Quantcast


CURATOR
A pinboard by
George Ng

Ph.D who studied Biotechnology, now part of an eCommerce startup based in Hong Kong.

PINBOARD SUMMARY

Will the use of machine learning to predict professional basketball championships be the norm?

In 10 seconds? General managers of sport teams spout stats like PER, Pty/36, PPG, APG, win shares and VORP, what do they mean? Predicting winning teams sounds complicated, but machine learning gives everyone a shot at becoming Jerry Maguire.

Show me the money! Free and open access to teams, game stats and players past game data, has allowed data scientists to aggregate game stats, analyse game trends and surface predictor factors. 3 different research groups have published how machine learning has been used to accurately predict players performance and ultimately a teams' chance of bringing home the trophy! (see the papers)

But aren't games won by star players? Of course, star players attract a lot of followers and they do win games, but exceptions do happen. Stephen Curry of the Golden State Warriors was voted the Most Valuable Player (2015-2016) and won the NBA All Star 4 times from 2014-2017, but his team were edged out in the 2016 finals by Cleveland Cavaliers. (read more)

We know for sure who will be the next championship team? Research groups have applied machine learning approaches to predict basketball team playoffs and championship wins, but the verdict is still out about the accuracy of predictions. (read more)

10 ITEMS PINNED

Finding Common Characteristics Among NBA Playoff and Championship Teams: A Machine Learning Approach

Abstract: In this paper, we employ machine learning techniques to analyze sixteen seasons of NBA regular season data from every team to determine the common characteristics among NBA playoff teams. Each team was characterized by 42 predictor variables and one binary response variable taking on a value of "TRUE" if a team had made the playoffs, and value of "FALSE" if a team had missed the playoffs. After fitting an initial classification tree to this problem, this tree was then pruned which decreased the test error rate. Further to this, a random forest of classification trees was grown which provided a very accurate model from which a variable importance plot was generated to determine which predictor variables had the greatest influence on the response variable. The result of this work was the conclusion that the most important factors in characterizing a team's playoff eligibility are the opponent field goal percentage and the opponent points per game. This seems to suggest that \emph{defensive} factors as opposed to offensive factors are the most important characteristics shared among NBA playoff teams. We also perform a classification analysis to determine common characteristics among NBA championship teams. Using an artificial neural network structure, we show that championship teams must be able to have very strong defensive characteristics, in particular, strong perimeter defense characteristics in combination with an effective half-court offense that generates high-percentage two-point shots. A key part of this offensive strategy must also be the ability to draw fouls. This analysis will hopefully dispel the rising notion that an offense geared towards shooting many three point shots is a sufficient and necessary condition for an NBA team to be successful in qualifying for the playoffs and winning a championship.

Pub.: 09 May '16, Pinned: 07 Apr '17