Rejected Puzzles?

[aesf]

Hi,

This is a really long shot and I don't expect anything more than a rejection, but I'd still like to put the idea forward.

Chess.com, in addition to having online matchmaking much like PS, also has an additional "Puzzle" format. I think it'd be cool to have a similar mode for Pokemon games, given the strategy components of the game.

Brief summary of what puzzle mode would entail

Many times, in both Chess and Pokemon there simply is one best play, or sequence of plays, that progresses the player to either an advantaged state or an outright victory. Chess.com utilizes this by presenting mid-game or end-game scenarios with one correct sequence of plays to the player. The player "wins" if they input all the correct moves perfectly (and the computer's moves, of course, can be predetermined without too much difficulty). The transition to Pokemon should be fairly smooth. We could, for instance, choose various scenarios from casual matches, ladder matches, or forum tournament matches as the puzzles. Users could then "solve" these scenarios and submit them as official puzzles for all of Pokemon Showdown to play. Puzzles would, of course, be of varying difficulty and types, and, with the endless possibilities within Pokemon, it's hard to imagine that there wouldn't be some puzzles that could stump even the greatest Pokemon battlers, just as grandmasters are often stumped by user-submitted puzzles on Puzzle Rush.

Cons:
= Possibly a big hurdle to implement
= Getting puzzles is hard -- unless it becomes very popular, a daily puzzle (or like daily easy/medium/hard puzzles) would probably be all that's practical for PS (Lowkey if everyone on PS saw these puzzles, people would be super incentivized to post puzzles bc of likes lol). And yeah, in the end, the playerbase of competitive Pokemon is still much smaller than the chess playerbase.

Pros:
=PS has generally been the same for the past 6 years I've used the site, and this would be a cool way to spice things up.
=Would also help with general playerbase skill
=Could help introduce more players into different metagames,
=Can increase interest in the tournament scene among both casual and competitive players.

 

[FriendOfMrGolem120]

I like the idea and believe it could also just start off as a forum project that doesn't require any programming knowledge. One could simple describe the situation and post the solution later or in a Spoiler tag. Compared to chess puzzles are less optimal for pokemon because 1. in pokemon players select their moves at the same time which means very often prediction is involved so we could only do puzzles for forced wins 2. pokemon has many variance moves and if they are involved, even the best line of play could not work out in practice.
In many gen1ou games I've played, really complicated situations occured and it takes some time to figure out which odds are better. What I'm trying to say is that even with variance moves the goal of the puzzle could still be to figure out which option gives you the best odds or even how high these odds are exactly for each considerable option.

 

[pre]

This reminds me of Earthworm's excellent 'Situation Analysis' article from The Smog. Puzzles for Pokemon are difficult because they involve mixed strategy Nash equilibria. I believe creating such puzzles is out of scope for Pokemon Showdown, though I agree with FOMG that it could be a fun and valuable forum project. :)

 

[Marty]

I've never played Pokémon XD, but this idea sounds a lot like the Battle CDs from that game, where you just have to win against an AI where you both have a predetermined set of Pokemon and moves, sometimes in a certain number of turns. Was that also inspiration for this suggestion?

That kind of thing certainly makes for interesting situations, and plenty of neat stuff was already possible back in Gen 3 so there would be a really wide variety of puzzle-like options with current-gen mechanics. The main thing I would be concerned about is making the scenarios too simple or so complicated and specific that no one really wants to play it. I'm not sure it's entirely out of scope for PS, but I agree this should be shelved for now, at least until someone comes up with a system that makes sense within PS. And an initial set of scenarios, assuming said Battle CDs are anything like what the suggestion was going for.

 

[pre]

To clarify, if the community can work out an established format/spec for puzzles and scope it properly I agree its something that could find a place on Pokemon Showdown if it gains enough traction, but I think the idea needs to be workshopped around a bit as its a little open ended at the moment. 'Rejected' now is not to say 'Rejected Forever', this time I just mean to say 'Needs More Work'.

EDIT: Upon further thought, 'Rejected' maybe was premature as I was unaware of the Battle CD example Marty brought up. Perhaps you could flesh out the proposal more so we're all on the same page?

 

[aesf]

I like the idea and believe it could also just start off as a forum project that doesn't require any programming knowledge. One could simple describe the situation and post the solution later or in a Spoiler tag. Compared to chess puzzles are less optimal for pokemon because 1. in pokemon players select their moves at the same time which means very often prediction is involved so we could only do puzzles for forced wins 2. pokemon has many variance moves and if they are involved, even the best line of play could not work out in practice.
In many gen1ou games I've played, really complicated situations occured and it takes some time to figure out which odds are better. What I'm trying to say is that even with variance moves the goal of the puzzle could still be to figure out which option gives you the best odds or even how high these odds are exactly for each considerable option.

The forum project idea is reasonable. I've submitted it to the OU forum moderator, and hopefully it'll gain some traction
As for the inconsistencies between chess and pokemon, I agree that there are differences, but I do not think these issues will be a big deal with regard to Pokemon puzzles:
1) First, puzzles submitted shouldn't have true 50/50s. In chess games, there are turns where more than one move works, and the puzzle engine doesn't account for these situations in their database. There are a lot of situations where if you present 10 good players with that situation, they will all make the same play.

2) For the sake of puzzles, as well as general pokemon playing, then it'd be easier to make the objective to maximize the probability that you'll win. This means that if you have a 30% chance to win with a scald burn and any other decision results in a 10% chance to win, and you don't get the burn, it shouldn't matter; the solution will recognize that scalding, despite not always working out, was the best play possible. Of course, if this scald play was part of a sequence of moves, then we get into more nuances, but the general principle stands.

Still, I understand the concerns that pokemon wouldn't work the same as chess. It's definitely possible that I'm wrong; we'll have to see how the forum projects kick off the ground first. Thanks for the feedback!

I've never played Pokémon XD, but this idea sounds a lot like the Battle CDs from that game, where you just have to win against an AI where you both have a predetermined set of Pokemon and moves, sometimes in a certain number of turns. Was that also inspiration for this suggestion?

Haha nah, I just stumbled upon Chess.com one day and thought the puzzles were pretty fun despite being garbage at chess.

 

[Hecate]

Closing due to inactivity.