Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[blueblimp]

Hypothetically, if you had 30 of the same card in a deck, what would be the best deck? Or would there be no one best deck because of a rock-paper-scissors effect? I'm sure this has been discussed on other Hearthstone forums but it's a fun puzzle.

Here's the all-around strongest I can think of:

30 Soulfires.

Because of life tap, starting turn 2 you can draw soulfire, play soulfire, draw soulfire, play soulfire, for 8 damage every turn.

As player 1, on turn 1 you have 4 soulfires in hand and can play 2 of them, so every turn of the game you do 8 damage. As player 2, on turn 1 you have 5 soulfires and 1 coin, so on turn 1 you can: coin, soulfire, soulfire, soulfire, tap, soulfire, to do 16 damage. Then 8 damage on each subsequent turn.

Absent opponent healing, that gives player 1 lethal on turn 4 and player 2 lethal on turn 3.

Counters: Because the soulfire deck has a per-turn damage cap, decks focusing on healing have hope, but they need to be very fast.

Healing touch is a certain counter, because beginning on player 1 turn 3 or player 2 turn 2, you can restore 8 health once per turn, completely negating the soulfires. If the warlock chooses not to soulfire then you win easily via hero power plus fatigue. If the warlock collects soulfires in hand, the most burst possible is playing 6 soulfires in a turn for 24 damage, not enough.

[Twistedarcher]

I think soulfire will win almost every matchup, but I think it would lose to lightwells (which in turn would lose to most other decks, which in turn would lose to soulfire

[nkirbit]

On the play, soulfire barely outraces claw, with the warlock player tapping down to 2 for the last soulfire needed for lethal.

[nkirbit]

Soulfire on the play mindlessly going to face doesn't outrace timberwolves, either.  Not 100% sure who wins if the soulfire player knows what he's up against and plays optimally... May be the wolves, though.

[blueblimp]

I think soulfire will win almost every matchup, but I think it would lose to lightwells (which in turn would lose to most other decks, which in turn would lose to soulfire

Interestingly, from the P1 seat, lightwells are too slow to counter the soulfire deck, but they work from the P2 seat.

P1: Lightwell
P2: Soulfire

P1T1: pass
P2T1: does 16 damage; P1 -> 14, P2 -> 28
P1T2: play lightwell
P2T2: does 8 damage; P1 -> 6, P2 -> 26
P1T3: lightwell heals; play lightwell; P1 -> 9
P2T3: does 8 damage; P1 -> 1, P2 -> 24
P1T4: lightwell heals x2; play lightwell x2; P1 -> 7
P2T4: does 8 damage; P1 -> -1, P2 -> 22
P2 wins.

The direct to-face soulfiring below is clearly too slow.

P1: Soulfire
P2: Lightwell

P1T1: does 8 damage: P2 -> 22
P2T1: play lightwell
P1T2: does 8 damage; P2 -> 14, P1 -> 28
P2T2: lightwell heals; play lightwell; P2 -> 17
P1T3: does 8 damage; P2 -> 9, P1 -> 26
P2T3: lightwell heals x2; play lightwell; P2 -> 15
P1T4: does 8 damage; P2 -> 7, P1 -> 24
From this point on, P2 is healing 9 or more per turn, which more than negates the soulfiring.

Soulfiring the lightwells can give you a chance to do slightly better. There's no point trying to kill the lightwells, because you need to spend your whole turn to kill one, but if you hit it once, then it may (depending on RNG) soak up two lightwell procs, denying 6 points of face healing for your 4 damage soulfire, for 2 points net damage improvement. It's not enough to give any chance to win, though.

[markusin]

Big necro, but I just realized that Voodoo Doctor beats Soulfire both in first player seat and second player seat. How come nobody brought that up? I expect Voodoo Doctor can win a much broader set of matchups because it fights back. It also beats Healing Touch and Lightwell for example.

