GAME DIFFICULTY MECHANICS - BRAIN STORMING

Posts

Pages: first 12 next last
It's been a while since I've made a topic here, but I'm currently brain storming a few game mechanic ideas and would like other views on them.

The proposed game would be a role playing turn based strategy similar in some ways to fire emblem. I'm looking for thoughts and opinions on the below ideas as gameplay elements; all comments appreciated.

Mechanics I'm currently thinking about:
1. Perma-Death
A hallmark of fire emblem for a long time - BUT in fire emblem many people play with this as if it were a "game over if anyone dies" mechanic; I've been trying to think how I could implement perma-death without it just being a reset inducing feature. One idea I've had is that your army would be comprised of two seperate types of unit:
- characters: these would have a story and would level up and you would put effort into improving them etc.
- troops: these units would not have any significant story
The game would try and force you to sacrifice some units to complete most maps, the idea being that you'd probably sacrifice your troops, not your characters. Problem - would perma-death as a feature not be that relevant with these troops? How could it still feel like an actual cost?

2. Long term injuries
If a unit takes major injuries in one battle they should not be available for a certain period of time which may include the next battle. This would add a minor edge of realism and an extra point of strategy - "I can clear the map by doing X but it may mean I don't get my best unit for the next battle".

3. Kamikaze techniques
Powerful abilities that inflict either long term injury (see above) or perma-death on the unit who uses them in exchange for doing something incredible to the current map state possibly killing multiple enemies. For storyline reasons I'd like the game I've been working on to have special attacks of this sort but I cannot think of any good way to balance them. Can anyone think of any time when a mechanic like this has been good?

4. No healing
I'm looking to strongly discourage attempting to complete missions only with your best units and instead encourage players to use their entire army, so what about having no healing items of any kind with the only healing being between missions.
1. Check out this discussion on permadeath in XCOM and Darkest Dungeon

What stands out to me is the importance of clearly laying out risks/rewards for actions and giving people tools to plan, even when the actual outcomes are determined mostly by RNG. And making it easier to get stronger units, or having the "strength" of a unit be something you can transfer (like equipment in XCOM, for example)

I do think that separating "characters" and "troops" will make permadeath meaningless for the troops.

If you combine "characters with a a story that progresses", or even "characters you think people will feel attached to", then a lot of people will reset when they die rather than lose them forever. I say let them do it.

Someone mentions including checkpoints from Enter the Gungeon, and while you can't reload at them there, but maybe it's worth considering? Then once someone reached a checkpoint, they'd be committing to whatever their status was at that time... but this would introduce a lot of potential headaches. The only way I can think of to 100% prevent reloading to avoid a unit's death would be autosaving once the game knows the unit will die but before the graphical representation of that is displayed. But I think this would feel super punishing, and discouraging people from doing something with punishment without also encouraging them in some way just Feels Bad.

On that note, you could always provide some kind of bonus when a unit dies. Not just if they died through some kind of suicide technique, but whenever someone died in general. I dunno, if there's levels maybe dead units can drop something that can be used as currency to transfer some of their character-specific things to someone else, or saving it to some kind of inventory for later use.

For something like that, Final Fantasy Tactics' crystal system was really good. I loved being able to learn things from fallen enemies, buuut I still always reset when one of my units died. It was a lot of effort to train and raise them!

Speaking of FFT, the difference between "knocked out" and "dead" was really nice, and I always appreciate that kind of thing. Having a period where a unit is at risk of death and can't defend themself, so you gotta drop everything and rush to save them, can be very exciting!

2. Yeah, Darkest Dungeon has that too. idk if you've played/seen that. Some of the implications of it come up in the permadeath thread I linked, but in general I think that while this adds a nice risk, the reward for risking a long term injury should be pretty good. If it's something that can happen from an enemy getting lucky, that's totally valid but will definitely add frustration

3. I don't know of any examples where this has been combined with permadeath. Maybe someone else will!

4. I don't think "no healing" will do much for discouraging people from using only their best units. Again, I personally prefer to look at encouraging players to do something before I think about ways to discourage them. "no healing" sounds potentially exciting, but also sounds like a way to speed up the "shit something went wrong I gotta reset" process.

HP of any kind is usually like a "margin of error", with more HP = more times/ways you can mess up. When that "messing up" is out of your control (say, an enemy gets a critical hit or the RNG decides to bum rush a single character), and you can't recover from that messing up, then it's a recipe for feeling very bad.

If you include healing, there's ways you can make it risky. Off the top of my head, maybe...
  • Units that can heal are weak, creating the dynamic of "babysit this healer" that you see in uh, almost anything with squishy healers. I don't mind it but that depends entirely on the range of the healer and how much I can influence whether or not they're targeted by enemies.

