Abstract game structures of elite male squash under
The first forms of analysis of squash were devised by Sanderson and Way (1979). A notational system enabled analysis of the patterns of play in the game of squash. The study investigated whether or not there are tactics of how to win rallies and if there is a pattern of losing rallies. The amount of winners and errors were notated along with the position in which they were played from. This analysis was notated by hand, using seventeen symbols for different shots. This hand notation system was initially modified for use in a PC, and then converted into a system in which the data was then processed in a main frame computer by Hughes (1983). Hughes (1984, 1985) used this system to produce studies showing patterns of play for all standards. Notation Analysis can also be used to analyse the effect of rule changes in various sports. A study by Pritchard et al. (2001), investigated whether the use of a different scoring system for Badminton would provide a more attractive game for television. audiences. The study showed that the games were reduced in length and more ‘critical points’ were played per game, therefore providing a better suited game for television. In May, 1988, the game of squash had its most radical change, with the new PPR scoring system introduced to the Professional Men’s Circuit. The aim was to make squash a more attractive game to television audiences. The game of squash is traditionally scored up to 9 points, with only a point awarded if the server wins the rally. This English scoring system was changed in 1988 for the professional men’s world circuit to a point per rally (PPR) up to 15, to try to captivate the viewer by making it a more explosive game. This has yet to work, with matches lasting on average the same amount of time.
The first forms of analysis of squash were devised by Sanderson and Way (1979). A notational system enabled analysis of the patterns of play in the game of squash. The study investigated whether or not there are tactics of how to win rallies and if there is a pattern of losing rallies. The amount of winners and errors were notated along with the position in which they were played from. This analysis was notated by hand, using seventeen symbols for different shots. This hand notation system was initially modified for use in a PC, and then converted into a system in which the data was then processed in a main frame computer by Hughes (1983). Hughes (1984, 1985) used this system to produce studies showing patterns of play for all standards. Notation Analysis can also be used to analyse the effect of rule changes in various sports. A study by Pritchard et al. (2001), investigated whether the use of a different scoring system for Badminton would provide a more attractive game for television. audiences. The study showed that the games were reduced in length and more ‘critical points’ were played per game, therefore providing a better suited game for television. In May, 1988, the game of squash had its most radical change, with the new PPR scoring system introduced to the Professional Men’s Circuit. The aim was to make squash a more attractive game to television audiences. The game of squash is traditionally scored up to 9 points, with only a point awarded if the server wins the rally. This English scoring system was changed in 1988 for the professional men’s world circuit to a point per rally (PPR) up to 15, to try to captivate the viewer by making it a more explosive game. This has yet to work, with matches lasting on average the same amount of time. Hughes and Knight (1995) studied the changes in playing patterns between English to 9 and PPR to 15, and concluded that there were no significant differences in the average length of rallies. Furthermore, the additional 6 points required did not reduce the length of each game, so that it still took 17 minutes, on average to reach a critical point. In tennis, each game lasts on average 3 minutes bringing the critical points up more frequently (Hughes, 1995). The women’s professional association has not changed the scoring system to PPR to 15; showing that they feel the new system has not improved the game. The apparent failure of PPR to 15 led to another new scoring system in the men’s professional circuit, which was introduced in 2004. The games are still PPR, but have been cut to 11, thus trying to shorten the game and again trying to make it more entertaining and exciting to watch. Hughes (1995) used notational analysis to create a more attractive scoring system for squash. Elite squash players competed in Zurich, Switzerland using a form of tennis scoring. The aim of Hughes (1995) analysis was to create models of the different scoring systems gathered from this tournament along with data collected by Hughes and Knight (1995). The study showed that rally lengths did not significantly alter, but there was a decrease in the time it took to reach a ‘critical point’. Under the ‘tennis’ scoring, it took less than 3 minutes (every 7 or less rallies), compared to English scoring and PPR scoring which were found to take 17 and 12 minutes respectively (Hughes, 1995). This ‘tennis’ scoring method created a more exciting exhibition for the non-squash viewer, but was not recognised by the PSA, so was not trialled or tested any further. The most recent change has been in the Professional Men’s Circuit where the games are still played PPR, but reduced to 11. There was a belief that a goal of 15 led to no real urgency in the battle for the first few points, whereas condensing the game meant a sprint to 10 was almost a necessity. A further requirement for a two clear point margin if players reach 10-all, was aimed at creating more ‘critical points’ in matches. 1.1. Aims of the Study The main aim of this study was to analyse any changes in the game structure or differences in the patterns of play occurring amongst the elite of men’s squash (inside top 12 in the world), whilst playing in competition under two different scoring systems.
