Quest Decks Discussions

[GandoTheBard]

Really what i play forge for is just the quest mode. Things i like about it is just playing with random decks and trying to beat the other decks. Things i dont like is the balance of colors. I dont think red is even playable considering there creatures are just unplayable and there is hardly any burn. I think Green has been done really good and i like all there cards. Black im ok with atlot of generic stuff. White is done good also. And i think Blue could be done better maybe more counters or bounce or just tricks. I think that maybe we should up the pools maybe 2 or 3 more boosters. I dont like AI stack land (seems silly and i know i can turn it off). I dont like the mulligan just being for you and i can see why it would be hard to do for the computer but you could just use the old mulligan rule and say if you have all lands no lands or one lands you can do it and that should be easy for the computer right? I would not mind if you just changed quest mode so every 2 games you get a pack. Also with no sideboarding i would make it so it was only one game. I hope that does not seem like a lot because i do love playing it and do it ever day.

I concur about the colors. Red and Blue are usually the most shafted with Blue having too many poor cards and red just not enough cards period. Though using my shortened rarities files might help a little in regards to quality. I have excluded the worst of the worst and tried to keep only the cards I find fun or potentially fun.

I also agree about the Mulligan (see my comments above).

The one game thing is ...well I like having 3 game matches when you know you will be screwed at least 1 of those games. it never fails so having 3 games means I get to win a match instead of losing it. Probably wouldn't matter so much if you won every game.

[Chris H.]

Though using my shortened rarities files might help a little in regards to quality. I have excluded the worst of the worst and tried to keep only the cards I find fun or potentially fun.

Thank you for your help in this area. Having a set of play balanced rarity files can help to make a difference in the starting pool.

[Chris H.]

This is just a few thoughts that I have had over the last week or so. I would be interested in hearing your opinions on this.


Prior to the quest mode we had an ongoing deck archive project with the intent to create some strong AI decks for the computer to play against us humans. Rares thoughtfully imported these AI decks into the quest mode.

In the past we had gotten used to constructing strong decks for ourselves to play. We would often times plays these decks with the computer set to use a randomly generated two color deck. We grew accustomed to beating the computer by a wide margin. These stronger AI decks for the computer helped to give us a more balanced contest.

The quest mode has us now playing weaker constructed decks from a smaller pool of cards. We are playing against the stronger AI decks rather than the weaker randomly generated decks. We now are not winning as often as we were just a few months ago.

We may still decide to tone down a few of the included AI decks. I'm not sure that we need to tone down all/most of these decks. We may need some time to adjust to playing against a stronger opponent.

Some of the hardest AI decks can be a real challenge for us to play against. If more decks were added, this in turn would mean that these hardest AI decks would appear less frequently. I don't mind occasionally facing a Black Vise or a Nicol Bolas, Planeswalker deck. Not every third game, no. Once in a while is OK.

[GandoTheBard]

Turn 1 Black Vise, Turn 2 Boomerang, Turn 3 Regress, turn 4 Regress turn 5 Evacuate is game over. I dont care how you do the math. The fact is this deck is NOT appropriate for severely limited play. And I disagree with your assumption about play vs weak AI decks. I tried to avoid doing that because the Generate Decks would give absolutely horrible results for the AI. Who needs to play something you know you will auto win? Even for play test purposes. (Especially for testing really) You want the opponent to attempt to foil your plans, otherwise how can you see the flaws in action? So I for one am not used to a weakened AI. On the other hand using a limited pool of cards means that even with excellent play choices and the best deck building possible if you run across a deck with black vice, and bounce (even bad bounce) you are going to lose. Badly. Unless you have the amazing good fortune to start with a Naturalize or Disenchant or Shatter and get lucky enough to draw it in your opening hand.

Nicol Bolas at least has the limitation of being ridiculously expensive and not too hard to defeat if you have a decent starting pool. (You might have to alpha strike it to get it off the table, which could put you at a disadvantage, but since the AI can't use the 2nd ability all is not lost.)

[Vecc]

Very interesting points being raised here. You guys really get to it, eh?

One thing I'd like to add is... even if this is temporarily an issue, I don't think toning down the stronger AI decks needs to be done just yet. After all, the plans for the future of Quest Mode include separating decks by 'difficult' category, right? I'd be interesting to leave those decks untouched for now, getting severely bashed every now and then (oh well, it's happened in real life!), keeping things fun by messing with the rarity files... meanwhile, with some valuable experience we've all got about how the AI plays stronger decks, we could try and suggest more low and mid-level decks that are still challenging and fun to beat - that'd double as more material to work with in the future and as a semi-solution to the insane AI decks as of now; like Chris said, with more decks implemented, chances of getting Bolas'd goes down.