[blueblimp]

Big necro, but I just realized that Voodoo Doctor beats Soulfire both in first player seat and second player seat. How come nobody brought that up? I expect Voodoo Doctor can win a much broader set of matchups because it fights back. It also beats Healing Touch and Lightwell for example.

Neat. Checks out. I tried it on paper and it seems to work for both P1 and P2, assuming the Soulfire player doesn't kill any of the Voodoo Doctors, which I think is correct play if the doctor player is warlock. If the doctor player _isn't_ warlock, maybe it can work to kill the doctors to run them out of cards for certain classes, not sure. I guess the intuition for why the voodoo doctors work well is that a single doctor only needs to get in a single attack to match the health differential of a Soulfire, and playing a doctor doesn't discard a card.

A hard counter to that: Elven Archers. It's not even necessary to calculate it out: archer costs the same as doctor, but easily 2-for-1s it, with the battlecry allowing an archer to reactively remove a doctor. Just use the battlecry until you have 7 archers on board, then for subsequent turns you can trade the 7th archer to make room and play another for the battlecry. The doctor player will never even be able to hit your face, while your archer army deals 6 damage per turn, which the healing doesn't match. You also won't need to tap as much because of the card efficiency.

[Grujah]

Against Voodoo and Archer: Weblord ? or Ghoul ?

[markusin]

Against Voodoo and Archer: Weblord ? or Ghoul ?

Some crazy rock-paper-scissors action going on here.

[KingZog3]

What about a deck of all Young Priestess? It would beat the archer or voodoo only I think, but not the soulfire deck.

[popsofctown]

The new Healbot 3000, used by a priest, has a 50/50 matchup against Soulfire and beats Archer and Weblord and Healing Touch.  Yay priest/warrior

[markusin]

The new Healbot 3000, used by a priest, has a 50/50 matchup against Soulfire and beats Archer and Weblord and Healing Touch.  Yay priest/warrior

Do you mean Antique healbot. If so, how does that beat Weblord when Weblord's effect stacks.

[popsofctown]

It doesn't, I just forgot what the card did

[popsofctown]

But at least my matchup got buffed above 50/50 against Soulfire!

[nkirbit]

Light of the Naaru has a lot of good matchups. 

[ycz6]

The Soulfire nerf definitely makes the metagame a lot more interesting.

Annoy-o-Tron, Micro Machine, Mistress of Pain, and Seal of Light all seem playable, though I think they all still lose to Soulfire.

Light of the Naaru loses to any minion with 5 or more attack, because the other player can just not attack until they have a full board. It does probably win against almost anything else, though...

[popsofctown]

How would Mistress of Pain be anything but utter garbage in the format?  A 1 mana 1/4 isn't going to get you much board position against anything besides voodoo doctor

[ycz6]

That's true. I guess I was imagining it getting to turn 5 and being able to heal for 4+ health every turn while also doing damage, but most decks will be able to remove or outright kill the Mistresses without losing much.

Is Unstable Portal any good? What percentage of the time does it win against, say, Bloodfen Raptor?

[popsofctown]

That's true. I guess I was imagining it getting to turn 5 and being able to heal for 4+ health every turn while also doing damage, but most decks will be able to remove or outright kill the Mistresses without losing much.

Is Unstable Portal any good? What percentage of the time does it win against, say, Bloodfen Raptor?

I think it would beat Bloodfen Raptor, but I think that's only because Bloodfen Raptor is susceptible to Mage hero power.

[HiveMindEmulator]

The Soulfire nerf definitely makes the metagame a lot more interesting.

I think the nerf must have been targeted at the 30-of metagame.

[popsofctown]

Was flare too strong against 30 ice armor too?

[popsofctown]

Hm, that explanation doesn't work, Hunter beats Ice armor by default.

[markusin]

The Soulfire nerf definitely makes the metagame a lot more interesting.

I think the nerf must have been targeted at the 30-of metagame.

I don't think it even beats eye-for-an-eye anymore. There are more Silver Hand Recruits than Soulfires.

[popsofctown]

The Soulfire nerf definitely makes the metagame a lot more interesting.

