O que os treinadores podem fazer? A relacao entre as substituicoes e o resultado de jogos profissionais de futebol.
The use of new technologies in football, such as notational analysis, data analysis software, Global Positioning System (GPS), new training materials and metabolic analysis, are essential to improve performance in the playing field. Indeed, performance indicators must be able to reveal essential aspects to coaches, such as interactions between two or more athletes, rehearsed plays, tactics, and techniques (Hughes and Bartlett, 2002).
In this context, predominant factors should be differentiated from determinant factors for the game result. We understand the determinant factor as the most important data to be collected in a game, such as goals scored, assists and substitutions. These variables may represent what really matters in understanding how a game was played.
The increase in the number of studies in the past few years has made possible unimagined advances in this field (Debanne, Laffaye and Trouilloud, 2017; Del Corral, Barros and Prieto-Rodriguez, 2008; Fernandez-Navarro and collaborators, 2016; Flores and Balsan, 2016; Njororai, 2013b). However, even today there are several gaps and questions to be answered on how different variables can influence teams' performances during competitions. The player substitution is one of them.
The substitution of a player during a match could be understood as a determinant factor to the final result (Bradley, Lago-Penas and Rey, 2014; Del Corral, Barros and Prieto-Rodriguez, 2008; Myers, 2012), and coaches need to be prepared to do it wisely. On a large scale and with the accumulation of games during the tournaments, all the substitutions (as determinants of the result) can influence performance. Different studies showed that players' performance are strongly influenced by the distance covered in the field (Barros and collaborators, 2007; Di Salvo and collaborators, 2007) and by a congested calendar (Lago-Penas and collaborators, 2011; Rey and collaborators, 2010).
According to Barros and collaborators (2007), players run, on average, 10 km per game. The number of matches and the small period for physical recovery add up to the time in the field, physically tiring the players. Thus, the substitution has great importance in short, medium and long-term, by improving the performance during the season.
The demand of new and better information and the need for actualization by coaches, players, and researchers, makes the research involving new visions on the understanding of the game, vital for the evolution of sport. The insert of a new player in the field through substitutions, together with the physical aspects, can be an essential tool providing tactical changes through the different characteristics of the available athletes. Although there are some studies addressing the influence of substitutions on football (Bradley, Lago-Penas and Rey, 2014; Conte and collaborators, 2015; Del Corral, Barros and Prieto-Rodriguez, 2008; Myers, 2012), we did not find studies that compare and analyze the influence of substitutions in Championship and Cup games.
The present paper intended to examine how the replacement of a player can be a determinant factor for the outcome of professional football matches.
Thereby, the aim was to analyze the relation between player substitutions and matches' final results in Championships and National Cups. It is expected that player replacement influences the number of goals scored and final results, and that substitutions tend to be carried out differently in different types of competitions.
MATERIALS AND METHODS
The data include a sample of the sixteen major leagues of world football. For the analysis, 258 professional football matches from different competitions were observed (Argentina, Belgium, Brazil, Colombia, England, France, Greece, Italy, Netherlands, Portugal, Scotland, Spain, Turkey, Ukraine, United States, and Uruguay). It was observed, on average, seven matches from each country (see the teams at supplementary data).
A sample of 129 matches from sixteen national Championships and 129 of sixteen national Cups were analyzed. During the championships, 737 player substitutions were made (a mean of 5.71 per match) and 511 goals were scored (a mean of 3.95 per match) while in the national cups, there were 721 substitutions (a mean of 5.58 per match) and 390 goals scored (a mean of 3.01 per match).
The sample included matches from the 2016-2017 season, and the data were collected intentionally, according to the following criteria: i) the analyzed games were selected in order to choose games at the beginning, middle and end of the competitions (reducing the influence of the number of games played); ii) matches or teams without substitutions were not analyzed; iii) to reduce the influence of the quality of the teams, two matches of each team, at most, were analyzed; iv) in order to control the individual influences (i.e., teams or players with greater technical or financial conditions), only games played on equal terms between teams were analyzed (i.e., games between the first and last places in the classification table were not analyzed); v) all the cup games were in the playoff system, choosing randomly between games at home or away. Other studies were conducted similarly (Aquino and collaborators, 2017; Conte and collaborators, 2015; Myers, 2012).
