MTG Forge 03/10 - with Quest Mode

[Huggybaby]

Clicking "Random Page" on the MTGForge wiki links into other wiki's....

NP, Those are all pages on our wiki.

[GandoTheBard]

You have some interesting ideas, and I have given a little thought to the difficulty levels. I don't have your experience in playing sealed deck and I am having a hard time winning on the easiest level.

With my limited play testing at this point it appears that all of our opponents are constructed deck. We want the quest to be a challenging experience for the better players, but I think we may need to consider the needs of those whose skill set does not raise to a high level.

We could consider at EASY level, alternating constructed decks with random decks on a 50 - 50 % basis. And with MEDIUM level this could be changed to 2/3 constructed and 1/3 random. The HARD and EXTREME DIFFICULT levels would be constructed deck only.

When I first started my quest I was pleasantly surprised to see that I had both the Mox Jet and the Ancestral Recall. A few minutes later I realized that black and blue were my two weakest colors. I am getting close to finishing the easiest quest and now the five colors are roughly equal.

I am noticing that I have far more creatures than I do have spells. This in turn causes me to purposely create decks that are designed to swarm the computer with an army of creatures.

The Anti could also be be handled differently via the difficulty levels. Low level the anti comes from your non-deck cards and you are playing common vs common ... rare vs rare. At high level the anti comes from your deck and you could face losing a rare while the computer is only offering a common.

The number of cards that we start with could also be determined by the difficulty level.

If any of this look interesting, then you can consider adding to the list that you have started. Good idea, BTW.

Hehe Im not even a good player, but I see your point. The constructed decks can be hard since your pool is so limited. On the other hand that IS kind of the point. Congrats on nearing completion btw.

The decks will need to be seriously tuned for different difficulties. 75 card sealed decks probably wouldnt be a good idea since the computer doesn't build good decks from a limited pool but maybe something in between is called for. maybe make it a megapool of cards (150 or so) with special construction rules so that the decks actually have some cohesiveness. The problem of course would be someone taking the time to program this and making it balance out.

We also need a set of criteria of what comprises an easy deck vs what isn't an easy deck. My ruie of thumb is: any deck that plays lots of weenies without any special abilities or that are overcosted is a bad deck. Decks that have cards the AI has no idea how to use well are also bad. Decks that play cards that are hard to play (super expensive or seriously out of color) are likely bad too. The computer had difficulties when it has to deal with more than 2 colors.

The hardest deck imho (Id call it Project Godzilla btw not its current silly name) in the gamut at the moment is hard mainly due to random luck. The one with pipers + big creatures. Because it can throw down a piper turn 3 and search out a big fattie. If you dont have a way to kill the pipers quickly you are toast. Particularly against Akroma. That deck doesn't really belong in this set because it is too hard because of the unpredictability of the limited environment. We do have plenty of types of pingers but they are not common for the most part so you might get one. If you are smart you will play with it because pingers take care of all kinds of inconvenient 1 toughness critters and also can break a stalemate or provide tactical advantage. The random luck I refer to vs the Pipers is that you have no idea what your card pool will look like. Eventually you will get a direct damage spell even if you lose but it might take a while. Also you may need more than one. I had a game where I ended up scraping through because I managed to get a pinger in time to kill the piper before it could get Akroma on the table. The other times I've played that deck I have either crushed it too quickly for it to get the combo or lost to a 6/6 flying hasted vigilant first striker. :/ (lets not forget that the card is actually not correctly implemented nor is it's red counterpart.)

I highly recommend starting over if you end up with a dull cardpool with no pingers or direct damage. Though such a pool could present you with a serious challenge if you are looking for one...in which case restart until you get the worst pool you can imagine. If you luck into creature removal you sort of have to play it even if it is off color and badly costed. Anyway we do need more decks as Rares said to give the quest mode a sense of variety.

[Chris H.]

I think that we might have two different quest projects ongoing at this time.

1) Play balancing the existing quest decks and adding additional quest decks which could be done by the non-coders.

2) The list of difficulty-level adjustment ideas that you started which would require additional coding work.


I'm currently trying to finish the "more decks" folder with about 100 individual deck files.


Once I am done I would like to take a closer look at what you have done with # 1 above.

[mtgrares]

With all the ideas about quest it is hard to decide specifically what to change. I'll try to summarize.

1. Difficulty changes the number of cards that you begin with: Easy starts out and wins more cards, Medium has less cards than Easy, etc...

