New South Wales
Is this the real life?
Is it just fantasy?
For Lucidity: Six-Sided Nightmares, I have to produce a ton of custom dice (okay, maybe not 900kg of dice - probably closer to 150kg). As such, I have been doing a bit of investigating into the different options available. I learned a lot from the process of having custom dice made, and thought it would be worth sharing those things here for anyone planning to make custom dice in future (whether the graphic design for them, playtesting with them or producing them). Especially since custom dice for games seems to be all the rage lately (I'm looking at you, Dice Throne: Season One and Grimslingers (2017))!
This will be a bit of a cross-post from one I did earlier on my website, with some updated information and thoughts, now that I've had time to think and review. I've reordered it a bit too, to help artists rather than being a bit "advertising".
First of all, here is what I ended up ordering.
They look like delicious hard candy, and I don't know whether that is a good thing or not. I may see about experimenting with swirls of opaque black within them, to look like little wisps of shadows trapped inside the dice. But that's something I will be doing over the next few months - working out the proper materials and ensuring it fits the theme, looks vibrant, and works for colourblind players.
Every copy of Lucidity will have 80 dice; an equal mix of 4 colours, representing each type of nightmare that players will encounter in the game. Each of the colours has its own unique mix of icons, and each icon was custom made for the game.
The early prototype online icons (for the Tabletopia demo) were by Lorc from game-icons.net (a fantastic resource for iconography in games, and one relied on by games like 5-Minute Dungeon). I highly recommend looking there when you are starting. They have really nice, clean, clear at any distance icons that are made freely available under a Creative Commons license. Give credit and you can use them freely.
I used a bleeding eye to represent Shadows, a moon to represent Exhaustion, and gnashing teeth to represent Hunt. The 1 and 2 Points were represented using numbers.
My hand-made icons (drawn onto stickers placed on the dice) were a little different! They had to be simple enough to draw 400 of quickly, but recognisable when they started to smudge (which they did - I had to replace them every 50-100 playtests in the end). My initial design also needed to be modified as the "Z" I used for Exhaustion was often mistaken with the "2" I used for the 2 Points. The second batch I made used a thick crescent moon next to the Z to distinguish it from the 2.
When it came time to make the icons for the final dice, I went through a couple of different ideas. I wanted an eye-catching eye (okay, that isn't even a "low-hanging fruit" pun, that's just a "fruit that's fallen on the ground and been trod on" pun - my apologies) with smudged mascara as the main Shadow icon, which was the focus of my design.
For the Hunt icon, I started by looking on game-icons.net again for inspiration, then tried drawing teeth like an orc is usually represented with. Eventually, I started playing with the idea of teeth that also looked like flames and, through that design process, created the goat skull.
Scanning in those pages, I used the GIMP (as I cannot afford Photoshop) to lay out the icons. The Shadow didn't change at all, while the Exhaustion, Hunt and Points had a couple of different options. I sought feedback from a number of artist groups and slowly made improvements to the design over time.
In this first iteration, I was already thinking about placement of shapes, widths, etc. The reason the Points and Exhaustion all had circles around them was so that players wouldn't be able to feel out the symbols by touch - drawing dice from the bag must be random!
Comments were generally positive across the board, but the thinness of the Hunt icon's nose and the line around the Exhaust was commented on. I also noticed that the Shadow's secondary smudge mark looked odd and didn't fit the curve of the main smudge mark (so I fixed that, and increased the eye dot so it would look clearer at a small scale). I redid the icons and, this time, did some photo-manipulation magic to see how they would look on the dice.
In the second run-through, I used the symmetry that people loved about the Hunt icon's horns; but many commented that they loved the weirdness that came from the asymmetrical horns. So I applied that to the nose, resulting in a symbol that (upside down) looks more flame-like than lotus-like. I think it also has that "tentacle weirdness" without being too over the top.
Now that I was getting down to the fine details, the comments started changing. Primarily, people pointed out that the symbols were all very asymmetrical, except for the Exhaustion and Points, which were very even.
So I revised the Exhaustion icon by adding an asymmetrical moon and zZ instead of Zz to balance out the icon better. I tried something radically different for the points, based on an artist's suggestion. King of Tokyo uses icons instead of numbers for its "points" (ie the Lightning bolts). I hoped doing the same would unify the symbols on the dice.
But after countless tries, nothing seemed to fit. The icons looked like ninja stars or eyes or starbursts. The theme didn't make sense (perhaps footprints would have represented escaping Nightmares better, but that was too close to the claws on the King of Tokyo dice). And I started thinking, "Why did I replace the numbers anyway?"
One of the designs I really liked was eyelashes that looked like tentacles. It didn't work thematically, and people thought they were fern fronds, but it inspired the crescents that ultimately surrounded the numbered icons.
I went back to using numbers, customising a font and adding tentacle shadows to fit the theme better and to give it that unified curve shape, so that players could not sense which dice was which by touch alone.
Here are the final icons, and below are them translated to the dice (with production issues discussed below, if you are interested).
Now I just need to work on colour: making those greens a little yellower, adding less orange to the yellow, and darkening the red a bit more. Again, I want to experiment with swirls of black in the material: giving a feeling of captured shadows.
