Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
Moderators: timmermac, Blacksmith, KrazyTheFox, Agetian, friarsol, CCGHQ Admins
7 posts
• Page 1 of 1
Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by HarePaiR » 11 Dec 2014, 16:46
TL;DR - As a pet project, I am working on editing Alpha/Beta/Unlimited to make it more fun to draft, and will post my musings for interested parties here. I've attached edition files and improved card rankings if you'd like to try it out!
Hello beautiful people! For some time now, I've been wondering what Alpha/Beta/Unlimited might look like if it were optimized for draft in the way that contemporary large sets are. I'm pretty new to Forge, but it seems like a perfect platform for this kind of project, so first let me say thanks to the developers for this awesome program! I intend to contribute, and will hopefully be able to use this project to teach myself the ropes.
Inspired by lujo's truly fantastic thread on Fallen Empires Draft (http://www.slightlymagic.net/forum/viewtopic.php?f=26&t=15834), I'll be posting here to keep my thoughts in order and to get feedback from anyone that's interested.
In order to draft the set, add the following line to your blocks.txt folder in /res/blockdata/blocks.txt
You might want to make backups of the originals first!
If you'd like to play with the temporary new cards I added (pseudo-implementing Power Leak and Chaos Orb), they're in LimitedEditionGammaCards.zip, and can be placed in any folder under /res/cardsfolder/, though I recommend something like /res/cardsfolder/custom_cards/.)
Current Set Stats: 283 Cards (95 Commons, 102 Uncommons, 71 Rares, 15 Mythic Rares)
Ideal-ish Stats: 283 (116 Commons, 91 Uncommons, 61 Rares, 15 Mythic Rares)
Comparison (M15): 249 Cards (101 Commons, 80 Uncommons, 53 Rares, 15 Mythic Rares)
Redesigned Cards
Card Rarity Changes
Messing with card rarity is the tool I feel most comfortable with when editing the set for drafting. It doesn't require me to remove or add cards, but allows for an adjustment of power levels and matches the way Wizards now designs for draft. A potentially controversial change is the addition of Mythic rarity, but again I'm striving for a contemporary take, and Mythic rares also make regular rares more common, which can be healthy for a draft environment. These changes are a work in progress.
Future Work
Once I'm somewhat satisfied with the card list and rankings, I can see some areas for improvement that I'd like to work on which should let me get into actually developing little bits of code for Forge proper.
Archetype Drafting
I haven't looked at the Forge draft algorithm, but from what I can tell it tries to pick a card based on the rankings list, giving bonus points for staying in color and subtracting points if it can't use the card. I think the ranking list is really fantastic because it means that a doesn't-know-anything-about-coding user can adjust the AI draft behavior just by editing a text file. It's surprising how well the AI drafts just following this approach! However, there are obviously problems, too.
Ideally, there would be a parser for a similar archetype-based rankings list, that bumps cards up in points if they fall into an archetype that the AI has already drafted some cards for. I think big improvements could be made even with pretty simple algorithmic changes, and different sets could have different user-editable archetype lists. Later we could implement more sophisticated ideas, like some of the things lujo has been suggesting (http://www.slightlymagic.net/forum/viewtopic.php?f=26&t=15834&start=6), all of which could in theory be done in a set-specific text file rather than on a card-by-card basis.
Deckbuilding Improvements
Whoever wrote the deckbuilding algorithm did a fantastic job, as even though a lot of cards are marked unplayable by the AI the decks end up being surprisingly competitive. However, there seem to be a couple areas that are too rigid. In particular, the AI seems to be restricted to building decks with exactly 17 land, which occasional leads to it 'splashing' Counterspell off of zero blue mana sources because it doesn't have enough playables.
Also, I should look into adding AI improvements that let it use some more of these cards!
Banding Combat Math
The AI just doesn't make very good decisions when attacking into or blocking bands. Okay this is probably a really low priority considering banding's fate in modern Magic, and whereas the other suggestions would probably benefit everyone, this would only help out old guys like me that want to draft older sets.
