estimated worth of the global:

• box office market was USD 21.3 billion;

• music market was USD 26 billion;

• sports market was USD 354.96 billion;

• video streaming market was USD 419.03 billion;

• automotive manufacturing market was USD 2.7 trillion; and

• toys market was USD 156.5 billion.

Other interesting comparisons include Spider-Man: No Way Home, 2021's largest box office release, grossing approximately USD 1.892 billion globally and the highest grossing box office release in history, Avatar, grossing, USD 2.847 billion globally. Grand Theft Auto V grossed USD 1 billion in just three days and is now reported to be well clear of the box office figures.

Adding to this is the esports market, which we consider to be a subset of the broader gaming market. The global esports market is estimated to be worth in excess of USD 1 billion for 2021 and is expected to generate nearly USD 1.38 billion in revenues globally by the end of 2022. Viewership of esports events is expected to increase in 2022 as well which is impressive when you consider that League of Legends, one of the most popular esports titles, world final had 73,860,742 peak concurrent viewers in 2021. Again, for context, the Euro 2020 final between England and Italy was watched by 328 million people.

Yes, the statistics and what they cover versus what they don't cover can be questioned but the important point is that the numbers illuminate that there are plenty of opportunities for participants and investors to be encouraged by. We haven't even covered the extent and expectations surrounding the metaverse here, which the authors of this article strongly believe has its roots in gaming.

Opportunities

If reading the previous section has whetted your appetite for getting involved with gaming and esports, we list a few ways below:

• Sponsorships, strategic partnerships and advertising

• Considering the use of AI and data in gaming

Sponsorships, strategic partnerships and advertising

Gaming and esports presents a great opportunity for businesses to grow their brands.

Esports is the obvious area of focus here - it's a global spectator activity which now has professional teams, leagues and competitions, with growing pots of prize money making regular sports a natural comparator for esports.

For example, the Asian Games 2022 this year will debut esports with medals being awarded for eight games. The Commonwealth Games this year will also pilot an esports event called the "Commonwealth esports championships" which will be held in conjunction with the Commonwealth Games (28 July - 8 August). It is not an official Commonwealth Games event, but it will have its own medals and branding.

Within the regular sports sphere, sponsorships and "partnerships" are a large source of revenue for sports teams. In the esports sphere, the position is no different - it's estimated that up to 60% of global esports earnings in 2022 will come from sponsorships. The viewership and reach of esports is the attraction to sponsors as demonstrated by the League of Legends viewership statistic above. Notable gaming/esports partnerships between companies not directly related to gaming and gaming/esports participants include:

• Fashion retailer ASOS being appointed esports organisation Fnatic's official retail fashion partner;

• Rocket League developer Psyonix's partnership with Lamborghini;

• FaZe Clan's partnership with comic book publisher DC Comics (to create a limited-edition comic book, featuring a number of FaZe Clan content creators in the form of superheroes and related merchandise); and

• esports organisation TSM's partnership with cryptocurrency exchange FTX (a 10-year naming rights where TSM rebranded to "TSM FTX").

An interesting trend from 2021 seems to be that crypto companies are flocking to involve themselves with esports.

Sponsorships, partnerships and advertising aren't just limited to esports - the viewership of gamers playing games through platforms such as Twitch and YouTube is on the rise. Elden Ring was released on 25 February 2022 and, at the time of writing, has reportedly amassed more than 50 million hours of watch time. Sponsorship of these gamers and/or gaming channels, including the possibility of advertising on these channels.

Advertising and targeting esports fans

Talk of in game advertising has been around for a while now in the industry and is particularly prevalent in mobile/tablet and web based gaming but the PC and console games have been slower in their take up. But, there are examples and it's been reported that Sony is considering putting ads within its free-to-play games. Therefore, the possibilities here are growing. Arguably, using esports and gaming as a means to grow your brand offers the ability to target different and a more varied type of individual than regular sports allows

For example, the profile for esports fans seems to generally be male, young (around 26 years old), educated, working professional with above average incomes and an interest in technology. Looking closer, esports fans are not all homogenous and differ across different games. Not only is there an opportunity to advertise products or services, but companies could look to focus the product / service to the particular audience group.