I think the nerf must have been targeted at the 30-of metagame.

I don't think it even beats eye-for-an-eye anymore. There are more Silver Hand Recruits than Soulfires.

If you hoard them in hand the right way I lean towards Soulfire but it seems unclear without doing the calculations

[heron]

Ice Block seems pretty good against non-healing decks.

[Grujah]

Ice Block seems pretty good against non-healing decks.

How? It loses to everything.

[markusin]

The Soulfire nerf definitely makes the metagame a lot more interesting.

I think the nerf must have been targeted at the 30-of metagame.

I don't think it even beats eye-for-an-eye anymore. There are more Silver Hand Recruits than Soulfires.

If you hoard them in hand the right way I lean towards Soulfire but it seems unclear without doing the calculations

In order to hoard them, you'd have to Tap while not using your Soulfires, but everytime you tap you lose 2 health. You can only use at most 5 Soulfires, 3 of which will activate Eye-for-an-Eye. the extra 2 Soulfires get cancelled out by your need to tap just to maintain the same hand size as laast turn, all the while there are Silver hand recruits that need to be addressed. The calculations are tricky.

[heron]

Ice Block seems pretty good against non-healing decks.

How? It loses to everything.

Well you have to use hero power too.

[KingZog3]

Ice Block seems pretty good against non-healing decks.

How? It loses to everything.

Well you have to use hero power too.

You will lose to fatigue against every deck. You can only have 1 of each secret in play, so Ice Block protects you once, then the fatigue just kills you since you can't trigger ice block on your own turn.

[markusin]

Ice Block seems pretty good against non-healing decks.

How? It loses to everything.

Well you have to use hero power too.

You will lose to fatigue against every deck. You can only have 1 of each secret in play, so Ice Block protects you once, then the fatigue just kills you since you can't trigger ice block on your own turn.

Riiight! So then Soulfire totally beats eye-for-an-eye. I was thinking you could have 3 stored at a time.

[Grujah]

We really should to a play-by-forum Tournament in this.
Probably Round Robin, or play-offs I guess. Theoretical only, of course (everybody submits decks, we play them out in optimal scenarios (and you probably know what your opponent plays before the match)).

3 points of you win as both player as P1 and P2, 1 each if it is a tie. Some exceptions if neither can win somehow. (like 2-0, and 1-1).

[markusin]

We really should to a play-by-forum Tournament in this.
Probably Round Robin, or play-offs I guess. Theoretical only, of course (everybody submits decks, we play them out in optimal scenarios (and you probably know what your opponent plays before the match)).

3 points of you win as both player as P1 and P2, 1 each if it is a tie. Some exceptions if neither can win somehow. (like 2-0, and 1-1).

To get the best results out of such a tournament, we'd want to have more than one tournament. Like the iterated prisoner's dilemma, the 30-of deck format becomes really interesting if we allow the local meta to evolve.

[Grujah]

Of course.
Its kinda like MTG 4-Card Blind.
http://www.starcitygames.com/article/25627_Bringing-Theory-To-Your-Local-Gaming-Store-4-Card-Blind.html

[markusin]

Of course.
Its kinda like MTG 4-Card Blind.
http://www.starcitygames.com/article/25627_Bringing-Theory-To-Your-Local-Gaming-Store-4-Card-Blind.html

Random effects always having the ideal result for your opponent is a good policy for this format as well. Applied to Unstable Portal, drawing Ancient Watcher should be ideal for your opponent pretty much all the time.

[Drab Emordnilap]

Weird. I always knew that as three-card Magic.

[popsofctown]

Does Soulfire beat 30 Mech mechwarper still? I'm not sure it does.

[markusin]

Does Soulfire beat 30 Mech mechwarper still? I'm not sure it does.

I think it loses against a Hunter with Mechwarpers. The question is, does ANYTHING beat Mechwarper?