This study analysed 6 matches from both scoring systems played by male players ranked in the top 12 in the PSA World Rankings. The analysis of the data gathered entailed such comparisons as rally lengths and the amount of winners and errors played. This study attempted to determine if matches are shorter and if more ‘critical points’ are created through the new scoring system, hence making it more attractive to the television media. 1.2. Research Hypotheses The new scoring system should produce shorter matches due to the points being reduced by almost a third. The new scoring system may create a race to 11 points, causing the players to play more attacking shots. This may increase the amount of winners played in a match. However, due to the games only being to 11, the players will not want to give away points, which should reduce the amount of unforced errors in a match. At the same time the players will be playing at a higher pace to try to put their opponents under pressure, hopefully causing them to commit forced errors. Point per Rally scoring to 15 in comparison to Point per Rally to 11: 1. Significantly shorter average length of games under the new scoring system. H° = There will be significantly shorter length games. H¹ = There will not be significantly shorter length games. 2. Significantly more winners played under the new scoring system. H° = There will be significantly more winners. H¹ = There will not be significantly more winners. 3. Significantly less unforced errors played under the new scoring system. H° = There will be significantly less unforced errors. H¹ = There will not be significantly less unforced errors. 4. Significantly more forced errors played under the new scoring system. H° = There will be significantly more forced errors. H¹ = There will not be significantly more forced errors. 1.3. Limitations The computerised notation system designed by Hughes (1995) placed certain limits on the degree of accuracy which could be obtained. The main inaccuracy is the recording of the position at which a player strikes the ball. The system gives the option of 16 cells, but to notate the correct cell when the ball is struck at high speeds is very difficult, even with the capability of freeze frame. Another limitation is the variation of the squash courts used. Different courts are used from all around the world, which have varying effects on the speed of play. This can lead to differing patterns of play from court to court. 1.4. Delimitations A delimitation of the study was that 6 matches were analysed for each of the new and old scoring systems. The problem is the old scoring system requires a third more points per game, with the implication that games of PPR to 15 will produce higher total figures than those of PPR to 11. As both the new and old scoring systems are used exclusively for the Men’s Professional Circuit, this does not allow the analysis of different standard players or female players.
2.1. Introduction A computerised squash notation system developed by Hughes and Knight (1995), was used to determine the pattern of play adopted by elite squash players whilst playing under two different scoring systems. The system analyses and produces both statistical data and produces 3-dimensional displays, which aid the understanding of the results. This system is used in conjunction with video recordings of elite squash matches played under two different scoring systems. The matches were analysed post-event using video, this was due to the speed of live match play being too fast to gather all relevant details. 2.2 Validity and Reliability The system had already been validated and used in a number of published studies (Hughes and Knight, 1995; Hughes, 1995). As an intra-operator study, it was decided to analyse a game three times and the results of these analyses were placed into a spreadsheet, and compared using chi-square(P>0.95). An absolute measure of error in the study was also calculated, using percentage differences (Hughes et al., 2004). 2.3 Data The players analysed were all within the top twelve in the world at the time of recording. All were analysed for both scoring systems apart from one player, who only competed under the old system. The recordings were taken from within the last two years so to avoid any possible changes in the patterns of play that may occur over time.
Results and discussion
3.1 Reliability Edita: Alto Rendimiento Three reliability tests were performed (T1, T2 and T3) to show how reliable the analyst is on two different components of the use of the analysing system. The two components are the selection of what type of shot a performer uses and which cell on the court the performer strikes the ball from. The figure below represents a squash court, displaying the total shots played from each individual cell for T1, T2 and T3.
