PokemonRL 1.2 released

Oh my… this project is surely a lot of work. I’ll probably release another version with some usability improvements and leave it alone for a while, else I’ll stick to it forever (and I need to go back to generating revenue)…

Enough ranting! this versions comes with HUGE changes, you can play the latest version HERE. Changelist below:


  • Show stats of nearby visible monsters


  • Use speed stat and modifiers to decide turn order.
  • Use evasion and accuracy (with modifiers) for battles
  • Fix type damage modifiers, complete type x type table and use secondary types
  • Add attacks and moves modifying stats temporarily.
  • Implement FOCUSED status (increases critical attack chance)
  • Make native evolved versions of monsters inherit moves from their pre-evolutions
  • Prevent monsters from learning moves already on their list


  • Fix issues with gym definitions

PokemonRL – v1.1

The game has now gone beyond the 7DRL scope into a more full fledged development. The original 7DRL and the 7DRL++ versions have been archived for reviewing and historical purposes. All advancement has been pushed to the github repo and is also available to play online here.

Implementing the full generation one move-set is a lot of work! (ORLY?) I just managed to add the data for all 151 pokemon, their 165 moves and the 735 combinations of pokemon/move/level which represent the moves they can learn. I used data from all around the Internet for this, specially from Bulbapedia and this gamefaqs FAQ. I could have instead used the veekun pokedex database for this but was too lazy to set it up. Seriously, still can’t understand why I overlooked this for the 7DRL planning. I should note that I’m restricting myself to Gen1 and even so, it’s not something to take lightly.

I’ve done good advancement on implementing the moveset, initially focusing on the attack moves and their different parameters. So far there are 41 out of 165 moves covered. (There are a bit more which are partially covered).

Some of the things to do in this regard are implementing Stat penalties and bonuses with the “stages” model, and implementing effects for the different status modifiers.

That being said, I don’t think I’m going to focus on completing that initially… I’m more interested in the procedural aspects as well as the adaptation of some particular skills into a “single-mode” roguelike (no battle screens). Some interesting things to do include:

  • Simulate pokemon trading somehow, since some require that for evolving.
  • Make “Dig” let the pokemon actually go underground.
  • Make “Fly” let the pokemon soar the skies.
  • Affect terrain with skills such as flamethrower and earthquake

As for procedural generation, now that all pokemon are available I added some more route stereotypes (which should allow more varied wild pokemons to show up). I made these manually, using the wild pokemon spawn data from bulbapedia for the different routes. I decided to stop doing that and instead work in generating this higher level metadata mixing the pokemon data, route difficulty and biome/weather selected for each road, on a next release.

I did some small tweaks on the town level generator to add a “fence” around them. I also modified the metadata generator so that the first town has a single exit to route 1, since people was getting stranded into high level routes and killed by deadly rattatas of death. This should make starting the game much easier.

PokemonRL 7drl++ and continued development

Even tho I’m not completely happy with the results and lots of things could have been done better, I decided to mark PokemonRL as a “Finished” entry for the 2017 7DRL Challenge, which means it’ll get feedback from the evaluation committee and will be more visible to players.

The main reason for this is I feel it’s stable enough and I managed to implement a roguelike with a winning condition which is interesting to play. The one big thing I missed, adding all required data, makes the experience feel incomplete, but it still meets the challenge criteria.

I have released the 7drl++ version which fixes 3 critical issues:

  • Stat calculations (Specially HP which was off by about 100 HP)
  • Remove test move sets from some pokemon races
  • Prevent messages from disappearing on the text box
  • Fix amount of damage recovered by potions
  • Make starting monsters be level 5

You can check it out here

The original 7drl submission without these fixes is also accessible, of course.

Screen Shot 2017-03-14 at 7.52.38 AM

Now that the challenge is over, I decided to continue working on it to finish my initial vision. I’m gathering data for all Gen 1 pokemon in order to load it into the game.

2017 7DRL, PokemonRL…. FAILURE?

Alright… this year I just couldn’t make it. You can play the current version here if you want to take a look tho.

(Edit: Later on I decided the game was worth considering as a Finished entry, even if didn’t meet my whole expectations)

I made so many mistakes during the challenge that it seems I deliberately ignored every single piece of my own advise:

  • I created an almost complete roguelike engine from scratch instead of creating something lighter.
  • I didn’t have a clear goal and I spent too much time wondering what to do and switching plans.
  • I aimed blindly for an overwhelmingly big scope and didn’t think on an initial increment.
  • I didn’t make time to playtest the game nor gathered playtesters for the task.
  • I worked alone instead of building a team.
  • I didn’t have a plan!

