This is the second guest article from Olivier DEGRENNE, who works in the Sports Science department at University Paris-East Creteil.
You can read his first piece here: How Can Video Analysis Improve Performance in Sport?
Competition is fierce between the different video analysis software providers in the market, with each solution having its own strengths and weaknesses. From high end solutions for professional sport, to more affordable software for amateurs, to biomechanical analysis software, individuals, teams, and organisations must make choices, which are often dictated by their purchasing power.
The purpose of this article is not to promote one software over another, but rather to explain that it is a tool to support whichever video analysis approach that those in the sports world wish to use. Throughout this article, we will switch between considering the technical elements of software and other more macro-level considerations related to the process of observing and understanding performance. We will therefore go through each step involved in video analysis, from the creation of a video file to the editing of clips for the athletes, to provide us with a common thread and logic to follow.
We will rely on BEPRO's Editor software (available through their Basic Package), to illustrate the different stages in the analysis process. We chose this software because of its accessibility (500 €/£/$ a year for a team), its versatility (it can be used online or offline and on Windows or MacOS) and its ease of use.
Creating a Video File to Analyse
This process is often the same for different solutions. When returning from a match or competition, you need to remove the memory card from the camera and upload the video onto your computer. Once the file is on the computer, we can launch the analysis software and create a project or package depending on the solution (there will often be software-specific terminologies).
The software interfaces are often similar. They are presented in the form of a video viewer, a timeline under which the clips will be displayed, by type, and a pane on the side that allows you to create your own buttons (these are the different names of clips that you will add to the timeline during your analysis).
For example, after performing these initial actions to set up analysis for a football match video, a critical question arrises – what do I do now? Should I watch the entire video?Be careful, there’s a rule that says, “if I look at everything, I observe nothing”! Observing is a cognitive process that allows us to analyse our surrounding environment. Observation is different from contemplation in that it invites us to understand the interactions between a subject and its environment. Therefore, there are two central questions that should then be asked to help structure our analysis:
· What should be observed?
· To what level of precision?
What Should be Observed?
When we analyse a sports practice, we must observe it with precision and identify the observables we wish to focus on, or in other words the indicators which we will be looking for. For each indicator, it is necessary to identify the qualitative variables that will allow us to judge the actions being performed, which are known as success criteria.
For example, when we make a pass in football, the pass is the indicator that is being observed. We will then look at the direction, the type of pass, its result, etc., which are the success criteria (qualitative variables) that we will assess to inform our judgment of the pass.
To What Level of Precision?
Firstly, you will have to create the buttons (clip types) you will use to breakdown and sequence the full video. Depending on the software, we can either create “event” buttons specific to each criterion (for example a “pass successful” button and a “missed pass” button) or create a general “event” button and descriptor buttons, or labels, which will describe the performance of an action (for example “pass + successful” or “pass + missed”). The difference lies in the level of analysis. For example, coding all the passes will allow me to compare them all, maybe depending on the result. However, if I code each type of pass separately, the comparison will be more complicated since they will no longer belong to the same action category.
Using an “event” button will create a clip which will have a beginning and an end on the timeline. We can then choose to create very short clips (all the shots of a handball match for example) or longer clips (each phase of possession of the ball by a team during a football match). The analyst is the master but must be guided by a principle called the “Atomicity Principle” (Davidson, 1996): it is impossible to analyse at a lower level than that at which we are coding.
In short, all approaches are good providing they are guided by an analytical objective.
Once the analysis project is identified and the buttons created, sequencing can begin. We can then play the video and click on the different buttons (or use keyboard shortcuts according to the preferred working methodology) to create the clips. For example, as each pass is observed, a new clip can be added to the timeline.
If an error is made, it is possible to go back and delete the wrong sequence to replace it with another. This is the great advantage of post-match analysis. Once the match video has been fully analysed and all the clips have been created, the real job begins.
All this work only makes sense if it aims to improve an athletes’ performance. We have several options. We can produce a statistical report that allows the athlete to evaluate their performance, but be careful not to fall into the trap of providing meaningless statistical data! Indeed, performance evaluation, whether technical or tactical, is rarely binary, that is, either all good or all bad. It will therefore be necessary to deepen the analyses by going beyond the usual “shots on target / shots off target”. We may also make video highlights to explain or demonstrate the athlete's performance. But what should be shown? The positive (to reinforce good behaviours), the negative (to correct errors) or both, and in what order? We could use the 'sandwich' method (to put a criticism in between two compliments) so that the athlete or team remains open to corrections without getting the impression of having done everything wrong. Unfortunately, there is no single right answer. The transmission of information must be done as in all human relationships, using language that can be understood by the recipient. It will then be a question of them understanding what they did and evaluating their own performance to help them improve. However, this subject is very broad and will need to be discussed in another article.
