Picking the first 13 generators

When I started putting Rollforge together, I almost made forty generators before launching. That sounds ambitious for about ten seconds. Then you remember that every generator needs good data, a sensible page, and testing.

I would probably still be adding elf middle names if I had followed that plan. So I picked thirteen.

That felt like the right amount of “this is a real site” without becoming a promise that I had to finish every idea in my notes app before anyone could see it. It also gave me enough variety to learn which kinds of tools were actually worth expanding.

Why thirteen, not forty

The first version needed to feel like a place you could browse, not one lonely name generator with a very optimistic navigation bar. At the same time, it needed to be small enough that I could finish it without hating every word list by the end.

Thirteen gave me a useful constraint. I could cover a few obvious fantasy and TTRPG needs, add some non-medieval stuff, and include a couple of tools that have nothing to do with dice. More importantly, each one had to earn its spot. If it sounded funny but had no clear use, it went back in the pile for later.

That is a decent way to pick first generators for any small tool site. Start with enough range to test the idea, then leave yourself room to react to what people actually use. A launch list is not a lifetime commitment.

The categories that made the cut

Fantasy names were the easy call. People need elves, dwarves, taverns, kingdoms, and general character names all the time. I use them myself, and there is a nice difference between a broad Fantasy Name Generator and a more specific tool that has a stronger sound.

TTRPG tools came next because they solve an extremely real mid-session problem. Someone opens a door, captures a scout, searches a crate, or asks what the inn is called. Suddenly you need an answer before the table goes quiet. Random loot, NPC helpers, and dice tools are built for that exact kind of panic.

I also did not want the whole site to look like it only works if you own a cloak. So there are sci-fi names, usernames, band-name type ideas, passwords, and a decision spinner. The spinner is not deep technology. It is mostly for people who cannot pick lunch, which includes people I know.

You can see the full mix on the generators page. It is deliberately a little uneven. That is part of the fun.

One engine, many word lists

The useful technical choice was keeping the generator interface shared. The individual tools mostly differ in their lists and assembly rules, not in how a person clicks Generate. That means adding a new idea does not require rebuilding the page from scratch every time.

There is no AI model quietly inventing results behind the scenes. Most of the work is arrays, weighted choices, and enough hand-picked pieces that the output has a recognizable flavor. A dwarf name should not sound like a sci-fi ship by accident. A random loot result should still give a GM something they can use.

That setup also lets me improve a generator without touching the rest of the site. If a list starts feeling repetitive, I can add better material there instead of redesigning everything. It is boring in a good way.

What I would add next

I have plenty of ideas, but I am trying not to turn that into another giant launch list. More race-specific names make sense. So do more quick TTRPG tools, especially things that help a GM recover when players immediately ask about the one detail nobody prepared.

For now, I am more interested in making the first set solid than racing to inflate the number on the homepage. If a generator is fun once but annoying by the fifth click, it needs better inputs, not a new neighbor.

Thirteen got the site out of my head and into a browser. That was the actual goal. The rest can show up when it is ready.