  • Units can "trade" HP between each other, so healing one unit would cost HP from another. There's a lot of variables you could introduce (questions like "what is the range for this ability?" or "Do you have to be close or can you swap HP between any units?" as two examples), so someone with high HP and a lot of movement would be more like a healer in this kind of system.

  • Healing comes from killing enemies. This would encourage people to actually go out and fight, and I think makes them more likely to engage in risky combat. It can be very exciting! Think the Glory Kills from DOOM, and Bloodborne does this too
Thanks for the comments PentagonBuddy. In particular thank you for the XCOM comparison, it's not a game I've played but I've just started taking a look at it and it seems to have a lot in common with the sort of game I've been thinking of making, so will likely provide useful insights in other areas too.

I think it's important to differentiate between difficulty, punishment, and general game play mechanics.

Punishment is facing consequences which far exceed the severity of a mistake. Punishment is not being allowed to correct for mistakes. Punishment is not being given an opportunity to understand a situation before suffering from it.

Some games are built around punishment. The Darkest Dungeon relies on it to create atmosphere, and the actions of the player are based largely around avoiding punishment. This is what makes Darkest Dungeon a fun and interesting game.

But punishment also discourages risk taking. And in many cases punishment discourages trying out new things. A player's decisions will be based primarily around avoiding punishment rather than seeking enjoyment. When done right, punishment creates a level of tension that is impossible to get otherwise in the virtual realm of video games. But when done wrong, it is frustrating and takes away from the overall experience. When thinking about punishment, ask "am I adding to the game play or am I just wasting the player's time?" If the player is just going to reset all the time, you're wasting their time.

When designing Civ4, Sid Meier actually eliminated the randomness of fights and made them predetermined so that you couldn't reset to change the outcome. The punishment mechanic inherent to Civilization had gotten out of hand and started wasting people's time. The change prevented that (while, oddly enough, making it even more punishing).

Difficulty means that a player needs to be of a higher skill level to win. Even the most skilled Darkest Dungeon players will, at times, lose characters. It is unavoidable. But a player skilled enough will never lose at a difficult game.

Another way to contrast punishment and difficultly is that if you are punished, you can reset (if resetting is allowed) and succeed the next time. But resetting doesn't solve difficulty, as you still need to have the skill required to complete the challenge.

Difficulty is not necessarily preferable to punishment. The Darkest Dungeon is a difficult game, but it does not require anything abnormal to beat it. But a difficult game's skill requirement may be simply too high for people to participate. A punishing game can simultaneously create tension with "will I win" while maintaining "I know I can win," while tension for a difficult game is simply "can I win?"

Punishment can also create some fun new avenues for game play. I come from a Shining Force community, and one thing we do is get a group together and rotate beating battles (I take battles 1, 7, 16, for example). By adding perma-death, it forces people to play with characters they otherwise would not, and creates a nice amount of variety for a game that is over 20 years old. It also improves the sense of community, since others' actions directly alter your experience.

Finally, you can create light tension with light punishment. The Final Fantasy Tactics example PentagonBuddy mentioned (where people have a "knocked out" period) is a perfect example of this. Only in very rare cases will your character ever die, and of course you'll reset if it happens. But the very fact it could creates tension that would otherwise be absent from the game.

Game play includes elements that are balanced around, but are not necessarily punishing or difficult.

There is some overlap between all three types, so to illustrate, let's look at Long Term Injury. On the surface, Long Term Injury is a game play mechanic that forces the player to swap players in and out between fights, depending upon who is off the injury list. Think of sports where coaches and teams are shuffling lineups based upon injuries. It's just part of the game. It can be punishing, however, if you run out of characters to fulfill certain roles, or if you simply run out of characters entirely and can't complete the next battle.

Perma-death can be an element of difficulty and game play, such as with your troops. The same is true in a game of Chess. Or it can be punishment if incorporated with the appropriate amount of random elements or abusive enemy AI.

Kamikaze techniques are 100% game play, since they are not punishing (as they are your decisions) and actually could make the game easier (since it allows you to do an abnormally large amount of damage).

My opinions on the four items you suggested:

1. I've never liked perma-death in Fire Emblem. As you said, people just reset rather than take the death. And coming from Shining Force, I see no problem with people imposing perma-death upon themselves. I guess you could have it a toggle, if you really wanted. I also mimic the belief that having "troops" negates any value in perma-death.