Sharing data and video together seems to be the best way to present information. The data makes it possible to point out what to correct, to continue or to stop doing. The video then provides a better understanding of what the data has highlighted.
After performing the statistical analysis, we can start editing the video that will add visual context to the data presented. Note that a video debriefing must be brief and precise. For optimal effectiveness, debriefing should ideally take place the day after the performance and last less than 20 minutes (Wright et al., 2012). When playing videos, it is best to go back and forth between the video elements and verbal feedback. Indeed, it has been shown that a player retains less than 20% of the elements presented during a video session(Middlemas et al., 2017). This can be explained by the fact that the analyst, or coach, has access to an immense amount of information via video and tends to want to transmit a large amount of information to his athletes. Longer video sessions and a lack of clarity in take-home messages can lead to information overload (Nicholls et al., 2018).
It is therefore imperative to structure the debriefing around a clear message by identifying some key videos that would constitute the future basis for work. Be careful, a 20-second video played once at real speed, once in slow motion (x0.5) and once at real speed with 5 freezes of 5 seconds becomes a sequence that lasts 1 minute and 45 seconds! It is therefore important to select the right sequences and show them in the right order to ensure effective communication.
The Choice of Example Clips
To maintain efficiency and to keep the attention of athletes, it is preferable to avoid “funny” clips or examples that are too extremely positive or negative. The choice of the right sequence is not always obvious, especially when you want to highlight a situation that does not happen very often. Ideally, it is good to include sequences that include details that you will be able to analyse in more depth using your video drawing tool.
The Video Drawing Tool
Sometimes referred to as “telestration”, this process consists of integrating animations, drawings, the use of slow motion and more to highlight the key points to observe. When presenting images to athletes, focus on simple and effective analyses with a few key elements highlighted to support them in the observation. The video presentation aims to deliver a clear and precise message. Animations should be used to clarify the message and facilitate the processing of information. Indeed, the video tool requires the use of short-term memory. If the message is supplemented with too much information, there’s a risk of overloading short-term memory since it has limited processing capacity (Sweller, 1988). It is therefore important to limit stimuli to focus athletes' attention on the key aim of the message.
Editing the Video
Once the video has been fully edited, we now need to decide what we want to share as a final presentation. It is then a question of choosing the order of diffusion of the images. Few rules exist in the literature on this topic, but we nevertheless hypothesise that the use of a “zoom” approach is enough to highlight the elements to be observed. It is then a question of starting from a classic vide oat real speed and then refining the playback to focus the eye on increasingly precise elements using the features offered by the various solutions (zoom, drawings, slow motion, freeze-frame, etc.).
Finally, almost all software allows you to add title thumbnails between videos. This feature makes it possible to clarify the focus for the upcoming observation and to make a clear transition between the different themes being addressed during the video feedback session. Nevertheless, it must be kept in mind that an effective video session should have a limited number of themes to optimise the message.
The purpose of this article is to provide a guideline for using video tools to support debriefings. The approach is in no way prescriptive. The purpose here was to identify the knowledge (practical and theoretical) that are useful to optimise the handling of the tools and the management of video feedback sessions. Given the time required to analyse a full video, it is vital to make every effort to ensure the edited video is as effective as possible.
For this we followed the process of using video analysis software (here we used BEPRO Editor), from the importing of the video to the exporting of the final presentation content. We stressed the importance of being precise when analysing the full video to create clips and the need to convey a clear message when building the content to improve the effectiveness of video debriefing.
For anyone keen to try video analysis for themselves, our Basic Package is the perfect place to begin.
Davidson, F.(1996) Principles of Statistical Data Handling, Thousand Oaks, CA: Sage Publications.
Middlemas, S.G., Croft, H. G., & Watson, F. (2017). Behind closed doors: The role of debriefing and feedback in a professional rugby team. International Journal of Sports Science andCoaching,13(2), 201–212.
Nicholls, S. B., James, N., Bryant, E., & Wells,J. (2018). Elite coaches’ use and engagement with performance analysis within Olympic and Paralympic sport. International Journal of PerformanceAnalysis in Sport, 18(5), 764-779.
Wright, C., Atkins, S., & Jones, B. (2012). An analysis of elite coaches’ engagement with performance analysis services (match, notational analysis and technique analysis). International Journal of Performance Analysis in Sport,12(2),436–451.