IBM's artificial intelligence "Watson" tells us who should be drafted at NBA

ByKeith Allison

Artificial intelligence (AI) developed by IBM "WatsonHas been active in various fields in recent years,Publish a strange recipeOr,Use for future cancer treatmentOr answer with a humorous answer to the question attack "Why or why?" Of the childCogniToysIt is being used for. Watson of such AI is how American professional basketball leagueNBAIt is said that it is being used to select a player who will nominate a draft.

The Toronto Raptors Are Using IBM's Watson to Draft A Winning Team | Motherboard
http://motherboard.vice.com/read/toronto-raptors-nba-draft-day-ibm-watson


In the 2015-16 season, Toronto Raptors won the Division championship for the third consecutive year and entered the playoffs. In the first round, I rejected Indiana Pacers with 4 wins and 3 losses, and I will proceed to the conference semifinal from 2001. In addition, at the semifinal, Miami Heat played a hot battle with Miami Heat 4 wins and 3 losses, and entered the first conference final. However, in Conference Final, I lose 2 wins and 4 losses to Cleveland Cavaliers. Such Toronto Raptors is actively reinforcing the team for the new season, and it will be done before the opening season which will become the centerpiece of its reinforcementNBA draftThen, I am about to start a new approach. Although the draft nomination to win promising young athletes to strengthen the team, the athlete acquired by Toronto Raptors here is neither a director nor a competent scout man, but a player elected by Watson, IBM's AI It may become.

This new initiative started in February 2016, when Maple Leaf Sports & Entertainment, the parent company of Raptors announced "Raptors will be the team that will incorporate the player analysis using the first Watson at NBA" It became clear.

"Watson does not answer the question of who is best to trade with whom, we are comparing players in a different dimension from those that are," said the sports analyst Do researchIBM Sports Insights CentralIt is Mr. John Lentina. For example, if Raptors is conducting a survey of college basketball, Watson can quickly calculate and display statistical data on the player's shoot, assist, rebound, etc. Moreover, it can compare with past college life data of nominated draft player.

Until now, Raptors' scoutmen have judged which player to acquire based on data and statistics specialist affixed on the whiteboard, but with the advent of Watson far more than before It is said that it will be able to make that decision efficiently.


However, before Watson was introduced Raptors was one of the best tech in NBA. Raptors' analysis team, for example, boasts the best track record among NBAs, and is a tracking system for modeling the best movement of playersSportVUIt has also been known that we have developed various analysis tools such as. Attention is drawn to the draft nomination which can combine this superior analysis team and IBM's technology.

In addition, the average annual salary of NBA player is 5 million dollars (about 500 million yen), and when becoming the top player, the average annual salary is extremely high, 25 million dollars (about 2.5 billion yen). As annual salary is expensive, it is also a big bet whether to add it to the team, and it is normal to think that "I want you to fit the team somehow", but at the moment Watson recommends " It is said that it can not be measured accurately up to? However, it is possible to select unique players different from conventional selection methods.

Normally, NBA attempts to acquire athletes who have strengths in specific fields and players who are likely to fit the team. As to how to reinforce such critics, "Although the analysis team only sees the" numerical value "of the athlete, he is missing the individuality of the athlete," he says, in the case of Watson, It is said that they will decide which player to acquire after understanding the individuality of the players by analyzing the players' SNS etc. by making use of it. In other words, it is possible to make an election that emphasizes both the data on the player's play which may have been analyzed so far and the human nature of the athlete individually.


Although Watson gives an evaluation to search for the players the team's management desires, "This system does not make a decision in the end, we expect synergistic effects between humans and the system "Linciner says.