2. I like long-term injury. It requires diversity in cast and means that you can't play each battle exactly like the last. The only concern is running out of characters to use. If it were me, I'd inflict long-term injury on characters who died in battle (rather than perma-death), and then not have healing in battle. That way you can strategically sacrifice someone to lower the risk of injury to others. With perma-death, "strategic sacrifice" doesn't exist.

3. I only see Kamikaze techniques being valuable if there is no perma-death. Otherwise they'll never be used until the last boss or something.

4. Abandoning healing will make your game significantly harder to balance, but I encourage you to try. Healing is something we've all taken for granted is games and removing them really changes things. Removing healing also speeds up battles because there will be no advantage in sitting around waiting for everyone to heal up before engaging the next group of enemies.

Finally, just quoting some things from PentagonBuddy I agree with:

  • Speaking of FFT, the difference between "knocked out" and "dead" was really nice, and I always appreciate that kind of thing. Having a period where a unit is at risk of death and can't defend themself, so you gotta drop everything and rush to save them, can be very exciting!
  • I don't think "no healing" will do much for discouraging people from using only their best units. Again, I personally prefer to look at encouraging players to do something before I think about ways to discourage them. "no healing" sounds potentially exciting, but also sounds like a way to speed up the "shit something went wrong I gotta reset" process.
  • Healing comes from killing enemies. This would encourage people to actually go out and fight, and I think makes them more likely to engage in risky combat. It can be very exciting! Think the Glory Kills from DOOM, and Bloodborne does this too
Perma death ideas:
1) Unit loss is mandatory even if you won the battle flawlessly. The game makes you sacrifice units after the battle. Each battle could have a sac "cost" so the player knows that they have to perma-lose at least that many units.
2) Make there be no way to reload a previous save.

Unit injury is a great alternative to perma death. Obviously you wouldn't put both in the same game.

Kamikaze is basically an extreme version of breakable items, which is a terrible idea to begin with. There's no way to balance it. Any time a player would use this they would theoretically be better off saving it for an even more difficult battle down the road. The only way to make these work is to make it mandatory to use them. (see first paragraph) Maybe like some kind of limit break where once a unit's kamikaze gauge has been filled, they MUST use it at that instant, or one of their existing attacks is replaced with it. Or you could have it triggered when the unit it at critically low HP or when they would die anyway, meaning the player wouldn't have direct control over it.

No healing sounds like a good idea. Battles would probably be much shorter as a result unless the number of units per battles is higher to compensate for it.

edit: additional thoughts:

- THEORY: Character driven games can't have permadeath, that's like oil and water. There's just no sense in having a storyline only to have the player experience the first portion of it. The player isn't going to sacrifice that just to "beat" the game. That's almost like fast forwarding through a VHS movie in order to "beat" it. Not to mention that even if beating the game was all that mattered, losing that character only makes things harder.

- FFT's countdown system is a great idea, it adds urgency to reviving characters, but the crystals they drop once they actually do die hardly make up for the death. It would've been largely the same if they had dropped nothing. Actually the best thing about the crystals was you can replenish HP/MP with them, increasing the chances of winning the battle, which adds tension to the choice between resetting and continuing. It's easy to reset the game if the battle is a lost cause anyway but the HP/MP restore crystals can make all the difference. So in general I guess any benefits to having units permanently dying is a good idea since it's like a consolation prize.

- Healing from killing is a cool idea. It might be even better in a game without permadeath since you'd be compelled to see how far you can survive on a shred of health.

EDIT:
Just thought of an idea, what about something like FFT's countdown mechanic but the character is still alive.
So let's say when a character goes to 0 HP, you can choose to have that one character retreat but if you don't, the character stays alive for 3 turns only and then dies from blood loss. (permanently)
Bonus points: The character is invincible during this phase.
Bonus points: The game is really difficult unless this countdown state is regularly exploited.
Come to think of it, there's no logical reason the character wouldn't be able to retreat during every turn during this state, making permadeath nearly impossible unless the player arbitrarily chooses to kill off that character. So I guess the best solution is to have it so triggering this state kills off the character even if the battle is won before the 3 turns are up.
Then just adjust difficulty so that the player realistically has to trigger this state in order to win difficult battles.

A permadeath with an advantage! That's the way!
Thank you for all the comments so far, lots of good ideas here to get me thinking - I don't agree with all of the suggestions but I'll think more carefully about certain points than I would have done without the above input.

I'm still interested in any further comments; also I have a thought, what if a unit on getting below a certain HP value lets say 5 for now would be rendered unconscious being unconscious would mean they cannot move or act for the rest of the current fight AND they will be unavailable for a certain period of time after the fight missing one or more subsequent fights. BUT if they're hit again whilst unconscious they will then die and never return.

