Plato said: "You can discover more about one person in an hour of play than in a year of conversation".
That was the first thing we remembered at BGA when we got contacted by the famous service RealClearPolitics for a study about the links between the way you behave during a board game, and the way you behave in real life.
The results of this study - to be published next week in Social Science Research Journal - are quite astonishing since they came to confirm what Plato said 2500 years ago! You can really guess many things about someone just by watching him/her playing a game: personality type, general attitude, social skills, but also wishes, secret thoughts, consumer profile and ... political vote!
This unusual study leads of course to extraordinary new possibilities for BGA. This is why we'd like to share with you some exchanges we had with Tom McIntire from RealClearPolitics about this.
**BGA: What is the main result of this research?
**
Tom McIntire: Being a boardgamer myself, I was pretty convinced that the link between the way you play and the way you behave in real life is important. I was expecting to discover a good correlation between certain game habits and certain psychological patterns. The main result of this study is to confirm that this link exists ... but what was very surprising is how strong the correlations are.
The most astonishing results concern politics. One of our main objective was to find a correlation between the way players behave in a game, and their political opinions. In most cases, we managed to get a success rate of 80% in predicting the vote of a given player on BGA. In certain circumstances, for example if a player played more than 100 games with at least 30 different opponents, our success rate is even over 90%!
**BGA: How is this even possible?
**
TM: We created an algorithm that detects some specific game situations on Board Game Arena and surveys the reaction of the players. The moves, but also the reaction time, and what they write in the chat.
Some game situations are really close to real life situations, and like in real life, you can have many different reactions: aggressive, passive-aggressive, diplomatic, and so on. We used an artificial intelligence based on a neural network that has been trained to recognize these different reactions and categorize them. This way, if you watch the same player playing different games during a few hours, you are able to build a pretty accurate portrait of his/her mind.
Afterwards, we contacted directly a sample of 500 users of BGA to confirm or infirm what we found out about them.
**BGA: Why did you choose BGA for this study?
**
TM: BGA is hosting +16.000 games each day, and all games take place on servers where we can install our algorithms and our AI. If we had tried to do this in real life, it would have taken hundred of years to reach the same results.
**BGA: What other things did you manage to know about the players?
**
TM: The results are incredibly accurate in many domains.
For example, we managed to get good results at predicting basic information like the age of the players, their gender, their level of education or their revenues.
But what is very interesting is that the best results we got was on the personality aspects. We got extraordinary accurate results for predicting for example the mood, the sociability, the rationality of a player. These informations are very difficult to gather otherwise, even with personality tests. What is exciting is that unlike with a personality test, a player on BGA doesn't even know that he/she is being analysed, which makes the results even better!
**BGA: what could be the possible applications for this?
**
TM: There are practically no limits to the possible applications.
Of course, our main interest at RCP is to predict votes for the next elections. We already created a joint-venture company with Board Game Arenea named "Meeple Mind". The goal is to analyse data from all games played on BGA, categorize players depending on their political opinions, and sell the possibility to political candidates to target people they need to reach. Typically, a candidate from a left wing party will want to send a different message to the right-wing people he/she needs to convince, and to the left-wing people he/she needs to reassure.
But we won't stop at that. For example we are capable to predict who in the whole BGA database would be your best husband/wife according to your personality (or to predict your probability of divorcing your current partner if he/she is also a player). We are also studying the possibility to detect the personalities who are the most likely to commit a crime and automatically send a police surveillance drone when the probability reaches a certain treshold. These two examples show how great it will be for mankind to exploit such a tool!
To conclude, Board Game Arena would like to thank RealClearPolitics for partnering with BGA for that study. As players, we want to promote the board games by any possible ways, even the unusual ones, and it was a really good surprise for us to see that board games and technology can make the world even better.
To be continued!
愚人節：我們可以藉由桌遊來猜測你會投給誰！
柏拉圖說: "想要深入了解一個人, 與其跟他談話一整年, 還不如跟他一起玩樂一小時"。
這是我們在BGA記得的第一件事, 當我們被著名的網站RealClearPolitics聯繫在一起研究你在棋盤遊戲中的行為方式和你在現實生活中的行為方式之間的關係。
這項研究的結果將於下週發表在社會科學研究雜誌上, 這是相當驚人的, 因為他們確認柏拉圖在2500年前所說的話!