[blueblimp]

Mechwarper vs Unstable Ghoul is interesting. Card-wise, one ghoul trades with one warper, two warpers just barely beat two ghouls (with the remaining warper at 2/1), and three ghouls kill any number of warpers. Mana-wise, warpers cost 2 for 1 and 3 for any greater number, while ghouls cost 2 each. Ghouls have taunt, so the warper player can't always choose to hit face.

It's complex even without fully calculating. Suppose there's one ghoul out and as the warper player you have two on board. If you don't hit, the ghoul player can just sit back because your warpers won't do face damage the next turn. If you do hit with one warper, then the ghoul player can trade off the ghoul, then but down another, which will finish off the remaining warper.

I think the ghoul player would have advantage in a long game, but before turn 6 and especially before turn 4, the warper player's ability to spam might be usable to take a decisive life advantage.

[markusin]

Mechwarper vs Unstable Ghoul is interesting. Card-wise, one ghoul trades with one warper, two warpers just barely beat two ghouls (with the remaining warper at 2/1), and three ghouls kill any number of warpers. Mana-wise, warpers cost 2 for 1 and 3 for any greater number, while ghouls cost 2 each. Ghouls have taunt, so the warper player can't always choose to hit face.

It's complex even without fully calculating. Suppose there's one ghoul out and as the warper player you have two on board. If you don't hit, the ghoul player can just sit back because your warpers won't do face damage the next turn. If you do hit with one warper, then the ghoul player can trade off the ghoul, then but down another, which will finish off the remaining warper.

I think the ghoul player would have advantage in a long game, but before turn 6 and especially before turn 4, the warper player's ability to spam might be usable to take a decisive life advantage.

I thought about Unstable Ghoul, but I'm not sure about it. It's even more complex because Ghoul's deathrattle hurts other Ghouls.

And then what if the Mechwarper player is a Mage? and what of the Ghoul player's hero?

[popsofctown]

Mechwarper vs Unstable Ghoul is interesting. Card-wise, one ghoul trades with one warper, two warpers just barely beat two ghouls (with the remaining warper at 2/1), and three ghouls kill any number of warpers. Mana-wise, warpers cost 2 for 1 and 3 for any greater number, while ghouls cost 2 each. Ghouls have taunt, so the warper player can't always choose to hit face.

It's complex even without fully calculating. Suppose there's one ghoul out and as the warper player you have two on board. If you don't hit, the ghoul player can just sit back because your warpers won't do face damage the next turn. If you do hit with one warper, then the ghoul player can trade off the ghoul, then but down another, which will finish off the remaining warper.

I think the ghoul player would have advantage in a long game, but before turn 6 and especially before turn 4, the warper player's ability to spam might be usable to take a decisive life advantage.

I don't think the Unstable Ghoul player can win.  It seems like at best you take a huge bullrush at the start, then put the board into stasis after that then die first to fatigue due to the early burst phase.

It's a 1/3 with a symmetric effect versus a 2/3 with no symmetric effect, I don't see why unstable ghoul would beat crocolisk.

[blueblimp]

Ghoul has taunt and takes two turns to die to a croc, killing the croc in the process. Meanwhile, the ghoul can hit face. That's why ghoul has a chance. Against croc, I'd think that ghoul always wins although I haven't worked it out, simply because I can't see how the croc player can ever hit face.

[popsofctown]

Ghoul has taunt and takes two turns to die to a croc, killing the croc in the process. Meanwhile, the ghoul can hit face. That's why ghoul has a chance. Against croc, I'd think that ghoul always wins although I haven't worked it out, simply because I can't see how the croc player can ever hit face.

But the croc can hit the ghoul once, then kill a totally unrelated ghoul, and the ghoul dies, but the croc lives.  It seems at least complicated to me.  And landslide for the Mechwarper.

The taunt is relevant, but the AoE 1 damage doesn't seem like it should really help much.  Anodized Robo Cup wins for sure.

[markusin]

Ghoul has taunt and takes two turns to die to a croc, killing the croc in the process. Meanwhile, the ghoul can hit face. That's why ghoul has a chance. Against croc, I'd think that ghoul always wins although I haven't worked it out, simply because I can't see how the croc player can ever hit face.