This would give a potentially interesting dual mechanic, someone is down can you protect them whilst completing the objective? (even though you're not going to have them available in the next fight) Do you sacrifice them permanently (which won't impact your next fight but may impact the one after) to complete the current mission more easily.

Is that kind of long term thinking interesting? It feels interesting to me but I'm the kind of odd individual who enjoys looking at spreadsheets and sums...
I'd first have to know whether AI units will intentionally try to kill the unconscious unit. But then again why would they bother? After all it's not like you would intentionally go after their unconscious units since if you win then they all die anyway.

Actually what about this version: The unconscious unit can't be attacked at all. So really the only way to die is to be hit by an attack so powerful that your HP goes from non-unconscious range to 0.

I thought of that after I was about to suggest a mechanic where units can play dead. This is a weird idea though because in any game where such a mechanic existed, there'd be no reason not to go after lying-down units with extreme prejudice, thereby defeating the point of playing dead. The only way for such a mechanic to remotely make sense is in a game without visible HP, because at least then there's no way to tell if a lying down unit is dead or not.
author=zeello
I'd first have to know whether AI units will intentionally try to kill the unconscious unit. But then again why would they bother? After all it's not like you would intentionally go after their unconscious units since if you win then they all die anyway.
What if some of the enemy are blood thirsty monsters that enjoy killing and have an AI that makes them act as such?

Meanwhile other enemies could be more tactical. (I may be getting beyond myself here I haven't written a single line of AI code yet)
Please don't make blood thirsty monsters that ignore winning moves just to perma-death your characters. It's fine if they kill someone if there are no other targets, but it's stupid if someone suicides just to screw over the player.
author=hedge1
Please don't make blood thirsty monsters that ignore winning moves just to perma-death your characters. It's fine if they kill someone if there are no other targets, but it's stupid if someone suicides just to screw over the player.
What if it was something you can learn about in advance, e.g. a specific enemy unit type that always prioritises injured targets - some kind of blood hound or something that the enemy uses.

Idle thought - in general having different AI settings for different enemy unit types could be very interesting... ARRRRGH I always get way too far ahead of myself with ideas verses actual script written...
author=hedge1
Abandoning healing will make your game significantly harder to balance

Indeed, if you've built your game with a healing feature and abandon it later, you'll have to rebalance your whole game. If you start without healing, the game is actually easier to balance (no need to choose the amount of HP healed, the range of the healing spells or the cost of potions).

Healing is a tool in the player's hand to fight against bad odds of the RNG. The smaller the impact of the RND, the easier the player can predict the outcome of an attack and know how many hits a unit can withstand. I built a small strategy game where there is no healing but also no RNG so the player never gets a frustrating unlucky roll that ruins his/her plan. If your game features critical hits that inflict 3 times normal damage at 2% rate, you need the healing feature. If the random factor increases or decrease damage by 20% maximum, you can forget healing. The actual "amount of randomness" that is best for your game will come out of testing.
author=Irog
author=hedge1
Abandoning healing will make your game significantly harder to balance
Indeed, if you've built your game with a healing feature and abandon it later, you'll have to rebalance your whole game. If you start without healing, the game is actually easier to balance (no need to choose the amount of HP healed, the range of the healing spells or the cost of potions).

Healing is a tool in the player's hand to fight against bad odds of the RNG. The smaller the impact of the RND, the easier the player can predict the outcome of an attack and know how many hits a unit can withstand. I built a small strategy game where there is no healing but also no RNG so the player never gets a frustrating unlucky roll that ruins his/her plan. If your game features critical hits that inflict 3 times normal damage at 2% rate, you need the healing feature. If the random factor increases or decrease damage by 20% maximum, you can forget healing. The actual "amount of randomness" that is best for your game will come out of testing.
I haven't coded any of the stats yet, there's no healing, no criticals and no percentages of any kind, yet.....

My work in progress battle system currently has support for moving characters around on a grid (with a path finding system so you can select where to move to but it only allows possible moves), and an attack function that lets you select an enemy you're next to and attack them; currently anyone who's attacked instantly dies - a placeholder action until I script HP and attack stats etc...

I'm provisionally planning on having QTEs instead of most RNG - so RNG screw won't be a massive thing, but bad reaction screw could be... I'm unsure on critical hits I probably wasn't going to have them or at least not as a massive factor, certainly not Fire Emblems 3 times damage - as I was intending a lot of characters to go down in 2-3 hits, a triple damage critical would be crazy (the provisional idea is to use numbers and movement and choke points so you don't get an individual hit many times)
Are you making a custom engine for your game?

