Volume 3, Issue 4

While it is generally recognized that the relative importance of different skills is not constant across different positions on a basketball team, quantification of the differences has not been well studied. 1163 box scores from games in the National Basketball Association during the 1996-97 season were used to study the relationship of skill performance by position and game outcome as measured by point differentials. A hierarchical Bayesian model was fit with individual players viewed as a draw from a population of players playing a particular position: point guard, shooting guard, small forward, power forward, center, and bench. Posterior distributions for parameters describing position characteristics were examined to discover the relative importance of various skills as quantified in box scores across the positions. Results were consistent with expectations, although defensive rebounds from both point and shooting guards were found to be quite important.

In papers on basketball, it is standard practice to treat the home-court advantage in terms of percentages or point differences at the end of the game. This practice leaves out of account how the advantage develops during the game, when it accumulates most strongly, its course and the in-course dynamics. This study analyzes all games played in two seasons of the NBA by quarters and overtime periods. The main result is that home advantage in the NBA is strongly front-loaded. In both years studied the home team accumulated two thirds of the home advantage it had at the end of the game in the first quarter. It accumulated less of an advantage in the second and third quarters, and still less in the fourth quarter. Further, the home team does not on average lengthen its lead in quarters which it enters ahead, but gains strongly in any quarter which it enters behind. The paper concludes with a discussion of theoretical issues raised by these results and next steps in research.

The aim of the present study was to present the playing characteristics of the teams in complex I and to attempt to determine which of these characteristics led to victory and to the final ranking of the teams. The subjects were 38 Olympic Volleyball men's games. In every game, teams were characterized according to the result of the game (win or lose). The games were video-recorded and analyzed with the "Data Volleyball Project Sport Software" program. It recorded every skill of the complex I of the game, for each player of each team. There was a five level scale protocol according to the effectiveness of the skill (Eom & Schutz, 1992; Eom, 1989). Discriminant analysis was conducted to select which subset of the measured variables significantly contributed to the prediction of winning or losing in Olympic Volleyball teams. Among the five variables of service-reception: two variables "best reception, 1st set attack" and "Good reception, high set attack" were entered into the final model yielding, standardized coefficient 1.22 and .78 respectively. Among the five "attack from reception" categories, only one variable, "ace-point" was selected to enter into the final model. These results conclude that the "best" and "good reception" and the "ace-point" in the attack remain powerful aggressive tools for high level teams and were predictors to win.

Position play is a key feature of carom billiards: on easy shots, players can manage to score while ensuring that the next position will be favorable. The difficulty of a shot therefore depends on the previous shot; e.g., an easy shot generally follows an easy shot. We introduce a Markov process that accounts for such correlations. This model can explain the long series of easy shots and the high scores that ensue. It can also identify differences in the scoring patterns of players at different skill levels. Players can use this model via the web page http://billiards.mathieu.bouville.name/biMar/.

The 10 disciplines in the decathlon can be broadly characterised as running, jumping and throwing. However, these simplistic characteristics may not represent the groupings defined by performances in the decathlon. The identification of groups may reveal a recondite advantage for athletes who excel in particular disciplines. Therefore this study used cluster analysis to determine the groupings inherent within the decathlon disciplines. The data set was derived from the top 173 decathletes between the years 1986 to 2005. Six clustering methods were applied to a Euclidean proximity matrix. The highest number of clusters common to all the methods was accepted as the solution. All six methods produced the same 3-cluster ([100m 400m 110H LJ PV HJ][SP DT JT][1500m]), 4-cluster ([100m 400m 110H LJ PV][SP DT JT][HJ][1500m]) and 5-cluster ([100m 400m 110mH LJ][SP DT JT][PV][HJ][1500m]) solutions. Stability tests confirmed the consistency of all the solutions. The 10 disciplines of the decathlon form into five groupings, which can be adequately explained from a physiological perspective. The clustering suggests that athletes who perform better at the sprint/track disciplines may obtain an advantage in the decathlon.

In the 2004 Ryder Cup Matches, US Captain Hal Sutton decided to pair his two best players, Tiger Woods and Phil Mickelson, in the opening four-ball match. The popular press characterized this pairing with the adjectives ``bold" and ``risky" since it is generally thought that balanced teams are preferable to unbalanced. In this paper I compare a Balanced strategy (matching stronger players with weaker players so that the resulting pairs are balanced) against an Unbalanced strategy (matching stronger players with stronger players and weaker players with weaker players) and find the Balanced strategy to be superior, but only slightly so.