Elements the Game Forum - Free Online Fantasy Card Game

Elements the Game => Card Ideas and Art => Design Theory => Topic started by: joebob777 on January 30, 2011, 04:27:36 am

Title: Random Effects
Post by: joebob777 on January 30, 2011, 04:27:36 am
we will discuss what kind of cost decrease we should have on random effects here
Title: Re: random effects
Post by: joebob777 on January 30, 2011, 04:30:22 am
on your previous comment, doublecross, about the types of effects being different as a variable, that is a good idea, but we cant have something as different as "add one purify counter" and "play miracle"  to be categorized under "heal
Title: Re: random effects
Post by: Hyroen on January 30, 2011, 04:32:00 am
I think a good card to consider for this thread is Pandemonium, or possibly even Chaos Seed.

EDIT: Also, would you mind capitalizing the thread title?
Title: Re: random effects
Post by: OldTrees on January 30, 2011, 04:34:06 am
Good idea to make a new thread for this topic.

For effects that do not decrease in value I think we should not have a cost decrease.
For those effects that do decrease in value I think the probability of their uselessness multiplied by their value at such a time (might be negative) be used instead of the normal value for those probabilities.
Aka calculate at more and more detailed levels until the events considered get insignificant.

Lets start with Chaos seed.
Title: Re: random effects
Post by: doublecross on January 30, 2011, 04:34:58 am
http://elementscommunity.org/forum/index.php/topic,20492.msg278035#msg278035



Can you guys link to the other cards where we have already argued this some?
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 04:36:17 am
http://elementscommunity.org/forum/index.php/topic,20448.0.html  is the original arguement
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 04:39:56 am
The categorical approch as a 2nd tier more detailed analysis is a good idea Doublecross

Hard CC
Drain Life
Fire Bolt
Gravity Pull
Ice Bolt
Infection
Lightning
Shockwave
Sniper

Soft CC
Freeze
Lobotomize
Reverse Time

Not CC
Parallel Universe

Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 04:41:34 am
OldTrees, I take it this means you read my analysis of Wish|Wish (http://elementscommunity.org/forum/index.php/topic,20492.msg278035#msg278035)

Do you agree with my logic about why the cost should be marginally reduced?


Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 04:43:18 am
Chaos Seed | Chaos Power analysis

it cost one because it can give many different statuses to different creatures
ie: give momentum to enemy creature, or freeze your lava destroyer, stuff like that
that is why it was justified with a cost of one
chaos power can go anywhere from +1|+1 to i think up to +4|+4 which should cost (based that improved blessing gives +3|+3 for 2 quanta) the average of .6, 1.3, 2 and 2.6 because the boosts are most likely only going to go to your own side, which is why it gets little cost reduction
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 04:44:00 am
OldTrees, I take it this means you read my analysis of Wish|Wish (http://elementscommunity.org/forum/index.php/topic,20492.msg278035#msg278035)

Do you agree with my logic about why the cost should be marginally reduced?
Yes occasionally the increased desire for one effect should raise the price (see shockwave) and the occasional decrease in desire of other effects (chaos seed as cc) should decrease the price.
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 04:45:51 am
Shockwave is not a random effect card though.

The randomness should always force the cost below the average, unless every effect is always fully useful.

Edit: I broke out my arguing profile pic.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 04:48:31 am
so if whatwe go by is number of categories, then we should first figure out the different categories before finding a determining cost factor
oldtrees stated above the types of cc
we could split up major and minor buffs
poisons in a group
split up major and minor heals
anything else?

@doublecross, he was stating a category, just like in wish, miracle is not random, but part of a random effect
hehe didnt read full post
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 04:50:17 am
I thought he was saying that shockwave costs more than just it's damage, because of it's shatter effect. Which would be true, but not directly relevant.
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 04:52:12 am
Shockwave is not a random effect card though.

The randomness should always force the cost below the average, unless every effect is always fully useful.

Edit: I broke out my arguing profile pic.
Contrary to popular belief Shockwave is random. There is a variable chance of it being used to target a frozen creature.
Randomness that creates added value (like shockwave's instant kill) should raise the cost above tier 1 averaging.


so if whatwe go by is number of categories, then we should first figure out the different categories before finding a determining cost factor
oldtrees stated above the types of cc
we could split up major and minor buffs
poisons in a group
split up major and minor heals
anything else?
Um Chaos seed only does the things I listed and each card probably should be judged by itself.
Title: Re: Random Effects
Post by: Hyroen on January 30, 2011, 04:52:38 am
[offtopic]

[snip]
Edit: I broke out my arguing profile pic.
[/snip]
Totally off topic but I just wanted to say I found that funny, because before you changed your profile pic I had right clicked the image, copied location and pasted it in a new tab...