Strategy and QTE... interesting. I picture Fire Emblem with Legend of Dragoon attacks and counters.

author=Rhuan
the provisional idea is to use numbers and movement and choke points so you don't get an individual hit many times
Or you DO hit the enemy multiples times this way. This is actually the best way to win battles in my game.
author=Irog
Are you making a custom engine for your game?

Strategy and QTE... interesting. I picture Fire Emblem with Legend of Dragoon attacks and counters.

I'm working in Sphere, not RPG maker and writing all the scripts myself - so I guess that means the answer to the first question is yes.

The key inspirations I was running from for the combat were:
- Fire Emblem
- Shadow Hearts (a turn based game with QTEs)
- Fallout (the original from the 90s - turn based strategy with multiple actions per turn based on "Action points" and each action having a cost)

Irog
author=Rhuan
the provisional idea is to use numbers and movement and choke points so you don't get an individual hit many times
Or you DO hit the enemy multiples times this way. This is actually the best way to win battles in my game.
Well yes, hit the enemy multiple times whilst not being hit multiple times yourself, I'm potentially going to make this trickier than in fire emblem by not allowing your units to move through each other BUT easier by allowing movement after attacks if you have AP left.
Sphere is an engine? How does it relate to programming languages like C, Python, Java and Ruby?

author=Rhuan
turn based strategy with multiple actions per turn based on "Action points" and each action having a cost)

I'm potentially going to make this trickier than in fire emblem by not allowing your units to move through each other BUT easier by allowing movement after attacks if you have AP left.

This is exactly what I made! But inspired by:
* table-top Warhammer fantasy battles
* Heroes of Might and Magic 3 (hence the hex grid)
author=Irog
Sphere is an engine? How does it relate to programming languages like C, Python, Java and Ruby?

author=Rhuan
turn based strategy with multiple actions per turn based on "Action points" and each action having a cost)

I'm potentially going to make this trickier than in fire emblem by not allowing your units to move through each other BUT easier by allowing movement after attacks if you have AP left.


This is exactly what I made! But inspired by:
* table-top Warhammer fantasy battles
* Heroes of Might and Magic 3 (hence the hex grid)

1. Sphere is a somewhat simple engine - it's basically a javascript interpreter with built in simple graphics/sound/file handling functions and a pre-made map engine.

If using sphere you have to script all your game logic: combat, menus, stats etc yourself (though various people have made scripts they'll share).
See spheredev.org for more details

2. Oh dear I thought my system would be semi-original, I hope it doesn't turn out identical to yours, is yours fully custom or based on someone else's?
1. So Sphere is the equivalent of what I use: C and the SDL2 library (to handle mouse, graphics and sounds).

2. Don't worry about originality too much. Just make something that is fun for you to build. Mine is fully custom and somewhat based on my personal experience in table-top and LARP battles. It has some really unforgiving elements and unusual controls scheme. You can try it, see for yourself and learn from my design mistakes.
I realise we've run rather off topic - thanks again to everyone for the comments, lots of useful things to think about - I won't necessarily do what people have suggested or not do what people have said I shouldn't do. BUT I'll ensure I have a good reason before diverting from advice - any more thoughts still appreciated.

Irog:
1. Sphere is similar in some ways to working in C with SDL but different in ways too, working in Sphere your functions are much higher level, there's a preset folder structure and a lot of useful high level functions e.g. to put a picture on the screen in a brand new project I can just do:
var picture = LoadImage("picture.png");//look in a folder called images inside my game's directory and load "picture.png" from there

picture.blit(10,10);//draw that picture to the back buffer 10 pixels down and across from the top left corner of the screen
FlipScreen();//exchange the back buffer with the current screen
GetKey();//wait for a key press

Also as it's all run time interpreted you don't compile it - a sphere game is basically a standardised folder structure, one folder of scripts, another of images, another of sprites another of maps etc, with a document at the top level telling Sphere which script to load on launch and what screen resolution to set. The engine is already compiled for Mac, Linux and Windows so there's no porting or cross compiling to do for a finished game.

2. I had a quick glance at your game - it looks very different to what I'm thinking of/working on - some similar concepts but a very different final result, interesting how two things can sound so similar to begin with and end so different.
1. That's a lot more compact than SDL basic functions. And portability is definitely an interesting feature. Thanks for the info.

2. A lot of concepts are similar on paper but implementation choices makes them totally different games. Best of luck for your project. Feel free to contact me for feedback on a demo.
Permadeath and no healing seem to be good options.
Pages: first 12 next last