Gizmos (2018)

Gizmos is something of an awkward blend. At first glance it’s a light, breezy and intuitive game of picking up marbles and spending them to build ever greater contraptions that yield ever greater effectiveness. It plays though like Potion Explosion’s brainy older sibling – the one that graduated top of their class and whose accomplishments are never far from their proud parents thoughts. The disconnect between impression and experience has yielded some of the most contradictory play I’ve had in doing this project. For something that involves playing with marbles, Gizmos sure is an intimidating game for a lot of the people I’ve played with.

In many ways, Gizmos reminds me a lot of Wingspan – not in terms of its design so much but in terms of the learning curve associated with play. I think the thing Gizmos most has in common here is an expectation of an awareness, familiarity and comfort with the jargon and conventions of board game play as an activity. Speciality requires literacy in the jargon and conventions of the specialism. That’s a roundabout way of saying ‘Not everyone is going to get Gizmos on an intuitive level’. They might need to work at it.

On to the game.

Gizmos doesn’t make a good first impression when you open the box. You’re presented with what is essentially a set of flatpack furniture and an assembly guide straight out of the IKEA rejects section. You need to construct the central marble depository before play, and while it’s not complicated it’s probably the least satisfying thing to see when you open a box for the first time. Somewhat important here is that it doesn’t even look particularly cool when you’ve put it together. Potion Explosion had a marble dispenser that looked nice enough to merit consideration as a coffee table installation The contraption in Gizmos looks and feels cheap. Maybe the best way to illustrate that is with the ‘lots of marbles in here’ texture used for the interior of the main body.

I hate the look of this so much that it almost made me hate Gizmos. This is the board-gaming version of having wood-panel wallpaper, or marble-tile effect linoleum. Actually no, it’s worse than that. It’s like someone covered the inside of the game with the wallpaper that is made up of pictures of shelves of books. It just looks so cheap and nasty. Things don’t get much better when you fit the last part of the device either, because the art on the final shield is just as bad.

Things do improve though once you’ve constructed the dispenser and set up the starting configuration itself. That’s when you see how the game, not the dispenser, fits together.

Here’s how it works. You get a control deck outlining the actions that you can take on your turn. Arrayed in front of you is a range of gizmos (title drop, boom) in a range of levels. You can hold a small number of marbles in your energy ring at a time, and to begin with you can keep only a single one of the gizmo blueprints in your hand before it is built to your console. Each time you get a turn, you can do one of your actions. You can pick up a blueprint (a file action), pick a marble from the front of the dispenser (a pick action), add a card you have in hand to your console (a build action), or draw a number of cards from a face-down deck and choose one to build if you have the marbles (a research action). Each card you play down gets you victory points, and the game ends when sixteen of these have been played to a single console, or someone builds the fourth of their ‘level three’ gizmos. Largest number of points wins.

What makes this pedestrian design interesting is that each of the gizmos you build will empower you in a range of ways. Some will let you upgrade your file, pick and build actions so that they have additional things they do. Some will increase the limits you have on marbles, cards in hand, or how many cards you draw with a research activity. Some let you convert marbles of one colour into marbles of another, or increase the value of marbles of a particular colour. Essentially what you are building from these gizmos is increased capacity, and that’s what’s going to allow you to progress from cheapo every day gizmos to the more expensive and zeitgeist defining Nobel-prize winning masterpieces.

Consider the cards on the top row in the image shown above. The first costs a single yellow marble to build, gives you a single victory point for building it, and once you have you’ll get a victory point every time you build a blue blueprint. The next two along let you hold an extra marble in your energy ring, and draw an extra card when you research. The last card lets you draw a random marble from the hidden dispenser every time you pick a yellow marble. As you go down the levels, you see more powerful blueprints. The second level II blueprint there for example lets you pick a marble from the visible energy trail whenever you build a blue or yellow gizmo.

Your starting console contains an upgrade for the file action – whenever you take a blueprint into your hand, you get to grab a random marble from the dispenser. And that’s how it works – you collect up blueprints and marbles, and spend those marbles to build those blueprints and upgrade your capabilities.

And it’s actually here that Gizmos starts to become problematic for a lot of people, because this is a kind of meta-mechanism that moves some people outside their comfort zones.

I play games with a lot of different people – Mrs Meeple pretty much constantly, but we have others that come around for games nights and they’re not all as deep into the hobby as we are. They’re normal people. It’s wonderful to draw these perspectives into reviews because it ensures that I’m not making the same mistake of thinking the conventions in which I am immersed are equally tractable to others. And, importantly, it forces some reflection on why that might be the case.

In one game of Gizmos, I called an end to it early and we went on to play something familiar instead. This wasn’t because the other players asked for that, but because I could see pretty clearly that there was no connection between what they did and what they expected to happen. Their willingness to play was driven by politeness and ‘not wanting to make a fuss’. I put that down to ‘I obviously screwed up teaching it’ – it happens.

Then we played it with another couple and while they, as far as I saw, grasped it after a few turns it was clear in our post-game discussion that they hadn’t at all felt comfortable with what the game was asking them to do. Again here we get an echo of Wingspan. ‘It’d make sense after a few plays’, said one of them, with the heavy implication there being ‘It certainly didn’t make sense after this one’.

These kind of responses are interesting because functionally there isn’t anything particularly complicated in Gizmos. With both groups I’d played more mechanistically heavy fare. So, what it is it in Gizmos particularly that was a stumbling block? And for those that didn’t stumble… what was the feature they had in common?