The match reports were analyzed to collect data. The data were provided by Wyscout [R], which is a well known and renowned system used by many teams around the world. Two trained observers collected the data. Both are football analysts who have been working with football for at least two years. Different studies also chose to carry out data collection in a similar way (Azli and collaborators, 2016; Castellano, Casamichana and Lago, 2012; Flores and Balsan, 2016; Hughes and Franks, 2005; Lago-Ballesteros Lago-Penas, 2010; Njororai, 2013a; Silveira and Flores, 2018). This research was conducted according to the Helsinki Protocol, and since the data is available in match reports and internet websites, the Statement of Consent was not required.
Descriptive analysis with mean scores and the standard deviation was used to characterize the data. The Shapiro-Wilk test confirmed the data normality and all statistical assumptions underlying ANOVA were met. After removing outliers, ANOVA was used to compare the Championships and National Cups. For follow-up analyses, we used Tukey posthoc test. The Software Statistical Package for Social Sciences TM (SPSS 22.0) was used and, in all analysis, the alpha level was set at 05.
In Championships, the substitute players scored 10,76% of all goals, representing 26,19% of the goals scored after the first substitution. In the National Cups, the substitute player was responsible for 10% of the goals scored, which represents 24,84% of the goals scored after the first substitution. Statistical differences were found when comparing goals scored after the substitutions in championships and national cups (p=0,000).
Teams playing in championships had more goals scored after the first substitution (F=6,223; p=0,013), after third substitution (F=7,038; p=0,008), and after all substitutions (F=24,375; p=0,000). No statistically significant differences were found after the second substitution, or before the first substitution.
When analyzing the championships, comparing win at home games and win away games, the results showed a statistical difference between goals scored before and after the substitutions (p=0,000), the same occurring in National Cups (p=0,000). Also, when each substitution was analyzed separately, results showed that the first one is more important when the team plays at home (p=0,009).
This paper aimed to enhance the professional football understanding by analyzing the relation between the player substitutions and the results of the football matches in Championships and National Cups.
The results show that when the coaches make good substitutions, it may change the course of the match. In other words, the staff of professional football teams must be prepared to change players when the game result is not as expected. Moreover, the results showed that substitutions could be understood as a Determinant Factor for goals scored and game results, mainly in Championships.
The analysis of the matches indicated that the first substitution is the most influent for a higher number of goals scored in comparison to the second and the third, both in Championships and in National Cups, with statistical differences (p=0.013). It was also noted that the second substitution had less influence on the results of the matches. The reason for this result can be explained due to the longer time to participate in the game. Our findings showed that during the championships the first substitutions allowed, on average, 15.08 and 25.14 more minutes to play than the second and the third substitution did, respectively. Also, the same pattern occur during the national cups (12.33 and 21.65 more minutes to play than the second and third substitutions, respectively).
The outcomes showed that when the home teams win a match, they make better use of substitutions, scoring more goals directly influenced by them (p=0.000). Despite that, few researches aimed to evaluate the importance of a player substitution (Bradley, Lago-Penas and Rey, 2014; Del Corral, Barros and Prieto-Rodriguez, 2008; Myers, 2012).
According to Frencken, Lemmink, Delleman, and Visscher (2011) during offensive moves, the attackers increase the distances traveled to make moves by opening spaces between markers. This strategy results in more fatigue and, consequently, the attackers are the first choice to be replaced in a 90-minute match. Bradley, Lago-Penas e Rey (2014) observed that the entry of a new player rises the levels of speed, intensity, and the distance covered. In general, the substitution could enhance the teams' performance, which were find in the present study.
