By Carl Valle
After thousands of people read the article on mistakes with video analysis in sport, I was flooded with a rush of requests asking how to do it correctly. Video analysis is a truly powerful tool, but it’s not appropriate to use for everything and you should only use it for specific reasons. If you’re developing athlete speed, be it a soccer athlete or a sprinter in track and field, video can help any athlete at any level.
Everyone can extract information and improve athletes with video, but it does require a coach to know what to do after they see the problem. In this article, I cover just the tip of the iceberg: how to do some simple analysis that will help a team manage large groups of athletes or take one athlete and push them to the limits with performance.
What Video Analysis Can Do and What Video Analysis Should Not Do
Coaches have a limited amount of time, as a successful program requires coaching, game or meet management, recruiting, and even fundraising. Team coaches and performance staff simply can’t spend all day behind a computer breaking down video, and as someone who does sports technology, I would rather not be glued to a screen either.
Most of the effort with video should be to make changes that are realistically possible later in training. It make no sense to analyze every step and every lift if you can’t use the information productively or address it later. Coaches should think “minimum effective analysis” when using video, as an athlete can only absorb so much information in training. Otherwise, we see the “paralysis by analysis” syndrome surface.
Here is what video does well for athletes:
- Allows an athlete and coach to share the same view or perspective together, literally getting them on the same visual page together.
- Captures specific measurements that you must modulate to improve performance in an objective and finite way.
- Discovers truths in training that you couldn’t extract by any other means with movement, skill, or sporting technique.
- Teaches how other performers execute a movement and build a movement strategy that may be effective for the athlete.
- Audits the consistency and the development process of training compared to competition.
The list above is not exhaustive or complete by any means, but just working with those few is enough to close the gap or widen the margin of victory. My problem with video analysis is that measuring for the sake of measuring isn’t a sound approach—it’s just busy work. If a youth athlete has poor knee lift, measuring the angle only adds unnecessary detail to the problem; it may be a symptom of something unrelated to technique.Knowing what tool is best for a job beats using one tool to do everything in a mediocre fashion, says @spikesonly. Click To Tweet
Knowing when NOT to video is just as important as knowing how to break down recordings. Any time a coach shares a video, they put a potential fool’s errand under the microscope if the analysis is arbitrary. Over and over again, I see coaches measuring angles that don’t mean much. Instead, they should be spending time on simpler analysis that gives instant benefit to both coach and athlete.
It’s valuable to capture splits of a race, but if you are getting splits in practice, you are wasting hours for little reward because electronic timing does this instantly. Never perform manual video analysis when a technology can automate the process. Biomechanical analysis from video can be incredibly precise, but only in snapshots of time as camera angles can only tease out so much. If you can glean one lesson here, knowing what tool works best for the job is better than knowing how to do everything in a mediocre fashion with one tool.
Why Joint Degree Measures Need Time and Distance Data
Body angles and takeoff angles are instrumental in knowing how elite performances are created, and those in research that provide such information deserve to be in a special place, in my book. Those measures collected by sport scientists take an enormous amount of time to acquire, and some of them are not really coachable. Indirectly, most measures will fall into place if the training is progressive and sound, and video can sometimes misguide a coach. Still, video is the ultimate truth between coaches and athletes, as the reference is not debatable. The retort of coaches bragging about their eyes and what they see in training is beyond old; it’s a way to hide from the reality of being accountable.
As mentioned in the last article on video analysis, time and space are the dual principles of sport. What you can do with the given time and area is what usually separates the good from the great. A world-class sprinter can cover 100 meters under 10 seconds, and a pitcher can throw 100 miles an hour. A basketball or soccer player can move in a small space with amazing elusiveness, but it requires video to understand how. All the sporting events have a time and space element, and video transforms the concept to a working solution for coaches.
My biggest frustration is seeing data that is ornamental, and that coaches can’t use immediately. Sometimes, data or video information is there to create awareness and that is fine, but if a coach doesn’t have a plan of action with the information, it’s just fantasy.
The best example of performance analysis with video is the hurdle events. Coaches can spend time looking at knee and takeoff angles all day, but those are descriptions and not useful for developing an athlete on the track. Asking an athlete to take off at a specific angle never happens, as the world of sport is happening so fast you can’t expect them to be able to address angles. What you can do is ask an athlete to take off earlier or farther away, and to stay closer to the hurdle so they save time.
We already know what the best in the world do, thanks to years of video analysis; coaches just need to know if their own athletes are using similar strategies. If a hurdler can run faster between the hurdles while having the same advantage over them, they will win every single time. Sprint coaches are nothing more than sculptors, removing time from an athlete’s performance little by little until the end is a finished product. Field coaches are similar, yet they focus on creating more distance with what resources they have, either vertically or horizontally.
