Hey Giantbombfolk,

For the past few months I've been working on a pet project that's loosely inspired by Google's April Fools joke this year. At the time I saw the video and thought to myself: "Huh, Self, that's neat and all, but what if there was an actual Pokemon game that took this idea and ran with it? I mean, Pokemon has been using real life data for a long time now... Why not take an extra step and use your location? You can do that, right Self?"

And so this project started in earnest.

My goal has been to create a true, full featured (Well, loosely speaking) Pokemon game for phones that allows you to walk around the real world and randomly encounter Pokemon based on the environment that you're currently standing in. Want to catch water types? Then go by the shore of a lake or better yet take a boat out to find the rarer types! Climb cliffs and find Rock types, explore woods and find different sorts of grass types.

I for one think it's a really cool idea.

At this point in development all of the random encountering based on GPS data is implemented and working quite well if I do say so myself! The accuracy of the position detection is highly dependent on the quality of the GPS data available, but there's precious little I can do about that. I plan on making a video demonstrating it next week.

I'm still working on the combat system. The core of it is in place... but It's going to be a very long time before I flesh out all of the moves for all of the Pokemon (I decided on just the original 150, so that the project stays feasible) the animations especially are going to take some time. After that comes item management, game economics, and finally network play.

Here's a video I shot of it today (Using my camera, because Android screen recorders are the woooorst.) The button I push is a debug button that I use for testing purposes (It's inconvenient to need to wander around the neighborhood every time I change a variable). Normally you would encounter Pokemon through just walking around.

Anyway, this post isn't just about me shamelessly flexing development muscles (Though it is that!) I had some questions that remain unanswered, and I'd like some feedback from you all!

So:

1: Where/Under what circumstances do you all think Legendary Pokemon should be encountered? Should there be a finite number of these Legendary Pokemon?

2: Should Gym Leaders, the Elite Four, and/or The Champion be incorporated in any way?

3: Without NPC trainers, the traditional source of income of Pokemon games is no longer an option. My current solution is to let players sell Pokemon on a shady-as-hell Team Rocket black market screen. Any better ideas?

4: Pokemarts and Pokemon centers don't currently exist as I have no way of placing them well globally. So right now they're simply a planned menu option. However there's some serious implications here... You shouldn't damn well be able to heal after every fight after all. Any ideas for checks and balances on this front?

5: What should determine the level of the pokemon you encounter?

I've personally got a lot of ideas for all these, but I wanted to open it up for discussion because working in a vacuum is a pretty awful idea. So discuss away gentlefolk! And do feel free to be as harsh as you want about anything here. I can take it/feebly defend it.

Legal infringements aside, this looks pretty cool! Just how in-depth can you get when determining the "environment" someone is in? Obviously you can check for bodies of water, but could you check to see if someone was, for example, in a wide-open field vs. a wooded area? That's probably a dumb question but I know nothing of what kind of information you can obtain from GPS data.

@cmblasko: Currently I augment the GPS data with satellite imagery and relative elevation data. So yes, I can tell if you are in an open flat plain by comparing the elevation point with all of the nearby points. However, I doubt I could accurately determine how wooded the plain is, (Or if your just in a very flat city) which is why I don't currently have a plains biome.

Most plains right now would likely fall into the "light forest" or "heavy forest" biome, depending on their distance from the nearest road or urban area.

This is very cool. Did you write this in Java / Web app / Something like Unity?

The NPC's could be replaced with real players, though since it is hard to build an audience it may be very quite. You could create non-playable-characters and place them around the real world map randomly. You could create large zones around each of these random locations that players can encounter NPC's. The large zones will help in the event the NPC is on the top of a skyscraper or in the middle of a cavernous valley.

Nintendo has been experimenting with this on a level local to the player in games such as Animal Crossing. Their new DS systems have a feature where your game can passively interact with other people's games if they are nearby. Obviously moving this from the local context to online is a challenge for them as they have to retain the safety of their players. In Japan I they have recently released a free downloadable game based on some Pokemon movie, and I think it uses the local system. Obviously they their games are valuable to their handheld business, but it will be interesting to see companies move into this space and create games like the one you are making (without Nintendo IP), and build that on top of smartphone messaging platforms like Snapchat, WhatsApp, and Kik.

Duuuuuuude this looks hella rad!

Roaming legendaries should roam. I'm not sure how stationary legendaries would work, but I like the idea of a limited number of them. Hell, limit it all the way and just have one of each.

There should be Gym leaders. Every city or somesuch. Champion.... I dunno. Maybe cycle through cities of a high enough population every week or month?

Income's a doozy.

Charge pokecash for the center?

