EFFECTS OF GAME-RELATED STATISTICAL PARAMETERS ON FINAL OUTCOME IN NATIONAL HOCKEY LEAGUE (NHL)

The aim of this study is to determine which variables affected on final outcome more in situational parameters. We anticipate significant influence of more variables on criteria that shows won and defeated teams. We have taken into consideration scored goals in each period and their difference, minutes of penalty, power play opportunities, faceoff percentage, total shots on goal, save percentage, hits, blocks, giveaways and takeaways. These indicators have been standardized by the National Hockey League (NHL). To investigate separate contribution of each game-related statistical parameter on the final outcome of the game, multiple regression analysis was performed. In regression model we can see that variable Save Percentage (SVS%) (beta = 0,591242, t= 8,07600, p< 0,0000) had the main influence on final outcome. Variables P2 (beta = 0,4878), P3 (beta = 0,3464), P1 (a = 0,2891) and OT (beta = 0,2292) can explain variance on this regression model was R-2= 0,78. To eliminate influence of scoring variables in second regression model only three variables were statistically significant; SVS% (beta = 0,6966, t= 6,0264, p< 0,0000), PPG (beta= 0,3525, t= 3,1925 p< 0,0024) SOG (beta= 0,2298, t= 2,2112, p< 0,03135). Explained variance in second model was R-2= 0,46. In conclusion, save percentage contributes the most on the final result. Good teams usually have better defense setup to eliminate shots in front of the net and slot position to help goaltenders to have open shot form distance.