One of my favorite things to do when watching Baalor is something I call "The Shop Game." I'll pause the video and try to guess how Baalor will spend his money. Then I get to see what he actually does. This has really helped me improve my STS shopping game. Thanks, Baalor!
I've watched FrostPrime play a pretty good PVP mod, you would climb a few floors getting rewards, then you would face off. You both have a heart buff so you can only so much damage per turn to help against the turn 1 infinites. The loser would get an extra buff of some kind to help them keep up with the winner and not have a snowball, and then you would climb some more floors and repeat, I think it was a 1st to 3 wins in the faceoffs to win the match
Ooooh that could be a great mod with some modifications: I'd like it to be act 4 layout from the get go, you'd have 2 combat rooms, 2 event rooms, 1 shop, 1 elite and 1 boss each floor in this order. First you'll expend the act 1 combat and boss pools, so it'll be like act 1 thrice back to back but only 7 floors total. After each boss you get to fight against each other with damage cap buff, the loser might get extra gold (or perhaps only loser gets the gold) but winner gets card reward and relic? It's like a combination of boss rush and pvp, I'm loving it!
PVP could put you up against someone to see who can get the most points on the same seed same character. A competition to see who’s pathfinding and deck building is superior
I was thinking straight pvp like starting deck and bonus then each floor you climb is a new player that also climbed by beating someone else. You get offered a card each win, get healed like a boss, and a relic every 5.
Timestamps for when Baalor picks up the cards in the thumbnail (Compile Driver x4): [Updated 9/19/24 for the new thumbnail. Original thumbnail was Barrage x2.] Compile Driver #1: (14:29) (Floor 10 Combat) Compile Driver #2: (39:16) (Floor 21 Combat) Compile Driver #3: (51:46) (Floor 29 Combat) Compile Driver #4: (1:17:58) (Floor 44 Elite) Original timestamps: Barrage #1: (20:50) (Floor 11 Elite) Barrage #2: (1:05:03) (Floor 35 Combat) A downpour of orbs has drenched the Spire!
I like the idea of pvp like Tetris did where you and several other people start at the same time and see who can win first, last the longest, and get the most points. Everybody is on screen and you can see how they are doin.
It would definitely have to be something like this, because directly fighting would just be down to whoever goes first. Characters have like 70 HP on a good day and even an act 1 deck could do half of that on the first turn.
bottled seek+ is like genuinely one of the strongest things in the game imho, search for any 2 card combo in your entire deck every single fight (or later if you get apotheosis you can search apo + any 2nd card)
I would've probably taken Jax, to be honest, with Upgraded Bites and Barrage. Yeah it has antisynergy with Buffer, but I think that it's not such a bad trade-off as to warrant skipping it.
Yeah, I think the Bites+ would have had me taking the Jax over the transform, though I think transforming one of the Bites for the remove he was already talking about wanting was also very reasonable.
I've studied the game state question in chess, and think I understand the parameters moderately well in the starter deck/jaw worm question, and I'd like to respectfully disagree. I think perhaps once you've added a few cards it ends up being true, just because permutations of deck order alone is a large number. Additionally, I've studied the AI somewhat extensively both from a programming perspective and from an operational perspective. Provided sufficient development resources, I believe a well trained AI could very easily play spire at an inhuman level. An AI approach wouldn't need to consider all possible game states in all situations, it'd be a question of training and understanding the statistical and numerical outcomes of cards/relics and play order. I'd argue i's not a very complicated task when compared with actual games AIs now perform better than all humans, like Starcraft for example.
Agree. In fact there's a spire AI already though mastered is... sort of far from what it can do. Really the question isn't whether an AI *could* master the Spire but rather how one would go about defining mastery.