I would put level based on population or development somehow. Like, in order to get the toughest pokemon you have to get way out in the boondocks. But that brings up the problem of urban pokemon like magnemite and grimer.

How are you gonna distribute this? Are you gonna distribute this?

@afrofools: Everything so far has been handwritten in Java with the Android SDK. It's actually been a really great experience for me to use a language that I used to just think of as a "beginners language" and stretch it to limits I never would have imagined back in first year of University.

I think that looking at new ways of using existing data, like GPS, is really exciting! And it's a shame that companies like Nintendo have their hands tied as much as they do when it comes to this sort of thing. My full time job is researching virtual reality technologies and exploring new means of human-computer interaction, and It's been my experience that when people think VR they tend to just think in terms of devices like the Oculus Rift or the Virtuix Omni... But using location based data is certainly a part of it too! So I think that as we go further down this VR track as an industry we're likely to see more games like this. Games that use any kind of real world data they can.

God that's a hard question to answer.

Legally, of course, i'm bound by copyright restraints and could never hope to do anything like publish it on the Android market or anything. Not only would that never go through, it would also be a one way ticket to Nintendo's Legal Department Hell.

That being said, small time Pokemon games do exist, and continue to exist on the internet seemingly free of legal entanglements. I think what I'm going to try is likely a small site that distributes the APK file to anybody who wants to manually install it. I'll need to make it obnoxiously clear that I'm not affiliated with Nintendo and that they own all of the rights to all of the characters and what-not... Hopefully that kind of transparency will protect against a cease and desist (Which would shut things down really fast.)

One way or another, I'll need to cross the distribution bridge when I reach it.

Be ready for a cease and desist, if your idea becomes popular enough.

I think it would be cool if the battles triggered without your input, like you wouldn't have a choice when a pokémon appears or not. You'd have to walk around and hope for it to beep. Like in the games.

Your the fucking man!

This is awesome. Ever since that april fools thing with google I thought it would be fucking awesome in concept to be able to grab some friends and go on a pokemon catching roadtrip.

While distributing this is going to be impossible, at least your rolling the ball. Props.

I agree! And it's actually how it works right now. I have a debug button for development use that forces an encounter wherever i'm standing, but normally I need to walk around. Every X(Variable depending on current GPS accuracy) distance moved triggers a "step", and each "step" has a small chance of encountering a Pokemon. The biome is then checked, and an appropriate Pokemon is chosen from a pool for that biome.

Edit: Do you all think that detection should work as a service? Like, should you be able to encounter Pokemon even when the app isn't active? Or should you need to have the app out and running? There's certainly advantages and disadvantages to each.

@shofixti: Oh, cool! This sounds like a neat idea that I would like to check out. What are you developing this for?

Android only. I'm aiming for compatibility with 4.1.2 and up, and currently support a limited range of resolutions. I doubt I'll ever get around to making it work on tablets. (It's really annoying to resize sprites and still maintain a crisp pixelated look)

@shofixti: Oh, well, that counts me out then. Are you sticking to that limited userbase because of complications in integrating Google Maps onto non-Android OSs?

@snail: It's mostly a practical consideration. Because I'm hand coding it in the Android SDK I don't have many tools available to me to convert to IOS or Windows Phone (Or god forbid Blackberry). Porting it over then would involve rewriting most everything in a new language. On top of that, I don't personally have the devices needed to test thoroughly on any other platforms. I'm truly sorry about needing to exclude such a gigantic number of people, but I'm just one man working on this on my off hours. (Okay that's not totally true. I do have another bloke doing a fantastic job of filling in some of the tedious details for me. Thanks Kyle!).

That's really cool, well, until Nintendo come and shut it down.

I was thinking though, chances are most people will just end up with whatever Pokemon you decide should populate a city. But I suppose that would encourage people to go explore, eh?

EDIT: Oh, and i was thinking, Legendaries should spawn at fixed GPS coordinates around the world, maybe at well known locations like wonders of the world or similar. You could not publish where they are and make it first to find 'em gets the legendary?

If you want to this to stay alive I highly recommend only distributing the APK and not putting it on any app stores. Looks cool though, will probably not get it though because it seems incredibly inconvenient.

I see having a large user base and staying under Nintendo's/Gamefreak's radar as mutually exclusive goals, so yeah, it would likely be impossible to build a large enough audience to make Player trainers a viable option as the game's primary source of income. Not to mention the farming problems that arise out of this... NPC zones is good though, I'll keep that one in mind moving forward.

