Getting Fallen Empires draft to work
Posted: 17 Oct 2014, 23:57EDIT - Progress!
Deep Spawn, the notorious giant Lobster is in the house! Half of the dreaded High Tide + Dr. Zoidberg combo is operational thanks to Sloth (and god knows what else has been done that I don't know about!)
Edit - Also edited out A few things which were solved
-----
So, I'd love to assist anyone with programing skills and as little time as they could afford to get this little gem of a set working a bit better in limited. No really fancy stuff for starters, just get the bugs out and the AI's life a bit easier and the whole thing working a little better.
Why is this set worth a bit of special care? Because it's really, really sweet, much better than people give it credit for, probably the deepest of the original small pre-block sets, and also loaded with bugs.
And I'd love to eventually build a Terrisaire questing world for forge (The Dark, Fallen Empires, Homelands, Ice Age, Alliances and Coldsnap). I've been playtesting it and it's actually pretty fun. Such a world would be a nice default place to either play with cards of that era or fish for core staples (which include Force of Will, Pyrokinesis, Hymn to Tourach, Merchant Scroll, Necropotence, Orcish Lumberjack, Tinder Wall, Brainstorm and a whole lot of other useful junk).
Card's of note which make FE a fine little addition to a quest pool:
That's quite a bunch of playables and fine cards to have - some are bloody strong. Since the set is cheap due to RL overprinting, and is also rather small, I'd love to see if as many kinks could be pruned from it as possible.
No unrealistic expectations, if there's heavy constraints on something - what can you do. But I'd love to know at least why some things which don't really work don't work, because maybe we could come up with a solution.
Current situation - 1.5.28:
Unimplemented:
Merseine - It's being worked on! (Maybe )
Raiding Party - Ok, I can see how implementing this one would be a pickle. I wouldn't badger anyone with implementing it, but having a card not appear in boosters is very bad for drafting a set. This one would probably not be played, but there might be a bomby uncommon appearing more often in it's place. Would be nice if, until this is implemented if it ever is, it gets replaced with a basic land or something. I'd advise that for any such situation.
Bugged:
Night Soil (currently) - it's bugged in conjunction with the interface which lets you select which cards to remove. It "selects" cards in the "keep in graveyard" part of the selection interface, so it creates tokens without removing cards which means you can make infinite tokens provided you have infinite mana.
AI touchups on some key cards:
Black:
Necrite - is identical to Mindstab Thrull, but sacrifices itself to kill a creature. The AI isn't putting it in decks which is a shame, as the set is small and tight and most cards are in it for a reason (and removal in it is precious). I am really curious about why the AI isn't using it, if it is using Mindstab Thrull (it certainly seems to be drafting it, because I rarely see it in boosters).
Ebon Praetor - is essentially The Black Bomb of the set, and works like a Lord Of The Pit. It's much bigger than most creatures, has first strike and pretty much nothing can handle him, but you need to sacrifice a creature to him every turn. If you sacrifice a Thrull he gets bigger instead.
I don't see the AI using him, and I see the AI passing him around, which certainly isn't what would happen IRL, and he's a very good reason to be in Black, the best one in fact.
Blue:
High Tide / Deep Spawn
This combo is what High Tide is in the set for.
Deep Spawn is the largest creature in the set. I don't see the AI drafting either (or the spawning bed) which makes drafting mono blue kind of easy for the Human but it also often forces the human to draft blue as it's always "open" and the boosters are small and flooded with it. The AI fix might be easy if it's possible to do:
Homarid Spawning Bed -
I've never seen the AI play or pick up Homarid Spawning Bed. This is very strange, as that's basically the key card to blue and using it on any blue creature which isn't a token is essentially never the wrong move (most of them have higher casting cost than makes sense because to combo with it better).