In reference to play states in Slay the Spire, I present: How to imagine 52!, an essay by Scott Czepiel: This number is beyond astronomically large. I say beyond astronomically large because most numbers that we already consider to be astronomically large are mere infinitesimal fractions of this number. So, just how large is it? Let's try to wrap our puny human brains around the magnitude of this number with a fun little theoretical exercise. Start a timer that will count down the number of seconds from 52! to 0. We're going to see how much fun we can have before the timer counts down all the way. Start by picking your favorite spot on the equator. You're going to walk around the world along the equator, but take a very leisurely pace of one step every billion years. The equatorial circumference of the Earth is 40,075,017 meters. Make sure to pack a deck of playing cards, so you can get in a few trillion hands of solitaire between steps. After you complete your round the world trip, remove one drop of water from the Pacific Ocean. Now do the same thing again: walk around the world at one billion years per step, removing one drop of water from the Pacific Ocean each time you circle the globe. The Pacific Ocean contains 707.6 million cubic kilometers of water. Continue until the ocean is empty. When it is, take one sheet of paper and place it flat on the ground. Now, fill the ocean back up and start the entire process all over again, adding a sheet of paper to the stack each time you've emptied the ocean. Do this until the stack of paper reaches from the Earth to the Sun. Take a glance at the timer, you will see that the three left-most digits haven't even changed. You still have 8.063e67 more seconds to go. 1 Astronomical Unit, the distance from the Earth to the Sun, is defined as 149,597,870.691 kilometers. So, take the stack of papers down and do it all over again. One thousand times more. Unfortunately, that still won't do it. There are still more than 5.385e67 seconds remaining. You're just about a third of the way done. To pass the remaining time, start shuffling your deck of cards. Every billion years deal yourself a 5-card poker hand. Each time you get a royal flush, buy yourself a lottery ticket. A royal flush occurs in one out of every 649,740 hands. If that ticket wins the jackpot, throw a grain of sand into the Grand Canyon. Keep going and when you've filled up the canyon with sand, remove one ounce of rock from Mt. Everest. Now empty the canyon and start all over again. When you've leveled Mt. Everest, look at the timer, you still have 5.364e67 seconds remaining. Mt. Everest weighs about 357 trillion pounds. You barely made a dent. If you were to repeat this 255 times, you would still be looking at 3.024e64 seconds. The timer would finally reach zero sometime during your 256th attempt. Exercise for the reader: at what point exactly would the timer reach zero?
Yeah, i think it would be cool to organise sts streamers world championship. Where top streamers would play on the same seed and see who get most points. First 4 rounda could be with 4 different characters, and then round 5 would be with character of choice. We would have defect winner, silent winner etc, and then the overall winner in a person who got the most points total. If you would organised something like this, i would be super excited
Frost prime messed around with the spire PvP mod for a bit and had some fun. Can’t explain how it works it’s been a while and it basically re worked the entire game but the characters and cards
Alright man, you seem to be on point with defect runs. He's my favorite character for sure. Got him to ascension 20 but keep dying on the double boss fight. Any good tips for a frustrated person lol?
You could do pvp in a segmented draft format and both players get the time eater mechanic. Health bars would need to be hyper inflated though. And you could still have some dumb stuff like burst+->bouncing flask+->catalyst+ for 216 poison. Also balance would be slightly wack like there's little counterplay against focus unlike strength Another cool idea would be having a sort of "spire director" player control to some degree the node encounters and enemy AI.
There is precedence for first turn wins in TCG games. Magic the gathering have certain formats where its possible to get those of. But not anyone trying to compete will strive to do as the odds of pulling those combos off are very slim even on the best of days.
@@413xmb You're talking about the typical case. The scenario Baalor was presented with was early Act 2, Membership Card, and Molten Egg. It was a terrific cost-value for HoG here. Ultimately, bottling Seek here turned out to be extremely valuable but I don't think taking HoG was a wrong move either.
@@JoshuaKIM-qd2vd I think given the decks reliance on seek and the turbo/echo form or core surge/biased cog turn 1 starts, the HoG was still the worse choice. I agree that it was high value comparatively, but even still the starting capacitor and barrage combo was being leaned into so hard it might be difficult to actually land HoG without taking unnecessary damage
I think Watcher also struggles against guardian, since her main strat is deleting defends and sometimes you don't get enough block density or consistency and she gets plapped.
