...But It's gotta be JUST Right.
Published on January 31, 2010 By ScottTykoski In Elemental Dev Journals

So the next big beta is scheduled for this Thursday, where evenone on the beta group will get their first taste of several new features.  One such feature (possible the biggest, from a gameplay standpoint) is in the implementation of the spellbook. You can now learn spells and, provided you have the mana and/or essence, cast them with a flick (click) of your wand (left mouse button).

One of the suprising things we found this Friday, during our end-of-the-week powwow, is that the cloth map makes certain spells..well...dull. Spells that should be super-awesome just come across as lame. Raise land, for instance. You're a sovereign, summoning your powers to rip the world asunder, pulling a rocky cliff-face form it's ageless slumber. The world shakes as dirt and ash fill the sky in a magical haze. The rumbling stops...your mana drained, will depleated...you look upon your creation... 

*ploop* A brown little mountain icon.

Now, over time it's cool to slowly shape the world to your needs, but without the 3d map the effect is anti-climatic, to say the least.

On the other hand, this has given spells such as 'Charm Monster' a chance to prove their worth. Using magic to build an arachnid and troll army has proven quite enjoyable...an instant gratification spell that could also turn the tide of future battles.

The problem I forsee is one of balance...and I know we've talked about making these spells fun and crazy in the beta sandbox, but I'd like to pilot this one a bit tighter, since it does have the power to be very unbalancing.

A balanced feature has solid logic as it's cornerstone, and that's what I'd like to discuss today. There are several different stats and countless viable equations that can be used to deterime the a sucessful charm, and this forum is a good a place as any to pick some brains for a solid solution.

--------------------------------------------
- Current Implementation -
Any 'CharmTarget' modifier (the XML data that get's attached to a spell) has 3 key values...the PARENT Unit (who cast the spell), the TARGET unit (who's being charmed), and the STRENGTH of the charm (0-100).

What the game currently does is this...
- calc the difference between the two units levels (TARGET LVL - PARENT LVL: a negative value meaning the target is weaker)
- subtract the above difference from the STRENGTH (a weaker target will result in a stronger strength)
- use the new STRENGTH value as a % chance the charm will work.

So, in the current game's implementation, the charm spell has a strength of 100 (just for fun). However, if your level 1 sovereign casts it on a level 8 troll, there's only a 93% chance that the charm will hit it's mark. Seems high, but it IS the strongest charm spell you could get.

----------------------------------

Perhaps it's a good-enough implementation, a more gameplay will determine if it's fun-factor dosen't overpower game-balance, but I'd like to open the floor: what do you guys think would be a fun and balanced way to deal with the suprisingly enjoyable art of 'Charming'.


Comments (Page 1)
on Jan 31, 2010

What the game currently does is this...
- calc the difference between the two units levels (TARGET LVL - PARENT LVL: a negative value meaning the target is weaker)
- subtract the above difference from the STRENGTH (a weaker target will result in a stronger strength)
- use the new STRENGTH value as a % chance the charm will work.

The problem with such a formula is that it makes differences in level almost inconsequential, except for huge differences (level 20 versus level 5) or with weak spells (STRENGTH = 20, but to a whole stack, for example).

But aren't units going to have a magical defense rating? If so, why not use it to determine the odds of charming? So you could have : charming probability = 1 - TARGET MAGIC DEFENSE / (STRENGTH + PARENT LVL).

on Jan 31, 2010

Meh sounds fine to me. Now on to more important matters, can I get in the beta yet.

 

I wonder how long I can ask the same question before I get smacked with a giant ban hammer.

on Jan 31, 2010

Wow, great post boogie!  I like the current scenario well enough, but what if we had to maintain control over units with mana upkeep? You know in order to keep the units will as ours. This would make our decision a little more important. Make us weigh our options more. Will we be able to charm enemy units as well? I think that would be an interesting implementation for spying   

I think units with a "powerful mind" should be able to resist these charms easier, while unintelligent beings (spiders and trolls) should be easy to take control of and have a low upkeep. But of course upkeep should probably be determined by the attack strength of the unit for balance sake.

on Jan 31, 2010

First we will send waves of spiders to wear them down... then the trolls, then a dragon, and finish them off with da bears! Only a few more days!

on Jan 31, 2010

Charming, as you describe it, lacks something : a "level" of charming : will the charmed unit obey your orders as a zealot ? Has it some brains remaining ? What will happen to the unit when the charm will stop ?

