한국프로농구경기 2000～2009시즌 PLAY-OFF 진출 팀의 경기 자료 데이터를 이용한 경기분석 (Analysis on data of teams entered PLAY-OFF in Korea professional basketball games among 2000～2009 season)

서지정보

· 발행기관 : 한국체육과학회
· 수록지 정보 : 한국체육과학회지 / 19권 / 4호 / 1405 ~ 1411페이지
· 저자명 : 김세중, 허종관, 이강웅

초록

10 seasons were held from 2000 to 2009 in Korea pro-league basketball team. In this study, 3186 playoff matches were analyzed by the CART algorism which is one of the decision making tree method. Based on this pattern, it provides more objective information to advance every teams go on a playoff with analyzing win & loss factors.
Through CART algorithm of decision making tree, we cut down 3186 total games to approximately 10% of the total games(300 games). And analyzed win & loss patterns that have more than 75% predictive value. As a result, the decision making tree that made of 9 causes and 10 patterns was made.
The highest main pattern for winning was as follows. More than 23.5 defense rebound, more than 16.5 assists, lesser than 11.5 fouls with free throw.
The highest pattern for defeats was Lesser than 23.5 defense rebound, lesser than 25.5 2pointers, lesser than 9.5 3-pointers, lesser than 17.5 free throw, lesser than 22.5 3pointers attempted.
The second highest pattern for defeats was as follows. Lesser than 23.5 defense rebound, more than 25.5 2-pointers, lesser than 15.5 defense rebound.
In conclusion, main factor that acted on results of games played by 10 seasons of playoff teams was defense rebound. Sub factors were assist, 2pointers, 3pointers, and free throw. The interaction of these factors influenced to result of playoff team's games.

영어초록

10 seasons were held from 2000 to 2009 in Korea pro-league basketball team. In this study, 3186 playoff matches were analyzed by the CART algorism which is one of the decision making tree method. Based on this pattern, it provides more objective information to advance every teams go on a playoff with analyzing win & loss factors.
Through CART algorithm of decision making tree, we cut down 3186 total games to approximately 10% of the total games(300 games). And analyzed win & loss patterns that have more than 75% predictive value. As a result, the decision making tree that made of 9 causes and 10 patterns was made.
The highest main pattern for winning was as follows. More than 23.5 defense rebound, more than 16.5 assists, lesser than 11.5 fouls with free throw.
The highest pattern for defeats was Lesser than 23.5 defense rebound, lesser than 25.5 2pointers, lesser than 9.5 3-pointers, lesser than 17.5 free throw, lesser than 22.5 3pointers attempted.
The second highest pattern for defeats was as follows. Lesser than 23.5 defense rebound, more than 25.5 2-pointers, lesser than 15.5 defense rebound.
In conclusion, main factor that acted on results of games played by 10 seasons of playoff teams was defense rebound. Sub factors were assist, 2pointers, 3pointers, and free throw. The interaction of these factors influenced to result of playoff team's games.

참고자료

· 없음