2. Difficulty changes what decks you play against, Hard decks will be harder to beat than Easy decks

3. For Medium through Very Hard - different decks appear after X wins, such as 20

[mtgrares]

I mean c'mon everyone seems to have their niche here... invaluable dedication... and now, I just have an inkling that this project just turned the corner onto a big road...

It would be the first software team I've been on. Maybe I can do Quest updates.

[GandoTheBard]

I think that we might have two different quest projects ongoing at this time.

1) Play balancing the existing quest decks and adding additional quest decks which could be done by the non-coders.

2) The list of difficulty-level adjustment ideas that you started which would require additional coding work.


I'm currently trying to finish the "more decks" folder with about 100 individual deck files.


Once I am done I would like to take a closer look at what you have done with # 1 above.

Strange, I dont recall you commenting on the Quest Mode Decks thread I have in the Decks section...perhaps I missed it?

==edit== pardon my sarcasm...Im just a little annoyed that I posted that thread a few days ago and no one even noticed.

[Chris H.]

I'm currently trying to finish the "more decks" folder with about 100 individual deck files.


Once I am done I would like to take a closer look at what you have done with # 1 above.

Strange, I dont recall you commenting on the Quest Mode Decks thread I have in the Decks section...perhaps I missed it?

==edit== pardon my sarcasm...Im just a little annoyed that I posted that thread a few days ago and no one even noticed.

I saw your message thread and I have been testing all of my AI decks by adding dual lands that do not come into play tapped. I happy with the results so far. Thank you Dennis.

Have you had a chance to test your white decks to see if the Flagstones of Trokair now work correctly? I have not yet had a chance to look at this card yet.

I am sorry that I have not yet responded to your thread in reference to the quest decks. I plan to get more involved once I finish the more decks project. I am getting a wee bit stretched out. I just need a few more days to finish this other project.

[Almost_Clever]

Okay, I just finished the quest on Very Hard with a record of 40 wins / 6 loses (sic) and wanted to give my impressions.
Starting with problems:

It is irritating having to resize the column widths in the deck editor every time and having the "Resizable Game Area" and "Stack AI land" defaulted to off.

As Rares stated, the AI lands aren't really stacked so the computer is occasionally mana screwed

Rares also noted that you can only earn common cards.

It seemed like I got too many duplicates in my new cards. While I haven't looked through the possible cards, I seemed to get too many duplicates without seeing a significant portion of the available card pool. I got three copies of three cards and two copies twenty cards in the 11 "packs" I earned. Of course, I haven't really worked out the statistics, and I might be looking for problems where there aren't any based on my impression from hundreds of drafts with poorly randomized packs.

The computer's decks seemed really over-powered and too hard at first (I earned my first pack with four losses in my first five matches).

The computer's decks seemed under-powered towards the end when I had a larger card pool to put together a better focused deck. I imagine this would be even easier if I had gotten some uncommon and rare cards in the packs I had earned.

I saw that I could choose my opponent, but I just took whatever choice I was presented with. I didn't really see the point to the option.

Okay, enough about the problems. On the positive side:

I had a lot of fun and will start another quest soon

Very fun and interesting mix of computer decks. The Kiki-Jiki, Mirror Breaker and the Elvish Piper decks were particulary well designed to frustrate the players and the Battle of Wits deck was totally surprising.

Suggestions (besides fixing the bugs):

Progressively more difficult computer decks as you work your way up the ranks

Require more match losses than wins before earning a pack -- let's not reward bad play, but at least make it possible to eventually work out of a really bad starting pool

I like Gando's suggestion of ante, although it should come from the unused card pool and the computer's ante should match the rarity

I wouldn't give the option of chosing the computer's deck. Or how about having a guantlet of decks for each tier (and difficulty) and the player must defeat all of them to earn the next pack (and advance to the next tier).

[DennisBergkamp]

Almost Clever,

I agree with all of these suggestions, even though I haven't played too many quest games yet... I was especially disappointed to see I would only earn commons after a few games. The main appeal for me in this type of quest game is to collect cards and build a stronger and stronger deck around the cards you obtain.

I don't know, but I'd rather have a much longer campaign as well... couple of hundred matches with progressively more difficult opponents would be much more fun.

Fortunately, some of the "frustrating" things you listed, I have already fixed The deck editor's column sizes, and also each time you win cards, now you'll get 6 commons, 3 uncommons and 1 rare (for the latter fix I just used rares' posted code).

