A Deeper Statistical Assessment of Athletes Performances at the Commonwealth Games 2018

In Interesting Stats #19, I undertook a Statistical Assessment of New Zealand Athletes Performances at the Championships in 2017. I have used the same protocol and methods to undertake an assessment of the New Zealand team performances at the Commonwealth Games in 2018.

In 2005 a former colleague of mine, Professor Will Hopkins, published a study, Competitive Performance of Elite Track-and-Field Athletes: Variability and Smallest Worthwhile Enhancements. http://jzbz.sportsci.org/jour/05/wghtrack.htm. In this study, he described the reproducibility of competitive performance of elite track-and- field athletes over 17 international competitions in a single competition season. Typical within-athlete variability from competition to competition was derived as a coefficient of variation by repeated-measures analysis of log-transformed times (for running and hurdling events) or distances (for jumping and throwing events).

The (adapted by Hollings) below table shows the typical within-athlete variation in performance from competition to competition for the various events.

Typical variability of a track-and-field athlete's performance between international competitions, expressed as a coefficient of variation (CV). CV (%) (90% conf. Event limits) Running <3000ma 1.0 (0.9–1.1) Running 3000m -10000m b 1.4 (1.2–1.6) High jump, triple jump 1.7 (1.5–1.9) Long jump, pole vault 2.4 (2.1–2.7) Discus, javelin, shot put, hammer 2.8 (2.4–3.2) a100m to 1500m run; 100m & 110m and 400m hurdles.

b3000m to 10,000m run; male 3000m steeplechase.

When the above table is applied to the New Zealand team members’ results at the , it is possible to see how each athlete competed in relation to their season best performance before the Championships. Whilst there is every expectation that an athlete will reach or get very near to their season’s best (their peak) performance at the major championship of the season, the reality is that they rarely do! In a study I undertook in 2013 on the performance outcomes at the 2013 World Championships, I found that generally less than 10% of all athletes managed to achieve either a personal best or a season’s best performance at the World Championships. Athletes in New Zealand are generally selected for a major championship on the basis that they will achieve or get near to their season’s best at the Championships. To quantify their achievement at the championships, compared to their season’s best performance we multiplied the above CVs by 0.3, 0.9, 1.6, and 2.5, which equates to the performance change required in an elite athlete’s performance in order to achieve small, moderate, large and very large differences between their season’s best and Commonwealth Games performances (Hopkins et. al., (2009) at http://www.sportsci.org/2009/prostats.htm). Further, using the works of Hopkins, it is also possible to make a qualitative judgement on the championship performance compared to the athlete’s seasonal best performance before the championships. This explanation and the following tables have been produced with the help of my colleague, Dr Sian Allen, a former Performance Analyst with HPSNZ, who now works in Canada.

Notes:

1. It is conceded that in the middle- and long– distance events, performance time at the championships, and also the personal best time may have been influenced by tactics and by pacing strategies. 2. (10000m) and (10000m) did not compete in their event in 2018 prior to the Commonwealth Games – thus they did not have a season’s best performance to compare with their performances at the Commonwealth Games. 3. Whilst seasonal best performance is used as the baseline performance in this study, it is accepted that much depends on: the circumstances surrounding the season’s best performance, the scheduling of that performance, the time-period between the season best performance and the championships, any injuries sustained in that time-period, any other interruptions to the training programme, and the level of competition able to be accessed in that time-period. All the fore mentioned can impact on the quantum of the season’s best performance. 4. I have not included (F46 Javelin Throw) in this analysis as there is no historical CV data that can be used in the analysis.

Table 2. Athletes’ results, and overall ranking at the championships and the change in their ranking (+ or -) from the personal best pre-competition ranking.