If you actually read this much then thanks so much, and I hope it wasn't a huge waste of your time! I would love to hear your feedback (positive and negative!) and if not that's okay, too
Hello beautiful people! For some time now, I've been wondering what Alpha/Beta/Unlimited might look like if it were optimized for draft in the way that contemporary large sets are. I'm pretty new to Forge, but it seems like a perfect platform for this kind of project, so first let me say thanks to the developers for this awesome program! I intend to contribute, and will hopefully be able to use this project to teach myself the ropes.
Inspired by lujo's truly fantastic thread on Fallen Empires Draft (http://www.slightlymagic.net/forum/viewtopic.php?f=26&t=15834), I'll be posting here to keep my thoughts in order and to get feedback from anyone that's interested.
In order to draft the set, add the following line to your blocks.txt folder in /res/blockdata/blocks.txt
- Code: Select all
Limited Edition Gamma, 3/6/A00/, A00
You might want to make backups of the originals first!
If you'd like to play with the temporary new cards I added (pseudo-implementing Power Leak and Chaos Orb), they're in LimitedEditionGammaCards.zip, and can be placed in any folder under /res/cardsfolder/, though I recommend something like /res/cardsfolder/custom_cards/.)
Current Set Stats: 283 Cards (95 Commons, 102 Uncommons, 71 Rares, 15 Mythic Rares)
Ideal-ish Stats: 283 (116 Commons, 91 Uncommons, 61 Rares, 15 Mythic Rares)
Comparison (M15): 249 Cards (101 Commons, 80 Uncommons, 53 Rares, 15 Mythic Rares)
- Design Goals | Open
- Cards Missing in Forge | Open
- Cards Removed From The Set | Open
Redesigned Cards
Card Rarity Changes
Messing with card rarity is the tool I feel most comfortable with when editing the set for drafting. It doesn't require me to remove or add cards, but allows for an adjustment of power levels and matches the way Wizards now designs for draft. A potentially controversial change is the addition of Mythic rarity, but again I'm striving for a contemporary take, and Mythic rares also make regular rares more common, which can be healthy for a draft environment. These changes are a work in progress.
- Cards Upgraded to Mythic Rare (15) | Open
- Cards Upgraded to Rare (Currently Empty) | Open
- Cards Demoted to Uncommon (21) | Open
- Cards Promoted to Uncommon (3) | Open
- Cards Demoted to Common (17) | Open
- Cards Doubled Up on at Common (5) | Open
- Improved Card Rankings | Open
- Archetypes | Open
Future Work
Once I'm somewhat satisfied with the card list and rankings, I can see some areas for improvement that I'd like to work on which should let me get into actually developing little bits of code for Forge proper.
Archetype Drafting
I haven't looked at the Forge draft algorithm, but from what I can tell it tries to pick a card based on the rankings list, giving bonus points for staying in color and subtracting points if it can't use the card. I think the ranking list is really fantastic because it means that a doesn't-know-anything-about-coding user can adjust the AI draft behavior just by editing a text file. It's surprising how well the AI drafts just following this approach! However, there are obviously problems, too.
Ideally, there would be a parser for a similar archetype-based rankings list, that bumps cards up in points if they fall into an archetype that the AI has already drafted some cards for. I think big improvements could be made even with pretty simple algorithmic changes, and different sets could have different user-editable archetype lists. Later we could implement more sophisticated ideas, like some of the things lujo has been suggesting (http://www.slightlymagic.net/forum/viewtopic.php?f=26&t=15834&start=6), all of which could in theory be done in a set-specific text file rather than on a card-by-card basis.
Deckbuilding Improvements
Whoever wrote the deckbuilding algorithm did a fantastic job, as even though a lot of cards are marked unplayable by the AI the decks end up being surprisingly competitive. However, there seem to be a couple areas that are too rigid. In particular, the AI seems to be restricted to building decks with exactly 17 land, which occasional leads to it 'splashing' Counterspell off of zero blue mana sources because it doesn't have enough playables.