For example, Fortnite attracts a younger audience (the bulk of the players being 18-24 year olds) with a majority of players coming from the US and then Brazil, Russia, UK and Germany. Compare this to Counter-Strike Global Offensive (a more mature game) where a majority of the players are 19-34 year olds and most of the players come from Russia, US followed by Poland.

Direct interaction with fans at each point of the ecosystem and the scalability of games

Two other features of esports make it an attractive investment proposal: the ability to directly interact with fans through various different channels and the scalability of games.

As mentioned, there are numerous examples of sponsorship partnerships between esports organisations and various companies. However, the esports ecosystem is not made up of only professional esports teams. There are the individual professional players, tournament operators, streamers (usually these are not professional esports players but are influencers that have a large following on online platforms) and the online platforms on which esports and casual play are streamed on. The interesting thing is that esports fans can directly interact with the game (as media entertainment) at each part of this ecosystem. This means that each of these channels could be used as a sponsorship asset e.g. to gain exposure to esports fans sponsoring an esports team is not the only way. Popular streamers or a partnership with an individual player could also be possible.

Esports is also an exciting product. Compared to normal sports like basketball, esports is all played on a digital platform. This means that the dynamics of a game can be changed and varied quickly and easily. It is also scalable so that a "battle royale" style of up to 100 players is possible (the limit on Fortnite's battle royale mode). This means that balance patches and the dynamics of the games can be changed to bring about a fresh and fast-paced nature to esports. New tactics and strategies need to be developed by players as new elements are introduced, making for entertaining viewing and the chance to make the product new and different each time. 

Use of AI in eSports and Gaming

Aside from the investment potential in esports, gaming has a natural affinity with data and artificial intelligence (AI) which presents its own opportunities and learning points.

AI is being used in various different ways in gaming. The natural affinity comes from the ability to capture different data very easily given games are played on digital platforms. AI products are helping games and professional teams to enhance their performance. Players can identify strengths and weaknesses by feeding AI game data. In order to successfully generate and make use of AI components, it is necessary to access large amounts of data for the training of an AI system. Such training data forms the basis on which a computer recognizes patterns and regularities by means of self-learning algorithms. The computer learns by example to find independent solutions to unknown problems without having to be programmed to do so in advance.

A quick summary: AI - Machine Learning, Deep Learning and Black Box

Machine Learning is one of the most important methods for creating AI. Deep learning is a sub- category of machine learning that uses artificial neural networks. These consist of a large number of nodes (so-called neurons), which are arranged in layers. Similar to a filter, they work their way layer by layer from the coarse to the fine. In the process, the self-learning system establishes or removes the connections between the nodes or adjusts their weights. This method increases the probability of a correct result of the AI system created by it.

The algorithmic logic of the decision making of an AI system created by Deep Learning is not comprehensible in detail. The decision-making processes of such a system are not explainable even for its developers. This circumstance is called a "black box".

By entering examples and comparing the respective results (output), it is possible to at least roughly understand which criteria were decisive for a decision (black box tinkering). However, it is not (yet) possible to trace the algorithmic decision in detail. Everything that happens in the black box thus remains a secret for AI developers, AI users and their customers.

Example use cases

An interesting use of AI in gaming is how algorithms are used to integrate advertising in recorded or live stream gameplay. 4D Sight is an AI platform that can provide native ads in real time on esports streams or recordings. Advertising in esports is just like with traditional TV or YouTube where in between the breaks (or during mid-stream) advertisements would pop up.

4D Sight provides an alternative where instead of unwieldy advertising interrupting the esports stream, advertisements are embedded so that they appear (for example) on empty walls or spaces in the gameplay. Counter Strike is a first person shooter which has many walls and corners for players to use during gameplay. All of these empty walls and corners could be embedded with advertising.

4D Sight also looks to track key moments in the gameplay so that ads are integrated at these important moments to give maximum value to the targeted advertising.

AI that was developed using games have translated over to real world application. DeepMind is an AI company that made AlphaGo, a computer program that was trained to master the board game Go (an ancient board game that was first played in China and still played particularly in China, Japan and Korea).