Ranking

Overall Ranking on Finishing Athlete Event Performance Finishing PB Change Position Position Pre-Comp 200 m 6s3 21.01 16 18 +2

Brad Mathas 800 m 5 1:46.07 5 15 +10 10,000 m 5 27:30.90 5 11 +6 Cameron French 400m h 6h2 50.60 14 12 -2 20 km Race Walk 5 1:21:47 5 10 +5 Nic Southgate Pole Vault 9 NH 9 7 -2

Tom Walsh Shot Put 1 22.41 m 1 1 0

Ben Langton Burnell Javelin 12 73.77 m 12 5 -7 800 m 5h1 2:00.62 9 8 -1 Camille Buscomb 5,000 m 12 15:55.45 12 9 -3 Camille Buscomb 10,000 m 14 33:07.53 14 7 -7

Alana Barber 20 km Race Walk 2 1:34:18 2 7 +5

Eliza McCartney Pole Vault 2 4.70 m 2 1 -1 Olivia McTaggart Pole Vault 9 4.30m 9 9 0 Shot Put 2 18.70m 2 1 -1 Discus Throw 4 57.16m 4 4 0 Hammer 1 69.94 m 1 4 +3

Six out of sixteen athletes improved their ranking in the competition from their (PB) ranking pre-competition. Three others maintained their pre-competition ranking position. However, only nine athletes in the 17 events contested achieved a “top 6 finishing position” which was one of the criteria for selection. Table 3. Qualitative Description of the Magnitude Thresholds for each athlete and their event.

Magnitude Thresholds Athlete Event Event Variability Small Moderate Large Very Large Valerie Adams Shot Put 2.8% 0.8% 2.5% 4.5% 7.0% 20km RW 1.4% 0.4% 1.3% 2.2% 3.5% Camille Buscomb 10000m 1.4% 0.4% 1.3% 2.2% 3.5% Camille Buscomb 5000m 1.4% 0.4% 1.3% 2.2% 3.5% Cameron French 400mh 1.0% 0.3% 0.9% 1.6% 2.5% Siositina Hakeai Discus Throw 2.8% 0.8% 2.5% 4.5% 7.0% Ben Langton Burnell Javelin Throw 2.8% 0.8% 2.5% 4.5% 7.0% Eliza McCartney Pole Vault 2.4% 0.7% 2.2% 3.8% 6.0% Olivia McTaggart Pole Vault 2.4% 0.7% 2.2% 3.8% 6.0% Brad Mathas 800m 1.0% 0.3% 0.9% 1.6% 2.5% Joseph Millar 200m 1.0% 0.3% 0.9% 1.6% 2.5% Angie Petty 800m 1.0% 0.3% 0.9% 1.6% 2.5% Julia Ratcliffe Hammer Throw 2.8% 0.8% 2.5% 4.5% 7.0% Quentin Rew 20km RW 1.4% 0.4% 1.3% 2.2% 3.5% Jake Robertson 10000m 1.4% 0.4% 1.3% 2.2% 3.5% Nick Southgate Pole Vault 2.4% 0.7% 2.2% 3.8% 6.0% Shot Put 2.8% 0.8% 2.5% 4.5% 7.0%

Table 4. Qualitative Description of the Magnitude of the Championships Peformance vs Season’s Best Performance

Commonwealth 2018 Best % Diff CWG Athlete Event Qualitative Description of Difference Games Perf. (SB) Perf vs SB Valerie Adams Shot Put 18.70 18.48 1.2% Positive difference by a small amount Alana Barber 20km RW 1:34:18 1:32:19 -2.1% Negative difference by a moderate amount Camille Buscomb 10000m 32:23.91 Camille Buscomb 5000m 15:55.45 15:49.55 -0.6% Negative difference by a small amount Cameron French 400mh 50.60 49.33 -2.5% Negative difference by a very large amount Siositina Hakeai Discus Throw 57.16 60.54 -5.9% Negative difference by a large amount Ben Langton Burnell Javelin Throw 73.77 78.20 -6.0% Negative difference by a large amount Eliza McCartney Pole Vault 4.70 4.75 -1.1% Negative difference by a small amount Olivia McTaggart Pole Vault 4.30 4.35 -1.2% Negative difference by a small amount Brad Mathas 800m 1:46.07 1:46.44 0.3% Positive difference by a small amount Joseph Millar 200m 21.01 20.60 -2.0% Negative difference by a large amount Angie Petty 800m 2:00.62 2:00.73 0.1% Positive trivial difference Julia Ratcliffe Hammer 69.94 68.39 2.2% Positive difference by a small amount Throw Quentin Rew 20km RW 1:21:47 1:21:55 0.2% Positive trivial difference Jake Robertson 10000m 27:30.90 Nick Southgate Pole Vault NH 5.35 Negative difference by a very large amount Tom Walsh Shot Put 21.41 22.67 -5.9% Negative difference by a large amount