I couldn't read the font. Funny stuff though.

Also, it would be a nice touch to the forum to include an actual "off topic function" which would work with the code presented.

^_^

[/offtopic]

EDIT: Goes to suggest it in FS&F.
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 04:53:44 am
Please explain. Shockwave is a targeted attack.

(Unless you mean shockwave as part of pandemonium)
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 04:54:14 am

so if whatwe go by is number of categories, then we should first figure out the different categories before finding a determining cost factor
oldtrees stated above the types of cc
we could split up major and minor buffs
poisons in a group
split up major and minor heals
anything else?
Um Chaos seed only does the things I listed and each card probably should be judged by itself.

i wasnt just talking about chaos seed i was talking about the categories of randomness all together
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 04:55:11 am
Please explain. Shockwave is a targeted attack.

(Unless you mean shockwave as part of pandemonium)
shockwave is targeted but the instakill cost increase was what they need to judge based on how often it was used, i think this is what oldtrees is saying
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 04:55:56 am
Please explain. Shockwave is a targeted attack.

(Unless you mean shockwave as part of pandemonium)
The reason why a random effect might deserve a lower/higher cost is the same reason why Shockwave has a higher cost than its damage.

I was referencing that shatter effect as an example of increased value in contrast to your examples of decreased value.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 04:57:06 am
are we all back on the same page now?
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 04:57:26 am
Right. I wouldn't consider that to be random though. That is, however, a situational consideration.



I still don't see any reason that randomness would ever take the cost ABOVE the average cost of each possible effect.


Tell me one scenario were that would make sense.
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 05:00:06 am
Right. I wouldn't consider that to be random though. That is, however, a situational consideration.



I still don't see any reason that randomness would ever take the cost ABOVE the average cost of each possible effect.


Tell me one scenario were that would make sense.
Photon + Chaos Power + Adrenaline
+2atk is worth much more than +3atk in this situation. +3 attack is not valued less than normal rather +2 attack is valued more than normal.
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:01:49 am
Yes.

That doesn't actually work as an example.




Chaos power costs 1 :entropy.


That is less than or equal to the cost of any other stat gain card.


Thus, it is not an example of randomness increasing the price.
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 05:05:16 am
Yes.

That doesn't actually work as an example.




Chaos power costs 1 :entropy.


That is less than or equal to the cost of any other stat gain card.


Thus, it is not an example of randomness increasing the price.
It is an example of the kind of situation that in great enough quantity (probability x value) might. It is probably offset by the negative of increasing HP to 5+ vs Gravity shield.
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:07:43 am
Right. However, the point still stands that chaos power is a non-example.
I would say that this is because randomness should only ever decrease cost.
This is because being able to choose an effect is always more valuable than not being able to choose.


Being able to choose between A or B should, logically, always be more valuable than being informed that the computer will choose for you.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:09:48 am
ok now that we have that cleared up, i think we can all agree that we should section random effects together so that we can better determine the cost decrease, and for that, we need to think of all possible sections
NOTE: i am not talking specifically about chaos seed/power but about random cards in general
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 05:12:33 am
Right. However, the point still stands that chaos power is a non-example.
I would say that this is because randomness should only ever decrease cost.
This is because being able to choose an effect is always more valuable than not being able to choose.

Being able to choose between A or B should, logically, always be more valuable than being informed that the computer will choose for you.
Good point but is the increase in value due to choice always >.5 quanta? I would claim that it typically is not unless wasteful targeting occurs.

@joebob
What type of random card should we talk about first?
Random between two beneficial options? Random target? Random between conflicting options?
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:14:10 am
Well, the sectioning will only ever work as an approximation.


Here is the reason why (several examples)
1 damage vs. 2 damage  Equal utility for a 1 HP creature, unequal utility for a 2 HP creature
Healing 10 HP vs. Healing 20HP   Equal utility at 90HP, unequal utility at 80HP.

Need I go on?
Trust me that the uniqueness of these abilities means that the grouping will help us only to the point of being a useful approximation. We are not going to produce a formula that has mathematical reasoning behind it.

We can figure out one that would be a useful guide, but it will still have to be adjusted for a given scenario.

EDIT: Not necessarily over .5 quanta. Like I said, this will ultimately have to be a case by case thing.   However, I was arguing the mathematical principle behind it, which says there should be some sort of decrease. Whether or not that gets nullified by rounding was not part of my original consideration
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:17:06 am
well we're going to have to get as close to the best case guide as we can because we wont always be able to judge stuff like 1 vs 2 damage on a 1 or 2 hp creature
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:19:44 am
@joebob777
If we are making an approximation, I rather liked your original approximation back on Wish|Wish.