Modeling Successful Performance and Key Performance Indicators
Some coaches think “key performance indicator” (KPI) is a buzzword, and the truth is that they are right. For the most part, it’s easy to list KPIs but harder to actually apply them. When I work with athletes and other coaches, my job is usually to do performance analysis and summarize what’s holding them back. I usually perform dozens of tests, ranging from biochemical analysis to biomechanical evaluation, all building a case for what my summary will suggest.
Nearly all of what I do is unlocking barriers in speed, and without a model that has solid feedback, suggestions are just a soft science. If I had to simplify what I do, I find out how much an intervention is truly worth in each circumstance. Why fight for a mythical hundredth of a second when a tenth is more easily found? Video analysis is often the guiding beacon for getting faster.
I wrote about modeling on swimming, but most sports on land care about leg speed on the ground. A good model is one that has enough detail and firm enough anchors with KPIs that a change in one variable for the better will likely lead to a better performance soon after. If you improve one area with an athlete and it rarely improves their performance, the efforts are misguided.Video is the ultimate truth between coaches and athletes, as the reference is not debatable, says @spikesonly. Click To Tweet
Nearly all the mistakes I see try to add up small “marginal gains” and hope the performance changes down the road. It’s better to look for “maximal gains” in simple areas first, and then look for the secondary factors. Easy wins or easy gains come from the more plastic areas in an athlete. I wasted years chasing unicorns when a good horse would have been a better use of my time. I have stripped down my workouts to what I used in the 1990s, just with a better sense of dosage and sequence.
For years, I followed different mentors in looking at very esoteric adaptations. But I really needed less neurophysiology and just the simple knowledge of what workouts helped get an athlete faster from point A to point B. Instead of making things complicated, know when to pick your battles over something less than trivial that needs to be put away for the sport scientists later.
My favorite example of modeling is college sprinting, namely because the athlete is young enough to see a lot of improvement, but mature enough to think about what I am saying. I did flying sprint tests with most of the teams I worked with, and all of the athletes that recorded times had improvement. Those that didn’t struggled for obvious reasons—they didn’t know the nature of the problem, so solutions were not successful.
How fast an athlete can sprint in practice and in meets can tell a coach what needs to change. An athlete who is in great shape, finishes strong, and is consistent usually has a speed issue if they are not winning. It makes no sense to try to squeeze more fitness out of them if they run a perfect race and are not in it to win it halfway through. Conversely, an athlete who has great speed but struggles out of the blocks and tightens up needs consistent training under their belt to resolve those issues, and video takes the obvious and quantifies it.
Problem-Solving with Video Analysis Practically
Trying to get athletes better is humbling, but the best way to reduce frustration is to stay focused on whatever plan you have and don’t give up too early. It’s easy to change course when the results are not coming easily, but only change direction when you actually reach the intended location and it’s not what you really wanted.
For example, look at the release angle for the javelin throw in athletics. It’s tempting to try to gather it from a sagittal point, even if the camera calculates a measure close to the 3-D motion capture data. The release angle is not the only factor to success, but if it’s safe to the body and effective with release velocity, it’s the right motion. It doesn’t take a genius to know that throwing at the perfect angle with orthopedically dangerous mechanics and slow velocities isn’t a good game plan.
What does make sense is knowing how to improve release speed and get to a reasonable release angle safely and repeatedly. Video bridges this chasm of unknown answers by bringing in the known variables that are easy to collect and allowing the coach to see firsthand what isn’t working. Lasers can help with ballistic release details with immediate feedback, but what is not quickly discovered is the underlying creation of those forces.
Note: If you want to do much of what this article explains, you need to invest in the Dartfish Mobile Package. For the cost of a pair of jeans, anyone can do effective analysis without much time or expertise.
Take the long jump or triple jump as an example of approach speed, technique, and eccentric power. An athlete with a smooth, controllable approach and great eccentric abilities is not using their speed and elastic talents in their takeoff. What usually is the case is that horizontal velocity is lost when the jumper is unable to execute the transition from sprinting to jumping, and video is instrumental in telling the story of what is not working.Video is instrumental in telling the story of what is not working in athlete performance, says @spikesonly. Click To Tweet
Basic slow motion and side-by-side comparison videos work wonders, while analysis that is too deep isn’t digestible by coaches and athletes. Limb velocities are so fast that not many clues can fix mechanics, so it’s better to set up a change in technique than force it verbally. We are talking milliseconds—a time frame the brain wants to prune and distill to the simplest and shortest path.