Also, I should look into adding AI improvements that let it use some more of these cards!
Banding Combat Math
The AI just doesn't make very good decisions when attacking into or blocking bands. Okay this is probably a really low priority considering banding's fate in modern Magic, and whereas the other suggestions would probably benefit everyone, this would only help out old guys like me that want to draft older sets.
If you actually read this much then thanks so much, and I hope it wasn't a huge waste of your time! I would love to hear your feedback (positive and negative!) and if not that's okay, too
- Attachments
-
LimitedEditionGammaCards.zip- New Cards
- (2.05 KiB) Downloaded 108 times
-
- LimitedEditionGammaRankings.txt
- Rankings List
- (6.64 KiB) Downloaded 123 times
-
- Limited Edition Gamma.txt
- Edition File
- (4.92 KiB) Downloaded 108 times
Last edited by HarePaiR on 20 Dec 2014, 19:40, edited 13 times in total.
Re: Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by HarePaiR » 12 Dec 2014, 12:09
I've been working on getting a reasonable facsimile of Power Leak scripted and am running into some trouble. I know that this card is considered unscriptable but I am trying to teach myself exactly what can't be done since it seems like most everything is there already. This is what I have so far:
I feel like there might be a way to get this to work if I use Effect$ and create a pseudopermanent under the EnchantedController's control that has the damage prevention trigger on it. That seems pretty roundabout, but could it work?
- Code: Select all
Name:Leaky Power
ManaCost:1 U
Types:Enchantment Aura
K:Enchant enchantment
A:SP$ Attach | Cost$ 1 U | ValidTgts$ Enchantment | AILogic$ Curse
T:Mode$ Phase | Phase$ Upkeep | ValidPlayer$ EnchantedController | Execute$ PayPrevDamage | TriggerZones$ Battlefield | TriggerDescription$ At the beginning of the upkeep of enchanted enchantment's controller, that player may play {X}. Prevent the next {X} damage that would be dealt to that player by CARDNAME. CARDNAME deals {2} damage to you.
SVar:PayPrevDamage:AB$ PreventDamage | Cost$ X | Source$ Card.Self | Amount$ X | SpellDescription$ The next time CARDNAME would deal damage to you, prevent {X} of that damage. | References$ X | SubAbility$ TrigDamage
SVar:X:Count$xPaid
SVar:TrigDamage:DB$ DealDamage | Cost$ 0 | Defined$ EnchantedController | NumDmg$ 2 | TriggerDescription$ CARDNAME deals {2} damage to you.
SVar:RemAIDeck:True
SVar:Picture:http://www.wizards.com/global/images/magic/general/power_leak.jpg
Oracle:Enchant enchantment\nAt the beginning of the upkeep of enchanted enchantment's controller, that player may pay any amount of mana. Power leak deals {2} damage to that player. Prevent {X} of that damage, where {X} is the amount of mana that player paid this way.
I feel like there might be a way to get this to work if I use Effect$ and create a pseudopermanent under the EnchantedController's control that has the damage prevention trigger on it. That seems pretty roundabout, but could it work?