And it’s a surprising one – the ones that didn’t stumble had a background in computer programming. That’s significant, because when you take a step back and think about it – Gizmos is basically a game about programming a computer system. Each of your actions is a kind of subroutine, and the core mechanism of Gizmos is a gamified version of feature creep. It’s like Star Citizen: The Board Game – except you actually get a copy when you buy it.

Every turn you’re adding new code and conditions to your functions, and then you’re feeding them the input that results in the outputs that you want. Everything you do in Gizmos is test-driven development, with the logical comparisons being abstracted down to the level of marbles. You’re constantly running your program to see if it produces the right kind of results from the right kind of data.

And really it’s no surprise people have some trouble with this. The key skill you need to have in Gizmos is the ability to visualise one or more threads of execution through a set of branching conditions as measured against a set of input data. It’s easy to think ‘Take a card, pick a marble’. It’s less easy to think through the implications when you have actions that conditionally trigger other actions. Which then reference back to your original actions where every single card can only be triggered once per turn.

Check the console above. Let’s say you have a blue card in your hand you can build, and you do that. Your build action then lets you do a research action, which you use to draw out a red card and build that too. You’ve now got essentially two threads going through your computer program and you’ll need to make decisions within each. Decisions you make in one branch will shut off those options in the other.

Building the red from the research blueprint lets you do this:

1. Because you built a red blueprint, you get to pick a marble from the visible part of the dispenser.
   1. You pick a yellow marble.
      1. Because you picked a yellow marble, you get to pick a random marble from the hidden part of the dispenser.
      2. Because you picked a yellow marble, you get to pick another marble from the dispenser.

But also, you have a second card that triggers because of the original building of a blue blueprint:

1. Because you built a blue blueprint, you get to pick a marble from the visible part of the dispenser.
   1. Since you already used the yellow marble actions, you pick a blue marble instead.
      1. Picking that marble lets you pick a random marble from the dispenser.

Making sure that these two threads of logic don’t bang against each other and the limitation on gizmo activation (once per turn) is a perfect application of systemic, programmatic thinking. As is the ability to see the opportunities for this kind of trigger effect in the cards available to you. And make no mistake, this isn’t overclocking the game systems – the presence and exploitation of chain reactions like this is a big part of the intentional design of the game.

Still not too bad right? Well, what happens if you build this blueprint:

Now if you build a black or red blueprint you’ll get to trigger a file action. And remember those file actions? They let you pick new marbles. So suddenly you’re in a position where one single action might trigger all of your cards across the tableau, and the order in which you do that will be significant. You’re building subroutines here, and they’re pretty poorly decoupled. They don’t separate cleanly from each other – they’re all heavily dependent. Your contraption of cards is a nightmare to maintain, but there’s no denying it sure can scale.

This is what I mean when I say Gizmos is Potion Explosion’s brainy older sibling. It’s Potion Explosion with a computer science degree. Speaking as someone that has been teaching first year university students to program for almost… Jesus, almost twenty years… the mental firmware needed to do this well does not come pre-installed in everyone’s brain. I’m not saying then that Gizmos is hard. Mechanistically it’s very straightforward. What I’m saying is that there is a real chance Gizmos is almost unplayable to any real level of skill for a lot of people until they’ve sat down and studied their functional programming. That’s because their brains just don’t approach problems in this way. They’ll get the rules. They’ll understand how the game works. They’ll take a lot longer to internalise how to wield those rules well. It needs someone to understand how to debug their cards, and to visualise what’s missing in the code to make everything work smoothly and effectively.

Have I really been teaching for almost twenty years? That can’t be right. Hang on, let me check.

Jesus. It is right. Next step – the grave.

I suspect regular board-gamers on the other hand almost certainly have developed their brains in such a way as to be able to parse the puzzle that Gizmos present. It’s a difficult game though to recommend for playing with those that haven’t. And the thing is – if your mind works this way it’s easy to miss the fact that a lot of minds don’t. Rules based thinking, the construction of systems for manipulating internal states in appealing way – it’s a very ‘gamey’ thing. But Gizmos makes it into a formal mechanism in a way that we don’t see often and I think a lot of non-gamers are just going to end up angry when playing it. ‘These rules are easy’, they’ll yell, ‘So why the hell can’t I understand this game?’.

That’s not going to be everyone though, and those that have more compatible mental hardware will find Gizmos is actually a little bit special. The combos you can build are really satisfying. The escalation of capability is very well paced and the game ends right at the point it should. The tactility of the marbles adds an enjoyable physicality to play. There’s a real sense of accomplishment that comes from putting together an effective tableau of cards, and sufficient options on the table that you can exercise a surprising amount of creativity within the confines of Gizmo’s systems. Combine that with tight hand limits and you get a game that is almost perfectly designed to emphasise both flexibility and optimisation. I’ve worried more about resource balancing in Gizmo than I have in any computer program I’ve written in the past five years, and had a lot of fun doing it.

Gizmos is, literally, not for everyone. I mean that in a very specific sense. Not the usual half-hearted ‘I liked it but I understand if others don’t’ conclusion that has taken the place of a judgement at the end of many reviews. I mean that I don’t think this game works for a substantial proportion of the non-hardcore gaming audience. The level of literacy it expects in terms of logical manipulation is surprisingly high for such a mechanically simple game. Some people are simply going to have to work hard at its design to get anything out of it, and the disconnect between the relatively trivial rules and the mental architecture needed to utilise them is going to leave some people frustrated. Work past it though and there’s a lot of magic in the design of Gizmos. This is definitely a try before you buy scenario.

Twenty years.

Jesus.