If anyone would care to properly make it's AI work at least with its set, the list of valid targets is fairly short:
Using the HSM on most of these nets a profit in power on the board, chump blockers, tokens to madly attack with etc. Deep spawn might be an exception, because it's significantly larger than most creatures in the set, so simply having it on the board and attacking with it is the other way blue reliably goes for the win.
There are several ways HSB could be coded to work with FE blue creatures, but those depend on what the AI can or can't register. There's also very little removal in the set, so having HSB be used in response to any of those is also a perfectly legit option.
Other Blue Stuff:
Well, if the AI could be thaught to use HT + DS, Merseine was ported, and maybe the Spawning Bed made to work with the overcosted commons it'd be quite an improvement already.
Red:
Goblin Kites - Since there's no fliers in the set, and all red creatures have =< 2 base thoughness this card is essentially Fling on a stick except against Combat Medic, Spore Flower or a Spore Cloud from hand. I've never seen the AI use it, but it should be fairly simple to code at least a simple and effective way to use it.
If making X creatures go through would kill the opponent and you have enough mana, you do it and kill the opponent. It effectively sayes: "R: Target creature is unblockable untill end of turn. At the end of turn flip a coin. If you lose, destroy the creature and it can't be regenerated."
I think the AI could make that calculation, unless this is already working. It would certainly lower the amount of ground stalls (it works on non-red creatures too so it's basically a complete bomb).
Brassclaw Orcs - I see situations where the AI is unwilling to attack with these guys as it's afraid to trade so it leaves them on defense. But the problem is that they can't block anything worth blocking, so the AI just loses a turn of offense, which it can't afford. For all practical purposes these guys can't block, and if it's at all possible they should always be attacking unless they would die without trading. Not too many things can block them and survive either.
Green
Thelonite Druid - I've never seen the AI use it, and it gets passed around in boosters, which would never ever happen in RL, as it's one of the strongest if not the strongest card in the set.
If the AI isn't using it it causes ground stalls to go on forever, as the Thallids are generally slow and only really serve to clog the ground until you have this guy up to animate forests and kill the other guy.
A very basic functionality for this guy could (if possible) work this way:
Thorn Thallid - The AI tends to waste these guys by pinging the opposing player. A much better strategy is to hold onto counters until it can shoot a 2 toughness creature. If used properly it's the one of the stronger cards in FE draft.
White:
Icatian Phalanx - The AI is ignoring this thing, and it's a complete bomb. I understand that the AI isn't the best with banding (and also that this is actually being worked on), but in FE draft, even without actually using banding on it, a 2/4 which gets to chose how damage is assigned on defense is still a really high pick.
Icatian Javelineers - The AI will waste the counter by pinging the player. In a removal light format, this is a big waste, as it could otherwise kill really important stuff with that or add it to damage done in a combat situation.
Ok, so any amount of work on those would improve the experience quite a bit, some might be easy, most would probably make the AI on AI matches devolve into ground stalls much less often, and the drafts would be more sensible. Some might already be working, too.
A few general things about FE draft, deckbuilding and ways to improve it:
1) For one thing, FE draft should most likely involve 6 boosters rather than 5. It's a tribal set which is sort of heavily dependent on uncommons and rares for it's archetypes to work.
2) It's kind of too complex for 8 card boosters and the way it's generally drafted in practice is to have 6 boosters, but open 2 at the same time. This then makes the 16 card packs deep enough for a much more sensible draft.
3) The AI will sometimes make really terrible 3 or 4 color decks as it's not evaluating cards right, and this is causing really lopsided games, so something, whatever it is, needs to be done about it to either get the draft supplying more playables or the AI capable of playing more cards.
4) The set is a tribal set. It's not just removal-light it's instant-and-sorcery-light, too (only 6 instants + sorceries all in all). Many creatures in it function as essentially "Seals" like Seal of Fire would. Goblin Chirurgeon is the broken grandfather of Welding Jar. Basal Thrull is a more balanced Dark Ritual. Armor Thrull is a pseudo Holy Strenght. Most effects you'd see on spells in other sets appear as either sacrificeable creatures, repeatable effects on creatures or enchantments. (This is why the sacrifice theme is so emphasized in the famous flavor of the set, it's very tied to the mechanical design philosophy behind it).
