Game designers have a word for the stuff that makes a game feel good without actually affecting its mechanics. They call it juice. Screen shake when you land a hit. Coins exploding out of a defeated enemy. The satisfying crunch of a button press. None of it changes the game. All of it makes you feel like it does.
A few years ago I shipped a 360-degree scenario that was absolutely dripping in juice. Spatial audio. Branching conversations. A photorealistic environment that took weeks to render. Learners were finishing it at rates we had never seen. Stakeholders were delighted. I was quietly proud of it.
Six weeks later, almost nobody could recall the decision frameworks it was supposed to teach.
We had built the best possible delivery mechanism for content that was not designed to stick. The juice was doing the engagement work. Nothing was doing the learning work.
What game design actually gets right
The best game designers are obsessive about the difference between engagement and progression. Juice drives engagement - it keeps you in the chair, makes the loop feel rewarding. But progression is what makes you better. And progression requires friction. Desirable difficulty. Moments where the game withholds the reward until you have actually earned it.
The distinction that matters
Engagement gets someone into the experience. Retention requires the experience to make demands of them. Juice optimises for the former. Good instructional design has to care about the latter.
The 360-degree environment was extraordinary juice. The problem was that I had let the production value become the design. Every cognitive resource my learners had was going towards processing the environment - the spatial audio, the visual fidelity, the novelty of the medium. There was nothing left for the actual content.
This is not an argument against high production values. It is an argument for using them deliberately. The environment should have served the decisions, not competed with them.
What I would do differently
The scenario needed stripping back to the decision points. Lower visual complexity in the moments that matter. Deliberate pauses. Retrieval prompts built into the branching logic rather than bolted on afterwards as a quiz that nobody connects to the learning.
The 360 environment could still be there - but as context, not as the experience itself. Let learners orient, then pull them into a decision before the novelty overwhelms the objective. Give the juice a job.
Related experiment
Decision swiper - play the mechanic this article is about
[ key takeaways ]
Juice and learning are not the same thing. High engagement does not imply high retention.
Production value can become a cognitive load problem if it competes with rather than serves the learning objective.
Desirable difficulty - friction that requires genuine processing - is where retention happens.
Game designers solved this decades ago. Check your juice-to-learning ratio before shipping.
This became an episode
Ep. 01 - Juice vs learning, the full breakdown
Got thoughts on this? The best L&D conversations happen on LinkedIn - come argue with me there.