Go is a complex game where previous computer programs could only master up to a level of an amateur Go player. The board is 19 by 19 with approximately 250 possible moves each turn. With a typical game depth of 150 moves, this means that there are approximately 250 potential moves. AlphaGo was trained using deep learning to first understand the game using amateur games and then it was made to play against different versions of itself thousands of times, each time learning from its mistakes. Over time it became better until ultimately it beat Lee Sedol, the winner of 18 world Go titles.

Once AlphaGo was able to mimic human intuition with the game of Go, DeepMind set out to apply this to predicting the shape of proteins in the body. Figuring out the exact structure of a protein is expensive and time consuming, but if there was a way to map this using AI, then this could open the door for breakthroughs for understanding how medicines and diseases work.

AlphaFold was the programme which DeepMind created for this and ultimately it won the 14th Critical Assessment of Protein Structure Prediction (a competition whereby researchers can test their methods of predicting protein structures). It is now being used by hundreds of thousands of researchers where a protein structure database has now been created to share the structure predictions made by AlphaFold of more than 350,000 structures.

Risks and Issues

As with any traditional sports, esports has its own issues with players cheating. Using Fortnite as an example, at any one time there could be thousands of games being played simultaneously by hundreds of thousands of players. It would be impossible to police each and every one of these games for any programs which gamers may use to cheat or gain an unfair advantage over others.  

Cheating has also become more prevalent with esports becoming a billion-dollar industry (as we've discussed above). Even if you are not a professional esports player, streamers can make a living out of playing video games. Cheats could be used to make more entertaining videos and clips which in turn attracts more fans to your channel.

Anybrain is an interesting product that uses a patented algorithm to detect cheating behaviour by comparing it to data it has already collected on how a gamer usually interacts with a game. For example, touch data as well as keyboard and mouse dynamics are tracked so that the machine learning algorithm is capable of learning the gameplay of a gamer and raising the flag when it detects abnormal behaviours.

From a data standpoint, in order to successfully generate and make use of AI components, it is necessary to access large amounts of data for the training of an AI system. Such training data forms the basis on which a computer recognizes patterns and regularities by means of self-learning algorithms. The computer learns by example to find independent solutions to unknown problems without having to been programmed to do so in advance.

Where data is accessed, the consideration around personal data will never be far away, as is the case with AI use within gaming. AI training often requires personal data as an underlying basis which fall under the UK and EU General Data Protection Regulation (GDPR). As such, the black box phenomenon is not only contradictory to the principle of transparency laid down in the GDPR, but also challenging when it comes to the identification of the appropriate legal basis for the training and application of AI.

To the extent that AI components aim at analysing a player's performance, this will probably require the player's consent pursuant to Art. 6 (1) (a) GDPR, but sufficiently granular consent is practically difficult to obtain. When the AI training commences its final purpose is often not yet entirely clear so any consent wording cannot sufficiently specify such purpose. In addition, full transparency of algorithms is technically not possible so that the player can only be provided with limited information. Any withdrawal of consent by  players / data subjects would raise the question to which extent the player's personal data subject to self-learning algorithms may be continued to be used.

To the extent that AI is used to prevent fraud or match fixing, related processing activities involving the players' personal data could be based on the legitimate interest of the related esports or game organizer making use of related AI components (cf. Art. 6 (1) (f) GDPR). However, this requires the processing to be "necessary" for the purposes of preventing fraud or match fixing.

Even though data protection authorities usually take a restrictive GDPR approach, only a broad interpretation of the GDPR taking into account the black box phenomenon can practically balance both AI specifics and certain GDPR requirements.

A possible solution to these issues would be to use anonymised data where ever possible, e.g. by means of Generative Adversarial Networks. A Generative Adversarial Network is based on two AI models, the generator and the discriminator. Whereas the generator creates fake content, the discriminator identifies the fake scenarios. Thus, it is possible to create artificial data for the training of AI and thus to avoid the processing of personal data under the GDPR.

Summing up

To sum up, gaming and esports is a large fruit that is ripe for further growth. As further digitalisation of society and commerce looms, these opportunities for growth will no doubt continue to expand. With AI only getting smarter and esports prize values increasing year on year, the risks will increase too.  It's also worth mentioning at this stage that the EU is in the process of introducing regulation around AI which providers and users of AI services should take note of.

But, if you are in need of a brand boost or of ways to target your offerings to varying audiences, you could do worse than exploring the opportunities and learnings from gaming and esports.