And finally, the biggest mistake I made was not devoting enough time to add quality content to the game (a critical step of which Jeff Lait, another veteran 7DRL warrior, has warned about time and time again). This was basically what caused me to consider this a failure… I worked a lot on it, and it’s actually interesting, playable and it has a winning condition and some nice mechanics. There’s a pretty awesome empty cocoon which will require a lot of work to fill in order to provide a fun experience.

Sum all the mistakes above, and add up the “real life excuses” (Lost a day recovering from my trip to California and almost 2 full days in the release of Ananias 2.2, and it wasn’t an easy week at the office…).

I may have been able to pull a mediocre “successful” release if I had neglected my family even more, but I feel better I could spend some time with them and bring them some comfort in times of conflict. 

Enough crying… on the plus side, I managed to:

  • Create an interesting procedural generator for pokemon-like worlds.
  • Adapt elements from the widely known Pokemon universe into a roguelike format.
  • Create a simple command set, a friendly UI for new players.
  • Create a Javascript roguelike template (“JSBoilerRL”) anyone can use to create their own roguelike (I’ll have to write a tutorial for that)

And probably for the first time I feel good about the source code of the game! it’s mostly clean, readable and extensible. It’s now open source… I’ll come back to it soon if enough people cares about it, to make a proper first stable release.

One of the think I liked the most is how I made tall grass work in the “route” maps… you have to dive into it in order to find pokemon 🙂

In any case, here’s a gameplay video of PokemonRL, enjoy!

Here’s what I managed to do in day 7

  • Add Power Points to skills
  • Add Pokemon center to recover all HP and PP
  • Add Lab building to pick starter pokemon
  • Allow giving up on gym battles
  • Pokemon level up, learn new skills, evolve
  • Select random town names
  • Allow winning by getting 8 badges
  • Add inventory limit


2017 7DRL – PokemonRL – Day 6

Day 6! time’s almost over and I wonder what’ll be the best way to wrap up 😐

Today I managed to implement high level map generation (so it now generates both a random overall structure as well as the actual detailed maps). They are both simple, of course, since there’s no time to go in-deep. This includes assigning the gym data, pokemarts inventory as well as wild pokemon on the routes.

Sample metadata (Click to enlarge)

Spent a lot of time debugging an issue with the validations to prevent roads from overlapping not working.

Pokemons also respawn now on the routes, and I started implementing the actual skills and the damage calculation. (Based on the information from Bulbapedia)

I also implemented an alternate display mode which looks a bit like a gameboy color. At least to me 🙂

I’m less than 24 hours away from the end so… what’s the plan?

I don’t know.

It frankly looks grim, but I knew it wasn’t going to be easy (that’s why I hesitated). There’s enough material here for a seven years roguelike. The challenge will be wrapping up decently.

Pending stuff (not all of it will make it):

  • Lab building to pick starter pokemon
  • Pokemon Center buildings to heal pokemon
  • Pokemon “fainting” and losing gym battles
  • Pokemon level up, learn new skills, evolve
  • Place items randomly on map
  • Sell items to earn money
  • Adding a good chunk of data for pokemon and their skillsets.
  • Special attacks
  • Trainer battles to earn money
  • Restock stores

2017 7DRL, PokemonRL Day 5

Managed to advance a lot today! there’s still a long road but things are looking much better now:

  • Added simple “route” type map generation (A straight road with some patches of tall grass).
  • Ported my Cellular Automata classes to JavaScript and applied to route map in order to make the patches of grass look irregular
  • Semiopaque tiles so you can look a little into tall grass but must walk into it to find wild pokemon.

  • Pokemons slots, selecting, releasing and pulling back.
  • Make pokemon fight
  • Catch pokemon using pokeballs
  • Use potions on pokemons
  • Pokemarts where you can potions and stuff

  • Pokegyms: Single room with leader, door closes when entering. Badge appear when all pokemon defeated.

Next up

  • Metadata generator (for map)
  • Spawn wild poke
  • Use poke skills (Ranged attacks, special attacks)
  • Poke level up, learn new skills, evolve
  • Trainer battles to earn money
  • Place items randomly on map
  • Sell items to earn money
  • Restock stores
  • Lab building to pick starter pokemon
  • DATA!

2017 7DRL Day 4, PokemonRL again

Decided to go back to PokemonRL because IT WILL BE AWESOME. Tomorrow I’ll change my mind again.

Scribbled a rough plan in my notebook on the morning:

  • Deploy poke
  • Make poke fight wild poke
  • Catch poke
  • Use potions on poke
  • Buy potions
  • Generate map based on pokemon red
  • Spawn wild poke
  • Town maps: Simple building placement
  • Gym maps: Single room with leader, door closes when entering. Badge appear when all pokemon defeated.

