Collectible Card Game Headquarters

Go team

Corwin72 wrote:Go team

Thank You for new opponents Corwin. Will test them this week.

Update your local copy to rev 8100 or later:

- Several of the new quest decks had invalid card names that would cause an error exception when the deck was loaded. Fixed the Atomic Robo 3; Colossus 3 and Predator 3 decks.

- Several of the new quest decks had cards that are not in forge and this would cause an error exception when the deck was loaded. Fixed the Lucifer 2 and Lucifer 3 decks.

Rev8103:

- The deck editors will now display a column named "AI". Cards with the SVar RemAIDeck will show a "No", cards with the SVar RemRandomDeck will show a "?" and the rest of the cards will show a "".

- I have not been able to figure out the code in the table sorter class. If you click on the "AI" in the title bar you will get an error.


If you find some time Corwin, you could try to load in the AI decks into the editor and check for cards that need to be replaced.

Rev 8104:

The deck editor will now sort by clicking on the "AI" column title and it will no longer generate an error.

I am having some problems importing the .dck files into the general constructed deck designer. What is the best way to do this?

Corwin72 wrote:I am having some problems importing the .dck files into the general constructed deck designer. What is the best way to do this?

`
There should be no need to use the export or import option. Copy your new AI decks to the /res/decks/ folder, launch forge and you should be able to use the "Deck Actions" -> "Oen Deck - Constructed" command to load the deck into the constructed mode deck editor.

I think the import and export commands are only needed for decks/cards that you want to add to your questData.dat file for your own use during a quest. The decks used by the AI are simple ".dck" files.

Oh, I had to edit several of your deck files to fix several problems. Use the decks on the SVN rather than your originals.

Ok. I think that I have it now.
What is the criteria for the No.
I see Trinket Mage. I think I remember that the AI will play it but it's choice is not optimal.
Should decks not include any "No" cards?

Cards that have a "No" in the AI column will not appear in a randomly constructed deck for the computer. These cards will however appear in a randomly constructed deck for the human.

If possible, the quest AI decks should not contain any of the cards that have a "No" in the AI column.

Cards that have a "?" in the AI column will not appear in a randomly constructed deck for either the human or the computer. Trinket Mage is likely a weak card in a randomly constructed deck as there is only a very small chance that the deck will also contain an artifact card with converted mana cost 1 or less.

Be careful when adding cards with a "?" in the AI column to quest AI decks. Many of these cards are very useful in a constructed deck environment. But, the AI's ability to make a good decision with these types of cards at times will leave a lot to be desired.

There are likely to be some cards that need or don't need a "No" or a "?" in the AI column. Forge is a work in progress. If anyone notices anything let me know and I will try to find some time to improve this area.

This should help people to create additional AI decks for quest mode and that is a very good thing.

I had a look at some of your decks Corwin and I'm afraid there is a lot to be done about them. The problem is, that the decks look good on paper, but the AI can't use some otherwise powerful cards very well. Examples:
- Aquaman 3 has the cards Mortarpod, Tumble Magnet, Mana Leak, Spell Pierce, Jace, the Mind Sculptor and Preordain. All of these cards will be used very ineffectively by the AI.
- Ash 3 has cycling cards, but the AI won't use cycling at all!

The AI of forge is not that great yet. So we need to fix some decks before we release the next beta. You can start with these two decks and I will post other problematic decks that I find.

While we're at it, other members can post decks to be added/replace decks.

EDIT:
- Bear: Caller of the Claw, Havenwood Battleground, Slippery Karst.
- Beast 2: The whole plan with Furnace Celebration is too complicated for the AI.
- Beast 3: Why is this a completely different theme than Beast 2? (see Aquaman 3)
- Cable 2: Chromatic Sphere, Chromatic Star.
- Cable 3: Why is this a completely different theme than Cable 2? (see Aquaman 3)
- Captain America: Feral Hydra, Bant Panorama, Bant Charm.

As for the decks being random and not following the previous version of the deck...
I just gave up at the end. I was just trying to get them out there to be checked.

I have not had any problem with Caller of the Claw. The system plays it as a 2/2 for 3. I think that it is on theme.
I have not played against Beast 2 yet, but that should probably go. I wanted to see what the AI would do with it. It sometimes surprises me.

Cable: Chromatic Sphere, Chromatic Star. Yeah, those should be replaced....

Cap: It uses the panorama just fine. It searches for the first land it finds but it searched on turn 2. It looked ok.
with the X spells it usually uses them as soon as they can be played. So it would be a 1/1 for G1 or when ever it draws the card. Bant charm should go.

Aquaman 3: It will play Tumble Magnet poorly, but I can attest to more than one occasion where Mana Leak has wrecked me. I wanted to see if\how it would use Mortarpod.

Ash 3: I could have sworn that I had seen the AI cycle before. Ok that one will need to be redone as well. I will see what I can come up with for a WB deck.

Rev: 8107

I made minor edits to some card files. This batch includes spell countering cards that use the term "unless its controller pays".

The computer can cast these types of spells and the human will be required to pay additional mana to put the card into play. The problem as I see it is, compy does not have any ability to determine if it is countering a strong or a weak card. And compy is probably not checking to see how much mana the human has available before the spell is cast.

Corwin, you may want to consider replacing Mana Leak with Counterspell. This way the human will get no chance to negate the spell being countered.

If we had an Aquaman 1 deck, we could make it a weaker deck by using Mana Leak type cards rather than using Counterspell. Just a thought.

My reasoning for using Mana Leak was that it was in standard.
My thought is that any card you want to cast you want it to hit. On the other hand I cheat and import decks. This causes all of my cards to be good counter targets. I will switch if you think this is that relevant.
I am finished with 14 decks currently and will continue to update.

Corwin72 wrote:As for the decks being random and not following the previous version of the deck...
I just gave up at the end. I was just trying to get them out there to be checked.

I think it would be better to give them different names, so people can remember what opponent has which deck.

Corwin72 wrote:I have not had any problem with Caller of the Claw. The system plays it as a 2/2 for 3. I think that it is on theme.

That's ok, since Caller of the Claw is only in Bear 1 I think.

Corwin72 wrote:Cap: It uses the panorama just fine. It searches for the first land it finds but it searched on turn 2. It looked ok.

You're right. My mistake.

Chris H. wrote:The computer can cast these types of spells and the human will be required to pay additional mana to put the card into play. The problem as I see it is, compy does not have any ability to determine if it is countering a strong or a weak card. And compy is probably not checking to see how much mana the human has available before the spell is cast.

This is exactly the problem with Mana Leak.

Corwin72 wrote:My reasoning for using Mana Leak was that it was in standard.

Why is this relevant?

Corwin72 wrote:My thought is that any card you want to cast you want it to hit. On the other hand I cheat and import decks. This causes all of my cards to be good counter targets. I will switch if you think this is that relevant.

At least the tier 3 decks should be as powerful as possible (while still being on theme).

Sloth wrote:

Corwin72 wrote:My reasoning for using Mana Leak was that it was in standard.

Why is this relevant?

I am trying to include more standard decks now the we have a standard card pool for quest mode. That and I would like to play against more of the current standard decks. Not testing but I want to see how my terrible deck will do against a competitive deck for the format that my cards are from.

I guess I would like Forge to have a gauntlet mode and I am trying my best to do that.