The analysis of substitutions was also observed in other team sports (Messersmith and Corey, 1931; Morato and collaborators, 2016; Prudente, 2006). Gomez and collaborators (2016) showed that the substitution in basketball is a fundamental process to optimize the teams' performance. The authors found that the team enhances the performance after replacing a player.
According to Ronglan, Raastad and Borgesen (2006) in handball, the use of substitutions occur in the same way. The replacement of a player may be made at any time, so this allows an increase in intensity. Due to the high intensity and the inherent characteristics of the football game, the number of substitutions is an important question that should be more discussed in the literature.
Del Corral, Barros and Prieto Rodriguez (2008), state that the most crucial factor for the first substitution is the game score before it, thus, the preferable substitution is by an attacker. It is shown in the literature that changing a player creates a new fact at the match, in which the new player increases the odds to score more goals (Barros and Prieto-Rodriguez, 2008; Bradley and Lago-Penas and Rey, 2014; Del Corral and Myers, 2012; Gomez and collaborators, 2016). The same was found in the present paper.
According to Conte and collaborators (2015), during a football game, many variables can influence the substitution of a player. The authors explain that one of the main reasons is to make a tactical change. Although the present study did not evaluate the main reasons for the substitutions, we agree with Conte and collaborators (2015).
Other studies corroborate with the authors' findings (Del Corral, Barros and Prieto-Rodriguez, 2008; Jankovic and Leontijevic, 2006; Rey, Lago-Ballesteros and Padron-Cabo, 2015). In addition, according to Jankovic and Leontijevic (2006), coaches prefer to change players using attackers, instead of players from other positions. Thus, to Jankovic and Leontijevic (2006) this fact confirms that coaches influenced the increasing of efficiency by tactical changes using substitutions.
To Myers (2012), when the team is losing a match, the substitutions must be made at specific periods of the game. The results showed that these should be done before the minutes 12, 27, and 33 of the second half of the match. However, if the match is being won or is a draw situation, there is no rule to follow.
To Anderson and Sally (2013), when the performance of the regular players begins to decline, it is necessary that something be done and then, the most correct attitude is to carry out the substitution. Similar results were found in the present paper (see table 2).
Gomez, Lago-Penas and Owen (2016), analysing the Spanish football league (La Liga BBVA) found that most of first and second substitutions occurred between 61 and 90 minute period, and the third occurred during the 76 and 90 minute. The results of the present study corroborate to the authors' results (see table 5). Thus, our data together with those provided by Gomez, Lago-Penas and Owen (2016) and Myers (2012) show that an effective substitution strategy may booster the team's performance especially during championship tournaments.
Myers (2012), also pointed out that in 40% of the opportunities that coaches followed these instructions, the teams improved their field performance. On the other hand, it was also verified that coaches who did not follow these instructions, obtained the expected results only in 22% of the matches. With the evolution of sport and sports' analysis, rules and game strategies are also being improved. In this sense, the International Board and FIFA adopted the fourth substitution, when the match has overtime (FIFA, 2017). This alteration can be beneficial to the sport because they can reduce the levels of fatigue and improve players' performance.
According to our results, the substitutions are essential in professional football matches. Although there are differences between how they are used in Championships and National Cups, the results show that they follow a pattern and, in most cases, the substitutions of a player in Championships has a more significant direct effect in goals scored than in National Cups. Therefore, when performed correctly, and at the appropriate time, substitution may be a factor influencing sports performance (Wilson, 2013).
The findings of the present research add new information to the literature and have practical applications for professionals who work with football. As far as we know, there is no study designed to analyze differences between championship and national cup tournaments. Therefore, knowing that are significant differences between the tournaments, these results can help coaches and sport analysts during training sessions and preparation to the games. Hence, we showed that the substitutions follow a pattern and they need to be understood as a determinant factor during professional football games.
