[Agility Lab] Do Flashing Lights Help?

Pretty much all of our evidence points to one key principle when it comes to developing perceptual cognitive expertise.

Specificity matters.

Tennis players who are good at anticipating in a tennis game aren’t any good on a rugby pitch. People who are great at making fast, complex decisions under pressure from playing speed chess, aren’t any good with decisions on a basketball court.

So it is weird to see that our default as S&C’s has become:

COD Speed > React to a cone colour or a flashing light

While I am starting by saying this isn’t any use. It doesn’t mean they are completely out of the toolbox. The can have some beneficial impact on sessions and are useful in a rehab context. But you need to think carefully about why you are using them.

What a stimulus like a light or a maths puzzle does give to a session is engagement, intensity and a bit of fun. Which is great, no issues with that. But, it isn’t developing anything of use for the perceptual cognitive component of agility. So use them, but don’t kid yourself it is optimal performance enhancement. Remember you should be able to justify everything in your programme for the right reasons.

The other impact they have, especially when we talk flashing lights is an impact on the physical component. Sometimes referred to as exposing the athlete to ‘worst case scenarios’. A light arguably does this becuse an athlete can’t ease the time demands by being better and identifying information earlier. They have to wait for the light to flash to move, having a large impact on the movement execution due to the lack of time to pre-prepare. But the question for me is while this does have an impact on the position an athletes end up in, is it really a worst-case scenario biomechanically? I am not sure, but there are a few reasons I would argue against it to specificity (a human stimulus):

1) No deception demands

When athletes try to deceive, they have a desire to keep the hips square for as long as possible and even end up with counter rotation rotate to opposite direction). This means the postures adopted may be even more extreme than if just waiting for a light to flash. As always it depends, as there is an argument that the counter rotation takes slack out of the system and makes the cut more stable. But there is certainly a lot of ACL’s which tear when that attacker is trying to be deceptive…

2) Reactive direction, but not angle size.

Angle size is a big influence of load in cutting based movements. A 45-degree cut is a lot less demanding than a 60-degree cut. When you approach a light, you know where they are and exactly what angle they are asking you to turn to, just not if it’s left or right. So still not ‘worst case’ as a human stimulus might require L or R or a cut anywhere from 45-70 degrees. In fact, in a human stimulus the attacker might decide/react, but choose the wrong side and need an even greater cut than they initiated because the defender is going to same way.

3) Intent and the desire to avoid

Similar to the above, when studies compare a light stimulus vs a human stimulus it is common for the human stimulus to result in greater frontal plane demands and larger angles. It would seem because a human stimulus is something bigger to avoid or maybe just more representative so encourages the athlete to avoid it with more space. In comparison, a light gives the athlete less intent to ‘get out of the way’. You could make the lights farther apart to solve the problem. But it still feels a little square peg round hole for me.

There are lots of it depends arguments to the above, but there are just too many variables and things potentially missed with a light vs a human. And that is just a rebuttal to the ‘worst case scenario’ physical demand. They still have no use in perceptual performance development. So I am not convinced yet.


The other area when light stimuli come up is in testing. I am much happier here as there is more utility. Don’t get me wrong, it still isn’t perfect, and I would encourage a human stimulus over a light stimulus every time. But, it is more justifiable and has more support in the literature.

There is variation in studies as to whether a light stimulus can tell the difference between playing levels from a talent identification point of view. Some show they can, some show they can’t. But even studies that show you can, if they have used a human stimulus as another condition, they usually report that the light is less sensitive. So useful, but still second best?

The other consideration here is similar to the correlation causation principle. If you make athletes better at reacting to humans and the subsequent movement demands, this may show up in a light reaction test as they are better at executing the reactive movements needed. But if you train them to be good a reacting to lights, they won’t become better at reacting to humans, which is what is athletes need to do in competition. So I think human stimulus training trickles down stream to lights, but lights don’t trickle up stream. A bit like the concept of sprinting will drive up your weights, but weights won’t necessarily drive up your sprinting.

I’ll expand on more details of perceptual skills in another post. But remember our key rule in developing perceptual skills (which I have hopefully justified why it is also king for optimising the physical stimulus too).

Specificity is king