The reason I feel the need to explain this is that as opposed to other early sets, where I'd encourage it, trying to make Fallen Empires draft better by adding a booster of, say, 4E wouldn't work. The whole design falls apart if you add fliers (there are none), creatures with intimidate (there are none), mass removal (there is just one and it's tricky to use), etc.
This is why either simply adding another FE booster or adding another booster but opening 2 "at the same time" really is the way to go.
5) A very, very cool thing to do which Forge could allow, and which I'd personally LOVE to do is to use the space there is in the set and fill it with period appropriate cards from other expansions (keeping the strict design rules of the set pertaining to creature types intact and only using cards from The Dark, Ice Age, Homelands, Alliances and maybe Coldsnap if I could make old frame art for it).
What that would entail is this:
So, would anyone with coding skills like to help out with making a cool little old set kick ass as much as it could? I'm not demanding anything, I'm not expecting miracles, what can't be done can't be done or can't be done yet, but anything would be really highly appreciated. I tried to make a you tube vid of forge drafting Fallen Empires but I kept running into the Mindstab Thrull bug, or stuff that looks stupid but can actually be fixed and look kick ass, and I'd love to promote this program and present it in the best possible light.
And, well, there's way more that can be done with the set than these glaring things here and the experience with getting IT to work is bound to be invaluable - it's the granddaddy of a million things that came afterwards, moreso than anything other than possibly Alliances.
I could probably reconstruct the design doc to a T, explain why every card is in there, how's what supposed to work, etc etc. I really hope someone'd like to make this work as much as I would. Or at least make the damned Mindstab Thrull not crash the game.
If it would help, I'll make some vids. What the heck, I'll make some anyway, but still... Help?
Deep Spawn, the notorious giant Lobster is in the house! Half of the dreaded High Tide + Dr. Zoidberg combo is operational thanks to Sloth (and god knows what else has been done that I don't know about!)
Edit - Also edited out A few things which were solved
-----
So, I'd love to assist anyone with programing skills and as little time as they could afford to get this little gem of a set working a bit better in limited. No really fancy stuff for starters, just get the bugs out and the AI's life a bit easier and the whole thing working a little better.
Why is this set worth a bit of special care? Because it's really, really sweet, much better than people give it credit for, probably the deepest of the original small pre-block sets, and also loaded with bugs.
And I'd love to eventually build a Terrisaire questing world for forge (The Dark, Fallen Empires, Homelands, Ice Age, Alliances and Coldsnap). I've been playtesting it and it's actually pretty fun. Such a world would be a nice default place to either play with cards of that era or fish for core staples (which include Force of Will, Pyrokinesis, Hymn to Tourach, Merchant Scroll, Necropotence, Orcish Lumberjack, Tinder Wall, Brainstorm and a whole lot of other useful junk).
Card's of note which make FE a fine little addition to a quest pool:
- Cards of Note | Open
That's quite a bunch of playables and fine cards to have - some are bloody strong. Since the set is cheap due to RL overprinting, and is also rather small, I'd love to see if as many kinks could be pruned from it as possible.
No unrealistic expectations, if there's heavy constraints on something - what can you do. But I'd love to know at least why some things which don't really work don't work, because maybe we could come up with a solution.
Current situation - 1.5.28:
Unimplemented:
Merseine - It's being worked on! (Maybe
) - Info | Open
Raiding Party - Ok, I can see how implementing this one would be a pickle. I wouldn't badger anyone with implementing it, but having a card not appear in boosters is very bad for drafting a set. This one would probably not be played, but there might be a bomby uncommon appearing more often in it's place. Would be nice if, until this is implemented if it ever is, it gets replaced with a basic land or something. I'd advise that for any such situation.