Comment

1. Five athletes (Valerie Adams, Brad Mathas, Angie Petty, Julia Ratcliffe and Quentin Rew) had positive differences between their season’s best performance pre- Games and their championship performance. i.e., they all peaked at the Commonwealth Games. Additionally, Jake Robertson ran a Personal Best performance in the 10000m. 2. Although Tom Walsh (Negative difference by a large amount), Alana Barber (Negative difference by a moderate amount) and Eliza McCartney (Negative difference by a small amount), each had negative differences between their Season’s Best performance and their Games performance, they each won a medal. This negative difference does show how World Class these athletes are. They were able to win medals despite not being at their absolute peak. Tom, in particular, was impressive with his 22.45m shot qualifying mark and he headed six men over 20m in the final. In the women’s pole vault the Games record had stood at 4.62m, so there were four new records set in this event. Further, Alana walked well in the trying climatic conditions where all times were down on both personal best and season’s best. 3. Both Camille Buscomb (5000m) and Olivia McTaggart (Pole Vault) each had negative differences by a small amount, therefore demonstrating that they were very close to their seasonal best performance at the Commonwealth Games. A small negative difference is acceptable in the circumstances of the competitive environment of a major championship. 4. Unfortunately for the other five athletes, the ability to peak at a major championship proved to be a challenge.

NEW ZEALAND TEAMS AT THE COMMONWEALTH GAMES 1930 - 2018

Year Venue Team Size Gold Silver Bronze Total Top 6 (other than medals) 1930 Hamilton (CAN) 4 2 0 0 2 1 1934 London (ENG) 3 1 0 1 2 0 1938 (AUS) 19 3 2 4 9 9 1950 Auckland (NZL) 64 2 6 12 20 30 1954 Vancouver (CAN) 14 4 1 0 5 7 1958 Cardiff (WAL) 19 2 3 4 9 8 1962 Perth (AUS) 27 5 4 2 11 8 1966 Kingston (JAM) 17 5 1 4 10 5 1970 Edinburgh (SCO) 20 0 3 1 4 7 1974 (NZL) 59 2 4 2 8 21 1978 Edmonton (CAN) 28 0 2 0 2 10 1982 Brisbane (AUS) 32 2 1 3 6 9 1986 Edinburgh (SCO) 25 0 2 2 4 11 1990 Auckland (NZL) 81 1 2 6 9 14 1994 Victoria (CAN) 25 0 3 2 5 9 1998 Kuala Lumpur 28 1 0 1 2 9 (MAS) 2002 Manchester (ENG) 16 1 0 3 4 5 2006 Melbourne (AUS) 23 2 1 1 4 7 2010 Delhi (IND) 11 1 4 2 7 3 2014 (SCO) 18 1 2 2 5 4 2018 Gold Coast (AUS) 16 2 3 0 5 4

The team performance outcome was the same as that at Glasgow in 2014 with five medals and four other top six placings. MEDALS AND POINTS TABLE – 2018 COMMONWEALTH GAMES (reproduced from Athletics International)

Points: 8 for 1st to 1 for 8th.

G S B Total Pts 2014 2010 2006 AUS 7 7 7 21 258 179 200 379.5 JAM 7 8 9 24 207 184 84 185.5 ENG 4 3 5 12 184 290 297 238 KEN 4 8 5 17 144 205 251 131 CAN 2 6 3 11 136 148.5 126.5 114 RSA 4 1 3 8 84 73 36 134.5 UGA 3 1 1 5 75 33 30 22 IND 1 1 1 3 57.5 31 142 39 NGR 1 2 1 4 56 70 34 54 NZL 2 3 - 5 54 55 56.5 71 TTO 2 - - 2 48 54 37 28.5

NZL finished 10th on the Placings Table with 48 pts, six points less than they scored in 2014, but scoring more points than they did in 2010 and 2006. NZL finished 7th on the Medals Table - equal with Uganda. 18 nations won gold medals, 24 medals of any colour and 42 had athletes placed in the top 8 for a record spread of nations (compared to 12, 20 and 32 in 2014), The above table does not include the various events for Paralympians that were integrated into the programme.