But the croc can hit the ghoul once, then kill a totally unrelated ghoul, and the ghoul dies, but the croc lives.  It seems at least complicated to me.  And landslide for the Mechwarper.

The taunt is relevant, but the AoE 1 damage doesn't seem like it should really help much.  Anodized Robo Cup wins for sure.

Against Mechwarper? I'm not so sure. Player 2 can flood the board with 6 Mechwarper by turn 2. Player 1 does it by turn 3. Maybe the cubs catch up in time?

[popsofctown]

Ghoul has taunt and takes two turns to die to a croc, killing the croc in the process. Meanwhile, the ghoul can hit face. That's why ghoul has a chance. Against croc, I'd think that ghoul always wins although I haven't worked it out, simply because I can't see how the croc player can ever hit face.

But the croc can hit the ghoul once, then kill a totally unrelated ghoul, and the ghoul dies, but the croc lives.  It seems at least complicated to me.  And landslide for the Mechwarper.

The taunt is relevant, but the AoE 1 damage doesn't seem like it should really help much.  Anodized Robo Cup wins for sure.

Against Mechwarper? I'm not so sure. Player 2 can flood the board with 6 Mechwarper by turn 2. Player 1 does it by turn 3. Maybe the cubs catch up in time?

No i meant robo cub beats unstable

[ycz6]

Started thinking about this again today, and tried mapping out Mechwarper Warlock vs. Murloc Tidecaller Warlock.

Mechwarper wins easily as player 2, since it floods with 6 minions on turn 2 and can trade down before the Tidecaller gets enough damage in.

As player 1, if Tidecaller always taps and goes face, and Mechwarper trades and taps aggressively, then Mechwarper stabilizes on turn 5 with 2 health, and is able to beat down the Tidecaller before card advantage becomes a problem. I don't think there's a better strategy for either player.

The Tidecaller wins as Hunter, though, and Timber Wolf is equivalent. But Mechwarper Mage wins against that.

[ycz6]

As for new cards...

BRM: Quick Shot beats Mechwarper Mage but loses to Warlock and Hunter. Twilight Whelp looks promising, but is really just a worse Mechwarper.

Revenge is actually quite strong, as it can stack armor and clear the board against almost any small minion. The only problem is that if the opposing deck can reduce it to 14 or less health on turn 30, which Mechwarper can, then it will lose the fatigue war, even against Warlock.

TGT: Power Word: Glory, like most removal, beats small minions. Tournament Attendee does too. Flash Heal is usually weaker than Light of the Naaru. Brave Archer can do 26 damage on turn 4 as player 1, but is usually just a vanilla 2/1 as player 2.

[Haddock]

Murloc warleader?

[KingZog3]

Murloc warleader?

Is it fast enough?

[popsofctown]

Murloc warleader?

Is it fast enough?

'

It's not very fast at all.

[Haddock]

Yeah I figured it would be slow, but maybe JUST fast enough against normal decks. Havent done the maths against other thirty-of decks, hoped someone might do it for me.

[Awaclus]

Yeah I figured it would be slow, but maybe JUST fast enough against normal decks. Havent done the maths against other thirty-of decks, hoped someone might do it for me.

Well, it loses against Bloodfen Raptor.

[Haddock]

Not in a face race it doesn't.
Can't be bothered to work out the trading possibilities.

Edit: forget that, you're right. Forget all of it.

[ycz6]

Here's a conversation I had on reddit with someone who seems to have really thought this through: https://www.reddit.com/r/hearthstone/comments/3ocf2h/tavern_brawl_idea_one_man_army/cvw5jdp

Some stuff I learned...

