Custom Tiles: Before or After?

For custom or commercial projects, do people lay down placeholder tiles (such as from the generic list) or is making the tilesets normally the first thing that happens?

It probably depends mostly on whether you are an artist or not. If you can make your own tiles and enjoy it then go for it, if you can't and are eventually going to have to shell out hundreds or even thousands of dollars to get your game looking exactly like you want it I'd say it's probably more economical to use default/placeholder graphics until the game is in at least a late alpha state so you can be confident that as a system it is going to be complete in short order.

I know a couple people here who have done the latter and then cancelled their projects, leaving them in the hole for a whole lot of $ for a game that will never see the light of day. That's got to be a devastating feeling, at least as an artist you'll have the learning experience to look back on and your time making custom assets won't feel like as much of a waste as a result. But as a patron with a failing game that's all just wasted money, so I'd definitely think it's better to wait until you have a stable game before investing heavily, especially if it is your first game.

This would be my first RPG Maker game, yes. I've done others in RenPy.

If it's your first rpg maker game, then definitely make good use of placeholders. It's easy while you're thinking up details of the game to go "oh, I need twenty different types of rocks, because reasons!", but by the time you get around to actually making the game you'll most likely end up realizing that you never even needed more than five. Even if you're the one making everything for your game, that realization that you spent time on something that you don't need can feel pretty awful.

By using placeholders while you make the game and keeping a list of necessary resources as you progress, by the time you're ready to start replacing the placeholders you'll have a list of exactly what you need (and for things like tile sets, then you just need to swap out the sheets and barely have to worry about remapping everything).

I don't use placeholder tiles nor custom tiles. I'm pretty happy with RTP and any tiles that has similar style.

However, one of my friend once admit that using map with placeholder tiles might be a bad idea. Because once the actual tiles is ready, there might be a case when you need to map everything! So this particular friend of mine now always suggest to use final tileset rather than using placeholder.

For my current project, I am making a custom tiles. But I mostly develop as I go.

I initially put together a number of basic components (e.g. some water, some ground, wall, roof options). But as I develop new regions, I either create as I go or make a note of something I need and use placeholders in the interim. The farther along I get, the less placeholders I need. If you're not making your own assets (i.e. paying), I would probably do differently, using far more placeholders, since tiles are expensive, and I'd want to be fully invested in a project before shelling out dough.

This lets me spread the workload over time since as a solo dev, I tend to work on music, art, scripting, eventing, etc. as I feel compelled rather than in any structured fashion. For example today, I was working on a section, realized I needed a new character sprite and battler. So in the middle of eventing I took a break and made those up. I'm not sure it's the most efficient method, but it works.

I've done both, at different times for different reasons. As mainly a developer of turn based strategy games all I need to do to test out a battle map quickly is lay down floors, walls and objects that block line of sight and then I have a workable, easily changeable if very ugly map for testing purposes. But I can only do that for strategy games, if I'm developing a more traditional map that the player has full movement control in, I wait until I have a complete tileset because the placeholders might look entirely different to the tileset I end up using because if I'm using something like Time Fantasy, I often don't know exactly what tiles I'll have to work with until I purchase them, so I can't possibly design a map with appropriate placeholders because I don't know what they'll be placeholders for until I buy the finalised assets, and then it's a moot point anyway.

I spent the first few months of my development cycle making my own map assets last year. It was a pain in the ass but it was worth it since now I can put together a basic, original map pretty quickly and then fill it in with features/details later on. Personally I have the following formula for my maps:

1. Create the map in its most basic, blocky, undetailed form using a handful of RTP tiles in the maker. REALLY basic. As in I just mark out where I want ground/water/walls/buildings, and that's about it.

2. I use the program's export function to open that map in Photoshop and lay down my own assets to create something that's still basic, but serves as a foundation for the map, but still without any real details, aside from laying out paths and other features, which will let me decide what areas will be walkable.

3. After I have that basic version of the map done with my own assets/art, I save it for parallax mapping and load it back into RPG Maker, so I can actually see it in-engine and make sure the proportions look right.

4. Finally, down the line after I'm ready to flesh the map out (usually when I arrive at a point in the story where I need the map to be mostly finished), I'll go back and do the majority of details like plants, items, objects, etc... to make it actually look like a real map instead of a prototype.

