PandaScore, where esports and AI meet

PandaScore
Pandascore
Published in
5 min readApr 3, 2019

Throughout this past short decade, the world of competitive gaming found itself drastically gaining momentum. Estimated to hit $1.1 billion in global revenues this year (first time breaking the three comma mark 💸💸💸), the very concept of professional gaming has grown beyond its niche and dispersed origins to develop into a highly profitable, professionalized and exciting new industry — esports. As esports viewership rises steadily and rapidly — The International 2018 finals peaked at 15 million viewers while global esports audience growth is projected to hit 645 million in 2022 — investments, partnerships and sponsorship deals continue to pour in. This also means that there are more opportunities, jobs and businesses. Esports is growing, every day.

Flavien pitching PandaScore at 50DAY 2019 conference

PandaScore is a part of this story. Fueled by our love for pro gaming, our company aims to expand and enhance the esports industry by providing fast, accurate, and reliable data📊 and odds🎲. We use our own computer vision software built on artificial intelligence and machine learning technology, PandaVision, to detect, extract and export real-time data from live esports matches. PandaVision currently covers the four big esports titles, League of Legends, CS:GO, Dota2 and Overwatch and we are working to add support for PUBG

Example of live and play by play data feed

PandaVision’s watchful eyes track hundreds of live data points from each and every esports match — 300+live data points for a League of Legends match alone. PandaScore simply provides the fundamental intelligence database and the tools to facilitate new ways of reading and using rich data.

PandaVision in action, tracking champion coordinates by “watching” the minimap.

For example, our data scientists developed a fast and efficient module in PandaVision that allows tracking and recording champion coordinates during a LoL match. Click here if you want to read an in-depth piece on how we achieved this.

Accessibility to such information opens the door to a multitude of possibilities, and it’s up to our clients to decide how they wish to best integrate our data into their businesses. Teams can use data like this to assess the effectiveness of their strategies; analysts can use the same data to examine game patterns and predict what’s going to happen next.

VPGAME.com uses PandaScore API to fuel and run its large esports coverage and betting website.

All you need to do is to get a hold of the PandaScore API. A calendar, a ranking system, a pre-match summary, a live update or a post-match report, the rest is up to you.

BUT WAIT, THERE’S MORE! A large chunk of PandaScore’s efforts is directed toward using our data collection to create odds for esports bookmakers, accessible through our trading dashboard and API. We designed deep-learning prediction models and trained them on extensive historical data from esports teams, matches, leagues and tournaments to produce intelligent odds that can keep up with the competitive unpredictability of esports.

PandaScore’s model for creating odds

In addition to offering pre-match odds, our technology allows us to calculate hundreds of live PandaVision data points from ongoing esports matches and deliver in-play odds. We are talking about real-time odds that update every few seconds, based on PandaVision’s record of what is happening. This gives our clients the ability to offer extensive catalogue of playable markets and odds(e.g. total kills per game, team fights, first kill, winners, etc), even while matches are taking place. There will always be something for somebody.

BEYOND ESPORTS. Most of what we’re talking about here (data, odds, bookmaters, etc) are things that esports carried over from traditional sports — where sports betting, as an industry, circulates far more customers, money and businesses than all of esports combined. While I boasted earlier that esports hit the one billion mark in revenues, the global market for sports betting is expected to reach 155.49 billion USD in five years. 🐜 We see a potential here. Before esports and PandaScore, machine-learning trained computer vision software that can deliver accurate, real-time data and odds to hundreds of customers at once, was never possible 💁‍♀. With us, it is.

The arena at LoL World Championships 2015 😍😍😍

Esports is home to us. However, we’re confident that our technology and everything we can create with it have the potential to reach 💫 , not just all of esports, but also, all of sports betting, data science, artificial intelligence research and potentially much more that we haven’t discovered yet 🌌.

Think about it… automated data. automated odds. There’s a lot on the PandaScore agenda.

Warming up for a code rush

From optimizing our existing algorithms to refining the development of our current services, our goal is to continue delivering our clients the most accurate and reliable data that can race against time. This means that we must train our artificial intelligence technology even more, reach zero errors as closely as we can and let PandaVision watch all of esports 👁️.

We want to be at the forefront on the day esports stadiums 🏟️ roar louder than traditional sports stadiums. We want to be the technological pioneers when lossless, instantaneous and infallible esports data become the norm.

Visit our site at pandascore.co

--

--