[Official] Dimensional Shield | Phase Shield

[eaglgenes101]

Don't forget the quanta accumulating behind those shields. Monodarks are faster than mono aethers, and they have more heal cards.

[OldTrees]

Use the code in my post on page 24. Actually, Dim shield at 30 is still about as good as dusk shield (actually slightly better). With more pillars and fewer creatures it can do more damage to a pure rush opponent. I used 20 pillars and 4 creatures for 3 turn 30 quanta dim shield, and it was slightly better than 15 pillar, 12 creature, 3 dusk shield. 

With only 4 creatures in a 30 card deck, what fraction of the time does it deckout?
Did you attribute half of the deckouts as losses?

[jawdirk]

Use the code in my post on page 24. Actually, Dim shield at 30 is still about as good as dusk shield (actually slightly better). With more pillars and fewer creatures it can do more damage to a pure rush opponent. I used 20 pillars and 4 creatures for 3 turn 30 quanta dim shield, and it was slightly better than 15 pillar, 12 creature, 3 dusk shield. 

With only 4 creatures in a 30 card deck, what fraction of the time does it deckout?
Did you attribute half of the deckouts as losses?

The comparison is solely based on damage dealt to an effectively infinite-HP opponent. The decks are 30 cards. So if the dim shield deck decks out before dealing 100 damage, that only reduces its average damage. Since its average damage is significantly above 100, this isn't the common case. You have a good point though; while the dim shield deck may have the capacity to deal a lot more damage than the dusk shield deck over 23 turns, it may not be as reliable. So the dim shield deck may sometimes deal 50 damage when it only draws 1 creature in the first 20 cards, but compensate by dealing 300 damage in other games where it gets its creatures early enough. I would say however, that dim shield has the property of lengthening games, which makes it inherently more consistent than dusk shield. And there is always the possibility of drawing 2 or more dusk shields before being killed on turn 11.

It might be interesting to see what the median damage of the dim shield 30 deck is.

All of this, of course, is ignoring that we will never have a dim shield that costs 30  , simply because it is unaesthetic. I believe the simulations illustrate that correcting the cost of the dim shield is not the right approach to nerfing the card.

Edit:

I went ahead and calculated the win percentage (how many times out of 10000 the decks dealt 100 damage to the opponent). The dusk shield deck is significantly more reliable than the dim shield 30 deck.

Here is a table of win-percentage results (each based on 10000 games):

Deck	Aprx. Win %
3 Dusk shield (12 creatures)	61
3 Dusk shield (11 creatures)	62
3 Dusk shield (10 creatures)	62
3 Dusk shield (9 creatures)	57
6 Dim shield 6  (9 creatures)	92
3 Dim shield 6  (12 creatures)	82
6 Dim shield 10  (7 creatures)	81
6 Dim shield 20  (5 creatures)	71
6 Dim shield 20  (4 creatures)	78
6 Dim shield 20  (3 creatures)	78
6 Dim shield 25  (4 creatures)	64
6 Dim shield 25  (3 creatures)	68
6 Dim shield 30  (4 creatures)	43
6 Dim shield 30  (3 creatures)	51
5 Dim shield 30  (3 creatures)	43
6 Dim shield 30  (2 creatures)	44

I also ran it so that I could see a distribution of the damage dealt. The dim shield 30 deck has a peak at 45 damage dealt. This happens in roughly 10% of the games. It has smaller peaks at 100 damage dealt, and 225 damage dealt. I'm not sure what the significance of the peaks are.

[eaglgenes101]

^Calculating total damage rather than damage speed isn't correct; It's like saying a 500 HP OTK will help kill better than a 100 HP OTK.

[jawdirk]

^Calculating total damage rather than damage speed isn't correct; It's like saying a 500 HP OTK will help kill better than a 100 HP OTK.

Total damage is a useful metric because it represents efficiency. If your deck is capable of doing 200 damage, then you can lose 100 damage to control and still win. You can also reduce the damage your deck does and substitute your own control or additional stall. If you are trying to ascertain the relative efficiency of two cards, it makes sense to count damage after 100 to avoid penalizing a card just because its deck "won." In a real game, there are all sorts of factors that could change the actual turn-to-win, and the extra damage would become meaningful (healing, different creature strength, control, stall).

