Printing your own board game accessories: what actually matters

We print board game pieces for a living, which mostly means we have made every mistake available. This is what we would tell someone setting up to print their own: the short list of decisions that change the result, and the longer list of settings people agonise over that do not.

PLA is the right material, and it is not close

For pieces that live in a box in a house, PLA wins on everything that matters here. It is dimensionally stable at room temperature, it holds fine detail, it comes in every colour a game could want, and it does not smell.

The usual argument for PETG or ABS is heat resistance, and it is a real argument for exactly one situation: pieces that live in a car. A box left on a back seat in a Spanish summer will soften PLA, and a warped organiser is a ruined organiser. If that is your life, print in PETG and accept slightly softer detail.

Everything else people say about PLA being fragile is about thin, tall, load-bearing parts. Game tokens are neither tall nor load-bearing.

Silk and matte finishes are worth a word. Silk looks wonderful in photographs and shows fingerprints and layer lines under a dining room light. Matte hides both. For pieces that get handled every round, matte is the better-looking choice in the room where the game actually happens.

Layer height: 0.2 mm, and stop thinking about it

For a token you look at from arm’s length, 0.2 mm is indistinguishable from anything finer once it is on the table. Dropping to 0.12 mm doubles the print time and buys detail that is visible when you hold the piece up to your eye, which is not how anyone uses it.

The exception is embossed faces — a number, an icon, a symbol moulded into the top of a token. There, finer layers do show, and it is worth the hours on the faces you will read constantly.

What actually limits detail is nozzle width, not layer height. A 0.4 mm nozzle cannot draw a line thinner than 0.4 mm no matter how thin you slice it, which is why a fine engraved line comes out mushy at any layer height. If the detail matters more than the time, a smaller nozzle changes more than a smaller layer.

The tolerance problem is the real problem

Here is the thing nobody tells you, and it is the reason printed accessories fail when they fail.

A container that has to hold a specific card, or slot into a specific space in a specific box, does not fail because the print quality was poor. It fails because your printer is a few percent off from the one the model was designed on, and a few percent on a card slot is the difference between snug and useless.

Every printer has its own dimensional signature: first-layer squish, filament shrinkage as it cools, belt tension, flow calibration. Two machines running the same file produce parts that differ by enough to matter on a tight fit.

So before committing eight hours to a full set: print one piece. The smallest one that has to fit something. Check it against the real card, the real box, the real sleeve. Then scale the model by a fraction of a percent if you need to and print the rest. That single test piece saves more failed sets than any other habit.

Settings that matter less than you think

Infill. Tokens want about 15% and are perfectly happy. What people mistake for weak infill is usually too few walls: perimeters carry the load in a small part, so a container that flexes wants more walls, not denser fill.

Print speed. Slower helps overhangs and hurts nothing else. It is not the lever people treat it as.

Supports. The better fix is orientation. Put the face you will read flat on the bed and most supports disappear along with the scarring they leave. A token printed face-down comes off the plate with a better top surface than any amount of support tuning would give it.

Multicolour without a multi-material printer

Most game pieces need two colours at most, and they need them in layers rather than side by side: a coloured body with a contrasting number on top.

That is a filament swap at a known layer, not a hardware problem. Insert a pause in the slicer at the layer where the top face begins, swap the spool, and resume. It is slightly fiddly and it is how a great many two-tone tokens get made, including plenty of ours.

What not to print

Cards. Obviously, but it comes up.

Anything that must be identical to an official component in a game where that matters. A printed token with a different weight or thickness is fine in a co-op with friends and is not fine in a tournament.

Large thin flat plates. They warp, and they warp after they come off the plate as well as on it. If a design needs a big flat area, it wants ribs on the underside — a design decision, not a settings one.

What it actually costs

This is the part that surprises people, and it is why we say it plainly: a large upgrade set is not expensive because of filament. Filament is the cheap part.

It is expensive in machine hours. A couple of hundred pieces is days of printing, plus the failures, plus the cleanup, plus the time spent finding out which orientation works. If you enjoy that process, printing your own is genuinely rewarding and we will happily tell you how we do it. If you would rather have the pieces, that is what our shop is for, and both answers are honest ones.