So, some things should be added :

When you cham someone, you can have 3 results : a weak, a normal and a strong.

A weak result will give you access to the unit, but not its combat skills (for instance),, a normal just give you control, and a strong result will give you a zealot unit that hjave enhanced stats.

 

Maybe even better results : the charmed unit will stay with you forever.

Or worse results : maybe you damaged the unit forever (Braaaaaaaaains ....)

And what would happen if you fail the spell ? A free attack from the unit ? That unit gets enhanced attack against units from the charmer ? Or maybe the charmed unit is "just" scared ?

on Jan 31, 2010

Everything seems fine, except that a level1 caster shouldn't be able to use max lv charm. Casters should have to gain experience and lv up before they can cast rediculously powerful spells. LV 100(100 strength?) charm should require a lv 50 caster or something of the like. If you did this, you wouldn't have to factor in the level of the caster at all. The charm could have a strength, the monster could have a level, and any other modifiers could also come into play. S=strength  l=monster level and x=any other modifier(magic resistance). So the new formula would be s-l(-, /, or *)x= probability. If s=20 l=1 and x=5% magic resistance then the probability would look like this. (20-1)*.95= 18.05% probability. Just raise the strength of charm and you simplify the equation by a lot, AND you solve the caster level requirement ordeal. In the old system, a level 1 caster with a level 1 charm(1 strength?), trying to charm a level one monster, would have a 1% probability. If a level 1 caster can use a level 1 charm(level 1 charm has 10 strength for example), you have a 9% probability. Then a level 5 caster can get a level 2 charm(15 strength), a level 10 can get a level 3 charm(25 strength). etc. IDK, it just makes more sense to me, and it works just as well. Why factor in the level of the caster? Why does it matter how many dragons he has slain, or armies he has lead. When it comes down to it, the only thing that matter is the strength of the spell. His experiences thus far determine if he can cast said spell, not how well he casts it. IF you were to take that route, the only experience related factor that should modify the probability of the spell should be how many times the caster has used the spell, and how effective those uses were.

on Jan 31, 2010

Units with massive willpower should be able to resist charm most of the time, things like paladins and demons and such. On the other hand, units like zombies(or equivlent if applicable) should be flat out immune, since there is no brain to charm.

Perhaps also add a thrall or gaes spell that you can cast on your own troops to permantly bind them to you? Not only makes them immune to charm but also decreses upkeep since you don't need to pay magically bound units. Or, could we charm our own units to keep from having to pay them?

Also, It would be cool if charm wore off or could be broken free of. Charming that mountain gaint might have given you a huge advantage in the battle, but you need to keep troops near it who can handle it, in case it suddenly breaks free and starts beating on your army.

Charm can easily be one of the strongest spells in the game, and should be, as there is nothing more dangerous then a foe using your own forces against you, but at the same time, turning the game into a charm fest would get old fast.

on Jan 31, 2010

This sounds great but i can see one possible source of imbalance in the fact the level difference does not "hurt" enough imo for low end spells.

Right now a difference of one lvl between mob lvl and caster lvl always has the same impact on your ability to charm the creature: 1%.  It makes it possible to charm creatures high above your level with early charm spells. It also gives you an incentive to charm only high end creatures (why charm a level 3 spider when you can charm a level 10 dragon for a 7% penalty?)

I would imagine higher levels of spells make the level difference between casteer and creature "hurt" less allowing you to charm mobs with higher level than yours. Imo two factors should be balanced: the base probability which would be the maximal success chance of your spell and a multplier so that only players who have invested in the charm spell can effectivly charm creatures above their level.

In formulas, a level one spell would be for example 50-Max(difference of levels)*10, 0) and a last spell 100-Max(difference of levels)*1, 0). Early game you can only enroll cannon fodder for your army, late game, if you have invested in charm spell, you can charm top notch dragons that are stronger than you are.

ALso i imagine all charm spells should come with a cetain probability of seeing the creature break loose and wreaking havok in your army -)

on Jan 31, 2010

Hmm. It is not fun if you cast a powerful charm spell, and it does nothing wasting the spell. It is overpowered if you can too easily charm powerful creatures. Why not make the charm result non-binary?

