Collectible Card Game Headquarters

A more serious concern is that the AI can use the fixed result to always make the right guess.

The reason is that the AI will simulate the result of choosing heads as well as the result of choosing tails. Since it is deterministic, the AI will know exactly what happens in each case and be able to choose the one it wants.

This is the AI analog of the undo cheat that the player could use.

melvin wrote:A more serious concern is that the AI can use the fixed result to always make the right guess.

The reason is that the AI will simulate the result of choosing heads as well as the result of choosing tails. Since it is deterministic, the AI will know exactly what happens in each case and be able to choose the one it wants.

This is the AI analog of the undo cheat that the player could use.

Hehe, I see now - It's not the player cheating, it's the AI...

That complicates matters.

- Activated abilities with an X cost that ask for a target with converted mana cost X (Gorilla Shaman, Linessa, Zephyr Mage). Magarena requires a viable target in order to allow the ability to be activated, but in this case having a viable target requires activating the ability.

Is this correct? I saw Detonate implemented recently which I previously thought wasn't possible, which made me question this. I think I tried to do some of these and they parsed but I couldn't activate the abilities at all, so I thought they weren't viable due to the reason above. Looks like I had done something else wrong instead. Please clarify so I can correct the information, if possible.

PalladiaMors wrote:Is this correct? I saw Detonate implemented recently which I previously thought wasn't possible, which made me question it.

It is. If you tried it, you couldn't cast Detonate. Thanks for spotting the problematic card.

I didn't playtest it at all, was assuming I was wrong about the list's statement.

This is unfortunate, I had previously counted about 40 cards that would be implementable if it weren't for this problem. Looks like sort of a fundamental issue in how picking targets works in Magarena. Any chance this could be solved in the future?

The likelihood is low.

PalladiaMors wrote:

- Activated abilities with an X cost that ask for a target with converted mana cost X (Gorilla Shaman, Linessa, Zephyr Mage). Magarena requires a viable target in order to allow the ability to be activated, but in this case having a viable target requires activating the ability.

Is this correct? I saw Detonate implemented recently which I previously thought wasn't possible, which made me question this.

This was my bad - I made the mistake of testing Detonate in a deck with artifact land. This allowed the card to be able to be cast as there was an artifact in play with a casting cost of 0. X could then be spent when casting to target something else. (At which point the targeting relative to mana spent works fine). It's just that first hurdle really.

ShawnieBoy wrote:X could then be spent when casting to target something else. (At which point the targeting relative to mana spent works fine). It's just that first hurdle really.

The key problem is not allowing certain X to be selected. Eg, there isn't a 2 CMC artifact on the field, so you can't select 2 for X. Right now there is no way to limit what values you can use for X, the only restriction is that X is at least 1 and at most all the free mana you have.

Another case is when there is only one 5 CMC artifact on the field and you don't have sufficient mana to pay 5 for X, then the option to cast the spell should not be available at all.

Oh yes I understand that - you get yourself in a position where you can only undo. I meant more as a first hurdle of many

Hey guys, bumping the topic because of some great news: cast increasers/reducers (Sphere of Resistance, Thalia, Guardian of Thraben) have been implemented. This should bring over 100 new cards into Magarena, many of which are quite popular, and was one of the biggest groups of unsupported cards that remained. Pretty certain it should have an impact over Firemind as well. Great job!

Edit: got a bit excited here, only cost increasers for the time being, so that's not over 100 cards yet! No idea how many cost increasers there are, but pretty sure it should be at least a couple dozens? Anyway, please note that my comment above was wrong!

So far only increasers are fully functional - there's still issues with mana reduction with spells and additional costs like Kicker or Buyback. But on the case.

^ Oops, I didn't see the issue, just saw some cards getting implemented among the daily updates and jumped the gun here. Correcting the update!

The interaction with cost got sorted out, so it will work with cost reduction. Additional costs (kicker, buyback, etc) will not interact with cost reduction though, at least not in this first implementation.

@PalladiaMors in your opinion, would it be worth releasing this mechanic and associated cards even though additional costs cannot be reduced?

I've once stated that personally I'd support adding every card that's functional the majority of the time, and then eventually sort out the bugs later. I've seen other MtG applications follow this design philosophy. At the opportunity we discussed this over a year ago, I remember Lodici commented that in that case you'd have to explain people exactly what are the differences between the Magarena implementation and what should be usually expected from regular MtG rules, because you shouldn't mislead people, which is something that I entirely agree with.

So I have a radical pro-including more cards stance, which I suppose is very biased. My level of involvement is noooowhere near the amount of work you guys put into this and I've not even been so active lately, so I seriously don't think my opinion should matter.

Edit: for people who follow this discussion: the whole "cost reducers & cost increasers" card group got implemented. Thanks to the devs for the hard work, that's a LOT of new playable cards. Quite an achievement!

PalladiaMors wrote:I've once stated that personally I'd support adding every card that's functional the majority of the time, and then eventually sort out the bugs later. I've seen other MtG applications follow this design philosophy.

Thanks for your sharing your thoughts on the matter. I think it is very reasonable position. I think in this case the gain (allowing casting cost modification) is greater than the drawback (additional costs cannot be reduced) as this mechanic allows for some interesting gameplay.

Only casting costs that depends on the card's properties can be modified. Except for the following five, the rest have been implemented.

Exiled Doomsayer
Gloom
Icefall Regent
Kaervek's Torch
Suppression Field