[mesaprotector]

^Calculating total damage rather than damage speed isn't correct; It's like saying a 500 HP OTK will help kill better than a 100 HP OTK.

THIS.

To clarify, jawdirk: I wasn't saying your program had a bug, I was saying it was just calculating completely the wrong thing, which it is. Total damage is quite useless as a statistic outside of OTKs.

If you really want to figure out stalling power, I suggest using these two decks against AI3 (rebalance quanta for different costs, this is for 6 dusk and 30 dim):

Hover over cards for details, click for permalink

Deck import code : [Select]

5f0 5f0 5f0 5f0 5f0 5f0 5f0 5f0 5f0 5f0 5f4 5f4 5f4 5f4 5f4 5f4 5f7 5f7 5fb 5fb 5fb 5fb 5fb 5gi 5gi 5gi 5gi 5uo 5uo 5uo 8pt

Hover over cards for details, click for permalink

Deck import code : [Select]

5f0 5f0 5f0 5f4 5f4 5f4 5f4 5f4 5f4 5f7 5f7 5fb 5fb 5fb 5fb 5fb 5gi 5gi 5gi 5gi 5gi 5gi 5gi 5gi 5gi 5gi 5gi 61t 61t 61t 8pu

[jawdirk]

If you really want to figure out stalling power, I suggest using these two decks against AI3 (rebalance quanta for different costs, this is for 6 dusk and 30 dim):

Those decks are needlessly complicated to simulate the relative power of dim shield and dusk shield. Also, a duo deck is obviously not the right choice for a card that costs 30 .

On the flip side, I am satisfied that my simulations have established the correct cost of dim shield at around 27 relative to dusk shield by the metric of win percentage (in other words, Dim Shield is OP, and it shouldn't be fixed merely by raising the cost). If you want to run your own tests and present data, be my guest. If you want to challenge the validity of my results with facts or a priori arguments, be my guest. I don't see any support for your suggestion other than your "Master of Light" title, so do it yourself.

[mesaprotector]

If you really want to figure out stalling power, I suggest using these two decks against AI3 (rebalance quanta for different costs, this is for 6 dusk and 30 dim):

Those decks are needlessly complicated to simulate the relative power of dim shield and dusk shield. Also, a duo deck is obviously not the right choice for a card that costs 30 .

On the flip side, I am satisfied that my simulations have established the correct cost of dim shield at around 27 relative to dusk shield (in other words, Dim Shield is OP, and it shouldn't be fixed merely by raising the cost). If you want to run your own tests and present data, be my guest. If you want to challenge the validity of my results with facts or a priori arguments, be my guest. I don't see any support for your suggestion other than your "Master of Light" title, so do it yourself.

I give up. You clearly don't want to consider any suggestion that your precious program might not be the last word on card balance, so go ahead and don't.

[jawdirk]

I give up. You clearly don't want to consider any suggestion that your precious program might not be the last word on card balance, so go ahead and don't.

Well that was quick. Not even a shred of evidence before you give up? Well so be it.

[Quantumage]

I give up. You clearly don't want to consider any suggestion that your precious program might not be the last word on card balance, so go ahead and don't.

Well that was quick. Not even a shred of evidence before you give up? Well so be it.

It's clear that even raising the cost of dimensional shield to even 20 would make it much less viable, dare I say nearly unplayable. Saying that the true value of dimensional shields is 27 is clearly wrong from a commonsense standpoint (i.e. your program is inherently flawed). Raising costs is always a nerf; the card can be balanced by raising the cost.

[eaglgenes101]

I give up. You clearly don't want to consider any suggestion that your precious program might not be the last word on card balance, so go ahead and don't.

Well that was quick. Not even a shred of evidence before you give up? Well so be it.

I assume you can't see sarcasm from there?
Dim shield's cost isn't 6 once, it's 6 every 3 turns. Compare to dusk, which is 6 once.

[storyteller]

dim shield is weaksauce, just ask the honey charger














he doesnt give a shit yo.