Definitely not far away from having ML capable of solving spire, I'm afraid. Mastered Go already, and this was what, 10 years ago? Sequential decision making in such a controlled environment is certainly not that much of a challenge as it was 20 years ago. Noteworthy here is that despite the amount of game states it has playing ofc a huge role, what a deep neural net will do is learn a sufficiently expressive representation of a decision space where it can interpolate to never-seen states with somewhat good precision. My opinion at least xD
Well its kinda difficult and rather meaningless to compare the complexity but it sounds fun so i'll try. Feel free to correct me if i'm wrong. Disclaimer: Now that I wrote all of this stuff I realised what he probably meant "DURING a game of chess" not "THE game of chess". But as each player has (without capturing pieces and without health) at LEAST 10 different options (100 per cycle), I'd argue chess has a bigger growing flow chart than the yawworm fight. Tl;Dr: It takes approx. 36 turns or 18 deck cycles (250 turns forthe naive approach) for the Ironclad starterdeck vs jaw worm for the amount of possible gamestates to approach the amount of all possible gamestates in chess without capturing. Chess: I guess you can take the naive approach for chess. 32 pieces 64 squares makes it to 64!/32! (yikes) possible gamestates. Keep in mind that with this approach impossible situations are included (pawns on the first rank, impossible mates, passed pawns without any captures etc.). Further more all pieces are on the board (no captures and no promotions) plus identical positions are included (each pawn is considered differnt) whereby this is a VERY bad estimator. The other approach would be to consider pawns first. Each pawn has 6-x (x being the opposite pawns position so xe{1,2,3,4,5}) possible squares. So, there are 15 different states and 15^8 possible states for all pawns. With the pawns out of the picture there are only 16 pieces, 48 squares remaining and consequently 48!/32! gamestates. So the new estimator ceils the amount of possible gamestates at 15^8*48!/32! (if you want unique possible gamestates, divide it by 4 as rooks and knights are interchangable). Dont get me wrong this number is still too high as impossible state are still included but its not wrong anymore by several orders of magnitude. Jaw worm vs Iron clad starter deck: Yes jaw worm cant thrash 3 thrice but lets say he can. Additionally, lets consider the case for the starterdeck without ascenders bane as it breaks the two turn cycle and gone after the first. 10 cards (5 strikes, 4 defends and 1 bash) with 5cards drawn each turn. there are therefore 10 possible cycle orders. Meanwhile the jaw worm has 9 different states per deck cycle (3 each). So all in all you get the amount of possible gamestates of 90^x/3 as the jaw worm can only chomp on turn 1(x means per deck cycle or every two turns). This approach doesnt take health into account but it still gets big fairly quickly. Comparisson: So how many cycles does the starterdeck need to approach the complexity of chess? ln(3*15^8*48!/32!)= approx. 82 ln(90)= approx 4.5 => x=82/4.5=18.2 So after 18 cycles or 36 turns the yaw worm fight has more possible game states than there are possible positions on a chess board and respectively 125 cycles for the bad estimator. Again, this doesnt take health into account but we also never considered the possibility of capturing pieces. So yeah ... unlikely that you reach that, however, with branching paths and card rewards you should get there around mid Act 1.
If a 10 year old ai can beat professional dota players, i dont see why it couldnt play spire at at least a competent level. Maybe if you think about every card and possibility as an individual component, but if you just simplify cards potions and relics down to their basic mechanics: damage, damage over time, block, draw, etc. i think an ai could determine the relative likelihood of values (cards) to lead to success after a few million games, which is pretty feasible.
I dont understand the love for that FTL+. It's not a bad card but its upgrade has to be one of the least impactful upgrades in the whole game. +1 damage if its zero cost i guess is ok but it sure isnt pommel strike +. With the upgrade FTL is basically the same card. I dont believe being able to draw after one extra card played impacts the card much at all. It would need to gain at least 3 extra damage to be a good upgrade imo.
