One of the thing that volleyball coaches seem to debate a lot is the value of blocked training as opposed to random training. During these arguments there is sometimes a confusion of terms. In this post I want to try to clear that up and provide definitions. I will do so based on how these terms are used in the academic research from which we get them.
Blocked training is doing something with very little variation. The goal is for the athlete to execute a specific skill in a specific way. For example, let’s say a coach wants a player to work on shuffling to pass a ball. They toss 10 balls to the player’s right such that they have to shuffle to the ball. That is a blocked exercise. So too is something like hitting off a tossed ball. Also, repeatedly setting a certain set from a coach’s toss when the coach doesn’t move or change the location of the toss.
The whole concept of blocked training is to repeat one specific skill until the athlete develops mastery. Many people think of this sort of thing as how the athlete creates “muscle memory”. That is an entirely false term (see The Talent Code). Conceptually it gets us thinking about developing motor programs. It only goes part of the way, however, and what it leaves out can be quite important – as you can read in the More is not better. Better is better post.
The term “random” here is one often used in coaching discussions. In the academic literature, though, they tend to use “variable” or “distributed” more so than random. That’s because you don’t truly need the repetitions to be random. You just need them to not be the same repeatedly.
Here’s an example from a paper I read involving kayakers learning to roll. Blocked training involved simply rolling one way (say to the left) over and over again. Random training mixed up rolling left and right.
Using the example above of teaching a player to shuffle to the ball to pass, blocked would be training one direction at a time. Random training, in contrast, would be mixing up the tosses between left and right.
Which is better?
So is blocked or random training better? At this point there isn’t a clear answer. The academic research strongly suggests random training is superior for training simple skills, and seems to uniformly advocate for random training with more expert performers. The evidence when it comes to complex skills is less clear, especially when it comes to novices. Mainly that’s because there simply hasn’t been enough of it yet.
Contrary to what some blocked training advocates would have you believe, this DOES NOT mean blocked training is superior for training complex skills in beginners. As the saying goes, a lack of evidence for something is not evidence of the lack of that thing. In other words, just because we don’t have strong evidence in favor of using random training with novices does not mean blocked training is preferred (and similarly vice versa). It simply means we don’t know yet which is better.
Also, I strongly suggest check out this Coaching Conversation I had on the topic of Motor Learning.
Training in a game-like fashion is a separate concept from blocked vs. random, though coaches often conflate them. I wrote a full length article on the value of game-like training, so I won’t talk benefits here. Instead I’ll show how it is a separate consideration from blocked/random.
The blocked vs. random consideration has to do with how you distribute repetitions. The question of those reps being game-like or not relates to how specific they are to the way the skill gets used in the game.
There’s a really good discussion of specificity and making things game-like in an episode of Coach Your Brains Out featuring Steve Bain. Steve is the coach at Northwest University and a professor of orthopedics and sports medicine at Washington University. It’s a 2-part interview. The first part is the meatier one with respect to this topic. Steve has a very interesting observation in there about the dimensions of specificity.
Random is not the same thing as game-like
The thing I want to make clear here is that game-like and random (variable/distributed) are not the same thing. The belief that they are causes problems, especially in some of the debates I see. An exercise can be highly random, but not game-like in the least.
Returning to our example of shuffling to pass, as I noted, mixing up whether the player must go left or right makes the drill random rather than blocked. There’s no way, though, that you can call passing a tossed ball game-like.
Why does this matter? Because if the exercise is not game-like then the expected transfer to use in an actual game is low. Again, this isn’t something I want to go far with here. See the article I linked above.
Game-like tends to mean random, but not always
Now, while random doesn’t inherently mean game-like, game-like often does mean random. Why is that? Because game-like repetitions tend not to repeat the same thing over and over.
Continuing with our passing example, let’s say you start an exercise with a serve. Chances are you don’t have servers who can make exactly the same serve each time. That means even if the ball goes to just one passer each time, they are unlikely to pass the ball the same way each time. Sometimes they will move left or right, or forward or backward. Sometimes they won’t move, but will take the ball in different locations relative to their mid-line. Those are random repetitions.
Flipping things around, you can also have very blocked type repetitions in a game-like framework. Serving is the prime example. All it takes to be highly blocked and still game-like is a server attempting to repeatedly serve the same location against a set of receivers.
My point in all this is that you should be clear on your definitions with respect to these terms. They have different meanings, and those differences can be quite important. Know them and you’ll have a better handle on motor learning theory and practice.