On a side note, I think it's awesome that there are some dirty-cheap decks like Black Vise for us to toy with. How great it is to pull off a win against that, or to Edit your deck into a sideboarded version of itself just for that matchup...

[Chris H.]

Very interesting points being raised here.

... meanwhile, with some valuable experience we've all got about how the AI plays stronger decks, we could try and suggest more low and mid-level decks that are still challenging and fun to beat - that'd double as more material to work with in the future

Vecc, you may want to visit this link. It has some info in reference to accessing the hidden deck editor menu that is included in quest mode.

http://www.slightlymagic.net/forum/viewtopic.php?f=26&t=1062

And yes, we can always use some more decks for the quest mode. Rares has mentioned that he doesn't mind if some of the decks have strange/uncommon themes.

[mtgrares]

GandoTheBard wrote:

Fun things...
-- getting new cards in pool is fun
-- random rare is fun too
etc..

I agree 100% with everything that you listed.

Chris H. wrote:And yes, we can always use some more decks for the quest mode. Rares has mentioned that he doesn't mind if some of the decks have strange/uncommon themes.

The wierder the better This also includes "unplayable" cards like 3 R B U for a 3/4 creature or a spell that does something interesting or unusual. Quest decks can be slower than regular, competitive decks so it is not uncommon for the AI to have access to 7 lands. (You may want to turn off the "Curve AI Land" checkbox because it limits the AI to 6 or 7 land, I forget which.)

[mtgrares]

And to revisit an earlier topic, decks should be divided up into: easy, medium, hard or very easy, easy, medium, hard, very hard. (I don't mind dividing decks up into the 5 categories if you can decide which deck goes which, because I'm not really good at stuff like that. I only know that fast aggro decks are hard or very hard.)

[Chris H.]

And to revisit an earlier topic, decks should be divided up into: easy, medium, hard or very easy, easy, medium, hard, very hard. (I don't mind dividing decks up into the 5 categories if you can decide which deck goes which, because I'm not really good at stuff like that. I only know that fast aggro decks are hard or very hard.)

I would like to see some new and weaker decks added first. From there we can start to discuss how we want the coding to handle the separation between the deck pools/groups.

I am having a fun time playing quest mode. I had never tried sealed deck or drafting modes before. This type is new for me. I think that this may become my favorite over time.

[mtgrares]

The cards that you win and start with do need to be balanced and I can do this in the code if I know specifical what needs to be balanced. For example you win 10 cards, should there be 2 colors of each card and a minimium of 4 spells? (MTG Forge has more creatures than spells, so sometimes you are lacking any good spells.) I like some of the "randomness" but it seems too random sometimes, like you may get more green than red so you go with red.

One possible solution is to have MTG Forge balance out all of the colors and spells versus creatures internally after it has read from common.txt, uncommon.txt, and rare.txt. By balancing I mean either removing some cards or making duplicates of some cards so that colors have an even chance. This way when you get 10 cards it will have a good chance of being balanced even though the cards are randomly chosen and sometimes you might have much more of one color than another, but not often.

Another solution is to manually balance the files common.txt, uncommon.txt, and rare.txt manually using either my set editor (which doesn't give any statistics or rob's).

Dividing up decks in easy, medium, hard, or very easy, medium, hard, very hard would improve the quest experience by giving the user the appropriate challenge.

[GandoTheBard]

The solution I like which posted in the initial quest thread is to use sets. Determine which cards belong in which sets (by rating?) and then balance the sets so that there are an equal number of each number and a good amount of each type. The higher sets would not be available to start with. You would have to earn them. I haven't had the time to deal with rating the rest of the cards yet so this feesible yet.

[mtgrares]

If I understand you correctly, just having one "card set" is hard enough and I'll probably only stick with one. If you are winning cards for the 1st time or the 100th, you always get cards from the same set which is common.txt, uncommon.txt, and rare.txt (yes I mention those files alot but just for clarity).

I'm not sure should cards have an "enforced" distribution (all the colors would be even) or not? I could go either way, either would be an improvement.

[Rob Cashwalker]

If I understand you correctly, just having one "card set" is hard enough and I'll probably only stick with one. If you are winning cards for the 1st time or the 100th, you always get cards from the same set which is common.txt, uncommon.txt, and rare.txt (yes I mention those files alot but just for clarity).