Figure 1. Total shots played in each cell of the three tests.
Hughes (1995) system for squash was very reliable for the analysis of elite squash matches. The system requires the analyst to select the type of shot every time one is played by either performer. There are only six variations of shot options, which minimises the chances of making an error, thus increasing the reliability. For the shot selection to be deemed reliable, the chi square must produce (P>0.99). The P values for T1, T2 and T2, T3 were both greater than 0.99. The overall chances of making an error between all tests were minimal at less than 1.96%. The second selection the analyst has to make is the position on the court where the ball was struck with the racket by the players. The reliability test of this area produced P value figures above 0.9, which deems it as reliable. The P value for the test between T1 and T2 was 0.99, and 0.95 between T2 and T3. This showed that the results from the system are reliable. The percentage error overall for this test of shot selection was quite low, being 5.8% and 11.1%. The selection of a cell where the ball was struck can be difficult. This is because the court is divided up into 16 cells by the system, but not marked for the analyst to see. This means the analyst may perceive a shot to be in the front left cell, whereas next time it may be seen as being in the second row of cells.
Table 1. Total shot distribution for T1, T2 and T3.
Table 2. Reliability between tests of selection of shot by researcher for both players.
Table 3. Reliability of selection of cell as player strikes ball for both players.
3.2 General Data The data produced by the system were placed into Microsoft Excel to be analysed in greater detail. The first page of results which are produced by the system are the total amount of shots, total rallies and shots per rally for both players individually and collectively, as are total winners, forced errors and unforced errors. These figures were totalled for all matches under each scoring system. From these results an average amount of shots per rally was calculated for both systems showing very similar results. PPR to 15 averaged 12.67 shots per rally, with PPR to 11 averaging slightly higher with 12.99 (table 7). Performance indicators designed by Hughes and Bartlett (2002) were used to assess the changes in performance between the two scoring systems. The main performance indicators were the amount of attacking and defensive shots, shot selection, shot distribution, winners, errors and rally length. Attacking shots in squash are mainly linked with taking the ball early to put the opponent under pressure, which can be achieved by volleying the ball. The PPR to 11 shows 22% more short volleys played, compared to PPR to 15. There was also 18% more long volleys under PPR to 11, showing players are taking the ball earlier, placing their opponents under greater pressure. The amount of defensive shots, such as lobs were used 26% more in PPR to 11 (table 5). This could mean that the players are being placed under greater pressure so are required to use the lob more often to recover. The overall shot selection changed slightly under the new scoring system. PPR to 11 showed a greater number of straight shots, such as drives, lobs and short and long volleys (table 5). This showed that even though the game may be played at a higher pace, a safer game is adopted. A lower amount of cross-court shots were played, which may be due to the players not wanting to open the court up for the opponents.
The shot distribution does vary between PPR to 15 and PPR to 11. The backhand side of the court was used predominantly more than the forehand in both scoring systems, but the backhand back quarter is used more in PPR to 11 than PPR to 15 (figure 2). This may be due to the backhand side being the more consistent side for the players, thus reducing the chance of making an error or giving the opponent a chance. Using 10% change as significant, unforced errors decreased significantly, but the Mann-Whitney test produced the P value as 0.748 (table 7) showing no significance. Forced errors showed no significant differences. The amount of winners increased significantly under the new scoring system (% change >10), but the Mann-Whitney test produced no significant difference as (P>0.05). The last performance indicator stated by Hughes and Bartlett (2002) is rally length. Previous research by Hughes (1995) stated that the average length of a rally in PPR to 15 lasted 18 seconds, which differs from Hughes (1995) study, showing rallies lasting around 25 seconds. PPR to 11 rallies on average were lasting just below 26 seconds (table 10). This showed that the rallies are not shortened by the new scoring system. The idea that the rallies would be decreased due to the increase in attacking play has not worked. Instead, every rally has become more important, meaning that each player does not want to give away cheap points. In PPR to 15, the average number of shots played per game agrees with previous research by Hughes and Robertson (1998). There have been no significant changes in average amount of rallies per game or the average amount of shots per rally. The average time taken between rallies in PPR to 15 has increased from 10 to 11.55 seconds compared to Hughes and Robertson’s (1998) findings. This increase in rest may be attributed to average rally length being increased from 21 seconds to over 25 seconds. These sets of data compared to PPR to 11 are slightly different. The main difference between the two scoring systems is the average amount of rallies per game being cut from 28 to 18 (table 9). This is due to the game being cut by nearly a third, from 15 to 11. This in turn decreases the average number of shots per game and match. The average amount of shots per rally in PPR to 15 agrees with research by Hughes and Knight (1995), with PPR to 11 findings not producing significant changes. Average rally lengths show no significant changes between the two scoring systems. However, the average rest taken between rallies has increased from 11.55 seconds to 12.71 seconds in PPR to 11. Using 10% change as significant, this shows PPR to 11 creating a significant increase in rest time between rallies.
