Dedpool

General diagram
AI usage diagram
Browser application
Mobile application
During a stream!

Inspiration and Introduction

According to CNN, eSports describes the world of competitive, organized video gaming. Gamers are watched and followed by millions of people all over the world. Fans attend live events or tune in on TV or online. Even streaming services like Twitch, FB Gaming, and others allow viewers to watch as their favorite gamers play in real-time. On those sites is where popular gamers build up their fandoms.

The ESports market is growing;

But even from 2019 to 2023 it is expected to continue growing. According to Business Insider, it is expected to grow at a 9% CARG up from 454 million in 2019 to 646 million in 2023. The esports audience will grow on pace to nearly double over a six-year period, as the 2017 audience stood at 335 million.

Most projections put the esports ecosystem on track to surpass $1 billion in revenue for the first time this year. And revenue is expected to grow from here — Newzoo projects it to hit $1.8 billion by 2022. Money flows into esports through media rights, live event ticket sales, merchandise sales, and in-game purchases, but most of the revenue (69%) comes from sponsorships and advertising, per Newzoo figures cited by Statista.

Betting Market.

Sports betting is a form of gambling in a specific sporting event. Alongside with the odds are the prizes you can possibly win and the amount you need to gamble. As a result, a lot of people are encouraged to watch many sporting events today for the sole purpose of sports betting.

The global sports betting industry reached a market size of 203 billion U.S. dollars in 2020. Within this industry, there were approximately 197 thousand employees in a total of almost 31 thousand businesses. Many countries participate in legal sports betting, one of the more recent being the United States in 2019. Sports betting revenue in the U.S. was predicted to grow to as much as eight billion U.S. dollars by 2025.

Also, 45% of sports betting now takes place online; online gambling is available any time, providing more convenience and privacy.

But, there's one aspect of it that has grown substantially which is eSports betting. The eSports betting market wagers grew exponentially worldwide between 2015 and 2020. According to the source, the amounts wagered on eSports betting grew from 315 million U.S. dollars to 23.5 billion in the presented period according to Statista (6).

Among the biggest sites that also manage eSports betting we find the following:

Steaming platforms

ESports’ bets are becoming increasingly complex, there are sites with multipliers that change according to the statistics and the teams. But not even one site allows you to bet for specific circumstances of the game.

Even Twitch has introduced predictions. It is a new feature on the site that allows streamers to ask the possible outcome of certain scenarios. Viewers bet their channel points and all these aspects bring a new experience for the viewers. Streamers have a new engagement tool with their audience. The main problem with this feature is that it only allows ONE bet during a game. Also the streamer must create the poll every time.

What if we can bring that to Theta?

Improve it considerably and bring a new dimension of interactivity to the streaming ecosystem?

And instead of channel points, using Theta fuel that is collected by watching those same streams.

Solution

We decided to try and do just that, we present you Dedpool.

Dedpool uses IA that analyzes gameplay in real time while it is happening. Image Quality is required to analyze the images. It is a tool to automatizate those bets. So the streamer does not have to waste time on setting the predictions. Audience will experience different scenarios with every streamer. Lose or win are only one of many betting scenarios. Even between matches, every bet will be different.

Another advantage of Dedpool on theta is that you are betting Theta Fuel, which is collected through watching streams, viewers just need to keep watching the streamer to earn more. So the excitement of betting keeps on and adds a whole new dimension to the watcher's interactivity and adds a lot to the streamer's entertainment capabilities.

General Diagram

The Dedpool platform receives the data directly from the game through MQTT and by means of simple logic it manipulates the bets in such a way that they are dynamic throughout the game.

By employing two servers, one dedicated to run the Theta Ledger and another one runs the computer vision algorithm to do real-time analysis of the matches, we achieve this PoC.

You can find much more documentation of how we achieve this, and the logic we employ in this part of our Github: https://github.com/altaga/Dedpool-powered-by-Theta#backend

AI usage diagram

The events, score and game time are obtained by analyzing the characters on the screen, using a character recognition library such as Google's Tesseract OCR. For this we employ machine learning models based on mnist through the tensorflow framework, specially trained to recognize more specific characters such as score and starting time.

Here is a video of all the system running at once:

https://youtu.be/fRAGp-0K0b0

And here is the video Testing Deadpool while streaming!

Click on the image please to watch the video

https://www.theta.tv/video/vid9iig80weetgf30t6

Closing

In this first instance we are just improving Betting on League of Legends. League of Legends is the most watched eSport game with almost 149.5 M hours.

We want Dedpool to bring the best betting experience in League of Legends tournaments, but first we want to implement this technology to bring to theta community bigger streamers so the community grows with Dedpool.