• Mind Blast, Murloc Tidecaller (Warlock/Hunter), Grimscale Oracle/Timber Wolf, and Mechwarper (Warlock/Mage/Druid) are the benchmarks. Any viable deck needs to be able to beat at least one of them; in particular, it needs to be able to do disrupt the opponent somehow by turn 4, and in practice turn 3.
• Of these, Mind Blast probably has the fewest counters. The only way to win is to outrace it, which very few decks can do from both player positions. Minion-based decks, on the other hand, can lose to any number of things.
• Abusive Sergeant is the best of the 1 mana 2/1s.
• Healing-based decks are weaker than they look, because they can usually be played around in fatigue. For example, Light of the Naaru actually loses to Mind Blast.
• If we only allow minions, then there are actually a ton of viable options. Cards like Deathlord, Unstable Ghoul, and Millhouse Manastorm suddenly become quite good in certain "metas."
• Flame Leviathan is insane.

[popsofctown]

What even beats Mind Blast? Healing Totem and Lightwell? Surely Doctor is too slow.

[chairs]

Mind Blast is pretty impressive. Light of the Naaru loses to it, but only in some funky ways.

[ycz6]

What even beats Mind Blast? Healing Totem and Lightwell? Surely Doctor is too slow.

Mind Blast does lose to Voodoo Doctor, it turns out. Any 2/1 for 1 can kill on turn 5 from either position, and Voodoo Doctor's heal lets it survive on turn 5 even if Mind Blast goes first.

Also, from https://www.reddit.com/r/hearthstone/comments/3ocf2h/tavern_brawl_idea_one_man_army/cvwivvp:

The following are all able to beat mindblast when going second: Abusive sergeant, dust devil, grimscale oracle, tidecaller, timber wolf, voodoo doctor, mechwarper with any hero power except warlock's, shaman's or paladin's (though priest, warrior and mage have trouble when going first instead), whirling zap-o-matic; there are a few others that also come very close (and would beat soulfire) though fall just short of beating mindblast. There might also be some I missed (like maybe a lucky unstable portal)

[chairs]

It seems like Hunter with timber wolf might be the best "general case" one, though it in turn loses to things that can disrupt it early.

[chairs]

I was wondering if, as "tech" against the TW/etc decks that are x/1 minions, you could run novice engineer? You'd be able to trade and board flood quicker than your opponent, but I'm not sure you could put enough damage on the field to finish out before fatigue kills you.

ED: with hunter, I'd imagine, to allow for increased face damage once you've hit flooded board.

[Awaclus]

I was wondering if, as "tech" against the TW/etc decks that are x/1 minions, you could run novice engineer? You'd be able to trade and board flood quicker than your opponent, but I'm not sure you could put enough damage on the field to finish out before fatigue kills you.

ED: with hunter, I'd imagine, to allow for increased face damage once you've hit flooded board.

Novice Engineer is only as good as Paladin's hero power.

[blueblimp]

Wouldn't Mind Blast only defeat Light of the Naaru if you know that's what it is? If you're P1 running Mind Blast, all you know on turn 2 is that your opponent is a Priest and didn't play anything on turn 1. For example, your opponent might also be running Mind Blast. MB mirror is normally a win for P1, but if P1 skips playing MB on 2, then it becomes a win for P2.

Code: [Select]

mind blast mirror (saving coin)

    2. mb; mb
    3. mb; mb coin mb
    4. mb mb; mb mb
    5. mb mb -> win

    or

    2. mb; mb
    3. mb; mb coin mb
    4. mb mb; mb heal
    5. mb heal; mb mb
    6. mb -> win

    so it's p1 win both ways

mind blast mirror (p2 saves coin, p1 skips first mb)

    2. ; mb
    3. mb; mb coin mb
    4. mb mb; mb mb
    5. mb mb; mb mb -> win

    so p2 wins if p1 skips a mb


[chairs]

I was wondering if, as "tech" against the TW/etc decks that are x/1 minions, you could run novice engineer? You'd be able to trade and board flood quicker than your opponent, but I'm not sure you could put enough damage on the field to finish out before fatigue kills you.

ED: with hunter, I'd imagine, to allow for increased face damage once you've hit flooded board.

