I really like the idea of permanent consequences, but this will be hell to design and test. :)

I agree with everyone here that the level design will have to be really thought out.

For your own sanity, I would start with only 2-4 characters!

Players will need to be familiarised with the enemies and bad things in the game so they don't feel cheated when encoutering such a thing. When a player encounters another "type" of character he can play (say a wizard that can shoot magic) he will need a safe space to play around with him so he doesn't get killed right away. The creator of PvZ has a really good video up about immersive tutorials which I think you will need in this game a lot! http://www.gdcvault.com/play/1015327/How-I-Got-My-Mom

We're talking a lot about negative consequences - what about positive ones? For example a player eats something and that turns him blue for the rest of the game.

I would love to see more about this game. Pace, view point? Sounds exciting. Good luck!

The problem is that I'd then be effectively rewarding the player for making mistakes

no you wouldn't, but you wouldn't be punishing them neither(gameplay-wise at least)
though you could make the main-personality slower in performing those actions, for example.(or jump just a little less far then the jump-personality)

to realy see what works and what doesn't you should start making the game(leveldesign) though, so you can tinker around with the ideas and see how they work out.
(aka i suspect you re gonna have to implement most ideas to a certain degree, just not sure to which degree)

I really like the idea of permanent consequences, but this will be hell to design and test.

I agree with everyone here that the level design will have to be really thought out.

For your own sanity, I would start with only 2-4 characters!

Players will need to be familiarised with the enemies and bad things in the game so they don't feel cheated when encoutering such a thing. When a player encounters another "type" of character he can play (say a wizard that can shoot magic) he will need a safe space to play around with him so he doesn't get killed right away. The creator of PvZ has a really good video up about immersive tutorials which I think you will need in this game a lot! http://www.gdcvault.com/play/1015327/How-I-Got-My-Mom

We're talking a lot about negative consequences - what about positive ones? For example a player eats something and that turns him blue for the rest of the game.

I would love to see more about this game. Pace, view point? Sounds exciting. Good luck!

Thanks so much for your interest. I'll definitely have a look at that video today, but I can't right now as I'm at work where I can't have noise for obvious reasons.
I'm generally thinking of having the first chunk of gamneplay being rather low risk and offer a variety of puzzles where you can experiment with and learn about your characters.

Maybe I'll even make sure you will have used every single character to overcome an obstacle for sure before you even encounter something lethal. I was also thinking that teh characters would cue you in about how they could help you overcome obstacles.

You could for example encounter a very deep drop and one character would get frightened at it claiming you can't survive falling that deep. Then another character would respond by saying that he could easily survive that without a scratch. As he looks like he can take a hit (design reflects abilities), that should be enough to convince the player.

As for positive consequences, I'm hoping the unraveling of the backstory and encountering new puzzles will be enough to drive the player forward, but it could indeed be fun to encounter things that change the characters in some way, I'll certainly think about a way of implementing it, because it sounds like fun and very rewarding.

I'm not sure what you mean by pace and as for viewpoint, I see it as a 2.5D sidescroller for now.

When I have öore things decided or ready, I'll definitely revisit this thread and share it with you guys.

no you wouldn't, but you wouldn't be punishing them neither(gameplay-wise at least)
though you could make the main-personality slower in performing those actions, for example.(or jump just a little less far then the jump-personality)

to realy see what works and what doesn't you should start making the game(leveldesign) though, so you can tinker around with the ideas and see how they work out.
(aka i suspect you re gonna have to implement most ideas to a certain degree, just not sure to which degree)

I'm assuming that having the characters merging would feel like a success to the player. But you are right, from a gameplay perspective it is neutral and having the powers reduced would indeed be a punishment. However, when you get many powers in one character, it becomes a reward again. Since a character that can run faster and fight better is more useful than the two seperately. I'm currently not sure if I want the main character to be present before you lose all of your characters, I'm leaning to no.

I see that thsi is a bit of an issue with the design and only play testing can really serve as a final verdict here I think. Again thank you for your input, highlighting what some possible pitfalls can be and even providing me with appealing alternatives, this forum has been incredibly helpful so far.

If possible I'd like to ask anyone with an opinion on this particular issue to share it with me, to see what the concensus on it is...