The Order of Prototypes
My initial order of dice was ~2500, produced so I could split them up into prototype copies to be sent out to reviewers. I spent a long time designing the icons to fit (as you can see above).
We had a couple of choices. In short though, we could either:
1. Engrave (costly for such a huge volume)
2. Create acrylic molds
3. Create resin molds
Resin molds are much cheaper (~$120) and take less time to produce (20 days for resin molds), but resin dice are individually double the price of acrylic dice (~$0.11 vs ~$0.06). Normally, that wouldn't be much of a problem. But with 80 dice per game, I was looking at 40,000 dice minimum and about 80,000 dice if all went well with the campaign. Acrylic molds are very expensive, take longer to produce, but the dice are the cheapest possible. (In numbers, we would need to produce ~750 copies of the game to break even between the two.)
However, these dice were going to be sent to reviewers, so they needed to look good and they needed to be done quickly. I went with resin molds in the end because it was the cheapest up-front cost. For only 2,000 dice, I didn't want to spent almost 50% more than that on acrylic molds.
In the campaign, I figured I could start the pricing as if the molds were resin, but if it hit enough sold to make it worthwhile to swap to acrylic, I could take that $4 per game saved on the swap and invest it in the acrylic molds. That way, any further savings could be used on stretch goals (expansion dice etc).
I prepared the final icons as carefully as I could, putting each collection of dice icons into its own folder, using a labelling system (above) for the files to enable the factory to produce the dice without any mixups. I designed the direction of the icons so that rotating one dice 360 would have all of the icons facing the same direction. (The pantone colours were taken from a normal pantone booklet, with transparency specified by pointing to Chessex samples.)
Communication and Assumptions
There was a bit of a mixup in communication between the manufacturer and the factory. Because the volume was so low, the factory assumed I would want the dice engraved instead of cast in resin. Why would I waste money on resin molds if I was using them as a one-off print run? Of course, they didn't know that I wanted to save the molds for later.
As a result, they appear to have cast blank 12mm d6's in the pantone colours I asked for, then used a machine to etch the dice and painted in the icons.
As you can see, there is a bit of "drift" on the icons. I am convinced this would never happen with a mold, because the icons are literally molded in place. If the icon is drifting, the whole dice would be messed up. This is a bit of a point of contention, and I need to sort this out before the final product. What is acceptable for review copies is not acceptable for a final product.
I was given a few options at this point. Did I want to scrap the dice and have the molds done and the dice redone? Or did I want a partial refund on the molds (that I no longer had available) but keep the dice? I went with what made the most sense: take what has already been made and sort out the quality issues in the meantime. I have gone through those 2,500 dice and put aside about 800-1000 that had significant enough drift (or missing paint) to matter.
As I mentioned, there were also issues with the size of some of the artwork (according to the manufacturer). Now, this is something I have to raise with them after they come back from GAMA, but I was told that the pointy bits of the 1 Point icons were causing issues, hence why many of those sharp tips were not filled in with paint. But of those 1000 or so misprints, many of the ones affected by the missing paint had paint missing from the thickest parts of the dice (around the Shadow's iris, for example). So I suspect it isn't just the thinness, and this is something I plan to get to the bottom of - especially as the sample pack from that manufacturer had perfectly filled in areas much thinner than mine! These areas are easy to fix though, with a paintbrush and some white paint. I have also been told that using larger dice would completely solve this issue. So that is something to consider.
The engraving process also left distinct "carve" marks in one direction along the icons, and made the edges of the icons a bit "fuzzy". It isn't particularly noticeable from a distance, but up close you can see it. So if you are planning to engrave your icons, you should consider whether that will be an issue for you!
How can you design dice icons to not have these issues?
First, where drift is concerned (if you are engraving), make sure your icons are not so big that there will be issues if the icons drift around the dice face. For moulds, it's worth considering that space too, just to centre the icons nicely on the dice.
For the filled paint issue, there are some widths that have interesting (positive and negative) consequences. The issues with the Points not filling in was covered above.
On the Hunt icon, the area in the top (around the curve of the horns) was surprisingly okay. I expected issues there, but it worked out perfectly. However, the area around the skull's cheeks was not so lucky! As it turns out, that area was too thin, resulting in paint overflowing onto the surface and no clear line. By happy accident, this looks amazing. The white paint fills the gap, but the raised texture effect remains. This adds a nice 3D quality to the skull. I'm not sure how that will look with a molded icon, and I'd love the certainty of having a mold done already, but at the moment I can't afford to.
So I'd say to keep about 1mm of thickness on areas you want to be painted in, and I think about 0.5mm of thickness on areas you want raised and not painted. That said, even the thinnest parts of my icons were painted in the best cases, so this could just be a QA issue rather than a graphical one. More on that once I hear back from the manufacturer.
I am really happy with the results, and the process has been lots of fun (even though I had to do a bit of QA and TLC and other acronyms on some of them). What will go to reviewers will be a great representation of what Lucidity will look like, but hopefully those reviewers understand that the material and quality is a prototype level. Meanwhile, I have fancy dice to use for demonstrations at conventions! (And to give away!)
Anyway, I hope that was useful. I'm happy to answer questions if you have any.