Kinda circumstantial but I like how the upgrade works with all for one. If you've got a ton of 0 cost cards and you draw all for one, to get the card draw from ftl (assuming it's in your discard) you can only play one other card before playing all for one. If you draw more than one zero cost card alongside all for one, you then have to chose between the ftl card draw and playing that second 0 cost twice. Not saying the upgrade is that great but I feel in this case it's giving you some decent value.
One of my favorite things to do when watching Baalor is something I call "The Shop Game."
I'll pause the video and try to guess how Baalor will spend his money. Then I get to see what he actually does.
This has really helped me improve my STS shopping game. Thanks, Baalor!
I agree. Except I kinda do this same thing with card rewards, removals, relics, etc. It really does help.
He tends to hover his mouse immediately over whatever his snap pick is - he doesn’t always go with it, but usually.
true, same for any decision made i guess.
Thats exactly how I beat the game.
Not a copy paste just learning.
I've watched FrostPrime play a pretty good PVP mod, you would climb a few floors getting rewards, then you would face off. You both have a heart buff so you can only so much damage per turn to help against the turn 1 infinites. The loser would get an extra buff of some kind to help them keep up with the winner and not have a snowball, and then you would climb some more floors and repeat, I think it was a 1st to 3 wins in the faceoffs to win the match
Ooooh that could be a great mod with some modifications: I'd like it to be act 4 layout from the get go, you'd have 2 combat rooms, 2 event rooms, 1 shop, 1 elite and 1 boss each floor in this order. First you'll expend the act 1 combat and boss pools, so it'll be like act 1 thrice back to back but only 7 floors total. After each boss you get to fight against each other with damage cap buff, the loser might get extra gold (or perhaps only loser gets the gold) but winner gets card reward and relic? It's like a combination of boss rush and pvp, I'm loving it!
PVP could put you up against someone to see who can get the most points on the same seed same character. A competition to see who’s pathfinding and deck building is superior
Will need to be multiple players because most likely 90% of the 1v1 games will end on one of the players leaving within the first 10 minutes.
Yeah also it's hard to do a meaningful ranking as everybody can pause his game anytime and start it again a week later, or never.
I think you are describing daily climb
And then at the end you could fight them using the decks you both made. Just for fun and some novelty, y'know.
I was thinking straight pvp like starting deck and bonus then each floor you climb is a new player that also climbed by beating someone else. You get offered a card each win, get healed like a boss, and a relic every 5.
Timestamps for when Baalor picks up the cards in the thumbnail (Compile Driver x4):
[Updated 9/19/24 for the new thumbnail. Original thumbnail was Barrage x2.]
Compile Driver #1: (14:29) (Floor 10 Combat)
Compile Driver #2: (39:16) (Floor 21 Combat)
Compile Driver #3: (51:46) (Floor 29 Combat)
Compile Driver #4: (1:17:58) (Floor 44 Elite)
Original timestamps:
Barrage #1: (20:50) (Floor 11 Elite)
Barrage #2: (1:05:03) (Floor 35 Combat)
A downpour of orbs has drenched the Spire!
Is my UA-cam messed up? The thumbnail is of compile driver?
Baalor changes the thumbnails so you forget you watched the video and watch it again @@ronansilke8104
@@ronansilke8104the thumbnail was changed
I like the idea of pvp like Tetris did where you and several other people start at the same time and see who can win first, last the longest, and get the most points. Everybody is on screen and you can see how they are doin.
It would definitely have to be something like this, because directly fighting would just be down to whoever goes first. Characters have like 70 HP on a good day and even an act 1 deck could do half of that on the first turn.
Baalor 13:38 " We should be fine now", 20:14 "That was spooky" XD
1:33:46 which relic was the most important this run
I would argue that it was bottled lightning, since having seek on turn 1 made this deck OP
bottled seek+ is like genuinely one of the strongest things in the game imho, search for any 2 card combo in your entire deck every single fight (or later if you get apotheosis you can search apo + any 2nd card)
a couple fights into act 2 and this is looking like how all my act 2s look (pain)
waiting for him to just break everything in a bit though
I would've probably taken Jax, to be honest, with Upgraded Bites and Barrage. Yeah it has antisynergy with Buffer, but I think that it's not such a bad trade-off as to warrant skipping it.