I agree with the other posters. While I like the idea of long-term consequences in a game, it simply doesn't work from a gameplay perspective in any game with a significant puzzle element. It might sound like a nice idea, with each character providing you extra options on puzzles and losing them forcing you to use the default means, but the player will usually figure out the default method first because they are the most familiar with the default traits. Further, once a character is gone they won't be looking to use them to solve puzzles anymore and once they are thinking without them (which is really easy if they die early) they won't feel the loss anymore. In fact, it might be easiest for them to just kill off all but one character so they don't get used to any of the others and they'll be less distracted through the game. The only hope it has is being a plot device, and what it'll amount to then (unless all the characters are really likeable and easy to keep alive) is a large section of the plot the player will miss because one or more of their characters is dead.

hmm, thus far i assumed you could just switch between characters at will(maybe after waiting for 2 sec or so)

is the player supposed to play out an entire level with just one character or something ?

I agree with the other posters. While I like the idea of long-term consequences in a game, it simply doesn't work from a gameplay perspective in any game with a significant puzzle element. It might sound like a nice idea, with each character providing you extra options on puzzles and losing them forcing you to use the default means, but the player will usually figure out the default method first because they are the most familiar with the default traits. Further, once a character is gone they won't be looking to use them to solve puzzles anymore and once they are thinking without them (which is really easy if they die early) they won't feel the loss anymore. In fact, it might be easiest for them to just kill off all but one character so they don't get used to any of the others and they'll be less distracted through the game. The only hope it has is being a plot device, and what it'll amount to then (unless all the characters are really likeable and easy to keep alive) is a large section of the plot the player will miss because one or more of their characters is dead.

I have to respectfully disagree with you there, As the default way will usually be a lot more risky than the more complicated ways to solve a puzzle.

Would you rather figure out you can reach a rope, cut it and kill a bunch of enemies that look pretty vile or just face those enemies in a fight, knowing that if you die, you lose a character. I'm not saying it's easy, but I think with the right intro level and overall level design, I can get the player into an anxious mood that makes them always stop and consider the situation.

The back story element, I think, is designed in such a way that it is enough of an incentive for the player to really want to keep his characters, but still fun enough when you lose a few. Every character adds less meat to the story than the next one, because some of their perspectives overlap. Same goes for puzzle possibilities. This means that losing your first character is likely not a very big deal.(relatively)

hmm, thus far i assumed you could just switch between characters at will(maybe after waiting for 2 sec or so)

is the player supposed to play out an entire level with just one character or something ?

No, you assumed right. What I meant is that in the intro level, I might include obstacles that only one of the characters can possibly overcome. This way you are forced to use that character. If I make you face nothing lethal before facing one or more obstacles for each character the player has, I can teach the player how to use the characters in a safe environment.

I've been thinking about some of the concerns you guys raised about the concept and hurdles for the level design, and came up with a possible solution.

What If the levels' content is based on the characters the player has left?

This means that the game would lay out the obstacles faced in the game semi-randomly a few steps ahead of the players. I'd throw in the random factor to make sure that the player still feels theb yeight of his previous mistakes throughout the entire game. But now he will get to use the characters he has left more often and the puzzles can be more geared towards certain characters. Every obstacle can still be overcome with every character, but this way I can make the default way a lot harder.

This also fits into the back story as the therapist would adapt the challenges he throws at the personalities based on which are left. Furthermore, it would seriously ad to the replay value, as each playthrough could be vastly different from the last one.

Obviously, there are some technical difficulties (especially since I'd prefer the game world to be possible to navigate without loading screens) to overcome and this means more content to design and create, but I feel like it could provide a significant improvement for the game.

Does anyone have any feedback on this idea?

What If the levels' content is based on the characters the player has left?

This means that the game would lay out the obstacles faced in the game semi-randomly a few steps ahead of the players. I'd throw in the random factor to make sure that the player still feels theb yeight of his previous mistakes throughout the entire game. But now he will get to use the characters he has left more often and the puzzles can be more geared towards certain characters. Every obstacle can still be overcome with every character, but this way I can make the default way a lot harder.