Roaming legendaries would involve having dedicated server processes that track and determine the location of the legendaries... Possible, but I feel like when you're dealing with an object the size of the planet earth that you'd pretty much just be guaranteeing that nobody ever has a chance to find them. I have been considering limiting them to just one individual, but then what happens when that one guy catches a Zapdos and then never plays again? It effectively removes Zapdos from the game. What I've been considering is more like: "Well, if you go to a X biome, and the weather is a snowstorm/blizzard, and it's below a certain temperature, then there's a rare chance you'll find Articuno"

I guess the next part raises a lot of questions... How would Gym Leaders be determined? What would their point be? It's an amazing idea to have that kind of prestigious position, but it needs to be formulated under some rules and automation to be feasible I think.

Charging cash for the Pokemon center is good, but there can't ever be a situation in which you run out of money and can no longer heal your Pokemon (Which would likely in turn also prevent you from making more money)

This was my first thought too. It could work, but it would mean that some Pokemon are just naturally a higher level than others. I suppose that's not necessarily a terrible thing though?

I like this idea rather a lot. I could even combine it with weather / time of day information to further limit occurrences.

Keep the ideas and suggestions coming, folks! And just as a bit of a response to all the "This can't exist for long" comments. You're right of course. But this project means more to me than simply the expectation of a finished and successful project. This project to me is a very small realization of a childhood dream, and the prime motivator for me isn't money, or success, or fame or any of the usual things that go with the expectations of a finished game. It's simply a work of passion. It's something that I believe is cool, and that I think is worth doing. I'd be lying if I said that the ultimate fate of the work is irrelevant to me, but it's certainly not the point.

The idea of limiting legendary pokemon to specific weather conditions or times of day is cool, but you would have to make sure not to make it too risky for the user. Thunderstorms are cool, but it's not worth going out in one to catch a Zapdos, you know? You set groups of legendary pokemon to roam in large geographical locations (Pacific Northwest, Rocky mountains, midwest, south east, north east, the great plains) and when one is caught, another one would spawn in that region. For example, If I caught a Moltres in Texas, I keep my Moltres, but another spawns in Arizona because they're in the same region (you could put in some latency for respawn if you want). Or someone in Seattle catches Ho-oh, and after an hour or so, another Ho-oh spawns in northern Idaho.

For Pokemon Centers you could use where hospitals or Red Cross centers are because most cities have their hospital locations submitted to Google maps.

Would you have the pokemon encounters be random, or would it be more like searching for the Dragon Balls with the dragon radar telling you where to look? It might even be more convenient to have active and passive modes. Active mode being, "I'm going on a hike through XYZ trail to look for Pokemon," and passive mode being, "I'm walking to the grocery store, and if I happen to encounter a wild pokemon, cool! But if not, it doesn't bother me." In active mode, you would turn up the frequency of encounters (although for rarer and higher level pokemon you would still keep their chance of encounter as low as possible). You could also enable NPCs in active mode, like some roaming Team Rocket members where you can't escape from battle. In passive mode TR would be turned off and the encounter frequency would be less.

Does that make sense?

I've been thinking about this too ever since the April Fool's gimmick, so kuddos to you for being the most developed out of all the projects I've seen for this!!!
Best of luck!

You had me, then you lost me....

But I actually do like the idea of limiting legendaries to certain weather... only the commited people would go out in a thunderstorm to a high elevated point to capture a zapdos, or high elevated point to capture an articuno, or on a sweltering day going far out to a deserted area like a dessert with a butte for a moltres. Sure people with a sense of mind wouldn't go get them, but hell, that's what'll make them special! (and without having to limit their numbers so much, as they'll be naturally limited)

Gym leaders at landmarks of the high population cities. Each city high population city would have like 8 of the same gym leaders. There's 151 pokemon that you could spread out across a player limited by travel would still not have much trouble getting a variety of pokemon in his area even if there are some he can't get, but it would be nice for the player to still be able to at least experience all the gym leaders. Ex: Oregon the cities would be like Portland, Salem, Eugene.

3: Without NPC trainers, the traditional source of income of Pokemon games is no longer an option. My current solution is to let players sell Pokemon on a shady-as-hell Team Rocket black market screen. Any better ideas?

I don't see a problem with this... there's going to be a lot of duplicates caught so this would be great, just don't limit how many pokemon a player can catch.

4: Pokemarts and Pokemon centers don't currently exist as I have no way of placing them well globally. So right now they're simply a planned menu option. However there's some serious implications here... You shouldn't damn well be able to heal after every fight after all. Any ideas for checks and balances on this front?

Player sets his home location which he can change once every 30 days, that home becomes the location of both his Pokemon Center and PokeMart.

5: What should determine the level of the pokemon you encounter?

All encountered pokemon match the level of the pokemon you encounter them with and the opponent pokemon gets the first move, that way if you just try sheeping a level 1 pokemon to encounter level 1s that are easy to catch, YOUR level one will die and you'll still have to place your high level as a starter in the end. This would add a lot of strategy as to how you place your pokemon.