Re: Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by elcnesh » 12 Dec 2014, 12:36
You've run into exactly the problem with the card, namely having an arbitrary player (in this case the enchantment's controller) pay any amount of mana. With Unless you can only specify an exact amount, so that's unsuitable here. An Effect might work, but would be very interesting scripting... I think making a PayEffect storing the amount paid is probably the easiest and most appropriate way to go. Maybe I'll try making one later today, then you can be the first to use it 
- elcnesh
- Posts: 290
- Joined: 16 May 2014, 15:11
- Location: Netherlands
- Has thanked: 34 times
- Been thanked: 92 times
Re: Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by friarsol » 12 Dec 2014, 13:52
Right. We have about 8 cards listed as "PayCostEffect" on the spreadsheet of why cards aren't scripted. That would allow a fair chunk of remaining cards to be added in, although Power Leak and it's cousin Errant Minion might still be tricky.elcnesh wrote:I think making a PayEffect storing the amount paid is probably the easiest and most appropriate way to go. Maybe I'll try making one later today, then you can be the first to use it
- friarsol
- Global Moderator
- Posts: 7593
- Joined: 15 May 2010, 04:20
- Has thanked: 243 times
- Been thanked: 965 times
Re: Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by HarePaiR » 12 Dec 2014, 14:02
I appreciate the interest! Though I certainly understand if newer sets are more relevant for the average user, so no rush 
Since you might make some progress on that front, I will move on for now, since this placeholder should behave essentially identically within the environment:
I think I'll move on to my idea for how to represent Chaos Orb next.
Since you might make some progress on that front, I will move on for now, since this placeholder should behave essentially identically within the environment:- Not Power Leak | Open
I think I'll move on to my idea for how to represent Chaos Orb next.
Re: Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by HarePaiR » 12 Dec 2014, 16:09
I've added Chaos Orb back into the set as a non-dextrous flipping a coin card. I assume that both players have been 'smart' enough to infinitely spread their cards out, so that the Orb couldn't ever hit more than one thing, and that a player has 50% aim (so, not Gambit but a decent shot). Given these assumptions, this card should behave pretty much the way it would on tabletops. You know, given infinitely-sized tables.
This was surprisingly easy to script appropriately (choose instead of target, destroy instead of sacrifice), so thanks for such a robust Card API! It seems to work correctly even in games with Slippery Bogle s and Darksteel Forge s.
I have to admit I still don't know how to use most of the AI hooks, though. Is the best place to learn that by looking at cards that I think should be played similarly, or by getting my hands dirty in the source code?
This was surprisingly easy to script appropriately (choose instead of target, destroy instead of sacrifice), so thanks for such a robust Card API! It seems to work correctly even in games with Slippery Bogle s and Darksteel Forge s.
I have to admit I still don't know how to use most of the AI hooks, though. Is the best place to learn that by looking at cards that I think should be played similarly, or by getting my hands dirty in the source code?
Re: Limited Edition Gamma - LEA/LEB/2ED Reimagined for Draft
by HarePaiR » 21 Dec 2014, 04:32
Ranking list has been updated for a second time after significant changes to rarities and a few more cuts (the 'lace' color-changing cards).
My current concerns are lack of creatures, especially in White and at common in Blue. I have already doubled up on some commons to try and fix this, so if it remains a problem I may actually need to add new cards. I'll probably allow myself a few drafts of upcoming sets to see if there's anything appropriate to steal there.
Before I do that, though, I'm going to try to make sure that the AI can use as many cards as possible. For now, I'll do that by trying to understand the AI scripting hooks based off of other similar cards. Eventually, I assume there will be cards that need to be hard-coded, or at least that will need new hooks. (For example, Basalt Monolith probably doesn't work because the AI uses it when it doesn't need to, and untaps it when it doesn't need to. There are other cards like this, and there should be a way to deprioritize their use unless the AI doesn't have enough mana.)
My current concerns are lack of creatures, especially in White and at common in Blue. I have already doubled up on some commons to try and fix this, so if it remains a problem I may actually need to add new cards. I'll probably allow myself a few drafts of upcoming sets to see if there's anything appropriate to steal there.
Before I do that, though, I'm going to try to make sure that the AI can use as many cards as possible. For now, I'll do that by trying to understand the AI scripting hooks based off of other similar cards. Eventually, I assume there will be cards that need to be hard-coded, or at least that will need new hooks. (For example, Basalt Monolith probably doesn't work because the AI uses it when it doesn't need to, and untaps it when it doesn't need to. There are other cards like this, and there should be a way to deprioritize their use unless the AI doesn't have enough mana.)
7 posts
• Page 1 of 1
Who is online
Users browsing this forum: No registered users and 164 guests