Novice Engineer is only as good as Paladin's hero power.

Not true once you hit turn 4, where you can put 2 down; nor on turn 6, where you can put down 3... etc.

[chairs]

Wouldn't Mind Blast only defeat Light of the Naaru if you know that's what it is? If you're P1 running Mind Blast, all you know on turn 2 is that your opponent is a Priest and didn't play anything on turn 1. For example, your opponent might also be running Mind Blast. MB mirror is normally a win for P1, but if P1 skips playing MB on 2, then it becomes a win for P2.

Code: [Select]

mind blast mirror (saving coin)

    2. mb; mb
    3. mb; mb coin mb
    4. mb mb; mb mb
    5. mb mb -> win

    or

    2. mb; mb
    3. mb; mb coin mb
    4. mb mb; mb heal
    5. mb heal; mb mb
    6. mb -> win

    so it's p1 win both ways

mind blast mirror (p2 saves coin, p1 skips first mb)

    2. ; mb
    3. mb; mb coin mb
    4. mb mb; mb mb
    5. mb mb; mb mb -> win

    so p2 wins if p1 skips a mb


Ayup.

[Awaclus]

I was wondering if, as "tech" against the TW/etc decks that are x/1 minions, you could run novice engineer? You'd be able to trade and board flood quicker than your opponent, but I'm not sure you could put enough damage on the field to finish out before fatigue kills you.

ED: with hunter, I'd imagine, to allow for increased face damage once you've hit flooded board.

Novice Engineer is only as good as Paladin's hero power.

Not true once you hit turn 4, where you can put 2 down; nor on turn 6, where you can put down 3... etc.

You still wouldn't want to spend your turn 4 doing nothing but using Paladin's hero power twice.

[blueblimp]

The Blast/Naaru game theory seems pretty complicated. Let's imagine a meta where everyone is running Mind Blast, and everyone knows that everyone's running Mind Blast, and everyone knows that, etc. In this meta, running Mind Blast gives you a 50% win rate (win as P1, lose as P2).

Now let's imagine you switch to playing Naaru instead. As P2, you now always win because P1 will unknowingly play Mind Blast on turn 2. As P1, because you can't make a play on turn 2, P2 infers that you are running Naaru, as if you were running Blast then skipping turn 2 would be a losing play in the current meta; so P2 plays as if you are running Naaru--I think this leads to a win for P2 (by bursting during fatigue) but I haven't checked.

So your winrate after switching to Naaru is still 50%, which isn't better, but it also isn't worse. So there is nothing preventing some Naaru players from infiltrating the meta, at which point the mind games get super complicated in each game.

[blueblimp]

Here's my attempt at solving the Blast/Naaru metagame. TL;DR: Given those two choices, the equilibrium is to run Mind Blast at least 50% of the time (including 100%), and Light of the Naaru the rest of the time; when running Mind Blast as P1, always assume your opponent is running Mind Blast too.

To start the analysis, I think there are only three potentially-viable strategies: run Naaru, or run Blast and, as P1, either hold it until fatigue or play it on turn 2. Although I haven't quite proven it, intuitively I don't think there's any point as P1 to skip playing Mind Blast on turn 2 but not hold it all the way to fatigue; you already give up your certain win in the mirror by passing turn 2, so you might as well go all-in and play as if P2 is running Naaru.

Call these three strategies Bp (Blast-play), Bh (Blast-hold), N (Naaru). Each strategy has a 50% winrate against itself (only depending on which seat you draw). Bp wins 100% vs Bh and 50% vs N. Bh wins 100% vs N. Usual 2-player game theory can then solve for all equilibriums via linear programming, using this payoff matrix:

Code: [Select]

    Bp   Bh   N
Bp  0.5  1.0  0.5
Bh  0.0  0.5  1.0
N   0.5  0.0  0.5

It's not too hard to reason out the equilibrium without LP, using the fact that the optimal winrate must be exactly 50%, because of symmetry. Bp wins >=50% vs everything and 100% vs Bh, so Bh can't be played or always playing Bp would get better than 50% winrate. But if Bh is not being played, both Bp and N have 50% winrate not matter what proportion they appear. However N must be played no more than 50% of the time or else Bh would have >50% winrate vs the field.