First of all, a quick update:

I've been continuing work on the battle system itself, and it's nearing a conclusion! It's going to take a long time and a lot of menial work to program in all of the moves for every Pokemon, but beyond that, the system itself is in place. The Pokemon switch and Item screens have been implemented and are mostly working, though individual item functionality will take some time... Here's a shot of the in-battle item screen that I could use a bit of feedback on:

I feel like the top half is too cluttered, there's a ton of information being bombarded towards the user. But maybe i'm too close to my work.

This, actually, is really excellent. The more I think about it the more I realize it's an amazing idea. At first I was wondering about things like: "But if you have 4 level 15 Pokemon, isn't fighting a single level 15 going to be too easy?" But I guess it's really not, as you still stand to take a lot of damage if your constantly switching Pokemon in and out. It also gives the player additional control over what they expect to find. I think though, that there should still be some wiggle room for chance. I believe a lot of the core fun of games in general comes from surprise... "Oh shit" moments, for example. But that really shouldn't be too difficult to tie into the idea. Thanks for this!

Me and my buddy Kyle (Who works on a lot of the tasks I'm too lazy or bored to tackle myself) were talking about this very idea, but we thought it might be a bit extreme. The compromise we came up with was letting your home zone have a free Pokemon center, whereas anywhere else there is a kind of steep "ambulance fee" to fully heal your Pokemon. He suggested that this service be on a cooldown, but I'm not sure how necessary that is if the fee is high enough.

1: That was my first attempt. Unfortunately Google only allows 100 geolocation requests per 24 hours unless I purchase a business license. I need a geolocation request every encounter to check to see if you are by a hospital. So that's a bit unfeasible right now.

2: I think right now things are strictly working in what you've termed the passive mode. There's no real indication to the player what direction he/she should walk to find a Pokemon. In fact it hardly matters, as currently the program does not care as much about how they moved as IF they moved at all. Let me take a second to explain the mechanics of it all:

Basically as you walk around, if you have moved X distance (Where X is based on the accuracy of your GPS) then the program registers a "step". Every step, there is a small chance of encountering a Pokemon (The exact odds are being played with right now because I am SOOOOOO sick of accidentally encountering bloody Rattatas every time I step outside and forget to close the app) once an encounter has been determined, players are dropped on a loading screen (See right) until my server receives their request, determines from their GPS coords, Elevation data, and nearby Sat imagery what type of biome their in, picks a random (Weighted by rarity of course) Pokemon from a list for that biome, and then the encounter begins.

As far as changing your encounter rate, I was planning on items that could lower or raise your overall rate. But an idea occurs to me...

The app totally works as a service. Meaning the app doesn't actually need to be on screen for you to encounter a Pokemon. Perhaps while the app is inactive I should lower the encounter rate, as it shows the user probably cares less about finding Pokemon.

Next up: Disabling encounters when i'm driving/on a road. I swear i'm going to kill myself with how often I get an encounter while driving.

@shofixti: I think the UI looks fine for a work-in-progress, but you have a typo on "Bulbasaur".

When you're further into development you might want to match the HP bars with the aesthetics of the rest of the UI (i.e. pixelate it -- or vice-versa), and you might want to liven up the color palette, in my opinion. Beige doesn't easily make for a great menu color.

And "Weakend". Both are fixed in the version implemented on the phone : p

You wouldn't be able to pm some of us the apk and send out a new one for major update so we could test it? If not, that's totally cool.

@shofixti said:

First of all, a quick update:

I've been continuing work on the battle system itself, and it's nearing a conclusion! It's going to take a long time and a lot of menial work to program in all of the moves for every Pokemon, but beyond that, the system itself is in place. The Pokemon switch and Item screens have been implemented and are mostly working, though individual item functionality will take some time... Here's a shot of the in-battle item screen that I could use a bit of feedback on:

I feel like the top half is too cluttered, there's a ton of information being bombarded towards the user. But maybe i'm too close to my work.

That looks really good. The art style fits in well, its believable that this is a real product from The Pokemon Company!

What I was thinking for question 5. Since every starts out at different places. You could add up all the party Pokemon's levels and divide them by the number are party Pokemon you have. it would also make it easier to train your Pokemon because you would get a decent amount of expirence and it would always be a challenging match.

Congratulations on Pokemon GO.

This is an awesome old find. And that Google joke kind of blew me away.

Crazy how this only received a few comments a few years ago, and now it's doing gangbusters for Nintendo.

Well...would you look at that.

Wait, is this still a thing? Because after playing Pokemon GO for a while, I'd be WAY in if it is.