A Review Of The Hunt Squash Accuracy Test

A Review Of The Hunt Squash Accuracy Test

by Dominic Benacquista

November 5, 2019

A recent article was released about the Hunt Squash Accuracy Test (HSAT) on the Australian Squash Forums. This is exciting due to the fact it exposes coaches and players to relevant literature surrounding squash and it also means people are reading the articles in the literature section of my site.

I am writing this article to further shed some light on the test itself, as all of the relevant information was not shared (there was also some mistakes within the original article) and it would be a disservice to not provide coaches with some of the grey areas relating to the study and test.

So lets start with the back story, this test originally designed by Australian squash legend Geoff Hunt during his time in Qatar at the Aspire Academy. The goal behind the designing such a test was relatively simple: to test the hitting accuracy of his players across a number of squash shots. This would allow Geoff to track his players progress and monitor any changes that occurred during their training. At the time, there was no literature available that assessed the accuracy of squash shots. 

Therefore, Geoff and the sports science team at Aspire got together to put the HSAT through testing via the scientific method. Now you may be wondering, what is the scientific method? As taken straight from Wikipedia:

The scientific method is an empirical method of acquiring knowledge that has characterized the development of science since at least the 17th century. It involves careful observation, applying rigorous skepticism about what is observed, given that cognitive assumptions can distort how one interprets the observation. It involves formulating hypotheses, via induction, based on such observations; experimental and measurement-based testing of deductions drawn from the hypotheses; and refinement (or elimination) of the hypotheses based on the experimental findings. 

To put it simply, the scientific method can be broken down into a few simple steps:

1. Observe/Research the current body of knowledge.
2. Develop a question/refutation surrounding the current body of knowledge.
3. Develop a hypothesis relating to the question/refutation posed.
4. Test hypothesis.
5. Draw conclusion from testing.

Now that the scientific method is understood, back to reviewing the HSAT. As this was the first test of its kind, the validity or reliability of the HSAT needed to be established. 

Validity and reliability are used in research to evaluate how well a method/technique/test measures something. Reliability relates to the consistency of the test i.e. does the test deliver the same results under the same circumstances. Validity on the other hand relates to the accuracy of the test i.e. does the test measure what it is supposed to measure.

The first of three papers assessing the HSAT was titled: Measuring Squash Hitting Accuracy Using The ‘Hunt Squash Accuracy Test’ . The aim of this initial study was to determine the reliability of the HSAT whilst also investigating the validity of the test against player ability as determined by tournament ranking and coach judgement. The HSAT comprised of 375 shots across 13 different squash strokes performed on forehand and backhand side. To assess the reliability of the HSAT ten male squash players from the Aspire Academy aged 17.3 ± 6.4 performed the HSAT twice in 7 days. The typical error was then determined by comparing the two consecutive trials. 

To assess the validity of the HSAT eight male squash players from the Aspire Academy aged 14.8 ± 1.9 performed the HSAT and then competed in a round-robin tournament (PAR 11, best of 5 games). The tournament ranking was then obtained following the conclusion of the tournament. Prior to beginning the tournament the coach also ranked each player. The results are as follows:

• The “Total Overall Score” typical error of the HSAT is 6.94 shots or 1.82% of the 375 shots. This simply means that in order to determine a change in a HSAT score, players would have to increase their score by >6.94 or <6.94 or 1.82% as a percentage of the overall score. If a player fell within that range, a change within the HSAT would not be determined due to the sensitivity of the test.
• The correlation coefficient compared to HSAT for the round robin and coaches rank was 0.93. 
• The most unreliable shot was the backhand boast with a test error of 10.08%.
• The most reliable shot was the backhand shot from the middle of the court with a test error of 3.32%.

The correlation coefficient is a statistical measure that calculates the strength of the relationship between two variables. The values reported range from -1.0 to 1.0. A correlation coefficient of -1.0 represents a perfect negative correlation whilst a correlation coefficient of 1.0 represents a perfect positive correlation. 

Following the results the authors concluded that the study demonstrated that the HSAT to be both reliable based on the test error and valid when compared to tournament ranking and coach perception. Although this study showed a great deal of support for the HSAT, the test does not come without limitations that must be considered, namely:

1. A small sample size. Sample size has a strong relationship with reliability, meaning the larger the sample size, the more reliable the test.
2. Zero match pressure as there were no repercussions for a loss (not a real tournament).
3. The tournament was round robin, meaning a player could lose a match and still place highly or even win.
4. No comparisons to variables that would largely effect a match such as winners/errors.
5. The demographic/ages of the players. Due to the sample size, demographic and ages of the players, it would be quite a stretch to extrapolate the results/accuracy of the test to other populations and standards.
6. No elite players were tested (both senior and junior), again making it to extrapolate the findings to other populations.
7. No evaluation on whether scoring well on a particular shot on the HSAT relates to performing the same shot accurately in a match. 
8. The HSAT is performed in closed environment and thus presents a non-chaotic environment, unlike a match. 
9. The HSAT does not account for tactics or tactical situations i.e. a player could perform well on the HSAT, though not perform well in a match due to inferior tactics compared to their opponent whom may not be as accurate but have superior tactics.
10. Players are fed by a “experienced” coach. There is inherent variability within coaches feeding ability (as determined by ball warmth, ball tightness/accuracy etc) as well as inherent variability between coach-to-coach feeding ability that is not accounted for in the test.

