Home » Post Mortem – Wizard Tower Defence

Post Mortem – Wizard Tower Defence

This post was rescued from my drafts in July 2018 and backdated to its current published date.
Wizard Tower Defence is available on itch.io, free for Windows (and probably also Linux and OSX, but I daren’t say so because I haven’t tested that).
This post is a retrospective for the game. At some point I should separately write a more general look back on two years as an indie dev.
It’s long, but I don’t see the point in serialising it. Skim for the headings you like, if that’s better for you.


For a game made mostly by one person, I’m pretty pleased with it. For a free game I think it’s probably pretty good. As a game in general (since consumers aren’t interested in its development constraints), I’m disappointed, which is one of the reasons for releasing it free.
The design works, but I think a simpler riff on the normal ‘tower types’ mechanic would have worked better than the ‘innovative’ ‘wizard combination’ that I tried instead.
The art direction goes a long way to make up for the lack of art (and certain tools were a great help); in a sense I think it’s the biggest achievement on the project. Overall the art clearly isn’t professional quality, but I think the game gets away with that quite well.
See below for more detail on each of these things, plus code, audio, business, etc.


First things first, since I can imagine this being an obvious question*: why ‘free’?
* I don’t actually expect this to be asked much, since a typical player’s assumption will be that its free because they deserve it. But as a development retrospective, the question of “why aren’t I trying to maximise short-term revenue” is a worthwhile one.
In fact I’m not trying to maximise any revenue, because I’m about to start another career, and will probably use my ‘brand’ just for some fun prototype stuff in future (assuming I even ever have time for that, which certainly isn’t a given). So it’s not a loss leader.
I have no marketing budget; I could have, if it guaranteed sales, but nothing ever does. I had very little time (and virtually no inclination or energy, not to mention not enough skills) to do a reasonable job of marketing myself. On that basis, I can expect to shift approximately no paid copies, or few free ones.
I also know what I’m doing next, and where the next money comes from. So while it’s painful to admit the extent to which I’ve failed as an indie dev (financially: having finished and shipped something was always a strong backup objective) I don’t need to squeeze every last dollar out of this project, even if it could bring in enough to make a difference.
There’s also a question of quality. I’ve looked around itch.io (the Somewhat Unlikely account owns no games, but my personal one has a fair few) and I know there’s some rubbish there; Wizard Tower Defence is better presented than that (and hopefully generally better), but it’s not at the point where I’d be comfortable asking for ‘proper’ money (a few quid, for example). I don’t think the market has really accepted really cheap games (less than that), which get the disadvantages of being free and those of having a price.

The Name and Other Marketing

The other headline WTD that I might get is about the name: some people will think it’s a terrible name, that I’ve put no effort into it. They’re half right.
A good product name is memorable, catchy, evokes the product’s values and ties back to it.
‘Wizard Tower Defence’ is not a memorable name in the normal ‘sticks the brand in your mind’ sense. However, it is a name you can guess at provided you remember anything else about the game; this is often not true with more abstract games. It’s not really a catchy name, working as part of the whole for a memorable brand, for the same reason, but without the resources to build the rest of that memorable brand, I don’t think it’ll suffer much from that.
I say ‘ties back’ because many names don’t describe or illustrate the product, but instead rely on marketing to raise awareness until the market makes that connection in other ways. See my above note about marketing budget; I clearly needed something that said what it did on the tin.
Besides that, the strength of the name is in evoking the values of the game. I think it goes well with the sharp, clean art style and the blunt (but somewhat pithy) narrative.
The most pressing reason was that as I approached release I really couldn’t be bothered to call it anything else. I’m not a marketer, I’ve had mixed success naming things in the past, and the intellectual exercise of finding the perfect name in the normal mould was more than I wanted to take on.
In hindsight, I should have checked better for competitor uses; there are a couple of similar things, but nothing I’m too worried about from a passing-off or copyright point of view.
I don’t actually have much to say about other aspects of marketing. I’m going to mention it on social media a couple more times, but otherwise not worry about it.