Table 4. Data regarding critical points.
Figure 2. Average shot distribution per 500 shots for both scoring systems.
This may be caused by the increase in intensity of the rallies. The overall average game time is 6 minutes less under PPR to 11, but the overall match averages are the same. This may be due to a number of factors, the first being two matches lasting in excess of 90 minutes in PPR to 11, compared to an average length of 56 minutes for the remaining four matches, thus causing the data to be skewed. A second factor is in the PPR to 15 matches there were only 22 games played in the six games, compared to 25 games in the six matches in PPR to 11. The average score of an elite match is 3-1 (Hughes, 1995), so extrapolated results were produced by multiplying the average game time by four. This produced results showing PPR to 15 matches lasted on average just under 75 minutes, as compared to 53 minutes under PPR to 11. This shows that the introduction of the new scoring system should create shorter match times, retaining audience attention.
The new scoring system was introduced to make the game more attractive to television audiences. One major failing in this respect has been the low number of critical points. Tennis produces a minimum of critical points every 3 minutes, which keeps the attention of its audience. This study has shown that a minimum number of critical points occurs every 17 minutes in PPR to 15, which agrees with Hughes (1995) and the new scoring system PPR to 11 every 11.24 minutes (table 4). The option to calculate critical points was decided post analysis, so a critical point has only been recorded at known game points in the matches. The main reason why pPr to 11 produces more critical points is the introduction of the two clear points should 10-all be reached. This is similar to ‘deuce’ in a game of tennis, where two points in a row are required to win the game. PPR to 11 has improved the frequency of critical points, but does still not create as many as in a match of tennis.
3.4 Analysis of Game Structure The results from this study are compared with results from Hughes’ (1995) research to make five different timeline graphs. These graphs represent a game of different scoring systems for squash and a game of tennis. The graphs are all proportional to each other, displaying the average length of rallies and rest time taken between rallies. For example, the thicker the bar, the longer the average rally and the thinner the space between the bars, the shorter the average rest period between rallies. A red line was inserted as a potential indicator of the excitement levels of the audience as the game progresses towards critical points. Figure 15 shows a timeline of an English scoring game to 9 points. This shows rallies lasting 18 seconds, rests between rallies lasting 10 seconds, with the game taking nearly 17 minutes (Hughes, 1995). The amount of rallies per game was decreased from 36 to 28, but the length of rallies were increased from 18 seconds to 25 seconds. As the gradient of the red line indicates, it takes a long time to reach the critical point in these games. Figure 7 represents the new PPR to 11 scoring system game. A clear decrease in overall game length increases the gradient of the excitement line, showing the critical points of the game are achieved quicker. The new scoring system was designed with the intention of creating a more attacking style of play, indicating shorter rallies. The length of the rallies in PPR to 11 has slightly increased by a second along with the rest between rallies. If the English scoring game and the two PPR scoring systems are compared to a game of tennis, the difference is obvious to see. A game of tennis (figure 9) shows a very steep excitement line, due to each game only averaging 3 minutes. This has been recognised by Hughes (1995) and analysis was performed on a tennis scoring game of squash (figure 8). The results from this show the rally length and rest between rallies was consistent with the PPR to 15 results (figure 15). The main difference was the length of the games being reduced to 3 minutes, which shows the critical points are reached quicker. The comparison between a game of tennis and a game of squash played under tennis scoring show major differences even though the games last the same amount of time. Tennis receives a large amount of television coverage, even though the rallies on average are only around 4 seconds, with 23 seconds rest between rallies. This compared to a game of squash under tennis scoring rules, which has rallies lasting 18 seconds, with only a 10 second rest between rallies. This should create a more exciting game than tennis for the television audiences.