5. With that mostly finished version I divide it up into the parallax layers and export it back to RPG Maker. Later, as development goes on, I finalize them with any specific details or changes I need made, but at that point they're like 95% done until I'm entering the end-stages of development.

(mind you, I don't do all those steps in a row. I jump back and forth, sometimes I do a whole batch of basic maps at once in a night, etc...)

I don't just use placeholder tiles, I use placeholder maps. To give an example, a beach-style town in my current project has minimal buildings (as in, just squares with a door and autotile roofs) and no decorations, with really nothing more than the interactions required to advance the story. The idea is to build everything as quickly as possible, then go back and fine-tune everything by figuring out what custom tiles I'll want for that particular map. That way, if I decide to cut that town or change its theme completely, I didn't waste time on custom tiles for it or mapping it in general. Besides, as someone else above said, when converting from placeholder tiles to "real" tiles, you'll probably be remapping everything anyway.

@Aesica that's brilliant! I don't know why I did not think of that, but that seems like the most logical approach given how tilesets are very time consuming, tedious, and if commissioning then costly as well to make that having placeholder maps is the way to go for an initial pass of the game. On top of that, it keeps the game a minimum product for helping get the point across when seeking feedback especially when your game goes from "start" to "finish".

These are all fantastic ideas, thanks!

I would also add there's a ton of dlc tilesets and graphics fairly cheap. The game I'm working on uses the Pop! Sets and I spent less than $60 for it all including a music set with 12 songs and sfx(they were on sale last Halloween). So you don't need to spend hundreds or thousands of dollars honestly unless you just really want unique graphics for your game only. But I have yet to see a game use alot of those sets that are any good. I believe with most you are able to use them commercially, at least, I know you can with the assets I got.

I think that if you're going full custom it's good to use a placeholder, but not necessarily something like the RTP's tiles. I'm talking about using original placeholders, in blocky, flat colored scribbles that take a few seconds to draw. Using the RTP as a base for your mapping will influence the art, and the end result of both the mapping and the art will be more static and generic than what you would get if you let the art decide how it looks best instead. Using roughed out placeholders would give you enough information to make maps and continue on with your game in a quick fashion, while still giving you the flexibility of working original. Then you just substitute the sheets with the real art and everything just works. This is great for the artist too, because the more you know exactly what you want, the easier it is for an artist to do. It also makes it easier to improve as you go, making the important stuff first, and working to the less important. If you're using the tiles you know which ones are the most useful and which aren't used as much, and so you can adjust and focus on the things that will give you the most value as you go. If you run out of time/money, you'll still have the key stuff.

As an artist, I hate making something amazing and having it never used. I hate not knowing what to do, drawing something blindly, hoping that maybe I'm close, and having to redo things over and over until it looks "right". I hate having something I worked really hard on be cut out and never used. So yeah, the less you can do those things, the happier the artist is.

There was a simple study done where they had a bunch of random people answer a questionnaire. Everyone knew that the questionnaire was pointless and would be thrown away. There were three groups, one who had their questions graded as soon as they turned them in, one who had someone take it and put it in a pile but otherwise didn't look at them, and one who would have their questions shredded right in front of their eyes. The end result of what happened to the paper was exactly the same, and they all knew it, but the happiness and satisfaction levels were very different. Group one was very happy. Group too was actually still pretty happy, but less than group one. Group too was upset and dissatisfied. So, the takeaway is that whatever you do, make sure the people you work with feel like what they do is valued and useful, even if it won't be used, and they'll be way happier and more likely to work with you again. (Slightly off topic, but I feel useful in context.)

The truth is, even if you're the one doing all the art, it's expensive, either in time or money. It's a lot of work and just takes forever. So it's better to be careful about it, treat your artist with respect (especially if that's you), and value that time and effort. Make it as little work as you can while still being awesome and your wallet and artist will be happier all around.

Do the placeholders first; add the custom tiles later.

Custom tiles expensive (unless you create them yourself), time-consuming (especially if you create them yourself), and liable to cause burnout (either way). Additionally, you may find during your testing with placeholders that certain layouts don't work or certain features aren't fun - which means you can figure out what needs to be reworked before you invest the time/effort/money to create custom graphics.

With that being said, if you're not going to be creating the tiles yourself, it may be a good idea to have a lead artist on your team early on if it's feasible - to discuss what kinds of maps, places, and feelings you are going to create. They may have ideas that you never thought of to really nail a general theme, or they may have a style which affects the kinds of map layouts that will look good in practice.