Yeah, I think the Bites+ would have had me taking the Jax over the transform, though I think transforming one of the Bites for the remove he was already talking about wanting was also very reasonable.
I've studied the game state question in chess, and think I understand the parameters moderately well in the starter deck/jaw worm question, and I'd like to respectfully disagree. I think perhaps once you've added a few cards it ends up being true, just because permutations of deck order alone is a large number. Additionally, I've studied the AI somewhat extensively both from a programming perspective and from an operational perspective. Provided sufficient development resources, I believe a well trained AI could very easily play spire at an inhuman level. An AI approach wouldn't need to consider all possible game states in all situations, it'd be a question of training and understanding the statistical and numerical outcomes of cards/relics and play order. I'd argue i's not a very complicated task when compared with actual games AIs now perform better than all humans, like Starcraft for example.
I made an extremely similar comment before seeing yours. Totally agree. Cheers!
Agree. In fact there's a spire AI already though mastered is... sort of far from what it can do.
Really the question isn't whether an AI *could* master the Spire but rather how one would go about defining mastery.
@@hieronymusnervig8712beating every beatable A20 Heart.
In reference to play states in Slay the Spire, I present:
How to imagine 52!, an essay by Scott Czepiel:
This number is beyond astronomically large. I say beyond astronomically large because most numbers that we already consider to be astronomically large are mere infinitesimal fractions of this number. So, just how large is it? Let's try to wrap our puny human brains around the magnitude of this number with a fun little theoretical exercise. Start a timer that will count down the number of seconds from 52! to 0. We're going to see how much fun we can have before the timer counts down all the way.
Start by picking your favorite spot on the equator. You're going to walk around the world along the equator, but take a very leisurely pace of one step every billion years. The equatorial circumference of the Earth is 40,075,017 meters. Make sure to pack a deck of playing cards, so you can get in a few trillion hands of solitaire between steps. After you complete your round the world trip, remove one drop of water from the Pacific Ocean. Now do the same thing again: walk around the world at one billion years per step, removing one drop of water from the Pacific Ocean each time you circle the globe. The Pacific Ocean contains 707.6 million cubic kilometers of water. Continue until the ocean is empty. When it is, take one sheet of paper and place it flat on the ground. Now, fill the ocean back up and start the entire process all over again, adding a sheet of paper to the stack each time you've emptied the ocean.
Do this until the stack of paper reaches from the Earth to the Sun. Take a glance at the timer, you will see that the three left-most digits haven't even changed. You still have 8.063e67 more seconds to go. 1 Astronomical Unit, the distance from the Earth to the Sun, is defined as 149,597,870.691 kilometers. So, take the stack of papers down and do it all over again. One thousand times more. Unfortunately, that still won't do it. There are still more than 5.385e67 seconds remaining. You're just about a third of the way done.
To pass the remaining time, start shuffling your deck of cards. Every billion years deal yourself a 5-card poker hand. Each time you get a royal flush, buy yourself a lottery ticket. A royal flush occurs in one out of every 649,740 hands. If that ticket wins the jackpot, throw a grain of sand into the Grand Canyon. Keep going and when you've filled up the canyon with sand, remove one ounce of rock from Mt. Everest. Now empty the canyon and start all over again. When you've leveled Mt. Everest, look at the timer, you still have 5.364e67 seconds remaining. Mt. Everest weighs about 357 trillion pounds. You barely made a dent. If you were to repeat this 255 times, you would still be looking at 3.024e64 seconds. The timer would finally reach zero sometime during your 256th attempt. Exercise for the reader: at what point exactly would the timer reach zero?