From this study it may be concluded through the use of Mann-Whitney U test that PPR to 15 compared to PPR to 11 produces: 1. No significant difference in the amount of winners played. 2. No significant difference in the amount of unforced errors played. IV Congreso Mundial de Ciencia y Deportes de Raqueta 3. No significant difference in shot selection. 4. No significant difference in average shots per rally. But using 10% change as significant, it may be concluded that PPR to 15 compared to PPR to 11 scoring produces: 1. A significant increase in the amount of winners played. 2. A significant decrease in the amount of unforced errors played. 3. A significant difference in shot selection. 4. A significant increase in shot distribution from the back backhand quarter. 5. No significant difference in the average length of rallies. Finally, it may be stated that PPR to 11 scoring has produced significant changes in patterns of play, creating shorter games and increasing the amount of critical points compared to PPR to 15 scoring. 4.1 Further Research 1. A study in time analysis of each shot and rally. This would allow analysis of the intensity and speed of each shot in the game. 2. A different form of analysis that could be studied is physiological side of the players, which could be performed by recording the heart rates of the subjects. The performer’s heart rate can be analysed at critical points, such as game point or an argument with the referee, to see if any changes occur. 3. A study on the effects of playing patterns when the use of PPR to 11 scoring on a variation of playing standards. If this shows successful results, the standardisation of world squash may be considered.
- Brown, D., & Hughes, M.D. (1995). The effectiveness of quantitative & qualitative feedback on performance in squash. In Science and Racket Sports (edited by T. Reilly, M.D. Hughes and A. Lees), pp. 232-237, London: E. & F.N. Spon.
- Hughes, M.D. (1983). Using a computerised approach to Squash notational analysis. Proceedings of the British Association Sport & Science Conference. University of Liverpool.
- Hughes, M.D. (1984). Using a microcomputer for notational analysis in squash. Proceedings of the British Association Sport & Science conference. Bedford.
- Hughes, M.D. (1985). A comparison of the patterns of play of squash. In: I.D. Brown, R. Goldsmith, K. Coobes and M.A. Sinclair (Eds), International Ergonomics, pp. 139-141, London: E. & F.N. Spon.
- Hughes, M.D. (1995). Using notational analysis to create a more exciting scoring system for squash. In G. Atkinson and T. Reilly (Eds) Sport, Leisure and Ergonomics, pp. 243-247, London: E. & F.N. Spon.
- Hughes, M.D. and Knight, P. (1995). Playing patterns of elite squash players, using English and point-per-rally scoring. In T. Reilly, M. Hughes and A. Lees (Eds), Science and Racket Sports, pp. 257-259, London: E. & F.N. Spon.
- Hughes, M.D. and Robertson , C. (1998). Using computerised notational analysis to create a template for elite squash and its subsequent use in designing hand notation systems for player development. In Science and Racket Sports II, (edited by A. Lees, I. Maynard, M. Hughes and T. Reilly), pp. 227-234. London: E. & F.N. Spon.
- Hughes, M.D. and Sykes, I. (1994). Computerised notational analysis of the effects of the law changes in soccer upon patterns of play. Journal of Sports Science, 12, 180.
- Hughes, M.D., Cooper, S.M. and Nevill, A. (2002). Analysis procedures for non-parametric data from performance analysis. eIJPAS, 2, 6-20.
- Pritchard, S., Hughes, M. and Evans, S. (2001). Rule changes in elite badminton. In (eds M. Hughes and I.M. Franks) pass.com, Cardiff: CPA, UWIC, pp. 213-225.
- Sanderson, F.H. and Way, K.I.M. (1979). The development of objective methods of game analysis in squash rackets. British Journal of Sports Medicine. 11, 188.