Due to the vast limitations of the initial research, it was deemed that further research was needed. In 2015 the second study on the HSAT was published, titled: “A Quantitative Analysis Of Squash Shot Accuracy”. As aforementioned, the initial research of the HSAT demonstrated positive results, though it had numerous limitations, one of which was that there was no evaluation on whether scoring well on a particular shot on the HSAT relates to performing the same shot accurately in a match. Therefore, the second study on the HSAT aimed to “evaluate the link between the scores from the HSAT and the corresponding accuracy of shots played within tournament match-play”. Similarly to the first study of the HSAT, only a small sample size was used comprising of eight male squash players from the Aspire Academy aged 15.5 ± 1.8. Again, the design of the study was relatively similar whereby players performed the HSAT and then competed in a round-robin tournament (PAR 11, best of 5 games). The tournament ranking was then obtained following the conclusion of the tournament. To determine the accuracy of the shots played during the match, every shot was tagged using performance descriptors using analytics software. The analytics software enabled each shot type to be converted into a percentage. The formula that was used was: total number of shots – unforced errors / total number of shots. This gave the % accuracy for each of the specific shot types. Shot types were then matched from the HSAT and performance analysis data, as well the performance analysis data and the tournament rank to allow for statistical analysis. Similarly to the initial study, a correlation coefficient was used (see above for correlation coefficient description). Results are as follows:

• For the combined values, strong correlations occurred overall for the HSAT Score when compared to accuracy as a percent for backhand (0.94), forehand (0.77), total shots (0.90) and tournament rank (0.95).
• The specific shots within match-play that showed large significant correlations to HSAT scores were the backhand drive (0.92) and backhand volley (0.97), which also both had large significant correlations to tournament (0.93 and 0.84 respectively). It should be worth noting that the backhand side comprised of 23% of all shots.
• The backhand and forehand boast, forehand volley, backhand and forehand volley-drop, the backhand drop did not demonstrate any significant relationship between variables and were weaker as it pertained to correlation coefficients between accuracy for specific shot type, HSAT scores and tournament rank. Although not significant, the forehand drive, forehand volley drop and forehand drop were still reported to show larger correlations between match-play and HSAT scores comparative to the variables aforementioned (it is worth noting that although this was mentioned by the authors, the results were not particularly strong as it is recommended that correlations should be around 0.9). 

Taken from Will G Hopkins when assessing correlation coefficients - Correlations of less than 0.1 are as good as garbage. A correlation 0.9 is very strong. Correlations have to be this good (Hopkins referring to 0.9) before you can talk about accurately predicting the Y value from the X value, especially when you want to use the result of the prediction to rank people.

• It was suggested that weaker correlations for the other HSAT shot results compared to the % accuracy was due to the players favoring their more accurate shot types in order to win/control rally outcome.

Before quoting the conclusion by the authors its worth noting that there were some notable limitations to this study as mentioned by the authors as well as previously aforementioned. Potential things to consider:

1. The authors noted that cross-courts were not considered in the analysis. This removed 36% of the match-play shots as they could not be compared to the HSAT.
2. The authors recognized that the HSAT evaluates a stationary technical aspect of squash, the shot hitting accuracy, however it doesn’t take into account the potential differences in the mechanics between each stroke.
3. Only 10 of the 13 shots in the HSAT were compared.
4. All serves and lob shots also had to be removed from the analysis as those shots are not part of the HSAT and therefor non- comparable for the purpose of this study. These could have severe repercussions to tournament play from tactical position (if we understand the level of players whom were analysed, a good serve and defensive lob could have major influences on the ability to set up and get out of certain scenarios).
5. Again, a small sample size was used. Sample size has a strong relationship with reliability, meaning the larger the sample size, the more reliable the test.
6. The HSAT is performed in a relatively controlled environment, with the majority of shots being hit continuously and the remainder being fed by a coach, where the player knows where the ball will be and therefore has enough time to set themselves to perform the designated stroke without much pressure. 
7. The ball velocity and direction, body position and swing kinematics could all potentially change for the same shot type under match pressure when interacting with another player and trying to win a point. 
8. The HSAT does not account for tactics or tactical situations i.e. a player could perform well on the HSAT, though not perform well in a match due to inferior tactics compared to their opponent whom may not be as accurate but have superior tactics.
9. The demographic/ages of the players. Due to the sample size, demographic and ages of the players, it would be quite a stretch to extrapolate the results/accuracy of the test to other populations and standards.
10. No elite players were tested (both senior and junior), again making it to extrapolate the findings to other populations.
11. Zero match pressure as there were no repercussions for a loss (not a real tournament).
12. Players are fed by a “experienced” coach. There is inherent variability within coaches feeding ability (as determined by ball warmth, ball tightness/accuracy etc) as well as inherent variability between coach-to-coach feeding ability that is not accounted for in the test.

