Experience Points (Page 2 of 4)

I don't know if I agree 100% with any of the systems, although I have to admit that I would have never thought of some of them (i.e. setting up attacks as a list with queues).  If I was making an attack system I would make it sort of a sliding tug of war thing.  I may be oversimplifying things and/or touching on ground already covered in these posts (long posts on small monitors tend to start fading into a series of blah blahs... when I read them)

But dig this little scenerio where a player attacks a monster with the ultimate goal obviously being taking the monster down to 0% health.

Monster (100%)  ---  Player 1 (0%)
10/10 hp            ---  Player 2 (0%)

after attack (let's pretend each attack does 1 dmg)

Monster (80%) - Player 1 (10%)
8/10hp             -Player 2 (10%)

monster heals 1 pt health

Monster (90%) - Player 1 (5%)
9/10hp            -Player 2 (5%)

now say another player comes in for the attack and all three attack.

Monster (60%)  -Player 1 (15%)
6/10                 -Player 2 (15%)
                        -Player 3 (10%)

Now say player 1 dies but other two get hits

Monster (40%)  - Player 1 (15%)
4/10                 - Player 2 (25%)
                        -Player 3 (20%)

And so on and so forth until the monster is at 0%, in which case the players get their respective percentages of the total exp the monster has.

As a monster heals, the percentages slide from players equally until the monster has 100% of it's health again and the player has 0% in which point the link is broken between the player and the monster. 

Again, I feel like I might be oversimplifying this, but please feel free to shoot down any flaws in this system.

Parties... Tricky thing about parties is that, as zammbi mentioned earlier in this thread, the levelling system's experience requirements are nonlinear. So it may not be wise to distribute experience equally for folks in a party, and it'd have to tie in somehow to a party's total contribution, but how to distribute within the party...

Well my party system doesn't have to that complex. When a Healer is partied with a warrior, then he gets the % of xp the warrior gets by the % he healed the warrior. Not that hard. Shield spells could work the same way. A mage casting shield on a warrior then he gets a % of xp that the warrior gets by the % of damage blocked. Damage increase spells could act similar to normal hitting, so the % of increase damage of the damage done to the monster is what xp the mage gets. Cursing spells can act just like damage increase spells but just casted on the monster instead.

Like? I don't see anything else they can help out in a battle, they could be mining in the corner but that has nothing to do with the battle. Healing, spells, damage is all I can think of to kill a monster. I see this system still fair in my eyes.

It's entirely fair for the computer to decide worth in battle. That's the whole point of any experience system. Also, don't worry about efficiency for experience - there are many hugely inefficient parts of my code, and that one will never even show up on the map. It will occur at most a few hundred times per second.

One issue I have with determining 'merit' based on shield, etc. is what do you do about, say, petrify spells? A mage doesn't have to run away, and then you have to factor in the speed at which they are running, etc., whereas a warrior doesn't need to run away and it's just the missing damage, which is not a function of player and monster speed.. Honestly for design tools it would be valuable to be able to quantify the effect of speed and what have you, but I'm not sure of an appropriate way to do it. While due to all the difficulties I don't think I like having merit based on hundreds of factors determining experience points (because of the enormous uncertainty -- if you can think of a way to resolve it, though, I'm all ears), it *does* have value in say, a tool that goes in the monster editor which could estimate an appropriate experience given who will be fighting said monster. That would eliminate a lot of the guesswork.

Experience and monster levels would only work if we could guarantee a certain difficulty for the level. One option would be to use the creature's stats to calculate what an equivalent player would be at, but that still wouldn't take into account resistances, equipment, etc. which make a big difference for players. It'd be useful to be able to quantify that.

Also, Visualcode something on code style, in C++ float has a constructor. To cast you can go /float(100). It's safe. float() constructs zero, which leads to the odd primitive-only issue

int a; // undefined
int a(); // zero

Nevertheless I'd really recommend primitive constructors instead of static_casts. Also that should be /float(LevelRange).

I'm quite aware, and it's not bad style, because that's not a C style cast.