10:13 yugi-oh says hi
You beat me to it! LOL
Yeah, i think it would be cool to organise sts streamers world championship.
Where top streamers would play on the same seed and see who get most points.
First 4 rounda could be with 4 different characters, and then round 5 would be with character of choice.
We would have defect winner, silent winner etc, and then the overall winner in a person who got the most points total.
If you would organised something like this, i would be super excited
Baalor throwing shade on Blizzard by not mentioning it as an AoE Defect card 😢
Frost prime messed around with the spire PvP mod for a bit and had some fun. Can’t explain how it works it’s been a while and it basically re worked the entire game but the characters and cards
Had the AI spire argument with my flatmate awhile back. V good to hear your take!
Me when I eat boba
you could do multiplayer like super auto pets, you play a spire run and then have to fight a clone of someone elses run at certain points for example
a new video in the morning to pair with breakfast couldn’t think of anything better
Alright man, you seem to be on point with defect runs. He's my favorite character for sure. Got him to ascension 20 but keep dying on the double boss fight. Any good tips for a frustrated person lol?
You could do pvp in a segmented draft format and both players get the time eater mechanic. Health bars would need to be hyper inflated though. And you could still have some dumb stuff like burst+->bouncing flask+->catalyst+ for 216 poison. Also balance would be slightly wack like there's little counterplay against focus unlike strength
Another cool idea would be having a sort of "spire director" player control to some degree the node encounters and enemy AI.
There is precedence for first turn wins in TCG games. Magic the gathering have certain formats where its possible to get those of. But not anyone trying to compete will strive to do as the odds of pulling those combos off are very slim even on the best of days.
Imagine how loud I sighed when he said there was 4 aoe cards for defect….. only for him to not name the best aoe!!!!! BLIZZARD!! 😂😂
man, i would've taken those upgraded Hand of Greeds as soon as he got Membership Card. I'm surprised Baalor didn't.
Too weak
hand of greed is almost never worth buying in shops. you make the money back too slowly if you get single enemy combats, even upgraded
@@413xmb You're talking about the typical case. The scenario Baalor was presented with was early Act 2, Membership Card, and Molten Egg. It was a terrific cost-value for HoG here. Ultimately, bottling Seek here turned out to be extremely valuable but I don't think taking HoG was a wrong move either.
@@JoshuaKIM-qd2vd I think given the decks reliance on seek and the turbo/echo form or core surge/biased cog turn 1 starts, the HoG was still the worse choice. I agree that it was high value comparatively, but even still the starting capacitor and barrage combo was being leaned into so hard it might be difficult to actually land HoG without taking unnecessary damage
Even with membership card molten egg buying hog in act 2 is not that good, especially compared to bottling seek
I think Watcher also struggles against guardian, since her main strat is deleting defends and sometimes you don't get enough block density or consistency and she gets plapped.
The Barrage deck...
...It's blappening.
@14:40 - insane to me that you got past Gremlin Nob with that deck, that health, and that potion. Then went ahead and dominated the run later on.
Definitely not far away from having ML capable of solving spire, I'm afraid. Mastered Go already, and this was what, 10 years ago? Sequential decision making in such a controlled environment is certainly not that much of a challenge as it was 20 years ago. Noteworthy here is that despite the amount of game states it has playing ofc a huge role, what a deep neural net will do is learn a sufficiently expressive representation of a decision space where it can interpolate to never-seen states with somewhat good precision. My opinion at least xD
Well its kinda difficult and rather meaningless to compare the complexity but it sounds fun so i'll try. Feel free to correct me if i'm wrong.
Disclaimer: Now that I wrote all of this stuff I realised what he probably meant "DURING a game of chess" not "THE game of chess". But as each player has (without capturing pieces and without health) at LEAST 10 different options (100 per cycle), I'd argue chess has a bigger growing flow chart than the yawworm fight.
Tl;Dr: It takes approx. 36 turns or 18 deck cycles (250 turns forthe naive approach) for the Ironclad starterdeck vs jaw worm for the amount of possible gamestates to approach the amount of all possible gamestates in chess without capturing.