By the way, do you guys agree with these numbers? Or would you guys rather have 15 cards after each win?
Bonus conditions would be cool too, winning 10 games in a row will obtain you a random rare, or something like that. I also like the Ante idea.

[GandoTheBard]

Okay, I just finished the quest on Very Hard with a record of 40 wins / 6 loses (sic) and wanted to give my impressions.
Starting with problems:

It is irritating having to resize the column widths in the deck editor every time and having the "Resizable Game Area" and "Stack AI land" defaulted to off.

As Rares stated, the AI lands aren't really stacked so the computer is occasionally mana screwed

Rares also noted that you can only earn common cards.

It seemed like I got too many duplicates in my new cards. While I haven't looked through the possible cards, I seemed to get too many duplicates without seeing a significant portion of the available card pool. I got three copies of three cards and two copies twenty cards in the 11 "packs" I earned. Of course, I haven't really worked out the statistics, and I might be looking for problems where there aren't any based on my impression from hundreds of drafts with poorly randomized packs.

The computer's decks seemed really over-powered and too hard at first (I earned my first pack with four losses in my first five matches).

The computer's decks seemed under-powered towards the end when I had a larger card pool to put together a better focused deck. I imagine this would be even easier if I had gotten some uncommon and rare cards in the packs I had earned.

I saw that I could choose my opponent, but I just took whatever choice I was presented with. I didn't really see the point to the option.

Okay, enough about the problems. On the positive side:

I had a lot of fun and will start another quest soon

Very fun and interesting mix of computer decks. The Kiki-Jiki, Mirror Breaker and the Elvish Piper decks were particulary well designed to frustrate the players and the Battle of Wits deck was totally surprising.

Suggestions (besides fixing the bugs):

Progressively more difficult computer decks as you work your way up the ranks

Require more match losses than wins before earning a pack -- let's not reward bad play, but at least make it possible to eventually work out of a really bad starting pool

I like Gando's suggestion of ante, although it should come from the unused card pool and the computer's ante should match the rarity

I wouldn't give the option of chosing the computer's deck. Or how about having a guantlet of decks for each tier (and difficulty) and the player must defeat all of them to earn the next pack (and advance to the next tier).

I agree with some of your points. Nice record...I haven't tried the most difficult setting yet as I get bored quickly and waiting 4 games to get a pack is just too annoying. Anyway clearly there are some serious bugs that need addressing asap.

The Land stacking needs to be fixed so that the computer gives a good performance. Having a decent hand yet knowing you are going to win because the computer didnt show up is not fun. Now if you get screwed too (happens enough that it is higher than real life probability it seems to me) the odds even out considerably.

The way the initial card pool starts is not good. I started a new quest and got 11 rares (not counting the 4 rares noted as uncommons by the game), thats way too many just as starting a game and getting 2 is too few. I also noticed that same cards show up in the slots for rare/uncommon/common...this is not good. The point of this is to have a random card pool with a decent number of good cards spread among 5 colors and the rest being filler. The one frustrating thing to note here is that currently you are very likely to get 1-2 bombs and be unable to splash them because of the complete lack of other cards to go with them (not counting cards you shouldn't play with.)

I think having 5 different sets (think teirs) each balanced with the 5 colors + arts + lands + multicolor cards so that there is some variety is not only doable but necessary.

Rares also noted that you can only earn common cards.

I guess I missed this but I did notice that the cards were all commons. This is definitely not cool. The whole point of quest mode is to make it interesting and exciting. Commons can be great but in Forge they tend to be lackluster. Uncommons are important too. Alot of good cards are uncommon simply because of their quality.

To me the difficulty curve vastly differs based on which choices you make about which decks to fight against. Once you realize that "Ghengis Khan" (or as I like to call it the Godzilla Project which actually makes more sense to me) is hard to beat if you have no removal or pingers then you don't play it unless you are up for the challenge. Now no sane player in the real world would take such a deck to a tourney and hope to win because people always sideboard (or maindeck) a little hate against decks like that. Plus real world decks tend to be much more synergistic and combolicious so bringing the deck with the fatties is just suicidal. Not so much in the Quest mode. So paying attention to this leaves the smart player with the option to avoid said deck until you have the cards to deal with it. Einstein is very beatable but you can just auto lose to a goddraw by the AI. There is a way to play against it succesfully but figuring that out is probably part of the challenge so I won't spoil it here. Anyway those two decks are definitely top of the heap, but the other decks are not so much. We definitely need more decks too. Check out my thread in the Decks forum.