I’m a game designer; I have been for 9 years now. Design was supposed to be the thing I’d get right, but inevitably it would be the part for which I’d set the highest standards. I’m disappointed and I know it’s not unusually good design, but rationally I think I it’s probably good enough.
The design of WTD is fairly simple, but it does what was intended: it hooks into the normal build-kill-earn loop of a tower defence game without too much in the way. The new features should have been better taught, but what game can’t be accused of that?
I tried to innovate and it didn’t pay off. The wizard combination mechanics are different to a normal tower defence, and hopefully that’s a little interesting, but they don’t add enough and I think that I could have made a more ‘fun’ game by sticking closer to the traditional ‘choose your tower type’ model. (Indeed, I’m now haunted by other ways to deviate from it, that I now don’t have the time or interest to make.)
There was a time when I considered trying that. It would only have taken a couple of days to prototype a ‘traditional’ system of buying towers by type, even with a branched upgrade system. It was late enough in the day though that I was already thinking of it as a non-commercial game, so I decided to keep the distinctiveness rather than make ‘just another’ TD even if that might play a little better.
Throughout my time as an indie dev (and so probably a thought for another post) I’ve been haunted by the question of whether I’m actually any good at being a game designer. Sure, I can design stuff to spec, and it’ll work well with the rest of the project, be safe and robust, and so on. I’m also a handy person to have around, for teamworking and other professional skills. But I don’t understand fun, I don’t know how to maximise success while innovating, and generally don’t feel like I could be the kind of visionary designer that the industry seems to idolise.
The only saving grace is that seeing how often those projects still fail, and how often a hit game will be a careful mix of proven things, I’m not convinced that anyone else is that kind of visionary designer either, not delivering with much more than a random chance of success.
One thing is abundantly clear: I can’t design well on my own. I’ve been [often unfairly] accused of designing in some kind of ‘isolationist’ way, and it is true that I want to finish a draft before anyone gets the wrong end of the stick, but I’ve also been a good team designer, giving and receiving valuable feedback, and without that I just can’t iterate quickly or effectively enough.


The game works, and there’s not much more to say. One playtester reported a bug that I never managed to reproduce, but everything else was fixable. I didn’t cut any features for want of being able to code them.
Script performance is acceptable: it might even be ‘efficient’, but the stuff I was trying to do generally pales in comparison to the graphics processing.


The art direction was carefully chosen to try to maximise the amount I could do on my own. Feedback has been good, and I think it works pretty well to deliver something reasonably attractive despite my lack of skills. There are obviously places where a professional artist could have done better (I don’t think the ‘sharp edges’ aesthetic needed to extend to crystals and other props, for example, and I never did remember to replace those trees), but it doesn’t look like a terrible amateur job from most angles.
The towers were made with Archimatix, which went a long way to help me fulfil my art needs using my technical abilities, since I have little manual skill (I’m not completely without an eye for composition, but my hands-on skill in art packages is basically nil). While this cut down on my dev time compared to traditional tools, I was still frustrated with the pipeline; not long ago Archimatix released real-time parameter modification, and if I were carrying on with development I would try a pipeline based around that, which might have let me have much more tower variety.
Other environment art was made ‘normally’, in Blender. It took me a long time (if you’re a competent artist you might have trouble imagining how slow I am; picture yourself learning to code), but it was an interesting learning experience and worth doing.
Unity makes particle systems pretty easy, and I found that once I’d practiced with them for a while I was able to do the 3D ‘character’ art I wanted with just spheres. They might have been better as little robots or (*gasp*) creatures, but I think that along with the background of ‘elemental motes’ that I put into the narrative, I probably get away with the spheres.
I was dreading needing character art and a logo, even to the point of getting a quote for outsourcing them. In the end I couldn’t really justify throwing money at it (since I knew it was money I wouldn’t make back), so I did it myself, using HTML5 canvas drawing commands to build them up in a non-destructive way that I had complete control over. As an approach this is incredibly slow, but probably (for me) still quicker (or better) than learning to do it traditionally. I would really love to break some of that code out into a reusable tool, but it’s a long way from being user-friendly.
As it is I know that’s not professional-quality art, but it’s not bad for designer/programmer art. The little logo/banner icon is better than many free games on itch.
Icons are free third-party ones, either from game-icons.net or Icons8. As always, it’s tricky using what is already available (and that probably doesn’t change when you pay a little, only when you pay enough for bespoke), and I wish certain icons (especially attack types) were a bit more illustrative or distinctive.
Overall I’m very happy with the art. The game would have been better with professional art, and more of it, but that would have been prohibitively expensive unless I’d had an artist as a partner, and by the time I got to this game is was already too late for that (as I’ll discuss in detail in Somewhat Unlikely’s overall retrospective).


The audio was from the asset store (see the game credits), so I take credit only for choosing it and writing a little system for playing it dynamically. It was hard to find asset store music in parts (not layers) that would give the dynamic effect I was after; as it happens the tracks I used were free, but I only found them by looking up an artist from whom I’d previous bought a paid pack.


As in the summary, overall I’m pleased with WTD as a one-person project, especially given all the skills that I have. But as always, I’m mindful that consumers won’t generally be aware of those constraints, and more so that they don’t owe me any concessions because of them. Because of that I had high aspirations for the game, and it doesn’t really meet them.
Although I’ll probably say more about this when I PM the whole enterprise, this has been a common theme throughout the last two years (and to an extent my studio career, and even my life in general). It’s not the only reason I’ve shipped only one game in two years (and it’s not two years’ worth), but it’s a major factor. In fact, without the imminent change of career, and the prospect of ending two years without shipping anything (which really would be a failure), I might not have released even this.

Name of author

Name: Matthew