Magnetmen

Welcome to the devlog (“developer diary”) for my game Magnetmen.

This article will explain the process behind the creation, the problems I faced, interesting lessons, and more. It will be a rather short devlog (relatively speaking, at least), because this was a very small project that I basically made “on the side”.

What’s the idea?

I’m always looking for ways to make games even simpler and more accessible.

That’s how the following idea popped into my head.

• What if the paper was just a grid of icons …
• And on each turn you do the same simple thing: cross off one icon to get it.
• And at the end you sum all icons you collected for your score?

That’d be a really simple game. In fact, I just explained the whole game to you.

Of course, it also wouldn’t be that interesting or tactical.

I knew I needed some limitation on which icons you could grab. That’s how the following rule came to life:

The more you have (of a particular icon), the more it repels you.

In other words,

• If you have 1 icon, you can’t pick horizontally next to that type anymore.
• If you have 2 icons, you can’t pick vertically next to that type anymore.
• If you have 3 icons (or more), you can’t pick any square next to that type in any way.

In my mind—before testing the game yet—this seemed like a good rule.

As the game progresses, more and more parts of the board get closed off for you. This is easy to check and to execute. (I can add a graphic to the board that instantly shows you when each direction is forbidden, kind of like a compass.)

Yet it requires you to plan in advance. To make sure you can still grab valuable squares later and don’t get stuck too early.

When I considered a name and title for this game, that idea of repelling kept coming back. That’s the most intuitive way to explain the core mechanic. So I wrote down: “Magnetmen?” and went to bed.

No time to think, let’s make it

Graphics

Initially, I wasn’t stoked about that name and theme. But it fits. And it’s something different from what I’d usually do. (My gut usually jumps to cute animals or food for themes :p)

So I decided not to waste any more time and just make it.

I immediately jumped to creating all icons using the latest image generation AI. Why?

• I knew I only needed a few icons for the whole game, which is doable with my low amount of free credits ;)
• Those AIs are awesome at creating images based on real-life materials and objects. (Which also does not infringe on any copyright.)

I can ask for a “boy playing the guitar, made of metal parts and magnets”, and it usually gives me that without so much as a pause. I created a list of common occupations and ran through them all to get ~30 different icons.

As this was supposed to be a quick, tiny project, I accepted the styles I received and did minimal editing. (Just removing some ugly parts, adding a bit more color to dull ones, removing the background of course, etcetera.)

Board

This game is so simple that I want all the rules on the page. I already did this a few times before and learned from those attempts.

• The page is split in two parts: the actual playing board (the largest) and a “sidebar” (a vertical strip on the side)
• The board is filled with a grid of random icons.
• Around this board are strips of empty squares: the player “inventories”. This is where you draw each icon you collect.
• The sidebar contains …
  • A general tutorial. (What to do on your turn, when the game ends, etcetera.)
  • The explanations of every specific icon used on this specific board. (Their illustration + a sentence about how they score points.)

Below is a quick sketch of this idea.

In the final generation I always make the dimensions of things dynamic, so the size of the board/tutorial/inventories can change on the fly as we go.

The biggest challenge is always to get enough space for the tutorial, even when we include many icons.

Wait a minute, I had more rules!

I started making this so quickly, that I completely forgot about two other rules I wrote down. (I was only reminded of these when I went to close my notes before going to bed.)

Rule 1: Encircling. If you manage to completely enclose a group of icons on all sides (with your move), then you get the whole group. (Similar to something like Go.)

Rule 2: Must pick adjacent. On your turn, you must pick a square adjacent to a square that’s already been used.

And I can see why my brain decided to subconsciously ignore these.

First of all, they don’t fit on the page. If one or both of these had to be added to the explanation, it would become too long.

But secondly, they have all sorts of nasty side-effects that would overcomplicate the game (and lead to more needed rules). For example,

• What if you enclose a particularly large group? The player inventories are quite small, so it probably won’t fit.
• What if a player decides to be annoying and start in a corner? This limits the options for other players so severely, for the entire game, that it basically ruins the game => we’d need extra rules about where you’re allowed to start (to make that “must pick adjacent”-rule work)
• The Repel rule is already interesting and limits placement. The Adjacency requirement would be so limiting that you’re likely out of the game before it has even started.

They are still good rules, just not for this game. So I moved them to a future One Paper Game which would focus entirely on encircling.

The final boards

So, I wrote the code to randomly generate all these parts and draw them on the page.

The icon picker uses my own system for “balanced randomness” behind the scenes.

• It picks a subset of all possible icons. (If you use more than 6 or 7 unique icons, it’d just be overwhelming.)
• Then randomly fills all the squares.
• While keeping requirements into account. (For example, some type can’t appear more than X times, another type can’t appear next to type Y, etcetera.)

I also added an “Evaluator” at the end of that chain (as always), which checks if the final board is valid and has no nastiness. As of right now, though, it does nothing because I don’t see any potential issues.