Bugged:
- It's been solved! | Open
Night Soil (currently) - it's bugged in conjunction with the interface which lets you select which cards to remove. It "selects" cards in the "keep in graveyard" part of the selection interface, so it creates tokens without removing cards which means you can make infinite tokens provided you have infinite mana.
AI touchups on some key cards:
Black:
Necrite - is identical to Mindstab Thrull, but sacrifices itself to kill a creature. The AI isn't putting it in decks which is a shame, as the set is small and tight and most cards are in it for a reason (and removal in it is precious). I am really curious about why the AI isn't using it, if it is using Mindstab Thrull (it certainly seems to be drafting it, because I rarely see it in boosters).
Ebon Praetor - is essentially The Black Bomb of the set, and works like a Lord Of The Pit. It's much bigger than most creatures, has first strike and pretty much nothing can handle him, but you need to sacrifice a creature to him every turn. If you sacrifice a Thrull he gets bigger instead.
I don't see the AI using him, and I see the AI passing him around, which certainly isn't what would happen IRL, and he's a very good reason to be in Black, the best one in fact.
Blue:
High Tide / Deep Spawn
This combo is what High Tide is in the set for.
Deep Spawn is the largest creature in the set. I don't see the AI drafting either (or the spawning bed) which makes drafting mono blue kind of easy for the Human but it also often forces the human to draft blue as it's always "open" and the boosters are small and flooded with it. The AI fix might be easy if it's possible to do:
Homarid Spawning Bed -
I've never seen the AI play or pick up Homarid Spawning Bed. This is very strange, as that's basically the key card to blue and using it on any blue creature which isn't a token is essentially never the wrong move (most of them have higher casting cost than makes sense because to combo with it better).
If anyone would care to properly make it's AI work at least with its set, the list of valid targets is fairly short:
- Sensible HSB targets | Open
Using the HSM on most of these nets a profit in power on the board, chump blockers, tokens to madly attack with etc. Deep spawn might be an exception, because it's significantly larger than most creatures in the set, so simply having it on the board and attacking with it is the other way blue reliably goes for the win.
There are several ways HSB could be coded to work with FE blue creatures, but those depend on what the AI can or can't register. There's also very little removal in the set, so having HSB be used in response to any of those is also a perfectly legit option.
Other Blue Stuff:
Well, if the AI could be thaught to use HT + DS, Merseine was ported, and maybe the Spawning Bed made to work with the overcosted commons it'd be quite an improvement already.
Red:
Goblin Kites - Since there's no fliers in the set, and all red creatures have =< 2 base thoughness this card is essentially Fling on a stick except against Combat Medic, Spore Flower or a Spore Cloud from hand. I've never seen the AI use it, but it should be fairly simple to code at least a simple and effective way to use it.
If making X creatures go through would kill the opponent and you have enough mana, you do it and kill the opponent. It effectively sayes: "R: Target creature is unblockable untill end of turn. At the end of turn flip a coin. If you lose, destroy the creature and it can't be regenerated."
I think the AI could make that calculation, unless this is already working. It would certainly lower the amount of ground stalls (it works on non-red creatures too so it's basically a complete bomb).
Brassclaw Orcs - I see situations where the AI is unwilling to attack with these guys as it's afraid to trade so it leaves them on defense. But the problem is that they can't block anything worth blocking, so the AI just loses a turn of offense, which it can't afford. For all practical purposes these guys can't block, and if it's at all possible they should always be attacking unless they would die without trading. Not too many things can block them and survive either.
Green
Thelonite Druid - I've never seen the AI use it, and it gets passed around in boosters, which would never ever happen in RL, as it's one of the strongest if not the strongest card in the set.
If the AI isn't using it it causes ground stalls to go on forever, as the Thallids are generally slow and only really serve to clog the ground until you have this guy up to animate forests and kill the other guy.