This also fits into the back story as the therapist would adapt the challenges he throws at the personalities based on which are left. Furthermore, it would seriously ad to the replay value, as each playthrough could be vastly different from the last one.

Obviously, there are some technical difficulties (especially since I'd prefer the game world to be possible to navigate without loading screens) to overcome and this means more content to design and create, but I feel like it could provide a significant improvement for the game.

Does anyone have any feedback on this idea?

The big problem is it means you have to do each and every level multiple times. And it's more than "Alright, I have 5 characters so I need to do each level five times." The issue is that there are a lot of different combinations, so even with just five variables you'd be looking at (I think) 25 possibilities. With 6, (I think) 36 possibilities. And so on. Do you really want to do each level dozens of times over? Because I can assure you nobody else on the team will.

Beyond this, it's going to encourage some players to kill off all their characters to make the game harder, which is another problem. It'll also result in some characters being killed to avoid certain level designs, and certain level designs never being seen because a player doesn't like a character and kills them.

he big problem is it means you have to do each and every level multiple times. And it's more than "Alright, I have 5 characters so I need to do each level five times." The issue is that there are a lot of different combinations, so even with just five variables you'd be looking at (I think) 25 possibilities. With 6, (I think) 36 possibilities. And so on. Do you really want to do each level dozens of times over? Because I can assure you nobody else on the team will.

Beyond this, it's going to encourage some players to kill off all their characters to make the game harder, which is another problem. It'll also result in some characters being killed to avoid certain level designs, and certain level designs never being seen because a player doesn't like a character and kills them.

I think I may have worded myself a bit poorly. I didn't mean that I wanted to design every level for every possible combination of characters that could be left (which is 32 for 5 characters and 64 for 6 by the way). What I meant is that I design a whole bunch of modular obstacles and assign some values to them representing how easily certain character combinations can get across them. Then I'd use those values to procuderally generate the game world as the player goes along. My estimate is that it would end up being about 2 to 3 times more work than just designing a static, linear game world.

It's interesting that you mention that this might encourage players to kill off characters to make the levels harder, I hadn't thought about it like that yet. I'm hoping that the player would be too scared and too attached to his characters to kill them off for a more challenging game. Don't forget that this would mean that he actively makes game content inaccessible for himself.

Thanks for your reply in any case, food for thought. :]

he big problem is it means you have to do each and every level multiple times. And it's more than "Alright, I have 5 characters so I need to do each level five times." The issue is that there are a lot of different combinations, so even with just five variables you'd be looking at (I think) 25 possibilities. With 6, (I think) 36 possibilities. And so on. Do you really want to do each level dozens of times over? Because I can assure you nobody else on the team will.

Beyond this, it's going to encourage some players to kill off all their characters to make the game harder, which is another problem. It'll also result in some characters being killed to avoid certain level designs, and certain level designs never being seen because a player doesn't like a character and kills them.

I think I may have worded myself a bit poorly. I didn't mean that I wanted to design every level for every possible combination of characters that could be left (which is 32 for 5 characters and 64 for 6 by the way). What I meant is that I design a whole bunch of modular obstacles and assign some values to them representing how easily certain character combinations can get across them. Then I'd use those values to procuderally generate the game world as the player goes along. My estimate is that it would end up being about 2 to 3 times more work than just designing a static, linear game world.

That's fine. (And I didn't actually do the math, anyway.) It's still a lot of extra effort and the rest of the team will take a lot of convincing to get them do to so much extra work.

It's interesting that you mention that this might encourage players to kill off characters to make the levels harder, I hadn't thought about it like that yet. I'm hoping that the player would be too scared and too attached to his characters to kill them off for a more challenging game. Don't forget that this would mean that he actively makes game content inaccessible for himself.

Thanks for your reply in any case, food for thought. :]

Well, that would be assuming they're intimidated by the game's difficulty (a lot of players aren't) any they like all your characters (which is highly unlikely) enough to keep them around. Not only that, that'd be assuming they like the mechanics attached to each of them. There are some players, for instance, who HATE stealth-based gameplay. (Bunch of gung ho morons, if you ask me.) They'd kill off a stealth-based character on principle just to avoid sections designed for stealth, even if they like that character.