For now, there is no other decoration. Just solid rectangles and images, which looks as follows:

Notice the little drawings in the corner of the tutorial icons. Those are for drawing them yourself, in your inventory. (You can’t expect people to draw those complicated icons every time :p) They might still be too complex right now—we’ll see about that.

Honestly, I don’t see too many issues here.

• The board might receive more variation. Slightly different background colors behind the icons, for example.
• The tutorial is similarly barebones. Perhaps decorate the headings and add some magnets here and there.
• (And of course the “ink friendly” variant that uses less ink.)
• But otherwise, this is the game.

So I quickly printed a few of these and tested the game.

Playtesting & Finishing

Initial Playtests (2 players)

Due to some last-minute issues (with scheduling), I initially tested the game with 2 players only.

Below is an image of two of those papers. (With some quick notes by me as I try to improve the rules, which I’ll explain below.)

The results?

The game works!

• The rules are indeed explained within 30 seconds.
• The magnet rule clearly locks down the board over time, but not so much that you can have games where you are done after 2 moves.
• The different ways that squares score is easy to explain, read and use.

At the very first playtest, we realized some practical consequence of this rule: once you have one of each type in the game, you can literally not pick any more squares. I never even realized this before now, but when playing this jumped out immediately. Because me and my opponent had both collected the 5 unique types, almost out of “gamer’s habit”, and realized we’d been stupid.

The next game we played, we scored double the points and got way further.

The game clearly pushes you to go for one or two types. The challenge of this game, therefore, is about quickly identifying which are the best ones to go for (for you specifically). And, of course, hoping that others don’t see how many points that will score you and interfere with that plan.

Another consequence of this, of course, is that games with more different types go on a little longer and are slightly easier to play. If your board only has 5 types, then you will collect 2 or 3 of a type much earlier, and thus be done earlier. If you have 7 or 8 possible types, they appear far less, and it takes longer before you even have “one of each type”.

All of these little ideas about strategy are automatically learned while playing it, which is great. It means the game, whilst far from being “deep” or “complex”, certainly does have depth and replayability.

My issues

My only critique, really, was the low amount of interaction and how quickly your choices dwindle.

• The Repel rule only applies to you. So you can’t make it harder to grab something for opponents—you can only obstruct them by picking the exact square they want for yourself.
• Squares that have been “picked” are still on the board, so even that doesn’t change much.

I played with the idea of squares, once picked, being completely removed from play. As in, they also don’t count when checking the Repel rule.

This, however, didn’t feel great initially.

• It required adding a sentence or two to the rules explanation.
• It probably added too much freedom as the game goes on.

I considered splitting the core rule of the game into two options.

On your turn, pick an unpicked icon.

• Either cross it out and add it to your collection.
• OR fully destroy it. (You don’t get it; it’s just gone from the board.)

This would provide a balance between actually removing squares from the board (or stealing them from opponents), and collecting them so you can score them.

It would also complicate the rules explanation.

The more I considered this, the more it felt like a variation on the rules for those who wanted a little more out of this game. (And a little more cutthroat player interaction. Which, if you know me, is always my goal.)

If I were to make these rules the core, I’d lose the one thing Magnetmen was made to do: be the most accessible game ever, take it with you anywhere and play with anyone within 30 seconds.

So this was my solution:

• The rules remain the same for the base set. However, I change the rules for one type to be the “destroyer”, and always include this one.
  • “Worth 0 points. When picked, destroy 2 squares. (Make it unrecognizable to signal it’s removed from the board.)”
• But when you pick a more advanced set, the new rules image is displayed.
  • In other words, destroying becomes a core part of the game.
  • At this point, I can be sure people have already tried the game at least once and are, you know, ready for 1 more rule.

Playtesting the slightly tweaked system

Some time later, I was finally able to test with more players, and with the new rules / advanced sets.

I wasn’t able to test all types and configurations, but that’s almost always the case with these randomly generated OPGs.

As I expected, the game just became (slightly) better. You play for slightly longer, with more options, and there’s more interaction as the board changes all the time. There is slightly more strategy and challenge. The powers connected to the icons are now a bit smarter (based on my experience playtesting), which means it’s harder to pick the best strategy on a new board, even after playing several games already.

That’s enough for me.

Conclusion

As usual with One Paper Games, the game doesn’t have huge depth or crazy mechanics that will blow your mind. Instead, the depth matches the simplicity. For such simple rules and instant play, the depth and “replayability” of the game match, in my eyes at least.

The idea is that you can print 10 Magnetmen papers and play some quick fun games anywhere, with anyone, with no hurdles or obstacles or unnecessary complexity. I think we’ve achieved that!

Also, my laptop is really at the end of its life. Even the simplest changes (in terms of code or graphics) can take minutes, or cause a crash, which means I basically rushed to finish this project before it all imploded.

So, to finish it off, a screenshot of the generated “Magnetmen” that did not make it into the game. (Adding another set seemed overkill anyway, but it would literally have killed my computer I think. Also, I decided that a game about Magnet people should probably not include animals. Still, I wanted to show these somewhere, because they were quite nice.)

Until the next devlog,

Pandaqi