Title | AI BASEBALL PROJECT 'ZUNO' |
Brand | JAPAN BROADCASTING CORPORATION |
Product / Service | BASEBALL COMMENTARY AI |
Category | B11. Use of Technology |
Entrant | DENTSU INC. Tokyo, JAPAN |
Idea Creation | DENTSU INC. Tokyo, JAPAN |
Media Placement | DENTSU INC. Tokyo, JAPAN |
Production | DATA STADIUM Tokyo, JAPAN |
Production 2 | QOSMO Tokyo, JAPAN |
Production 3 | TWOTONE Tokyo, JAPAN |
Production 4 | AIRCORD Tokyo, JAPAN |
Production 5 | TAKI CORPORATION Tokyo, JAPAN |
Additional Company | JAPAN BROADCASTING CORPORATION Tokyo, JAPAN |
Additional Company 2 | DENTSU TEC Tokyo, JAPAN |
Additional Company 3 | BEARD INC. Tokyo, JAPAN |
Additional Company 4 | TYO MONSTER Tokyo, JAPAN |
Name | Company | Position |
---|---|---|
Naoki Tanaka | Dentsu Lab Tokyo / DENTSU INC. | Creative Director |
Kaoru Sugano | Dentsu Lab Tokyo / DENTSU INC. | Creative Technologist |
Moe Goto | Dentsu Lab Tokyo / DENTSU INC. | Copywriter / Planner |
Yuki Otsu | DENTSU INC. | Copywriter / Planner |
Kohei Ai | Dentsu Lab Tokyo / Dentsu Tec Inc. | Producer |
Yuji Yoshida | Dentsu Tec Inc. | Producer |
Kenichiro Takase | Data Stadium Inc. | Producer |
Hiroyuki Nagamatsu | Data Stadium Inc. | Coordinator |
Kei Kanazawa | Data Stadium Inc. | Executive Analyst |
Kyohei Okawa | Data Stadium Inc. | Analyst |
Sagiri Uehara | Data Stadium Inc. | Analyst |
Nao Tokui | Qosmo Inc. | Technical Director |
Koki Yamada | Freelance | Data Analyst / Back-end Engineer |
Shoya Dozono | Qosmo Inc. | Programmer |
Ryuta Modeki | TWOTONE Inc. | Art Director |
Yuri Matsumura | TWOTONE Inc. | Designer |
Ken Hirose | TWOTONE Inc. | Designer |
Miyu Hosoi | Qosmo Inc. | Project Manager |
Shintaro Murakami | DENTSU INC. | Creative Technologist |
Ryo Suetomi | Dentsu Lab Tokyo / DENTSU INC. | Planner |
Kentaro Mito | Freelance | Front-end Engineer |
Yusuke Koyanagi | DENTSU INC. | Art Director |
Toshiyuki Hashimoto | aircord Inc. | Technical Producer / Director |
Seiya Nakano | aircord Inc. | Device Engineer |
Yumi Fukuyoshi | TAKI corporation | Designer |
Yasuhide Kobayashi | Beard | Art Director |
Yota Mizozoe | TYO MONSTER | Producer |
We developed a powerful predictive AI that learned from over 3 million pitches in at bat data to compete with viewers. Baseball fans live to put out theories about pitching. They may say, "The next pitch will be a curve on the outside part of the plate to induce a double play," or maybe, "He's just gonna challenge him with a fastball." The AI makes its own predictions so fans can compete with it to see who's more accurate.
We used the results from our earlier challenge to open an online platform available to any baseball fan on the day of Game 2 of the Japan Series, the championship series to determine the best baseball team in Japan. We asked them to join a contest against our AI. There have been many experiments pitting humans against AI like AlphaGo, Chess, Quiz and so on, but this was significant because it was a platform where all pro baseball fans could go up against an AI.
This content naturally boosted coverage in social media and the press, but also created the first new way to enjoy live baseball in over 50 years. This gives people other than just fans of the teams facing off in a game to watch a baseball game on TV.
Zuno, the AI baseball project, is an attempt to find new ways of enjoying live professional baseball broadcasts that have become stale. It lets people watching a game get involved, thereby transforming a passive media viewing experience into interactive content.
We made a pitch-predicting AI and laid down a challenge to all professional baseball fans. We started with an experiement to get attention. We asked who would be more accurate: a former MLB pitcher, a passionate fan of a team, or AI? We enticed the public with this question. The AI won the challenge by a slim margin.