Random walks and momentum shifts in the 2018 Super Netball Grand Final

Introduction

The Sunshine Coast Lightning defeated West Coast Fever in the 2018 Super Netball Grand Final.

Prior to the game, I had identified some performance bandwidth profiles for both teams. Post-game, the performance profiles of both teams were:

Sunshine Coast

The team’s actual performance delivered an exceptional third quarter and moved the team into a game-winning position.

West Coast

The scoring sequence in the game was:

Random Walks

In 2006, Martin Lames explored dynamic interactions in sport games. His discussion included an examination of handball possession in terms of interlaced random walks and the momentary strength of competing teams.

Martin conjectures in his paper:

There is evidence for the hypothesis that a team’s scoring rate is independent from the one of other team, but we see also phases with a seemingly strong dependence. Moreover, sometimes the momentary scoring probabilities seem to be negatively correlated (my team is good when the other is bad and vice versa), but sometimes there is a positive relationship (my team performs well when the other does so).

There have been subsequent discussions of random walk in the literature. See, for example: Alan Gabel and Sidney Redner (2012); Leto Peel and Aaron Clauset (2015); Dilan Kiley et al (2016); and Jaime Prieto, Miguel-Angel Gomez and Jaime Sampaio (2016).

In their discussion, Alan and Sidney noted:

There are three factors that determine which team scores. First, the better team has a greater intrinsic chance of scoring. The second factor is the anti- persistence of successive scoring events that arises from the change of possession after a score. The last is the linear restoring force, in which the scoring probability of a team decreases as its lead increases (and vice versa for a team in deficit).

Leto and Aaron propose:

Anti-restoration or momentum occurs when the leading team has a higher chance of scoring again.

Momentum is the reverse of restoration.

Dilan and his colleagues note:

Each game generates a probabilistic, rule-based story, and the stories of games provide a range of motifs which map onto narratives found across the human experience: dominant, one-sided performances; back-and-forth struggles; underdog upsets; and improbable comebacks.

The Final

I used secondary day from Champion Data’s record of the Final. I though I would look for random walks and momentum changes in the data.

I used RStudio, ggplot2, and ggrepel to visualise the data. My record of the game tracks the Sunshine Coast’s score difference performance. Green shaded areas indicate Sunshine Coast lead. The second quarter is in purple to indicate West Coast’s lead throughout that quarter.

The data show eight excellent examples of the tendency of a random walk to move to a central location. Each of them exists at a new equilibrium in the game.

I have included some time-in-game labels to indicate my perception of a momentum shift. The ability for a team to create these episodes (and respond to them when opponents are driving the game) resonates powerfully with some of the discussions about temporal (T) patterns and their critical interval relationships initiated by Magnus Magnusson.

It would be fascinating to learn how coaches from both teams addressed these shifts in the messages they shared with their players. It would be interesting to learn what was said at half time too. The half time break straddled a seven-goal run from the Sunshine Coast.

Discussion

I have really enjoyed this season’s Super Netball competition. The final was closely contested. I was particularly interested in the pivot in the game that occurred in the third quarter. The Sunshine Coast produced their best third quarter of their entire season. Their previous highest score in the third quarter was 18. West Coast had experience of teams lifting in the third quarter in recent games. Their opponents in weeks 13 and 14 of the regular season had both scored 20 goals.

My pre-game priors suggested that West Coast would win the Final by 3 goals. At 13 minutes and 05 seconds in the second quarter West Coast led by 7 goals. What happened thereafter provides a case study of the interlacing of random walks in the context of momentum shifts … for an away team.

Photo Credits

Super Netball (Sue Gaudron Q & A)

Lightning (Twitter)

A Box Plot Look at the Super Netball 2018 Grand Final

The 2018 Super Netball Final between West Coast Fever and Sunshine Coast Lightning takes place on 26 August at the Perth Arena.

I have followed this season’s competition results through the data supplied by Champion Data.

I have used BoxPlotR to visualise the data in an attempt to identify how the Final might be played.

West Coast

Centre lines show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, outliers are represented by dots. There are n = 15 sample points.

Data summary:

Sunshine Coast


Centre lines show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, outliers are represented by dots. There are n = 16 sample points.

Comparison Plot: West Coast and Sunshine Coast

Centre lines show the medians; box limits indicate the 25th and 75th percentiles as determined by R software; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, outliers are represented by dots. There are 15 sample points for West Coast and 16 for the Sunshine Coast.

The data suggest a close game. I am keen to see how the Sunshine Coast stay in contact with the West Coast as the away team, particularly in the final quarter.

Photo Credit

Super Netball (Sue Gaudron Q & A)

Re-view Super Netball 2018: Round 13 Lightning v Swifts

Some of the delegates from the 14th Australasian Conference on Mathematics and Computers in Sport attended the Sunshine Coast Lightning v NSW Swifts netball game on 28 July 2018.

I wrote about their visit in a previous post. I shared some netball data relevant to the game in a GitHub repository. I have written a postscript about the data in the READ.ME file for the repository.

My summary chart is:

As the game progressed, the probability of the Lightning winning moved from 0.63 after leading at the end of the first quarter of the game to 0.76 after leading at half time to 0.87 with their lead at the end of the third quarter.

The two teams met in Round 7 this season’s competition. I thought the Lightning’s first quarter performance reflected their determination after losing the first quarter of that game by 10 goals. Their final quarter performance was also better than their median performance for that quarter this season.

The Swifts underperformed in the game compared to their median scoring profile this season. It would have been interesting to have attended the game to observe the pressure the Lightning placed on the Swifts.

A report of the game included these observations:

The defence of the Lightning did the job at USC Stadium. Karla Pretorius finished with six intercepts, Madeline McAuliffe was a brick wall in the second half and centre Laura Scherian was the Nissan MVP to complete the defensive triumvirate.

The Swifts were competitive and fought to the final whistle, but they were left to rue a number of missed opportunities to score off turnovers.

Post-game I looked for some temporal patterns (T-patterns) to support evidence of strong Lightning performances in quarters 1 and 4. The Champion Data scoring flow for the game had these sequences of scoring within the overall alternating of possession:

Quarter 1

Quarter 4

Photo Credit

Courtside (Andrew Simmons, Twitter)