A very basic functionality for this guy could (if possible) work this way:
- Thelonite Druid AI | Open
Thorn Thallid - The AI tends to waste these guys by pinging the opposing player. A much better strategy is to hold onto counters until it can shoot a 2 toughness creature. If used properly it's the one of the stronger cards in FE draft.
White:
Icatian Phalanx - The AI is ignoring this thing, and it's a complete bomb. I understand that the AI isn't the best with banding (and also that this is actually being worked on), but in FE draft, even without actually using banding on it, a 2/4 which gets to chose how damage is assigned on defense is still a really high pick.
Icatian Javelineers - The AI will waste the counter by pinging the player. In a removal light format, this is a big waste, as it could otherwise kill really important stuff with that or add it to damage done in a combat situation.
Ok, so any amount of work on those would improve the experience quite a bit, some might be easy, most would probably make the AI on AI matches devolve into ground stalls much less often, and the drafts would be more sensible. Some might already be working, too.
A few general things about FE draft, deckbuilding and ways to improve it:
1) For one thing, FE draft should most likely involve 6 boosters rather than 5. It's a tribal set which is sort of heavily dependent on uncommons and rares for it's archetypes to work.
2) It's kind of too complex for 8 card boosters and the way it's generally drafted in practice is to have 6 boosters, but open 2 at the same time. This then makes the 16 card packs deep enough for a much more sensible draft.
3) The AI will sometimes make really terrible 3 or 4 color decks as it's not evaluating cards right, and this is causing really lopsided games, so something, whatever it is, needs to be done about it to either get the draft supplying more playables or the AI capable of playing more cards.
4) The set is a tribal set. It's not just removal-light it's instant-and-sorcery-light, too (only 6 instants + sorceries all in all). Many creatures in it function as essentially "Seals" like Seal of Fire would. Goblin Chirurgeon is the broken grandfather of Welding Jar. Basal Thrull is a more balanced Dark Ritual. Armor Thrull is a pseudo Holy Strenght. Most effects you'd see on spells in other sets appear as either sacrificeable creatures, repeatable effects on creatures or enchantments. (This is why the sacrifice theme is so emphasized in the famous flavor of the set, it's very tied to the mechanical design philosophy behind it).
The reason I feel the need to explain this is that as opposed to other early sets, where I'd encourage it, trying to make Fallen Empires draft better by adding a booster of, say, 4E wouldn't work. The whole design falls apart if you add fliers (there are none), creatures with intimidate (there are none), mass removal (there is just one and it's tricky to use), etc.
This is why either simply adding another FE booster or adding another booster but opening 2 "at the same time" really is the way to go.
5) A very, very cool thing to do which Forge could allow, and which I'd personally LOVE to do is to use the space there is in the set and fill it with period appropriate cards from other expansions (keeping the strict design rules of the set pertaining to creature types intact and only using cards from The Dark, Ice Age, Homelands, Alliances and maybe Coldsnap if I could make old frame art for it).
What that would entail is this:
- Making a draft friendly Forge Fallen Empires | Open
So, would anyone with coding skills like to help out with making a cool little old set kick ass as much as it could? I'm not demanding anything, I'm not expecting miracles, what can't be done can't be done or can't be done yet, but anything would be really highly appreciated. I tried to make a you tube vid of forge drafting Fallen Empires but I kept running into the Mindstab Thrull bug, or stuff that looks stupid but can actually be fixed and look kick ass, and I'd love to promote this program and present it in the best possible light.
And, well, there's way more that can be done with the set than these glaring things here and the experience with getting IT to work is bound to be invaluable - it's the granddaddy of a million things that came afterwards, moreso than anything other than possibly Alliances.
I could probably reconstruct the design doc to a T, explain why every card is in there, how's what supposed to work, etc etc. I really hope someone'd like to make this work as much as I would. Or at least make the damned Mindstab Thrull not crash the game.
If it would help, I'll make some vids. What the heck, I'll make some anyway, but still... Help?