With all things considered, the authors concluded that “the large significant correlation between the total HSAT score and both the total % accuracy and tournament rank reconfirms the HSAT as a highly valid method of assessing the accuracy and performance of junior squash players when compared to overall performance at a tournament”. If you have read through thus far, you probably feel this is a bold conclusion based off the authors results. Most people would tend to agree. If you are tuning out and not staying for the review of the final paper, it would be recommended that if you were using this test within your program, that you do so in conjunction with a variety of other metrics due to its aforementioned limitations. 

The final paper that was published on the HSAT was released in 2018 and titled: Validation Of The Hunt Squash Accuracy Test Used To Assess Individual Shot Performance. This paper was written by the previous authors of the first two papers. The aim of the final study was to “quantify the relationship between the scores from the HSAT and the success of shots played within tournament match-play and to assess whether the HSAT score reflects player tournament rank”. Interestingly, when reviewing the method in which this hypothesis was being tested, it was noted that the exact same participants/data/methods/statistical analysis were being used from the previous paper published in in 2015 (Quantitative Analysis Of Squash Shot Accuracy). A few points should therefore be noted: 

1. This leaves question marks to the efficacy of the final study as the authors in question already had the results from the previous data collection and statistical analysis, thus exposing all them to inherent biases. 
2. The also authors concluded in their previous study “the results of this study, specifically the large significant correlation between the total HSAT score and both the total % accuracy and TR reconfirms the HSAT as a highly valid method of assessing the accuracy and performance of junior squash players when compared to overall performance at a tournament” . With this being a prior conclusion in the published study published by the same authors, one must again question the nature behind the hypothesis i.e. was hypothesis/aim formed on basis of already knowing the results…? See the aim of the research paper in question…
3. No change in sample size, methods etc also means that the authors would not have addressed the inherent limitations that were mentioned from their previous studies. 
4. The authors did not recognize their second study on the HSAT (Quantitative Analysis Of Squash Shot Accuracy) within the results or discussion section and chose to only support the findings of the current study in question using the initial study that was published in 2014 (Measuring Squash Hitting Accuracy Using The ‘Hunt Squash Accuracy Test’). A reason for this is could be that much of the data being presented within the paper is identical to the one published in 2015.
5. Only minor changes were made to wording when publishing results. In the paper titled “Quantitative Analysis Of Squash Shot Accuracy” percentage accuracy was used to determine whether the shot type was executed successfully or not in game and was then correlated with the HSAT score and tournament rank. In the paper titled: “Validation Of The Hunt Squash Accuracy Test Used To Assess Individual Shot Performance” the wording of “percentage accuracy” was changed to “shot success”. Shot success was then correlated with the HSAT score and tournament rank and formed the basis of the results, discussion and subsequent conclusion. 

With the information above information noted, the subsequent findings and conclusions of the final paper should be taken with a grain of salt as it essentially repeats what was already stated in the “Quantitative Analysis Of Squash Shot Accuracy” paper. 

After all this information has been presented to you, you may be wondering, is it worth implementing the HSAT into your current coaching programs? 

This is totally up to you. If you see value in implementing in the HSAT, implement it, if not, don’t. As mentioned, the HSAT has shown validity and reliability in measuring hitting accuracy in a closed situation. It could also be used as a simple tool to aid in quantifying improvements over time, outline weaknesses and/or deficiencies in stroke-play or monitor accuracy changes occurring within a training block with players whom you coach. Just be aware that the HSAT is a tool in your coaching toolbox and that having players perform well on the HSAT will not make them the next Ali Farag. Again, note there are limitations to the HSAT and further research would be needed to really quantify whether the strong conclusions presented by the authors are accurate in larger populations of elite level squash players.

Hopefully this article has outlined and further clarified some objective information surrounding the Hunt Squash Accuracy Test. If you would like to use this test within your program, you can download the Hunt Squash Accuracy Test below for free. It has all the information you need to run the test (protocols etc) as well as the references if you decide that you would like to read up on the tests more.

******