If the charm is strong compared to the target it works permanently. If it is weaker, the charm works temporarily. If it is pathetic, it fails. The temporary duration could scale smoothly, making it not an all or nothing effect. If the chance of complete failure is high, even a very successful roll should result in a few turns of the target being charmed. If the chance of permanent charming is very high, even the worst case scenario should allow you to charm the target for a dozen turns. By balancing the spell this way, it can be balanced to be non-overpowered, without feeling weak due to having it do nothing a lot.

Next, it appears that with charm level, you mean how strong this particular charm spell is. That is, there are multiple charm spells. If that is the case, that sounds like a bad idea. Having multiple spells with the same effect makes individual spells less unique. Instead, allow the caster to choose how much power to invest in the spell, like you can in Master of magic with spells such as disjunction.

Furthermore, the idea of using the caster's level directly seems odd. While level should influence the casting skill, it should not be only thing to do so. Making items for example, should be able to help. But since such things are not yet implemented, using level should do for now. It is something that I think should change prior to release though. The target's magic/mental resistance should also not be solely depend on its level. But for now, that too is acceptable.

So how about the following formula:

(STRENGTH + 5 * (CASTER LVL - TARGET LVL) + D30) / 5 = Duration

If Duration exceeds 15, effect is permanent instead.

With D30 representing a random integer with an equal chance of being any number between 1 and 30, or the effect of simulated 30 sided die, and the duration being in turns. Or something like this formula, depending on the desired random factor and influence of Level.

 

on Jan 31, 2010

I scoff you 3D-obssessed deck junkies, but you get the point back for good onomatopoeia. Bring on the plooping terrain icons, and sing praises to the Horse That Made the World.

Re the first run at charm spells, I also think that some entities should be difficult or impossible to charm no matter how powerful a spell-caster is. And I have an uncomfortable twinge from my inner D&Der who wants separate spells for charming character types vs. monster types. That would be one way to work around the no-free-will=no-way-to-charm problem Cerevox mentioned.

The Level thing also seems wrong because it is too abstract. If the formula needs a stat, it should be something like Will, Wisdom, or Intelligence. Otherwise it seems likely that the game will end up treating things like young dragons and troll-kings as equally susceptible to charms, which would seem very wrong.

Making the results a scale of some sort rather than a binary success/failure also seems both logical and more fun.

on Jan 31, 2010

Charming, just charming..... I agree that the spell math system is a bit oversimplified as is... I'd like to see maybe a few other potential effects.. such as if the unit is stronger than you a failed charm spell could potentially turn back on the user and charming him, or that a unit trying to resist charm spells would suffer mental/physical damage if it and/or the channeler is strong enough... or a possible multi-turn mind battle a bit like Denethor and Sauron...

on Jan 31, 2010

Charm is a kind of spell that makes sense to have periodic checks on. Charm does not imply Dominate, so Charm is more of a suggestive spell rather than a puppeteer spell.

To that end, the effectiveness of Charm should in part depend on what you suggest the unit does. Movement is easy. Self-defense is easy. Attacking is harder. Attacking a related species (troll attacking another troll) is very difficult.

The strength of the spell and the level/stats of the caster would determine the chances of the charmed unit obeying your suggestion, each time an order is made (this way a multi-turn movement order doesn't need to be re-evaluated each turn).

Now if you were to put an actual Dominate spell, that really is just a one time cast to break the creature's mind and if successful lasts until the creature dies.

 

A couple interesting mechanics to consider:

1) As mentioned, creature 'will' resists. A strong-willed creature (sentient) should be tougher to charm/dominate than a critter (spider).

2) Maintained spells. You'd have to maintain your hold on the target creature, costing x points of mana regeneration (similar to the spell system in Guild Wars, where each enchantment you maintain slows your energy regeneration). This would work well for Charm/Dominate types of spells.

on Jan 31, 2010

In every single thread about gamemechanics there's always talk about logic and sense just like it mattered....

 

 

You (developers) want the most control to make the game as good as possible so a willpower attribute would be useful.

I also echo what Netriak said about lvls of control instead of success or failure.

on Jan 31, 2010

How does charm relate to animals that can be tamed or used as a mount? If I find a critter A for instance, can I keep it away from the battle and start up a stable eventually (several years and generations later) overcoming the need to domnate the animal?

on Jan 31, 2010

I just assumed that probability determined the probability of complete charm and that the respective variations of control would be present, too. Damn. *slaps forehead* naughty brain, naughty! *slaps forehead again* stop assuming brother, get a hold of yourself.

 

*crawls into corner of the room, rolling back and fourth in feeble position*