Volume 13, Issue 2

We propose and throughly investigate a temporalized version of the popular Massey’s technique for rating actors in sport competitions. The method can be described as a dynamic temporal process in which team ratings are updated at every match according to their performance during the match and the strength of the opponent team. Using the Italian soccer dataset, we empirically show that the method has a good foresight prediction accuracy.

Temporal characteristics have been a main area of focus in the study of pre-performance routines in elite sports. Ball and player tracking data is currently available that precisely measures service preparation times in professional tennis. However, this data has yet to be utilized for studying service pre-performance routines. In this paper, we present a Bayesian multilevel model of the time-to-serve in a professional tennis match, which includes heterogeneous means, variances and covariate effects. Applying the model to a sample of serves played at the 2016 Australian Open reveals that the typical time-to-serve was 19 s for male players and 20 s for female players. Point importance and the length of the previous rally account for approximately 15% of the within-match variance. However, even with this adjustment, within-match variation is notably larger than between-player variation, 60% greater for men and 30% greater for women. The proposed Bayesian modeling approach is demonstrated to be a useful tool for analyzing in-competition temporal data on preparation time for tennis and other sports.

Ultimate is a field sport played by two teams, each with seven players on the field. USA Ultimate administers nationwide leagues that consist of a regular season and post-season with Sectional, Regional, and National Championship tournaments. USA Ultimate ranks teams by applying an algorithm to the regular season results, and distributes the sixteen bids for the National Championship to the eight regions based on these rankings. Teams then compete at Regionals to earn the bids granted to their region. This article presents a novel score-augmented win-loss model for ranking Ultimate teams and distributing National Championship bids. The proposed approach facilitates predicting the placement of each qualifying team at the 2016 Club National Championships as well. The key innovations are the use of a pseudo-outcome called the win fraction that splits a win between the two teams based on the final score of their match, and a weighted quasi-likelihood function that facilitates discounting older results. The proposed approach is applied to the 2016 Club Division results. Rankings, bid allocations, and predictive placement probabilities are reported, as well as a comparative evaluation with the USA Ultimate algorithm, a win-loss model, and a point-scoring model.

The NCAA basketball tournament attracts over 60 million people who fill out a bracket to try to predict the outcome of every tournament game correctly. Predictions are often made on the basis of instinct, statistics, or a combination of the two. This paper proposes a technique to select round-of-64 upsets in the tournament using a Balance Optimization Subset Selection model. The model determines which games feature match-ups that are statistically most similar to the match-ups in historical upsets. The technique is then applied to the tournament in each of the 13 years from 2003 to 2015 in order to select two games as potential upsets each year. Of the 26 selected games, 10 (38.4%) were actual upsets, which is more than twice as many as the expected number of correct selections when using a weighted random selection method.