Of course, each card should be re-balanced, but this can still be a useful tool.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:20:24 am
Right. However, the point still stands that chaos power is a non-example.
I would say that this is because randomness should only ever decrease cost.
This is because being able to choose an effect is always more valuable than not being able to choose.

Being able to choose between A or B should, logically, always be more valuable than being informed that the computer will choose for you.
Good point but is the increase in value due to choice always >.5 quanta? I would claim that it typically is not unless wasteful targeting occurs.

@joebob
What type of random card should we talk about first?
Random between two beneficial options? Random target? Random between conflicting options?
what we should talk about is sectioning random cards in general, like if it does random cc then section that into heavy cc, mid cc, and low cc we keep going until we run out of sections and then we can judge how much we reduce cost for randomness
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:21:21 am
@joebob777
If we are making an approximation, I rather liked your original approximation back on Wish|Wish.


Of course, each card should be re-balanced, but this can still be a useful tool.
i liked that aproximation too but to make a more acurate aproximation, we need to section randomness as you hinted in wish
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 05:21:52 am
How about this:
-0(+0): Few rare cases of decreased (or increased) value
-1(+1): Some occasional cases of decreased (or increased) value
-2(+2): Lots or frequent cases of decreased (or increased) value

-0: Few options
-1: A handful of categorically different options
-2: Random assortment of options
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:23:10 am
I still have yet to be convinced that the value would ever increase due to randomness.


I feel it would be much clearer without the (or increase) and the (+number)
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:23:54 am
we could go by that, so down to -4 and up to +2 if we can find a case like that
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:24:32 am
I still have yet to be convinced that the value would ever increase due to randomness.


I feel it would be much clearer without the (or increase) and the (+number)
this is a backup just in case we ever run across something random that increases value
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:26:49 am
...



There is no way that removing control from the user makes it more valuable.


Either the randomness includes an ability that you cannot normally do (in which case, we can't do the analysis, because we have no reference cost), or it does something we can already do, and just removed our ability to choose (in which case, it is less valuable)


IT DOESN'T INCREASE
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:27:21 am
ok i say our first CIA card to discuss should be wish, then we should move onto the lucky series
wish: http://elementscommunity.org/forum/index.php/topic,20448.0.html
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:27:57 am
...



There is no way that removing control from the user makes it more valuable.


Either the randomness includes an ability that you cannot normally do (in which case, we can't do the analysis, because we have no reference cost), or it does something we can already do, and just removed our ability to choose (in which case, it is less valuable)


IT DOESN'T INCREASE
as i said before, it is a just in case, it probably wont happen, but just in case . . .
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:29:41 am
Sorry I am being OCD, but there is no probably.

For every case of randomness, one of the two things listed above will happen. Both make adding the (+) unnecessary.


That is the last I will say on that subject though.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:31:24 am
ok now back to wish, how should we judge its cost?
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 05:33:40 am
Sorry I am being OCD, but there is no probably.

For every case of randomness, one of the two things listed above will happen. Both make adding the (+) unnecessary.

That is the last I will say on that subject though.
How about this:
Situational/Fringe effects modifier
-0(+0): Few rare cases of decreased (or increased) value
-1(+1): Some occasional cases of decreased (or increased) value
-2(+2): Lots or frequent cases of decreased (or increased) value

Loss of control modifier
-0: Few options
-1: A handful of categorically different options
-2: Random assortment of options
I named the modifiers in a way that might explain it better. Situational modifier included for things like Shockwave.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:37:03 am
i see 4 different category options in Wish | Wish
1) minor heal, heal 10 hp or heal 10 hp plus purify
2) major heal, miracle
3) summon creature, summons crusader, and in upped's case, a nymph
4) draw a card,
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:37:39 am
Well, I am inclined to stick with 3|2

I would be confident about that decision if the activation of miracle drained  :light like it normally does.


However, with the heal-but-don't-drain- :light I am torn between 3|2 and 4|3.

Your average work said that, if all effects were always full utility, the cost should be 5 :light.

Healing is usually useful, to some extent.
Summoning a crusader, or nymph is usually useful, but more so early. Miracle is more useful later.
Drawing a card is probably not what the person needs if they are using a card that seems mostly focused around healing.


Thus, I think a cost decrease of 1 or 2 is called for.

However, the modification on the miracle effect would mean that the full utility cost should be above 5. Thus, a decrease of 2 would take it to 4(ish).