[Grujah]

The Blast/Naaru game theory seems pretty complicated. Let's imagine a meta where everyone is running Mind Blast, and everyone knows that everyone's running Mind Blast, and everyone knows that, etc. In this meta, running Mind Blast gives you a 50% win rate (win as P1, lose as P2).

Now let's imagine you switch to playing Naaru instead. As P2, you now always win because P1 will unknowingly play Mind Blast on turn 2. As P1, because you can't make a play on turn 2, P2 infers that you are running Naaru, as if you were running Blast then skipping turn 2 would be a losing play in the current meta; so P2 plays as if you are running Naaru--I think this leads to a win for P2 (by bursting during fatigue) but I haven't checked.

So your winrate after switching to Naaru is still 50%, which isn't better, but it also isn't worse. So there is nothing preventing some Naaru players from infiltrating the meta, at which point the mind games get super complicated in each game.

This is why this theorycraft things like 3card blind in MTG usually use "open information" i.e. you know what your opponent plays.

[chairs]

I was wondering if, as "tech" against the TW/etc decks that are x/1 minions, you could run novice engineer? You'd be able to trade and board flood quicker than your opponent, but I'm not sure you could put enough damage on the field to finish out before fatigue kills you.

ED: with hunter, I'd imagine, to allow for increased face damage once you've hit flooded board.

Novice Engineer is only as good as Paladin's hero power.

Not true once you hit turn 4, where you can put 2 down; nor on turn 6, where you can put down 3... etc.

You still wouldn't want to spend your turn 4 doing nothing but using Paladin's hero power twice.

YOU DON'T KNOW ME MAYBE I WOULD

[blueblimp]

The Blast/Naaru game theory seems pretty complicated. Let's imagine a meta where everyone is running Mind Blast, and everyone knows that everyone's running Mind Blast, and everyone knows that, etc. In this meta, running Mind Blast gives you a 50% win rate (win as P1, lose as P2).

Now let's imagine you switch to playing Naaru instead. As P2, you now always win because P1 will unknowingly play Mind Blast on turn 2. As P1, because you can't make a play on turn 2, P2 infers that you are running Naaru, as if you were running Blast then skipping turn 2 would be a losing play in the current meta; so P2 plays as if you are running Naaru--I think this leads to a win for P2 (by bursting during fatigue) but I haven't checked.

So your winrate after switching to Naaru is still 50%, which isn't better, but it also isn't worse. So there is nothing preventing some Naaru players from infiltrating the meta, at which point the mind games get super complicated in each game.

This is why this theorycraft things like 3card blind in MTG usually use "open information" i.e. you know what your opponent plays.

That makes sense to me in the context of this thread. (If nothing else, it makes it a lot easier to compare decks.) The linked Reddit thread though is considering a hypothetical Brawl using this ruleset, and I think in that case you wouldn't know which deck your opponent is playing.

[ashersky]

Does Living Roots have a place in this conversation?

[ashersky]

Does Living Roots have a place in this conversation?

This thread is relevant with the current Brawl.

[markusin]

Does Living Roots have a place in this conversation?

This thread is relevant with the current Brawl.

So then, Maelstrom Portal should do well against any weeny minion deck right?

[blueblimp]

Soulfire was nerfed after this thread started (from 0 mana to 1 mana), and it seems bad now.

[pingpongsam]

Murloc Tidecaller seems to work well

[blueblimp]

Murloc Tidecaller seems to work well

Just beat it with Voodoo Doctor.

[ashersky]

Timber Wolf also works.

[blueblimp]

Timber Wolf also works.

It does? I'd expect Warlock Tidecaller to beat Hunter Timber, because the minions trade 1-for-1 and life tap gives you more cards to play.