Chess:
I guess you can take the naive approach for chess. 32 pieces 64 squares makes it to 64!/32! (yikes) possible gamestates. Keep in mind that with this approach impossible situations are included (pawns on the first rank, impossible mates, passed pawns without any captures etc.). Further more all pieces are on the board (no captures and no promotions) plus identical positions are included (each pawn is considered differnt) whereby this is a VERY bad estimator.
The other approach would be to consider pawns first. Each pawn has 6-x (x being the opposite pawns position so xe{1,2,3,4,5}) possible squares. So, there are 15 different states and 15^8 possible states for all pawns. With the pawns out of the picture there are only 16 pieces, 48 squares remaining and consequently 48!/32! gamestates. So the new estimator ceils the amount of possible gamestates at 15^8*48!/32! (if you want unique possible gamestates, divide it by 4 as rooks and knights are interchangable). Dont get me wrong this number is still too high as impossible state are still included but its not wrong anymore by several orders of magnitude.
Jaw worm vs Iron clad starter deck:
Yes jaw worm cant thrash 3 thrice but lets say he can. Additionally, lets consider the case for the starterdeck without ascenders bane as it breaks the two turn cycle and gone after the first. 10 cards (5 strikes, 4 defends and 1 bash) with 5cards drawn each turn. there are therefore 10 possible cycle orders. Meanwhile the jaw worm has 9 different states per deck cycle (3 each). So all in all you get the amount of possible gamestates of 90^x/3 as the jaw worm can only chomp on turn 1(x means per deck cycle or every two turns). This approach doesnt take health into account but it still gets big fairly quickly.
Comparisson:
So how many cycles does the starterdeck need to approach the complexity of chess?
ln(3*15^8*48!/32!)= approx. 82
ln(90)= approx 4.5
=> x=82/4.5=18.2
So after 18 cycles or 36 turns the yaw worm fight has more possible game states than there are possible positions on a chess board and respectively 125 cycles for the bad estimator. Again, this doesnt take health into account but we also never considered the possibility of capturing pieces.
So yeah ... unlikely that you reach that, however, with branching paths and card rewards you should get there around mid Act 1.
ending with fairy and lizard tail. ezpz
What mod shows the mousover tooltip for ? Events and card percentages please?
If a 10 year old ai can beat professional dota players, i dont see why it couldnt play spire at at least a competent level. Maybe if you think about every card and possibility as an individual component, but if you just simplify cards potions and relics down to their basic mechanics: damage, damage over time, block, draw, etc. i think an ai could determine the relative likelihood of values (cards) to lead to success after a few million games, which is pretty feasible.
I dont know if i could ever agree that ai would never be able to master sts. Seems shortsighted to me
The thumbnail used to be two barrages
Fun fact Yu-Gi-Oh has a couple turn 0 wins so I think people will tolerate turn 1 wins
Setup / foreshadowing / the journey / payoff / ggwp
I dont understand the love for that FTL+. It's not a bad card but its upgrade has to be one of the least impactful upgrades in the whole game. +1 damage if its zero cost i guess is ok but it sure isnt pommel strike +. With the upgrade FTL is basically the same card. I dont believe being able to draw after one extra card played impacts the card much at all. It would need to gain at least 3 extra damage to be a good upgrade imo.
Kinda circumstantial but I like how the upgrade works with all for one. If you've got a ton of 0 cost cards and you draw all for one, to get the card draw from ftl (assuming it's in your discard) you can only play one other card before playing all for one. If you draw more than one zero cost card alongside all for one, you then have to chose between the ftl card draw and playing that second 0 cost twice. Not saying the upgrade is that great but I feel in this case it's giving you some decent value.
why is there no boss swap start offered?
Because he died last run not reaching Act I boss
First, I win, I'm the best
Congratulations 🎉
Thank you so much
6:55 Blizzard exists too?
The chat question was specific to act 2 and blizzard is a terrible act 1-2 card.
umm 49:50