The Annals of Applied Statistics

Competing process hazard function models for player ratings in ice hockey

A. C. Thomas, Samuel L. Ventura, Shane T. Jensen, and Stephen Ma

Full-text: Open access

Abstract

Evaluating the overall ability of players in the National Hockey League (NHL) is a difficult task. Existing methods such as the famous “plus/minus” statistic have many shortcomings. Standard linear regression methods work well when player substitutions are relatively uncommon and scoring events are relatively common, such as in basketball, but as neither of these conditions exists for hockey, we use an approach that embraces the unique characteristics of the sport. We model the scoring rate for each team as its own semi-Markov process, with hazard functions for each process that depend on the players on the ice. This method yields offensive and defensive player ability ratings which take into account quality of teammates and opponents, the game situation, and other desired factors, that themselves have a meaningful interpretation in terms of game outcomes. Additionally, since the number of parameters in this model can be quite large, we make use of two different shrinkage methods depending on the question of interest: full Bayesian hierarchical models that partially pool parameters according to player position, and penalized maximum likelihood estimation to select a smaller number of parameters that stand out as being substantially different from average. We apply the model to all five-on-five (full-strength) situations for games in five NHL seasons.

Article information

Source
Ann. Appl. Stat., Volume 7, Number 3 (2013), 1497-1524.

Dates
First available in Project Euclid: 3 October 2013

Permanent link to this document
https://projecteuclid.org/euclid.aoas/1380804804

Digital Object Identifier
doi:10.1214/13-AOAS646

Mathematical Reviews number (MathSciNet)
MR3127956

Zentralblatt MATH identifier
1283.62245

Keywords
Competing process models Cox model ice hockey Bayesian inference hierarchical models player ratings

Citation

Thomas, A. C.; Ventura, Samuel L.; Jensen, Shane T.; Ma, Stephen. Competing process hazard function models for player ratings in ice hockey. Ann. Appl. Stat. 7 (2013), no. 3, 1497--1524. doi:10.1214/13-AOAS646. https://projecteuclid.org/euclid.aoas/1380804804


Export citation

References

  • Beaudoin, D. and Swartz, T. B. (2010). Strategies for pulling the goalie in hockey. Amer. Statist. 64 197–204.
  • Berry, S. M., Reese, C. S. and Larkey, P. D. (1999). Bridging different eras in sports. J. Amer. Statist. Assoc. 94 661–676.
  • Brown, L. D. (2008). In-season prediction of batting averages: A field test of empirical Bayes and Bayes methodologies. Ann. Appl. Stat. 2 113–152.
  • Cook, S. R., Gelman, A. and Rubin, D. B. (2006). Validation of software for Bayesian models using posterior quantiles. J. Comput. Graph. Statist. 15 675–692.
  • Cox, D. R. (1972). Regression models and life-tables. J. R. Stat. Soc. Ser. B Stat. Methodol. 34 187–220.
  • Dawid, A. P. (1994). Selection paradoxes of Bayesian inference. In Multivariate Analysis and Its Applications (Hong Kong, 1992). Institute of Mathematical Statistics Lecture Notes—Monograph Series 24 211–220. IMS, Hayward, CA.
  • Gramacy, R. B., Jensen, S. T. and Taddy, M. (2013). Estimating player contribution in hockey with regularized logistic regression. Available at arXiv:1209.5026.
  • Hans, C. (2011). Elastic net regression modeling with the orthant normal prior. J. Amer. Statist. Assoc. 106 1383–1393.
  • Hirotsu, N. and Wright, M. (2002). Using a Markov process model of an association football match to determine the optimal timing of substitution and tactical decisions. Journal of the Operational Research Society 53 88–96.
  • Hoerl, A. E. and Kennard, R. W. (1970). Ridge regression: Biased estimation for nonorthogonal problems. Technometrics 12 55–67.
  • Ilardi, S. and Barzilai, A. (2008). Adjusted plus-minus ratings: New and improved for 2007–2008. Available at http://www.82games.com/ilardi2.htm.
  • James, W. and Stein, C. (1961). Estimation with quadratic loss. In Proc. 4th Berkeley Symp. Probab. Statist. Sympos. Math. Statist. and Prob. 1 361–379. Univ. California Press, Berkeley, CA.
  • Klein, J. Z. and Reif, K.-E. (2001). The Hockey Compendium: NHL Facts, Stats and Stories. McClelland & Stewar, Toronto, Ontario, Canada.
  • Li, Q. and Lin, N. (2010). The Bayesian elastic net. Bayesian Anal. 5 151–170.
  • Lock, D. and Schuckers, M. (2009). Beyond +/-: A rating system to compare NHL players. Presentation at joint statistical meetings.
  • Macdonald, B. (2011). A regression-based adjusted plus-minus statistic for NHL players. J. Quant. Anal. Sports 7.
  • Macdonald, B. (2012a). An expected goals model for evaluating NHL teams and players. In MIT Sloan Sports Analytics Conference 2012. MIT, Cambridge, MA.
  • Macdonald, B. (2012b). Adjusted plus-minus for NHL players using ridge regression. Available at http://arxiv.org/abs/1201.0317v1.
  • Morrison, D. G. (1976). On the Optimal Time to Pull the Goalie: A Poisson Model Applied to a Common Strategy Used in Ice Hockey. In TIMS Studies in Management Science 4. North-Holland, Amsterdam.
  • Park, T. and Casella, G. (2008). The Bayesian lasso. J. Amer. Statist. Assoc. 103 681–686.
  • Rosenbaum, D. T. (2004). Measuring how NBA players help their teams win. Available at http://www.82games.com/comm30.htm.
  • Schuckers, M. E., Lock, D. F., Wells, C., Knickerbocker, C. J. and Lock, R. H. (2011). National hockey league skater ratings based upon all on-ice events: An adjusted minus/plus probability (AMPP) approach. Unpublished manuscript. Available at http://myslu.stlawu.edu/~msch/sports/LockSchuckersProbPlusMinus113010.pdf.
  • Sill, J. (2010). Improved NBA adjusted +/- using regularization and out-of-sample testing. In MIT Sloan Sports Analytics Conference. MIT.
  • Thomas, A. C. (2006). The impact of puck possession and location on ice hockey strategy. J. Quant. Anal. Sports 2 18 pp. (electronic).
  • Thomas, A. C. (2007). Inter-arrival times of goals in ice hockey. J. Quant. Anal. Sports 3 17.
  • Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. J. R. Stat. Soc. Ser. B Stat. Methodol. 58 267–288.
  • Tibshirani, R. (1997). The lasso method for variable selection in the Cox model. Stat. Med. 16 385–395.
  • Zou, H. and Hastie, T. (2005). Regularization and variable selection via the elastic net. J. R. Stat. Soc. Ser. B Stat. Methodol. 67 301–320.