I'm not sure should cards have an "enforced" distribution (all the colors would be even) or not? I could go either way, either would be an improvement.

No, I'll agree, that the pool should be drawn from at least a single "set" which is designed with a specific distribution of cards by color and type. So pretend you're MaRo for a moment, and imagine that you're designing a set with 350 cards. 50 cards in each color, 50 multicolor, and 50 cards split between artifacts and lands. Each color should be represented by 25 creatures and 15 sorceries and 10 instants. Artifacts/lands represented by a 30/20 split... so choose either the original dual lands, the ravnica duals, or the pain lands, plus the man lands and other assorted lands.

[GandoTheBard]

Im not sure Id want to be MaRo for a living lol. Though I guess the money is good. Anyway I don't agree with the numbers per se but yeah the base set should be large enough to contain a balance between all the colors and also have room for nonbasic lands and artifacts.

I really think we need to try and make multiple sets that we can either configure into the game via an ini file or something. For one thing, drafting is different from the sealed limited you get with the League Style Quest mode. It should have a different pool of cards. But even with that issue aside, is it really a good idea to allow Wrath of God, Rafiq the Many and Armageddon (just to name a few) as average rares in the base set? Are we going to not allow them at all? I thought the idea of a quest mode was to progress from a small underpowered pool to a moderate power larger pool with rewards in between. As it is now you can take even a bad pool and just lose your way into a good pool. Seems rather pointless once the novelty wears off. Not saying its not a great accomplishment. It definitely is...But if you stop there it is just a novelty. If you makes plans to improve it and make it more interesting then it might be something genre defining (such as Apprentice32 was threatening to become in the 90s.) Well thats my thought on the subject anyway.

[Chris H.]

I would like to thank Rares for the quest mode and we realize how hard it can be for just a single person to take on the responsibilities for the entire project.

I would like to at this time suggest that I (meaning me! ) add the following decks to the existing decks in the questData04-08-2009.zip archive:

Rogue = Burning Dragons
Catwoman = Cats
King Edward I = Little Ellie
King Edward II = Ellie
Storm = Lucy Rainbow
Terminator = Machines Mind
Cyclops = Sally
Tarzan = Tarzan
Magneto = Vengence Ltd
Scotty = Yo Yo

I would like to replace Rares version of the battle of wits deck with Gando's version.


The new deck listing by name would be:

Code: Select all

Abraham Lincoln   =   AI-Angels-Hammer
Albert Einstein   =   AI-Garruk-PanWolf
Batman   =   AI-Exalted-Unblckble
Blackbeard   =   AI-ww-Soldiers
Boba Fett   =   AI-DragonRoost-Fire
C3PO   =   AI-GobBog
Catwoman   =   Cats
Comic Book Guy   =   AI-Mt-Dragon
Cyclops   =   Sally
Darth Vader   =   Rares version -> battle of wits
Data   =   Rares -> grandeur.deck
Frodo   =   AI-White-Weenie-v2
Genghis Khan   =   AI-Animas-Armageddon
Green Lantern   =   AI-Nicol-Bolis
Han Solo   =   AI-Chanters
Iceman   =   AI-Bounce-Control
Jabba the Hut   =   AI-Exalted-Walkers
James T Kirk   =   Rares -> black discard
King Edward I   =   Little Ellie
King Edward II   =   Ellie
Luke Skywalker   =   AI-Green-Weenie
Magneto   =   Vengence Ltd
Morpheus   =   AI-Elf-Pump-Mod
Napoleon   =   AI-Enchanted-Wall
Neo   =   AI-Ball-Breaker
Newton   =   AI-Aluran-Rats-CH
Picard   =   AI-Elf-Pump-UWG
Pinky and the Brain   =   AI-Assassins-Dagger
Professor X   =   AI-Artifact-Armada
R2-D2   =   AI-Black-Vise
Rogue   =   Burning Dragons
Scotty   =   Yo Yo
Silver Surfer   =   AI-GreenBeats
Spiderman   =   AI-White-Weenie
Spock   =   Rares -> elf singleton
Storm   =   Lucy Rainbow
Superman   =   AI-Repulse-Control
Tarzan   =   Tarzan
Terminator   =   Machines Mind
Uncle Owen   =   AI-Kill-and-Control
Wolverine   =   AI-Kodamas-Nightmare