By analogy, in Formula-1 racing (and other racing too) it is a very strong strategic element to determine after how many laps you should stop at the pit stop, as it determines the exact amount of fuel you want onboard, which influences the car's weight, etc. This is a strategic element. If you were to include pit-stop stops every 2 or 3 laps, it would just become boring, and attrition-intensive. It wouldn't be a racing sport, but mostly the addition of who has the most efficient engineer team to fuel up ;)

This is an excellent point and serves as a great introduction to my opinion on this: attrition is fundamentally related to time. The pit stops are only carefully managed because they waste time, which is in racing and war, the ultimate resource. I think just about any attrition mechanic could work provided it is related in an interesting and fun way to time. The King's Bounty game mentioned above is one example, as are many rogue-like games. These games are generally turn-based, but also have strong incentives not to waste time. In King's Bounty, the total number of turns was limited, in many rogue-likes, consumables and equipment are so tightly rationed that the player has to strive for optimal play, and minimal attrition, or wind up hopelessly under powered.

Dead Rising made the attrition-time relationship explicit. The player could inevitably find almost unlimited provisions, but if he wasted too much time doing so, he would fall behind the clock. I know for me, this gave the game an exciting urgency.

Of course, there is a drawback (which I've experienced in both King's Bounty and Dead Rising ) to inflexibly limiting time, which is that a player may realize many hours into play that they have no chance of success, because they've wasted too much time. This is a serious downer, and very negative feedback for the player. There are many perma-death RPGs that overcome this by making dying-and-restarting a core mechanic, though, so it can be done.

In any case, I think a jRPG probably does not need to be quite so strict, but the standard approach (which I would say is roughly 'we'll discourage the player from wasting resources, by making them waste a bunch of time on a boring walk back to town, then even more time fighting their way back to where they left off') is not terribly fun, and could do with a replacement. One mechanic I think might be fun to experiment with would be purely aesthetic feedback: the more time a player wastes not destroying a ravaging monster, the more of his village will be left in ruins. Ultimately, there is no negative feedback other than seeing that all those cheerful but useless NPCs are now gone.

For the project I'm working on, I decided to limit the number of health potions the player can carry and make them rare and very expensive items. This forces players to try to beat enemies strategically and minimize damage taken. I want players to be mad at themselves if they have to use 3-4 potions during an encounter. I want them to be tempted to re-try the encounter to try to come out with more health potions left. (Similar to Dark Souls)

his is an excellent point and serves as a great introduction to my opinion on this: attrition is fundamentally related to time. The pit stops are only carefully managed because they waste time, which is in racing and war, the ultimate resource.

I think the very challenge is to find a fun way to lose time. One is the actual combat system, but you can only put so much on this before the game looks like a "fight Fest" which I think the game needs others, and let's be honest, most jRPGs are weak at this and often add only menu-based interactions (crafting, etc).

purely aesthetic feedback

While I don't exactly like this specific example, I actually liked a moment in Chrono Cross where it was purely exploration. No fighting. It was actually fun, but you can only go so far before it gets repetitive.

Well, I've played a game where you constantly took small amounts of damage when you were walking on lava or out in an icy environment, then whenever you stood on good ground or paused to huddle for warmth or went into a cave you healed. It was sort of a maze challenge where some painful paths looked good but were just too long for you to survive walking from one end to the other. Not something I'd want to do for a whole game, but fun for one level out of 5 or 6. The most common arrangement is that you take damage in combat and heal or heal faster when not in combat. This works best with games where you can see the encounters and avoid them if you are too low on health.

I've devised a similar system to create an alternate losing condition not related to HPs actually. Whenever in hostile environment, a meter of your internal body temperature was shown and would increase or decrease based on the environment. There were variations to what was cold, colder, hot, hotter, or temperature (Allowing you to neutralize your temp). It sounded ok, but like you said, I can't see a full game of just that, not even 25% of it.

Hp is a short term resource that when lost you get an "instant gameover".

Attitition can also come from other long term "resources" that when lost you get a "delayed gameover".
A "Delayed Gameover" is one that If the player "messes up", they will be able to progress further but they will eventually die and will have to restart the game.

1) The player's bodyparts could take damage (not restorable by heals), potentially wounding them and risking to lose them, at which case you may end up with a character unable to equp items, and they would die from extreme bleeding.

2) Status effects that instantly kill you when their resource bar ends. e.g
a) deadly poison kills you at 10 tokens.
b) petrify kills you after 4 seconds of direct medusa look.

3) permanent stat loss, level downs, the player will be encouraged to dodge spider attacks that lower stats because else they will end with 0 str and 1 hp and instantly die.

4) item durability loss
Player Solution: a) keep multiple item sets. b) avoid getting hit.

You can convert the "Restart from new game" clause to :
"Restart dungeon from start" e.g you have to go back to town to reset your stats, durability, wounds, or to
"Restart from last checkpoint" : Thus the player has to conserve resources between 2 checkpoint.

In wow case, the checkpoints are [battle start, battle end] thus there is no attritition to that game, everyone starts with full hp/mana.
Minions in that game serve no purpose, only to delay you from getting to bosses (filler).

In permadeath, the checkpoint are [game start, game end], it leads to making the game feared (if you die its permanent), more difficult, however if you die you have repeat everything from start. Playing the first stage 400 times, when you have only played the last stage only once sucks.