Did a lot of boring work the whole day. But then managed to complete rudimentary inventory management, using and dropping items for JSBoilerRL.

Started tinkering around level generation. As usual will take a layered approach based on examples to generate a cool pokemap. Following is the generators cascade:

  1. World Metadata Generator: Towns to place, name and theme / weather
  2. World Grid Generator: Town and road positioning and linking.
  3. Map Generator: Town-Route maps
    1. Every town will have a gym where you battle the leader to get the badge.
    2. Wild pokemon will be placed on the routes.
    3. All locations will be accessible (no route blocking)

There will be no caves / dungeon / towers.

World grid generator will generate something similar to this (found on the internet, author unknown)

Managed to advance on map generation based on fixed metadata


This is the current status, next up is completing a first iteration of basic map generation (including routes), in order to jump into combat.


2017 7DRL: Concordia Reborn – ongoing

Day 1 – Tinkering with ideas

Didn’t manage to code anything. Was still recovering from my GDC trip, used the little energy I had to work in fixes for Ananias instead.

However, managed to tinker a bit around the idea of a generator for a classic cRPG scenario, generate a world using fractals, place some towns, generate some simple quests on each town and had the player complete them to get an artifact on each and win the game. Initial plans were to have this be a kind of sequel to Ananias, leaning more towards a classic cRPG similar to Ultima 4, but reutilizing the engine.

Day 2 – Change of plans, start work on JSBoilerRL

Came back to the office to find a lot of stuff going on so it’s not going to be an easy 7DRL week but when has it been? I’m also working on an update for Ananias at the same time by the way.

Ditched the original idea and instead thought on reviving a very old idea: a fully traditional classic roguelike with a Pokémon theme.

Checked some gameplay videos for the Mystery Dungeon tiles to make sure there’s no overlap. The game will be different, based on the first generation games and with little plot. you’ll be a trainer and you’ll catch pokemon and give them orders to combat in a tactical grid map.

Started work in JSBoilerRL, a template project other people will be able to use to create their roguelikes in JS. It’s powered by unicodetiles.js producing similar output as my 2013 7drl, Rodney. Managed to implement walk in the map with collisions, FOV, memory and enemies moving around.

Was stuck trying to debug an issue with transitions between levels with stairway tiles moving around weirdly.

Next up is fixing the bug and start working in the actual pokemon mechanics.


Day 3 – More change of plans, JSBoilerRL continued

Spent some time designing the game:

  • You are a pokemon trainer, your goal is to become a pokemon master by defeating the gym leaders on all 8 gyms and then challenging the elite four and the current pokemon master in the Indigo Plateau.
  • In order to do so, you’ll have to travel the land looking for powerful pokemon, capture and train them to become stronger.
  • You can have up to 6 pokemon in your team at the same time, you can command them to use their abilities against particular targets.
  • You can store your pokemon data in the computer in order to free slots for new pokemon.
  • You can use items in pokemon, either wild ones, these in your team or your opponent’s. Items can be bought from pokemarts or found in the map.
  • The following items types are available: pokeballs (used to capture pokemon), healing items (recover HP or PP), status recovery items

I realized the scope would be too big for a 7DRL, mainly because of the required data and balancing to make it good. So I changed plans again to the original idea, only using a traditional roguelike UI instead of reutilizing the Ananias engine.

  • Randomly generated map with towns and dungeons
  • Place a quest on each town which can be one of the following types
    • Kill x monsters on a given dungeon
    • Find item
    • Rescue NPC
  • Complete all quests to win the game
  • Simple roguelike gameplay, traditional fantasy setting

Listened to almost all the Roguelike Radio podcast made by Jeff and Darren… some pretty good points there not only related to 7DRL but roguelikes and game design in general. Decided to try and make a classic roguelike and then include an unique twist which may be the game being combat-less, with your character being an traveling scholar fixing the world with the pen instead of the sword. Still don’t know how that can be implemented.

Struggled hard to fix the issue with the level transitions on JSBoilerRL… turned out it was caused by myself defining the object attributes as part of the function prototype instead of them being defined for every object instance (d’oh!).

Decided to lay down a plan for the completion of JSBoilerRL so I can start working on the game itself, basically wanted to make sure I’m covered with the basic which leaves the following pending:

  • Items: including inventory handling, getting, dropping, wearing, using
  • Message box

Managed to advance on item handling and inventory. Still pending dropping, wearing and using items.

By now I’m getting close to Rodney, which makes me wonder… why didn’t I use Rodney as a base in first place? other than having a cleaner codebase there’s no advantage on what I’m doing and instead I’m wasting more dev time… a very non-practical choice :/

Going to the bed hoping tomorrow I can start working in the actual game.