This research showed the importance that the players' substitutions have in modern football, being an indispensable strategy to enhance performance.
When well used, it can change the performance and results of matches, as well as influence the rest of the season, especially in championship tournaments.
Although the information contained in this manuscript is of paramount importance for high-performance professional practice and the study of football, some limitations need to be considered when interpreting the findings.
First, the sample size in each analyzed country was small--this fact made it difficult to compare countries.
Second, we do not know what would have happened if the substitution was not made. In addition, due to the nature of the study, we cannot analyze the influence of the substitute player during the match in other important plays, such as the construction of tactical movements or assistances for goals.
Finally, the nature of the substitution was not observed, so, we cannot verify the main reasons that led to player substitution.
Despite that, future researches should seek to understand how substitutions are made in matches played at high altitude and in games where heat and cold predominate, as well as analyze the influence of the player who entered, if he participated effectively in the plays that resulted in goals. Moreover, analyzing the negative aspects of the substitution (i.e.: ball loses, fouls committed, goals suffered, etc.) may be fundamental to future research. Understanding how substitution patterns are made, in different situations, can aid in the organization of training cycles and pre-game preparation.
(1)-Anderson, C.; Sally, D. The numbers game: why everything you know about football is wrong. Penguin UK. 2013.
(2)-Aquino, R.; and collaborators. Effects of match situational variables on possession: The case of England Premier League season 2015/16. Motriz: Revista de Educacao Fisica. Vol. 23. Num. 3. 2017.
(3)-Azli, M.; and collaborators. Changing the ratio of area per player during small-sided football games shapes performance. In: Favero, T.; Drust, B.; and collaborators (Ed.). International research in Science and Football II: Routledge. 2016.
(4)-Barros, R. M.; and collaborators. Analysis of the distances covered by first division Brazilian football players obtained with an automatic tracking method. Journal of sports science & medicine. Vol. 6. Num. 2. p. 233. 2007.
(5)-Bradley, P.; Lago-Penas, C.; Rey, E. Evaluation of the match performances of substitution players in elite football. International journal of sports physiology and performance. Vol. 9. Num. 3. p. 415-424. 2014.
(6)-Castellano, J.; Casamichana, D.; Lago, C. The use of match statistics that discriminate between successful and unsuccessful football teams. Journal of human kinetics. Vol. 31. p. 137-147. 2012.
(7)-Conte, D.; and collaborators. Substitution patterns and analysis in men's Division I college football. In: Favero, T.; Drust, B.; and collaborators (Ed.). International Research in Science and Football II: Routledge. p.44-52. 2015.
(8)-Debanne, T.; Laffaye, G.; Trouilloud, D. Motivational orientations and performance in penalty throws during elite male team handball games. Scandinavian journal of medicine & science in sports. 2017.
(9)-Del Corral, J.; Barros, C. P.; Prieto-Rodriguez, J. The determinants of football player substitutions: a survival analysis of the Spanish football league. Journal of Sports Economics. Vol. 9. Num. 2. p. 160-172. 2008.
(10)-Di Salvo, V.; and collaborators. Performance characteristics according to playing position in elite football. International journal of sports medicine. Vol. 28. Num. 3. p. 222-227. 2007.
(11)-Fernandez-Navarro, J.; and collaborators. Attacking and defensive styles of play in football: analysis of Spanish and English elite teams. Journal of sports sciences. Vol. 34. Num. 24. p. 2195-2204. 2016.
(12)-FIFA. Regulations. FIFA Confederations Cup--Russia 2017. Switzerland. 2017. Disponivel em: <https://resources.fifa.com/mm/document/tour nament/competition/02/85/41/57/fcc2017regula tionsen_neutral.pdf>. Acesso em: 01/07/2018.
(13)-Flores, F. S.; Balsan, L. A. G. Scout analysis of football: new look on the Brazilian Championship. International Journal of Sports Science. Vol. 6. Num. 3. 2016.