This is a C style cast.
(float)100

This is a C++ primitive constructor. I imagine it fails quite swimmingly on void*.
float(100)

All compilers will support float a(). You use it in templates, because otherwise there'd be no practical way to initialize arrays to their default values and still support both primitive and non-primitive types. I don't mention it suggesting its use outside of templates. What makes it interesting is that unlike non-primitive types, type a(); differs from type a;.

I'm pretty sure what Zammbi is suggesting is for normal experience distribution not parties.
In a normal situation people absolutely should be sharing based on what they do. Higher level players wouldn't need to let lower levels into their party, the lower level could just walk over and help them. I'd really like a scheme of that sort. It'd encourage unprompted charity, admittedly for gain so it's really not charity, but that would encourage a more pleasant atmosphere if we could do it. As it stands nonpartied healers get no experience at all, so anything's an improvement as long as it can't be exploited.

As for restricting levels, I despise games that railroad players without any freedom, and I don't intend to write one. But fairly loose restrictions that improve balance and gameplay, those is ok. It's a fine line.

Zammbi: What if the monster is chasing someone? Perhaps we could call it 'if they were able to attack in that second in the first place'. But the real trick is making the system sufficiently generic that it doesn't restrict the types of spells we can make.

1.  Is it possible to do this upon partying ( I am not a master coder, I haven't done any for years, so my pseudo coding is lacking too):
Check for "entity a"
"entity a" exists"
Check for "entity b"
"entity b" doesn't exist
add charactera to "entity b Char"
add characteraLVL to "entity b TotalLVL"
add 1 to "entity b count"
"entity b LVL" = "entity b TotalLVL" / "entity b count"

Then you could get an average lvl for the group and implement that with Visual's idea of % of exp if CreatureA produces x% to an "entity" at lvl 50 and (x+1)% to someone lvl 49.  And then you could have a dungeon "suggested" for characters around lvl 40 to 50.  This way characters could get a % boost or drop in exp for going to "suggested" dungeons.

2.  I think a perfect dispurtion of xp based on number of characters in a party will promote power leveling.  Maybe somehow using a method that could compare a characters lvl to the groups average lvl (mentioned above) to districute percentages on lvl instead of "what you have done".  I agree with Visual on this.  Lets say Berserker can NOT kill a blue dragon by his self even with food and potions, and Mage can NOT kill a blue dragon his self even with summons.  So, if by Mage healing Berserker once or slowing the blue dragon once, or doing anything once allows the two of them to kill a blue dragon together, why wouldn't they split the experience evenly?

Lets even go to extremes with this.  Berserker can't kill a tiger by his self.  Mage can't kill a wolf by himself.  I think the same situation above should still apply.  Without Mage, Berserker wouldn't even be at a tiger spawn.  Mage's roll is just as IMPORTANT as Berserker's.  I think experience should be based more on the idea of IMPORTANCE and not the amount of actions you do in battle.

3.  Stemming off of IMPORTANCE.  As Visual was saying with his idea of FIFO of experience, I think this plays very well with IMPORTANCE of the persons attacks.  The longer a monster is a live the less IMPORTANT the attacks of the first person are.  The most recent attacks are the most important.  I don't think the amount of damage a person does compared to another person per hit on a monster should be held in as high of a regard as Zer, and Zammbi are putting it.  Even though I may only be able to do 10hp of damage and you do 50hp of damage, I am doing just as much work and as important of work as you because I am keeping it from regenerating.  So if you attack a monster and get down to whatever health and you die and I walk up and start attacking it and take it down to only a couple more points and die then you come up and finish it off I feel that my part was just as important as yours, because I kept the HP to a manageable amount for you to handle when you got back.  I may not be in your group or even know you are fighting it still, but I should get rewarded accordingly.  As for waiting for a monster to heal to get more exp, I think that is stupid.  It SHOULD be faster to kill the monster and let it respawn.

These are just a few ideas I threw around in my head while I was reading ALL of the posts on this topic.  What do you guys think?