For the first order of 'Unreal Summit 2017,' Keynote speeches were given by Tim Sweeney of Epic Games and Katsuhiro Harada, a leader and lead producer of Bandai Namco Games' Tekken project.
▲ Katsuhiro Harada came on the stage as a second keynote speaker.
Producer Harada, who went on the stage as a second speaker, had a presentation about 'Philosophy in the development of Tekken Project game'. As the most important factors when progressing Tekken project, he mentioned that the viewpoint of hypothesis and statistical data has to be considered significant, and each opinion which does not appear on data has to be accepted. What's more, he stated that brilliant creativity and luck are also needed.
Especially, in the case of Tekken, since the aspect of character game is strong, the values and evaluation standards are different according to the wide community culture of the country and region, and the diversity of preference is big. Here, the difference by age group and gender is added, so there are many difficulties. However, as the series continues, many case studies on the world market are conducted.
Harada said, "You can think about people, games, popularity, character goods, popularity lifetime, etc. about the 'definition of a good character,' which is asked while creating a character, but such definition is actually hard to apply." And, "We aim to be loved by millions of products for more than a decade in the world market, but first we need to narrow down the goal and choose our customer who likes the character and buy a product. But as customers are not 100% aware of their tastes, it is also necessary to try to read hidden needs."
On top of that, when creating characters, one needs to set the hypothesis that such character would make a hit with the target. Without the hypothesis, it is hard to determine whether it hit or not through verification, so he emphasized that one has to set it regardless of the outcome even for the accumulation of know-how. Citing the cases of Steve Fox in Tekken and Hikari Miyamoto in Summer Lesson, he introduced that, "These characters are intended for completely different users, so they contain completely different settings and stories, and the way to make a character concept changes as well."
▲ We had time to introduce philosophy in the game development of Tekken project.
▲ As each character has different preferences, the more case studies the more advantageous.
▲ When containing completely different settings and stories for each character, the way to make a character concept changes as well.
Meanwhile, they need to get a feedback after the content is produced. If external disclosure is not allowed during development, they use feedbacks from overseas offices or qualitative investigation on the sample. After the release, they can use a variety of methods such as usage rate, popularity vote, user postcard and questionnaire, Internet sample survey, active research by research company, group interview, etc. As an important point in securing feedback data, having to secure as much feedback data as possible is mentioned.
When hypotheses, statistical data, and feedback come to hand, it is necessary to combine this data to validate the results, so the process of quantifying all the data and documenting as much as possible which part of the developer’s idea goes against regardless whether positive or negative. This does not mean that good and bad of popularity lead to good and bad of the result. It only carries out a process of calmly verifying the 'difference from the hypothesis'. In the case of producer Harada, he hypothesizes about popularity order as well as usage order and then proceeds development. It means to start development after deciding popularity order of characters.
In that case, he would get a question of "Should not all the characters be popular?". However, he stated that it is only idealism, and in reality, everyone can not make the first place. There could be some cases that are deviated from hypothesis and expectations. For instance, in the case of Lili and Dragunov, Dragunov was planned to be created as a less popular character in order to double the popularity of Lili by aiming for the effects of 'Yin and Tang'. But, it became increasingly popular than expected and affected the next work.
In addition, in the case of Ganryu, its popularity and usage rate are low, but it has been used for a long time as it has worth in its own way due to having a solid position in the segment. However, the popularity and usage rate are not necessarily proportional. If the character has outstanding performance, its usage rate will be high, but related products are unaffected. Therefore, he said that the hypothesis for it has to be set.
Besides, Harada emphasized that in the case of simply giving character design contract without establishing concept to a designer who has a great skill, the concept needed for the game could be lost, or the target could be wrong. In order to minimize this, he suggested to hypothesize about a character concept and target needed for the game, and share it with statistical data after documenting, so that a designer could successfully work.
▲ It is necessary to collect various feedback data as much as possible.
▲ There are also cases that are deviated from hypothesis and expectations.
▲ Quantifying and documenting the result as much as possible through verifying with statistical data.
▲ It is suggested to hypothesize about character concept and target, and document them.