你真的可以通過觀看他/她玩遊戲來猜測關於某人的很多事情: 性格類型、一般性的看法、社交技巧，以及願望、秘密、消費者個人資料和.....。
政治選票!
這不同以往的研究當然會給 BGA 帶來令人驚奇的新可能性。
這就是為什麼我們想要與您分享一些我們與 RealClearPolitics 的湯姆 McIntire 所進行的交流。
BGA: 這項研究的主要結果是什麼？
湯姆麥金太爾: 作為一個桌遊玩家，我很確信，你的遊戲思考方式和你在現實生活中的行為方式之間的聯繫很重要。
我期待能發現某些遊戲習慣和某些心理模式之間有很好的相關性。
這項研究的主要結果是確認該關係存在..。
但令人驚訝的是, 這之間的關聯是多麼的強。
最令人吃驚的結果是關於政治問題。
我們的主要目標之一是找出玩家在遊戲中的行為方式與他們的政治觀點之間的相關性。
在大多數情況下, 我們在預測某位玩家在BGA上的投票時, 成功率達到80％。
在某些情況下, 例如, 如果某位玩家與至少30個不同的對手玩了超過100場比賽, 我們的成功率甚至超過90％！
BGA: 這怎麼可能呢？
TM: 我們創建了一個演算法, 檢測在BGA上的一些特定的遊戲情況, 並調查玩家的反應。
這些動作, 還有反應時間, 以及他們在聊天中寫下的內容。
有些遊戲情境真的接近真實的生活情境，就像現實生活中一樣，你可以有很多不同的反應：積極進取，消極抵抗，外交等等。
我們使用了一種基於神經網路的人工智能, 它已經被訓練來識別這些不同的反應並對它們進行分類。
這樣, 如果你在幾個小時內看著同一個玩家玩不同的遊戲, 你就能夠從他/她的想法中, 精確地構建一個相似的寫照。
之後, 我們直接聯繫了500位 BGA 用戶的樣本, 以加強確認或減弱我們對他們的發現。
BGA: 你為什麼選擇 BGA 進行這項研究？
TM: BGA 每天舉辦超過16,000 場的比賽, 所有的遊戲都在伺服器上進行, 在那裡我們可以安裝我們的演算法和 AI。
如果我們試圖在現實生活中做到這一點, 就需要花費百年以上的時間才能達到同樣的結果。
BGA: 你對玩家的其他方面有什麼瞭解？
TM: 在許多領域，結果都非常準確。
例如, 我們成功地預測了諸如玩家年齡、性別、教育水準或收入等基本資訊。
但很有趣的是, 我們得到的最好的結果是關於個性方面。
我們得到了非常準確的結果, 來預測某位玩家的情緒, 社交性和合理性。
即使在性格測試中, 這些資訊也是很難收集到。
令人興奮的是, 與個性測試不同的是, BGA 上的玩家甚至不知道他/她正在被分析, 這使得結果更加出色!
BGA: 這有什麼可能的運用？
TM: 對可能的運用幾乎是沒有限制。
當然，我們 RCP 主要感興趣的是預測下屆選舉的選票。
我們已經與 Board Game Arenea 創建了一家名為 "Meeple Mind"的合資公司。
其目的是分析 BGA 上所有遊戲的資料, 根據他們的政治觀點對玩家進行分類, 並將這種可能性出售給政治候選人, 以達到他們需要的目標。
通常情況下, 來自左翼派別的候選人會希望向需要說服的右翼人士以及需要保證的左派人士發出不同的信息。
但我們不會停在這裡。
例如，我們可以根據你的個性預測整個 BGA 數據庫中的哪些人是你最好的丈夫/妻子 (如果你目前的伴侶也是玩家的話，則可以預測你離婚的機率)。
我們也正在研究是否有可能發現最有可能犯下罪行的人, 並在概率達到某一界線時, 自動發送監視警報給員警。
這兩個例子顯示了人類利用這樣一個工具會有多麼的偉大!
最後, Board Game Arena 想要感謝 RealClearPolitics 與 BGA 合作進行該項研究。
作為玩家, 我們希望以任何可能的方式推廣棋盤遊戲, 甚至是特別的棋盤遊戲, 而且看到棋盤遊戲和技術可以讓世界變得更美好, 這對我們來說是一個非常好的驚喜。
未完, 待續!