Having a checkpoint system will allow the players to go back to the [checkpoint route] they messed up, to fix their mistakes. E.g finish the route with more resources thus allowing them to beat the next boss.

2) Status effects that instantly kill you when their resource bar ends. e.g
a) deadly poison kills you at 10 tokens.
b) petrify kills you after 4 seconds of direct medusa look.

That would be attrition at the scale of a single combat. Once the battle is over, the player has lost nothing if they've overcome this, so I'd say this isn't attrition.

Having a checkpoint system will allow the players to go back to the [checkpoint route] they messed up, to fix their mistakes. E.g finish the route with more resources thus allowing them to beat the next boss.

I like this, but I'm having troubles applying the idea of checkpoint to a narrative jrpg. When would these checkpoints be? Specific chapters of the story?
What about backtracking?

A big flaw of Diablo 3 is the checkpoint system. While it allows for quick recovery, if you're where you shouldn't be and die, you start very far from there. Should all dungeon "floors" or "areas" save as checkpoints? That way, every time you enter a dungeon, the game saves a checkpoint, and saves more as you go if necessary?

That would be attrition at the scale of a single combat. Once the battle is over, the player has lost nothing if they've overcome this, so I'd say this isn't attrition.[/quote]
No you can make status effects last permanently, at checkpoints they recover from them, if the player gets unlucky to get hit by 10 deadly poisons between [c, c+1] he instantly dies and restarts from last checkpoint.

if you're where you shouldn't be and die, you start very far from there[/quote]
Agree lets make a more detailed discussion.

1) What approach is better
A) "Balance trash minions based attritition for the route between 2 checkpoints [c, c+1]" or
B) "Balance trash minions based on that they will require 100% of your hp pool to defeat in 1 encounter" ? Afterwords you recover to full hp/mana.

Approach A: makes mobs too easy psychologically, possibly boring as you will only get to critical death state after 4 fights, with no imediate results to signify danger, that you failed as a player to adapt to them.

Approach B: enemies deal 90% of your hp attacks, almost 1shotting you. The healers have to watch the health pools, and get paniced, do i use my slow casting time healing spell, risking death? in approach A, healers spammed only 1 button (their slow heals). Now they have to decide.

Both approaches can lead to the same difficulty, however A is harder because you dont see the result, thus you aren't forced to play at your peek performance. At same arbitary point you "run out juice", which means that you instantly die totally unprepared and unable to do anything to fix it except restarting the game from last checkpoint.

Wow Priests: cannot exist in approach A as their playstyle gets tedius (1button)

Dnd3.5 Mages: cannot exist in approach B as mages are all about resource concervation, the mages deal 1000% more dmg than warriors, but they could only cast 1-4 spells in the whole dungeon. Thus a mage has difficult trash fights, where a warrior has a difficult boss fight. The mage can 1shot the final boss of the game in 1button but he will likely die to trash mobs.


Why did we chose attritution ?
1) Games were too easy without attritution, players would to start fight with full hp, thus
defeating the purpose of random trash encounters.

Well not according to wow psychologists.

Imagine a game that you only fight bosses one after another, challenging your party, is this fun ?
NO !!!

trash encounters make the game have a "relax" period that we dont have to worry about Fight Effectiviness. They are there to relax us from the big boss that we will meet in next battle and we have give all our resources in order to progress. But putting too match effort into attrition of those fights we are removing the "free period" of farming without caring about fight effectiviness, thus defeating the purpose of trash fights.

Questions to hear answers:

1) What approach is better
A) "Balance trash minions based attritition for the route between 2 checkpoints [c, c+1]" or
B) "Balance trash minions based on that they will require 100% of your hp pool to defeat in 1 encounter" ? Afterwords you recover to full hp/mana.

2) Is it fun for trash encounters to exist if they wont challenge your party ?

3) How much time should trash encounters consume for the players to relax, and how much for challenging encounters.
How do i mix them up. Following wow formula i.e Easy fights 5 min; Boss fight 5 min ; Easy Fights 5 min; Boss fight 5 min ;

n00b0dy, your approaches are not a dichotomy. In fact probably all elements of game design (as all elements of life) are a continuum. For the sake of argument, let's say approaches A and B are both bad as you claim, well what about all of the possibilities in between? That is, after all, why we call it game balancing, because its a matter of finding the most entertaining balance between unpleasant extremes.

Also, WoW is not particularly relevant to balancing a turn based retro rpg. I know we've already introduced plenty of outside examples to support the discussion, but there are limits to how far you can take an example. It might be useful to look at why the designers made the pacing decisions they did, but it would probably be fruitless to try and port them into this model.

However, I actually think your example of a petrify (or potentially death spell, or other, really negative things) effect which would increment over time, forcing the player to avoid engaging the Medusa until he was ready to fully commit, that might be fun. It's maybe not a good basis for the basic attrition system in a game, but in a dungeon or two I think it has a lot of promise.

Imagine a game that you only fight bosses one after another, challenging your party, is this fun ?
NO !!!

Not sure if I agree with this. If the purpose of easy fights is just to let the player relax why have them at all? I think a game should strive to make every encounter challenging. If an encounter doesn't challenge the player in any way, its worthless.

I think the level of challenge should change but all encounters should have some challenge. You should be able to die during every fight. If there is no fear of dying, there is no enjoyment.