Re: matchlosses to win cards. I am not thrilled with this idea myself. If you are losing you need to redesign your deck. Maybe your card pool is totally hosed and it takes some skill to recognize this too and change it. But since not everyone is at the same skill level it is NICE that you do get cards for losing. I see no reason to make it harder on the low skill players.

I am also not thrilled with the idea of risking cards in my Sideboard. Id rather risk a card I have in my deck because it gives me an incentive to win. It also means that if I lose I might be hosed which sucks but is the downside that accompanies any gamble. Yes ante is a type of gambling. Thankfully it can be done without paying a dime real money I really loved the way Shandalar worked where you might just get the crap cards anted up front or you might get bonus cards. Particularly against the harder dragon opponents. That added alot of fun to the game. (xmas in july right?)

I like the idea of a gauntlet or series of gauntlets to be more precise. With varying degrees of difficulty. And maybe you get a pack from a new set if you complete a gauntlet successfully. Then you can start a mode over with the same pool and do another gauntlet. This adds alot of replayability to the game imho.

[Chris H.]

Okay, I just finished the quest on Very Hard with a record of 40 wins / 6 loses (sic) and wanted to give my impressions.

Awesome!

It is irritating having to resize the column widths in the deck editor every time and having the "Resizable Game Area" and "Stack AI land" defaulted to off.
As Rares stated, the AI lands aren't really stacked so the computer is occasionally mana screwed

I think that Dennis mention that he would look into these two problems.

It seemed like I got too many duplicates in my new cards. While I haven't looked through the possible cards, I seemed to get too many duplicates without seeing a significant portion of the available card pool. I got three copies of three cards and two copies twenty cards in the 11 "packs" I earned. Of course, I haven't really worked out the statistics, and I might be looking for problems where there aren't any based on my impression from hundreds of drafts with poorly randomized packs.

I chose to overwrite the rarity files that Rares provided with my set of complete rarity files. I have not yet won enough packs to notice any trends.

Very fun and interesting mix of computer decks. The Kiki-Jiki, Mirror Breaker and the Elvish Piper decks were particulary well designed to frustrate the players and the Battle of Wits deck was totally surprising.

Gando has started a quest decks thread and he is looking for help in modifying and adding to the selection of quest decks.

I like your and Gandos suggestions. I'm sure that our crew will be able to make some improvements over time. I am pleasantly surprised to see the quest mode work as well as it does ... Rares developed this project in such a short time period. It was a monumental task.

Thank you Rares. Job well done.

[GandoTheBard]

I wasn't sure if Rares had changed the rarities files or messed with them in any way because he mentioned he wanted to do so. I figured playing with his rarities was part of the experience. Now that I've done so I think I can safely say those files need some fixing. No offense Rares.
We really need to agree with a consensus on which cards belong in the Quest mode and stick to that until we can work out sets. Sets would be ideal because theyd be much smaller than the current 1355 cards we can have and would allow for subdivisions based on quality of cards. A card might even be rare in one set and common in another.

[Chris H.]

Have you had a chance to test your white decks to see if the Flagstones of Trokair now work correctly? I have not yet had a chance to look at this card yet.

The computer can now play the Flagstones of Trokair land card. Awesome.

[Chris H.]

I have finally finished my first quest, and at easy mode at that. I am not going to tell anyone how bad my record was ... it was bad.

I see that once we finish the quest, we receive our title and we are allowed to continue with the quest if we wish. Nice touch.

I copied over the original rarities files in order to get a real test. I wanted to see what it was like to play with trash cards. I was surprised to see that my original pool of card included a Mox Jet, Ancestral Recall and a Darksteel Colossus.

I put together a three color deck for my first game and was able to get out enough cheap creatures that the computer failed to attack and finish me. I was able to get 11 mana in play, and out came my Darksteel Colossus. It was game over dude for the computer.

After several games, I switched to blue and white. The creatures were not big but they had fun abilities for me to use. I lost most of my remaining games, but it was still fun.

I think for my second quest game I will keep the original rarities files and I will also import Gandos three decks.

[DennisBergkamp]

I too finished my first quest (I didn't have that great of a record either ). Now, what do you guys think of the current balance in the game?

I have a few thoughts on this:

- quest should be longer (my deck didn't look that different from what I had at the beginning, save for a few cards).
- should present some random rewards at certain points in the game.
- a card store would indeed be very cool.
- should start out with more commons at the beginning.
- AI should start out with some weak-sauce decks, and get stronger decks as the game progresses.