(14)-Frencken, W.; and collaborators. Oscillations of centroid position and surface area of football teams in small-sided games. European Journal of Sport Science. Vol. 11. Num. 4. p. 215-223. 2011.
(15)-Gomez, M. A.; Lago-Penas, C.; Owen, L. A. The influence of substitutions on elite football teams' performance. International Journal of Performance Analysis in Sport. Vol. 16. Num. 2. p. 553-568. 2016.
(16)-Gomez, M. A.; and collaborators. Exploring the effects of substituting basketball players in high-level teams. Journal of sports sciences. Vol. 35. Num. 3. p. 247-254. 2016.
(17)-Hughes, M.; Bartlett, R. The use of performance indicators in performance analysis. Journal of Sports Sciences. Vol. 20. Num. 10. p. 739-754. 2002. Disponivel em: <https://doi.org/10.1080/026404102320675602 >.
(18)-Hughes, M.; Franks, I. Analysis of passing sequences, shots and goals in football. Journal of Sports Sciences. Vol. 23. Num. 5. p. 509-514. 2005.
(19)-Jankovic, A.; Leontijevic, B. Substitution of players in function of efficiency increase of tactic play plan in football. Fizicka kultura. Vol. 60. Num. 2. p. 165-172. 2006.
(20)-Lago-Ballesteros, J.; Lago-Penas, C. Performance in team sports: Identifying the keys to success in football. Journal of Human Kinetics. Vol. 25. p. 85-91. 2010.
(21)-Lago-Penas, C.; and collaborators. The influence of a congested calendar on physical performance in elite football. The Journal of Strength & Conditioning Research. Vol. 25. Num. 8. p. 2111-2117. 2011.
(22)-Messersmith, L. L.; Corey, S. M. The Distance Traversed by a Basketball Player. Research Quarterly. American Physical Education Association. Vol. 2. Num. 2. p. 57-60. 1931. Disponivel em: <http://www.tandfonline.com/doi/abs/10.1080/2 3267402.1931.10625026>.
(23)-Morato, M. P.; and collaborators. Development and evaluation of an observational system for goalball match analysis. Revista Brasileira de Ciencias do Esporte. 2016.
(24)-Myers, B. R. A proposed decision rule for the timing of football substitutions. Journal of Quantitative Analysis in Sports. Vol. 8. Num. 1. 2012.
(25)-Njororai, W. Analysis of goals scored in the 2010 world cup football tournament held in South Africa. Journal of Physical Education and Sport. Vol. 13. Num.1. p. 6. 2013a.
(26)-Njororai, W. Downward Trend of Goal Scoring in World Cup Football Tournaments (1930 to 2010). Journal of Coaching Education. Vol. 6. Num. 1. p. 111-120. 2013b.
(27)-Prudente, J. Analise da performance tactico-tecnica no Andebol de alto nivel: estudo das accoes ofensivas com recurso a analise sequencial. 2006.
(28)-Rey, E.; Lago-Ballesteros, J.; Padron-Cabo, A. Timing and tactical analysis of player substitutions in the UEFA Champions League. International Journal of Performance Analysis in Sport. Vol. 15. Num. 3. p. 840-850. 2015.
(29)-Rey, E.; and collaborators. The effect of a congested fixture period on the activity of elite football players Biology of Sport. Vol. 27. Num. 3. 2010.
(30)-Ronglan, L.; Raastad, T.; Borgesen, A. Neuromuscular fatigue and recovery in elite female handball players. Scandinavian journal of medicine & science in sports. Vol. 16. Num. 4. p. 267-273. 2006.
(31)-Silveira, T. C. W.; Flores, F. S. Analise dos gols sofridos no campeonato brasileiro feminino de futebol. Revista Brasileira de Futsal e Futebol. Vol. 10. Num. 27. 2018. Available in: <http://www.rbff.com.br/index.php/rbff/article/vi ew/577>
(32)-Wilson, J. Inverting the pyramid: the history of football tactics. Nation Books. 2013.