Another popular example for technique development is recovery mechanics, something that will be debated for years, by the way, to “fix” poor mechanics. Much of the issues with bad backside mechanics are because of the lack of frontside landing positions. The root problem is multifactorial, but teaching the athlete to step over usually improves timing so they are using positive forces to rectify this very common fault.
Sometimes manual therapy or general power development fixes a specific mechanical error, but only if somebody measures and records both the intervention and the result. I have had athletes that simply got better on their own from just watching better athletes perform during practice, and not everything in technique responds to coaching cues and instruction.
The Top Methods in Video Analysis You Should Use
Coaches need to divide their video analysis into two categories: qualitative summaries and quantitative measures. Knowing the difference is not enough; actually having a written table and checklist improves the speed and consistency of the information.
Qualitative is criteria or parameters that are more general and open than cut and dry points or actual measurements. The benefits to qualitative criteria are that they are quick and easy to collect, and allow coaches to efficiently move through a list to get the problems documented. The functional movement screen has excellent criteria for ranking of movement, but the movements have poor validity for performance and injury risk. You can find good qualitative examples in summarizing phases or parts of a sporting action, rather than a snapshot in time.
Quantitative measures are those that have context and value for making choices. Most of the time, errors happen in measurement because the values are not representative of the athlete, or they are arbitrary and have no reference or connection to performance. Style, or visual differences in technique that have little bearing on performance, is the most visible because it stands out to observers. Commonalities that are similar with all top performers are usually the same movements that make or break a lower-level athlete. Measuring the right variables is simply searching for bedrock, and those points are clear and timeless.
It’s also important to understand the difference between performance analysis and biomechanics. Coaches with shallow backgrounds in kinesiology will get unnecessarily nervous about their education on how the body moves, when they have to trust that foundational mechanics are enough. Stride length and stride frequency lost favor 10 years ago, but they are again growing in popularity because you can use them to see change over an athlete’s career.
The best example of this is knowing how many total steps an athlete takes in a 100m and seeing horizontal speed with splits. An athlete covering more ground and having better speed is a success, and changing those measures often determines what modality can help an athlete more. Contact times and contact length will enable coaches to eventually know how strength training is working or failing to transfer, and only if you have all of the simple data will narrow measures help.
For example, look at acceleration from the blocks. Too much air time and poor leg repositioning will increase heel recovery periods that don’t produce force efficiently and only hemorrhage time. Measuring contact times and air times in the first three steps is more useful than getting lost in angles, unless those measures directly equate with time gained or lost.
Reporting the Data: Getting Everyone on the Same Page
The final need is reporting, and you can do this at any level, but I don’t do it unless it’s someone who can make a national team. Usually, the elites are at the level where their data is stable, meaning they are consistent and likely at a plateau, so addressing the data points makes sense. The development of athletes is a moving target, and by the time you collect enough information, spending time correcting it is fruitless because they already changed too much to get much out of the process.One page is more than enough space to efficiently report the data from a video, says @spikesonly. Click To Tweet
Reporting is simply taking the information from video and putting it on a single page. I have done extensive reports that were gorgeous and expensive, but if the PDF was too long, it just floated around in cyberspace or collected dust. Detail doesn’t need to be long, it needs to be efficient, and one page is enough. Longer reports are fine for serious investigations, but for the most part, you can get enough actionable data on a single page.
My personal favorite sport analysis is the collection of reports from the 2009 World Championships in Athletics in Berlin. They are not only illustrative, they are very vivid with what matters when sharing information. Overzealous in pushing the envelope for innovation, I wasted effort trying to change reporting and I have now resorted to using what is freely available. I attended enough data visualization courses and read enough books to simply restate what the best advice is with reporting.
“Talent imitates, genius steals.” – Edward Tufte
Instead of telling you what to do, I am just going the show you what works best. My influences besides Tufte are Milan Coh, Stephen Few, Megan Jaegerman, Jules Marey, Eadweard Muybridge, and Jacques Piasenta. From a best practice perspective, the IAAF collaborations seem to be the most illustrative, as well as several researchers that use tools brilliantly.
No More Excuses, Coaches!
The biggest lie in sports performance is that the coaches don’t have time. The truth is that while coaches have very little free time, they have enough total time to get the job done. Every minute counts, so don’t waste time performing chores—invest time to get advantages that matter.
If an athlete needs to get better, teaching should use the best options available, and video is one of them. Video should be a primary tool, even if you can only do it quarterly. You never know if one moment in time can change the course of an athlete’s career, so investing in video is definitely worth it.