What Is Performance: Part 1 - COD Testing

Rich Clarke

The use of COD testing in sports is mixed. Some love it, some hate it. Of course, one of its limitations is that it is commonly pre-planned and therefore only assessing the physical components of performance. So for clarity, the writing below is discussing things in relation to the physical requirements, not the perceptual-cognitive ones. A debate about splitting these up or not can/will come abnother time!

In COD, we are faced with lots of different tests to choose from, most of which have limitations. Regardless of what you choose they all have one thing in common; if they are completed in a traditional format they are usually related to linear speed. More simply, if you are faster in a straight line, you’ll likely be faster in your chosen COD speed test. The longer the duration and distance of the test the stronger the correlation is likely to be. This creates limitations for athlete profiling resulting in the birth of the COD Deficit (CODD) and the Deceleration Deficit (DD). But there is a bigger over-arching question before we even worry about profiling.

When we are testing, what metrics are the best representation of ‘performance’?

Performance is of course sport and situation dependant so before we answer the question of how we might measure performance; we need to identify what tasks players are commonly challenged with.

Once we know the tasks which are important, how we quantify performance will depend what the athletes intent needs to be in that task. In my experience, athletes are challenged to do one of the following things:

  • Get to a new location via a non-linear path (speed from A to B)
  • Change path as fast as possible (deception)
  • Decelerate into an adaptable position with multi-directional demands

1: Get to a new location

There are lots of situations in sport where performance is about getting from A to B as fast as possible. For example, a turnover resulting in a full change of playing direction or a player needing to react to a change in the oppositions movements and re-position themselves on the field. In these situations, the sport just wants you to arrive somewhere quickly and doesn’t care how you do it. Therefore, regardless of what test you choose, ‘total time’ could be considered your performance outcome. However, one important limitation to this is that this number doesn’t provide much use in profiling individuals to learn what they need to improve their performance (this topic requires a different article).

2: Change path as fast as possible.

If you need to create space and separate yourself from a defender, you need to change your movement path or location fast enough that the defender cannot react and close gaps. This may take place in an initiation movement from a slow or nearly static start (football/soccer) or at high speed during a cutting action (Rugby/American Football). Therefore, in these situations, the athlete needs to get from one movement path to another as fast as they can, not just to a new destination. To quantify this we need to capture the time it takes to complete the COD and ideally, remove the influence of linear speed. This requires a consideration of the CODD or for the time capture to simply be just before and just after the direction change. The CODD for anyone not familiar is time to complete a COD test (with a single direction change) minus the time taken to cover the same distance linearly. However, one thing to remember about the CODD is even where there are situations where it may be a key aspect of performance, it still can’t be isolated as a ‘performance’ measure as the CODD is relative to the athletes own linear speed capabilities. For example, two players can have the exact same CODD, but if one is much faster than the other, their performance isn’t comparable. Therefore, this number always needs to be looked at in conjunction to linear speed performance.

3: Decelerate into an adaptable position

Performance in other situations (likely more defensive) may require a player to get to a specific location but arrive in an adaptable position. For example, a footballer closing down an attacking opponent at speed, needing to decelerate and cut off an attacking option. The athlete needs to react early, accurately and decelerate to a controlled movement speed where they can further adapt. They may not need to perform a maximal action once they have arrived, rather smaller multi-direction actions which are non-committal. Deceleration becomes key as excess speed isn’t adaptable and neither are the movements needed to apply braking force. Therefore, in these situations you need to reduce your speed as fast as possible potentially while partially turning to one direction or another. Performance therefore isn’t going to be effectively represented by total time or by CODD. CODD is certainly better as the ‘deficit’ will generally include some deceleration time (dependant on the variables used to calculate it) but it won’t always give you a clear picture. This is the reason I developed the DD (I have a paper under review about this at the moment so will expand in future).

How does this fit into the COD landscape?

Before we get into the specific sections of the landscape there is an important question: Does the situation which requires that type of movement also require deception? The answer to this question changes how we may assess the physical performance output.  A really general rule as we work through these examples and explanations is that the more deceptive something needs to be the more important CODD is in our performance testing.

High speed, large angle cuts

Let’s start with the most complex area of the landscape; high speed large angle cuts. If no deception is needed, the situation is likely similar to those mentioned when discussing the importance of time to get to a new location. For example, during a turnover or large change in direction of play you simply need to get from A to B as fast as possible and how you do that isn’t particularly important.

Deceptive tasks however result in a change in focus from WHAT they are able to do (total time) to HOW they need to do it. CODD is useful in this situation, but it is worth remembering that because we are talking about large cuts, players can’t rely on speed maintenance to evade their opponent. The have to decelerate in order to perform these cuts, hence why the ‘deficit’ in comparison to sprinting over the same distance is relatively large. This is important as when you are trying to evade another player and they are close to you, the need to change the direction of movement is increased and this COD needs to occur earlier (more space between attacker and defender) rather than later. But if it is to a large angle, deceleration needs to precede this. Meaning performance (successfully entering into open space without being intercepted by the defender) needs you to get somewhere quickly, but also needs the COD to be performed at optimum distance from the opposition and in sufficient time (hence the importance of the COD and the DD). These important components determine the likelihood of whether the attacker will create sufficient separation or not.

Cuts and curves

Moving to cuts and curves, there is still a deception consideration, but the distinction here is less important as this section of the landscape has two movement categories, shallow cuts AND curves. Meaning that if you need deception, it is because you have less space and therefore you would perform a cut and not a curve. Both cuts and curves require athletes to change direction WHILE running at a high speed. Performance is determined by being able to make the largest change in direction with the smallest deficit in time (Hader at al., 2015). The best way to do this is actually to just run a curve. This is really common, but sharper directional changes will also be required which results in more of a cutting action. General rules are that curves are used when a smaller angle change is needed OR there is lots of time and space to perform it. Cuts are used when there is a larger angle needed or there is less time and space to perform it. Therefore, using total time here is still a great ‘performance’ metric but remember that the more deception is needed the more important CODD is going to be.

Multi-directional initiations

Having already explained the two above movement actions, multi-directional initiations aren’t any different. If you need to accelerate into a new direction different to where you are facing and get to a new location quickly, total time will serve you well (no deception). But if you are performing these actions in order to evade a defender, the importance of CODD creeps in again as it is the initial separation facilitated by turning quickly which is key.

The conclusion here is that when we are considering ‘performance’ as being what’s useful on pitch, total time from a COD test is still useful. But CODD is needed to help us understand how the performance is produced, Just remember it can’t be removed from linear speed during interpretation.

The elephant in the room?

Everything that I have stated above is relevant for pre-planned performance testing and understanding physical output. But as we know, sports aren’t purely pre-planned. Therefore, there are a couple of things missing. Firstly, perceptual cognitive expertise. But secondly and more related to the question of deception is the athlete’s ability to disguise what they are doing. This is huge and not captured in pre-planned testing or developed in pre-planned training. This, along with some further discussion of performance profiling will come in parts 2 and 3 so stay tuned!


Rich is the founder of Strength Coach Curriculums and an S&C coach who specialised in multi-directional speed. He runs the S&C provision for Bristol Flyers Basketball and consults with clubs across the globe while also leading the MSc programme at the University of South Wales