Received for publication in 11/27/2018
Accept in 01/06/2019
Fabio Saraiva Flores [1,2]
Daniela Lopes dos Santos 
Guilherme Rosso Carlson 
Eugenio Zanette Gelain 
-Universidade de Lisboa. Portugal.
-Universidade Federal de Santa Maria (UFSM), Santa Maria-RS. Brasil.
E-mails dos autores:
Table 1--Percentage and average of total goals, before and after substitutions. Championships Total Goals before Goals after first substitutions substitution n 511 301 77 % 100 58,90 15,07 Mean 3,96 2,33 0,60 National Cups Total Goals before Goals after first substitutions substitution n 390 233 52 % 100 59,74 13,33 Mean 3,02 1,81 0,40 Championships Goals after second Goals after third Goals after substitution substitution substitutions n 62 71 210 % 12,13 13,89 41,10 Mean 0,48 0,55 1,63 National Cups Goals after second Goals after third Goals after substitution substitution substitutions n 56 49 157 % 14,36 12,56 40,26 Mean 0,43 0,38 1,22 Table 2--Percentage and average of goals, before and after the substitutions, according to the result of the matches of the Championships. Total Goals before Goals after first substitutions substitution Wins at n 231 127 45 home % 100 54,98 19,48 Mean 3,92 2,15 0,76 Wins away n 169 105 16 % 100 62,13 9,47 Mean 2,86 1,78 0,27 Draws n 108 64 16 % 100 59,2 14,81 Mean 1,83 1,08 0,27 Goals after second Goals after third substitution substitution Wins at n 25 34 home % 10,82 14,72 Mean 0,42 0,58 Wins away n 25 23 % 14,79 13,61 Mean 0,42 0,39 Draws n 13 15 % 12,04 13,89 Mean 0,22 0,25 Goals after substitutions Wins at n 104 home % 45,02 Mean 1,76 Wins away n 64 % 37,87 Mean 1,08 Draws n 44 % 40,74 Mean 0,75 Table 3--Percentage and average of goals, before and after the substitutions, according to the final result of the matches of the National Cups. Total Goals before Goals after first substitutions substitutions Wins at home n 177 90 28 % 100 50,85 15,82 Mean 1,37 0,70 0,22 Wins away n 125 78 20 % 100 62,40 16 Mean 0,97 0,60 0,16 Draws n 44 25 5 % 100 56,82 11,36 Mean 0,34 0,19 0,04 Goals after second Goals after third substitutions substitutions Wins at home n 29 30 % 16,38 16,95 Mean 0,22 0,23 Wins away n 13 14 % 10,40 11,20 Mean 0,10 0,11 Draws n 10 4 % 22,73 9,09 Mean 0,08 0,03 Goals after substitutions Wins at home n 87 % 49,15 Mean 0,67 Wins away n 47 % 37,60 Mean 0,36 Draws n 19 % 43,18 Mean 0,15 Table 4--Goals scored by the substitute player. Championships Wins at Wins Draws Total home away n 21 14 9 55 Goals scored by substitute 20,19 21,88 20,45 26,19 player after substitutions Goals scored by substitute 9,09 8,28 8,33 10,76 player (%) National Cups Wins at Wins Draws Total home away n 19 9 6 39 Goals scored by substitute 21,84 19,15 31,58 24,84 player after substitutions Goals scored by substitute 4,87 2,31 1,54 10 player (%) Table 5--Goals scored with substituition effect, comparing National Cups and Championships Substitution Championships Average time (min) Goals % First 55,26 77 36,67 Second 70,34 62 29,52 Third 80,40 71 33,81 Mean 68,66 70 100 Total -- 210 -- Substitution National Cups Average time (min) Goals % First 57,67 52 33,12 Second 70,00 56 35,67 Third 79,32 49 31,21 Mean 69,00 52,3 100 Total -- 157 --