Actually, I now think 4|3 is the most balanced cost.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:40:13 am
the 4 different types of randomness and the miracle without the drainage cancel each other out, hehe nice, ok 4|3 it is
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:44:42 am
ok very quickly, under oldtrees guide, there are 3 levels of categories
Loss of control modifier
-0: Few options
-1: A handful of categorically different options
-2: Random assortment of options

but what are few, a handful, and random
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 05:45:31 am
I think 4|3 or even 4|2 sounds fair.
Removing the draw a card effect and increasing the cost would be better IMHO.

Few 1-3, Handful 5-7, Random 9+
4 and 8 options are the borderline cases.
Remember this only counts significantly different options. heal 10 and heal 20 would count once.
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:46:58 am
Well, for the categorization, we should not necessarily go based on number of categories, but rather on how likely the different effects are to be useful in the same scenarios.

That is hard to quantify though.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:48:48 am
if we cant quantify the different outcomes then we at least need to be able to quantify something to make up for it, and we might need to do both for some ridiculous cases
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:50:14 am
Well, the numerical version that we were looking for about 3 posts back would be a good substitute.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:53:18 am
agreed, so the randomness/chance part of determining a cards cost has been made, good job guys
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:54:44 am
I would call this a good success.

+ Karma for both of you for being willing to work with me.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 05:55:58 am
same for you
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 05:56:51 am
Cheers, mate.


Good work everyone.
Title: Re: Random Effects
Post by: OldTrees on January 30, 2011, 06:00:20 am
This was a helpful exercise.
+1 karma each.

Be warned. You are both on my list for people to draft into design theory discussions. Especially you doublecross. Too few people think about what I say instead of taking it for granted. Even fewer argue back with good reasons. You are on a short list with Uppercut (when he is not silenced)
Title: Re: Random Effects
Post by: ZephyrPhantom on January 30, 2011, 06:11:27 am
Yay! New CI&A organization founded! I'll try to be as unbiased as I can.

Well, we need a formula. I've done several calculations based on other formulas and cards, as shown below.

Chance Bonus = Number of Chances +Abilities in question
Note :  :rainbow = Other quanta, While  :underworld = different quantas combined
Wish is the example card.
Unupped
Chances: 5, -1
Abilities in question : Holy Flash Player-Only (0), Weaker Precog (0.5  :time ) ,Heal w/ purify counter ( 1   :underworld ) , crusader (5  :light ) , Miracle ( 15  :light)
Percent Modifiers result in the following values:
0
.1
0.15
.5
0.75
____
1.5  -> Rounds to 2.
Final Value: 1  :light

Upped:
Chances 6, -1
Abilities in question: Holy Flash Player-Only (0), Weaker Precog (0.5  :time ) ,Heal w/ purify counter ( 2   :underworld ) , crusader (5  :light ) , Miracle ( 12  :light) , Nymph (7  :underworld )
Percent Modifiers result in the following values:
0
.1
.4
.5
.35
.6
____
1.95 -> Rounds to 2
Final Value : 1  :light
Card calculation
Spell
Unupped
Effect - Precalculated Chance Effect = +1
Bonuses : +2 as a spell.
Final Cost : 3  :light
Upped
Effect - Precalculated Chance Effect = +1
Bonuses : +2 as a spell, but -1 as an upgrade. Final Bonus = +1
Final Cost : 2  :light
I know the discussion may be over, but we should have something for reference the next time this happens.
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 02:00:33 pm
earlier in wish i determined that the average cost was 5|4, then using the formula, we determined that there was a  -1 point because it occasionally decreased in value, then we determined that there were 4 categories, which is borderline -0/-1 so the cost should either be 3|2 or 4|3, your choice
Title: Re: Random Effects
Post by: doublecross on January 30, 2011, 04:17:22 pm
Of course, we haven't perfected the formula yet.


Number of categories is a good approximation, but the deduction ultimately should be based on how far below maximum utility the card is likely to be at any given time.

For instance, causing poison is an early effect, healing a certain percentage of health is a late effect. Something that could do one or the other will be at diminished utility at all times, because at least one effect with be less useful then when at full utility.


For wish, one such pairing is creature summoning v. miracle.
One is more useful early, one is more useful later.



Title: Re: Random Effects
Post by: karis on January 30, 2011, 07:39:23 pm
i'm know it's not look fit 'ramdom effects card cost' topic at all.  but can i'm say something...

entropy have a title of randomness.   but all random thing can't be only entropy.....
Title: Re: Random Effects
Post by: joebob777 on January 30, 2011, 08:08:08 pm
i agree not all random things should be entropy, but we are talking about cards like Wish|Wish that have random things that might occur
Wish|Wish: http://elementscommunity.org/forum/index.php/topic,20492.0.html
blarg: