So I've been playing around with the archetype. Unfortunately, I did not pull Vek'Lor but I may craft him if he truly is a solid card. In the end, I did include Fandral, but did not go all in on him. Only Choose One cards which are good on their own are in the deck. For now the deck is great fun but I did play a lot of ladder (only up to rank 9) so I don't really know if it is actually competitive.
Also if you;re interested, there is now a deck description attached to the deck.
Thalnos is very good combined with Swipe and it's card draw which I feel Druid's may be lacking after the ancient of Lore nerf. But it's a tech card and can be replaced by Big Game Hunter or (indeed) the Black Knight depending on the meta.
I'm also still on the fence about Klaxxi, a 4/10 for 4 is insane but by the time the condition is met (likely late game), I'm afraid the 5 extra health is not that relevant anymore. In any case Innervating it on t2 as a 4/5 is still a good play in standard, before GvG and the piloted Shredder you would see this a lot with Chillwind Yeti.
I guess it's a bit too early to say anything for sure, but this will be the first deck I'll try once WoG hits
Yeah I was already doubting if the Geomancer would be worth it. I like the idea of adding Brann Bronzebeard quite a lot, so I updated the decklist.
For now I'll keep the Druid of the Claw but I might include Crazed Worshipper depending on the meta. The extra stats on Druid of the Claw and its use as a charge minion seem more important to me than the possible C'Thun buff
I also replaced Cenarius for Nourish, I'm worried that the deck does not have enough card draw
Combo Druid is dead with the release of Whispers of the old gods. It seems that Blizzard is pushing two archetypes for Druid in the new set, Beast Druid and C'Thun Druid. When the expansion hits I'd like to try out the latter. I'll lay down down my thoughts on some cards and will post my first attempt at a deck. Feel free to comment on my decisions and propose better picks, in fact that's the entire purpose of this thread :)
I'll try to update it with my experiences on ladder next week.
For now the deck will have three themes and it will be legal for standard:
- Ramp
- Taunt
- C'Thun and its followers
To be honest: I don't think this will become a very competitive deck on ladder. It depends on the meta but I believe a mid-range C'Thun deck will perform better. Nonetheless it seems like good fun :)
Ramp:
Innervate: I think this is staple for ramp deck's, so no discussion there
Wild Growth: Good ramp card, not useless in the late game
Darnassus Aspirant: Is only valuable if it lives for a number of turns, for now I think this deck will not have early board control and I don't think this is a good pick
Mire Keeper: At the four mana spot it has strong competition so it will probably not make the cut
Nourish: Perhaps it is useful seeing that Ancient of lore will be nerfed and card draw might be necessary. I don't believe the ramping part will be used often
Even if you don't plan on playing Wild, buy it. The dust per gold ratio is much higher than normal packs which is very important once you've filled out most of your collection. And if you don't buy it now, you can never buy it. That said hold off on the rest of the wings until you've bought more packs, but at least buy the first wing so you can buy the others wings for dust later.
And hey, if you ever want to play Wild a lot of the cards it has are pretty important.
Naxxramas does not have have a higher dust per gold ratio. According to a meta study, on average, a pack will give you around 98 dust (Source), this means packs actually have a better dust to gold ratio then Naxxramas. Telling people to buy Naxxramas for the dust is bad advice in my opinion. More on this can be read in this thread.
However, I would still advice to get at least the first wing of Naxxramas to ensure you can buy the others in the future (as others above already pointed out).
Once standard hits you can only craft captain's parrot. If I'm not mistaking this also means you can dust him when standard hits (I would double check this if I were you). This means that the best course of action is to craft Bloodsail Corsair to get your parrots. You can always disenchant them for 2x100 dust, giving you a net profit 100 dust (and a free pirate :))
Good point, I don't think there is enough statistical proof to confirm it 100%.
But I do track my pack openings (since closed beta) and I can confirm that I've never gone more than 39 packs without a legendary. Admittedly, I have had only one case of going exactly 39 packs without a legendary (which was with GvG) and this is a very small sample size. For what it's worth, similar experiences are reported in this reddit thread.
I don't really like the use of the minimum amount of dust per pack (40) to compare crafting costs to gold/money. I'd rather use the average dust value of a pack, which is ~100 (source: Card pack statistics). Just take the example of the golden legendary. Almost everybody who opens 80 packs will get enough dust to craft 2 golden legendaries.
Also, suggesting to buy Naxxramas with gold in order to disenchant all the cards for 3140 dust is just plain bad advice. This is discussed at some length in this thread.
I think the overall message is pretty straightforward:
Purchase at least one wing of Naxxramas if you haven't done so already
Get Old Murk-eye
If you need GvG cards, buy packs now
Don't disenchant any (classic) cards. There will be some nerfs before standard hits and you'll be able to disenchant them for full dust value
And for those of you who track their pack openings: remember the pity timer. You are guaranteed to get a legendary if you haven't opened one in the past 39 packs. So if your last 30 GvG packs did not give you a legendary you're guaranteed to get a legendary in the next 10 packs. In those cases it may be interesting to buy GvG packs untill you open that legendary.
Nice explanation, very clear. I must admit that I'm (no longer) skilled enough in statistics to check whether you've covered every win condition. Did you take into account that for instance (3,6,0) is also a win condition?
In any case I ran the simulation again, you'll find the source code in the spoiler below. I'm assuming that bouncing blades only stops when a minion dies and thus continues even when your opponent has 0 or negative health. I don't know for sure if this is the case but this is how it works with Coldlight Oracle (in fatigue): if the first drawn card is lethal, then the second card will still be drawn.
The results are as follows (100.000.000 game simulations):
Opponent health at 2 --> odds of winning 98.76%
Opponent health at 4 --> odds of winning 95.47%
Opponent health at 6 --> odds of winning 90.01%
Opponent health at 8 --> odds of winning 82.69%
Opponent health at 10 --> odds of winning 69.51%
Opponent health at 12 --> odds of winning 52.83%
Opponent health at 14 --> odds of winning 36.70%
Opponent health at 16 --> odds of winning 23.63%
Opponent health at 18 --> odds of winning 14.12%
Opponent health at 20 --> odds of winning 7.55%
I can verify the top two cases analytically quite easily, they seem to match which gives me quite some confidence that simulation is accurate. Unfortunately there seems to be a mismatch between the calculated value (15.05%) and the simulated one (23.63%). Any ideas what's the root cause? feel free to check my code for mistakes, I make those every now and then :)
Yes I noticed this, with 'I do believe that the now updated version is correct.' I was referring to your correction. Although I must admit that I can imagine that this was not completely clear from my post. In any case your initial post had the correct reasoning, which is more important than a small error.
I understand that you stated your degree to give your post more authority. In general, it is my experience that people are not impressed by these kind of statements so I usually refrain from stating my degree. I'd rather let the evidence speak for itself, it is a much more powerful tool to convince people.
you only miss lethal if you hit thaurissan 4 times(1) or hit thaurissan 3 times and an axe flinger 1 time(2)
the chance of (1) is 1/3^4 = 1/81 the chance of (2) is 4*2/3*1/3^3 = 8/81
so you miss lethal which a chance of 1/81+8/81 = 9/81 = 1/9. which means you have a chance of 8/9 (88.888%) to kill him
edit: the people before me who calculated ~98% ignored the fact that there are 4 different ways to hit thaurissan 3 times and an axe flinger once: attt, tatt, ttat, ttta
P.S. I have a B.Sc in computer science
Although having a degree in playing hearthstone doesn't impress me much, the calculation is correct, short and simple. Edit: I was wrong, the explanation below is correct. Here we forgot in the "chance of (2)" to hit the emperor 4 times to kill him.
I agree, stating your degree in these cases always sounds like a argumentum ad verecundiam to me. The initial post was incorrect, I do believe that the now updated version is correct. The odds of winning are 95.47%
To wrap this up I wrote a simple Matlab script to simulate 10.000.000 cases of this problem and determined a winrate of 95.4721%. below the source code:
ngames = 10000000; wongame = zeros(1,ngames); axehit = 0; for i = 1:ngames
a=4; b=4; c=4; axehit=0; while a>0^b>0^c>0
hit = ceil(rand(1)*3+0); if hit == 1 axehit = axehit+1; a=a-1; elseif hit == 2 axehit = axehit+1; b=b-1; elseif hit == 3 c=c-1; end
You guys are underestimating the 'skill level' of these bots, they have reached legend in the past. I used to play against them (it was mostly the shaman bot) quite a bit before the big bot ban wave and I can tell you that they can be a tough opponent.
Read the thread below if you're interested in the topic (it's a bit long)
So I just tried to kill 10 million faerie dragons with arcane missiles and managed to kill 4,998,271, that's 49.983% Although I'm quite positive the 3/7 crowd will still not be convinced...
I think GvG opens up some possibilities for a mech druid. The new mech cards combined with Innervate will make some crazy opening plays viable, especially when combined with Savage Roar.
The main drawback of this idea seems obvious, the deck will run out of steam too quickly. Some card draw is really necessary or this deck will be dead in its tracks before it can deal lethal. Ancient of Lore is of course premium card draw which is available to druids but this may not be enough.
The idea is to add Gadgetzan Auctioneer to draw cards of spare parts, even after the nerf this may be viable. The chances of getting to turn 6+ seem quite well seeing the aggressive openings which are possible.
I also added one Force of Nature for the good old Force of Nature + Savage Roar combo. Rationale is that it's likely that the opponent is on 14 health or less on turn 9 due to your aggressive opening. In addition you'll do some heavy drawing in turns 6-8 so drawing into the combo by turn 9 is a possibility.
I don't know whether this concept is at all viable and I'm pretty sure the deck itself can still be fine tuned in many positions.
Main goal of this topic is to have some fun with a mechdruid and to see if this concept can be tweaked to be competitive. What do you guys think? Any feedback or suggestions is always welcome :)
0
So I've been playing around with the archetype. Unfortunately, I did not pull Vek'Lor but I may craft him if he truly is a solid card. In the end, I did include Fandral, but did not go all in on him. Only Choose One cards which are good on their own are in the deck. For now the deck is great fun but I did play a lot of ladder (only up to rank 9) so I don't really know if it is actually competitive.
Also if you;re interested, there is now a deck description attached to the deck.
What are your experiences?
0
Thalnos is very good combined with Swipe and it's card draw which I feel Druid's may be lacking after the ancient of Lore nerf. But it's a tech card and can be replaced by Big Game Hunter or (indeed) the Black Knight depending on the meta.
I'm also still on the fence about Klaxxi, a 4/10 for 4 is insane but by the time the condition is met (likely late game), I'm afraid the 5 extra health is not that relevant anymore. In any case Innervating it on t2 as a 4/5 is still a good play in standard, before GvG and the piloted Shredder you would see this a lot with Chillwind Yeti.
I guess it's a bit too early to say anything for sure, but this will be the first deck I'll try once WoG hits
0
Yeah I was already doubting if the Geomancer would be worth it. I like the idea of adding Brann Bronzebeard quite a lot, so I updated the decklist.
For now I'll keep the Druid of the Claw but I might include Crazed Worshipper depending on the meta. The extra stats on Druid of the Claw and its use as a charge minion seem more important to me than the possible C'Thun buff
I also replaced Cenarius for Nourish, I'm worried that the deck does not have enough card draw
0
Combo Druid is dead with the release of Whispers of the old gods. It seems that Blizzard is pushing two archetypes for Druid in the new set, Beast Druid and C'Thun Druid. When the expansion hits I'd like to try out the latter. I'll lay down down my thoughts on some cards and will post my first attempt at a deck. Feel free to comment on my decisions and propose better picks, in fact that's the entire purpose of this thread :)
I'll try to update it with my experiences on ladder next week.
For now the deck will have three themes and it will be legal for standard:
- Ramp
- Taunt
- C'Thun and its followers
To be honest: I don't think this will become a very competitive deck on ladder. It depends on the meta but I believe a mid-range C'Thun deck will perform better. Nonetheless it seems like good fun :)
Ramp:
Innervate: I think this is staple for ramp deck's, so no discussion there
Wild Growth: Good ramp card, not useless in the late game
Darnassus Aspirant: Is only valuable if it lives for a number of turns, for now I think this deck will not have early board control and I don't think this is a good pick
Astral Communion: I believe its too unreliable
Mire Keeper: At the four mana spot it has strong competition so it will probably not make the cut
Nourish: Perhaps it is useful seeing that Ancient of lore will be nerfed and card draw might be necessary. I don't believe the ramping part will be used often
Emperor Thaurissan: Technically not ramp, but worth including
Taunt
I will not list all the taunts of course, but only the ones I think are usable, if I'm missing any let me know
Class specific:
Druid of the Claw: Solid card, almost auto include in this kind of deck
Dark Arakkoa: This seems an auto include in a deck like this
Ancient of War: Susceptible to silence, might be useful if silence becomes less prevalent
Cenarius: I'm not sure if he fits very well, but I like him
Ironbark Protector: I'd rather pick Ancient of War over this
Neutral:
Sen'jin Shieldmasta: The good old shieldmasta might make a return when Sludge Belcher leaves the meta
Cyclopian Horror: This might be good depending on the meta
Defender of Argus: Good value, will help you survive
Crazed Worshipper: I'd rather run Druid of the Claw at the 5 mana spot
Sunwalker: Also a popular card before the arrival of Sludge Belcher but competes with Dark Arakkoa
Twin Emperor Vek'lor: Seems good in this kinda deck
Soggoth the Slitherer: Quite expensive but hard to remove
C'thun and its Followers:
C'Thun: included for obvious reasons
Beckoner of Evil: Good buff, but I don't really think this card will fit the ramp theme
Twilight Geomancer: Not sure if it's worth it, but might be nice as a one-off
Twilight Elder: Might be nice if innervated on t1 and also give the option to go t1 coin+wild growth
Disciple of C'Thun: This might be worth including, depending on the meta
Klaxxi Amber-Weaver: Insane if the condition is met
C'Thun's Chosen: Good stats, nice buff
Skeram Cultist: I would not run this over Dark Arakkoa
Doomcaller: Might be good in the control matchup, depends on the meta
Misc
Bloodmage Thalnos: Spell power + card draw, include if the deck ends up having spells
Wrath: Early removal and/or card draw, seems solid
Keeper of the Grove: Depends on meta, no longer auto include
Azure Drake: Perhaps good replacement of Ancient of lore
Ancient of Lore: For now I feel that it has become subpar after the nerf
Swipe: always seems solid
Starfire: I think it's too slow
Starfall: Depends on meta, if AOE spells are needed it could be included
Fandral Staghelm: I think he is cool but I don't know if he fits the deck that well
Reno Jackson: Seeing that we have this many options, it might be worthwhile too include him. Although I think that it would destroy reliable ramping
Sylvanas Windrunner: Always Solid, might be a good inclusion vs Control
Big Game Hunter: Still a good tech card in the right meta
Eater of Secrets: I don't think Secrets will be that relevant in standard, but if I'm mistaking this is a good tech card
The Deck for now:
Let me know what you guys think!
2
0
Once standard hits you can only craft captain's parrot. If I'm not mistaking this also means you can dust him when standard hits (I would double check this if I were you). This means that the best course of action is to craft Bloodsail Corsair to get your parrots. You can always disenchant them for 2x100 dust, giving you a net profit 100 dust (and a free pirate :))
0
I personally like Aviana Malygos Druid:
Aviana+Malygos+Innervate+Faceless Manipulator+Faceless Manipulator+Moonfire+Moonfire.
Innervate can be substituted with Emperor Thaurissan (2 reductions on Aviana). And the Moonfire can be replaced by Living Roots if you have another Innervate or if they are reduced by Emperor Thaurissan. One Faceless Manipulator can be replaced by Alexstrasza.
0
Good point, I don't think there is enough statistical proof to confirm it 100%.
But I do track my pack openings (since closed beta) and I can confirm that I've never gone more than 39 packs without a legendary. Admittedly, I have had only one case of going exactly 39 packs without a legendary (which was with GvG) and this is a very small sample size. For what it's worth, similar experiences are reported in this reddit thread.
14
I don't really like the use of the minimum amount of dust per pack (40) to compare crafting costs to gold/money. I'd rather use the average dust value of a pack, which is ~100 (source: Card pack statistics). Just take the example of the golden legendary. Almost everybody who opens 80 packs will get enough dust to craft 2 golden legendaries.
Also, suggesting to buy Naxxramas with gold in order to disenchant all the cards for 3140 dust is just plain bad advice. This is discussed at some length in this thread.
I think the overall message is pretty straightforward:
And for those of you who track their pack openings: remember the pity timer. You are guaranteed to get a legendary if you haven't opened one in the past 39 packs. So if your last 30 GvG packs did not give you a legendary you're guaranteed to get a legendary in the next 10 packs. In those cases it may be interesting to buy GvG packs untill you open that legendary.
0
Nice explanation, very clear. I must admit that I'm (no longer) skilled enough in statistics to check whether you've covered every win condition. Did you take into account that for instance (3,6,0) is also a win condition?
In any case I ran the simulation again, you'll find the source code in the spoiler below. I'm assuming that bouncing blades only stops when a minion dies and thus continues even when your opponent has 0 or negative health. I don't know for sure if this is the case but this is how it works with Coldlight Oracle (in fatigue): if the first drawn card is lethal, then the second card will still be drawn.
The results are as follows (100.000.000 game simulations):
I can verify the top two cases analytically quite easily, they seem to match which gives me quite some confidence that simulation is accurate. Unfortunately there seems to be a mismatch between the calculated value (15.05%) and the simulated one (23.63%). Any ideas what's the root cause? feel free to check my code for mistakes, I make those every now and then :)
ngames = 10000000;
gamestat = zeros(ngames,3)*NaN;
axehit = 0;
a_init = 4;
b_init = 7;
c_init = 4;
for i = 1:ngames
a=a_init; % health axeflinger 1
b=b_init; % health axeflinger 2
c=c_init; % health Thaurissian
axehit=0;
while a>0^b>0^c>0
hit = ceil(rand(1)*3+0);
if hit == 1
a=a-1;
elseif hit == 2
b=b-1;
elseif hit == 3
c=c-1;
end
end
gamestat(i,:)=[a b c];
end
%damage = (start health A+start health B - health A and B at end)*2
ndamage = (a_init+b_init-sum(gamestat(:,1:2),2))*2;
bins=0:2:20;
[n,x]=hist(ndamage,bins);
nc = cumsum(n);
fac = 100/length(ndamage);
disvec=nc'*fac;
disvec2 = 100-disvec;
figure
hold on
bar(x,[100; disvec2(1:end-1)],'Linewidth',2)
ylim([0 100])
xlabel('Life of opponent')
ylabel('chance of winning')
set(gca,'XTick',bins);
set(gca,'YGrid','on')
0
Yes I noticed this, with 'I do believe that the now updated version is correct.' I was referring to your correction. Although I must admit that I can imagine that this was not completely clear from my post. In any case your initial post had the correct reasoning, which is more important than a small error.
I understand that you stated your degree to give your post more authority. In general, it is my experience that people are not impressed by these kind of statements so I usually refrain from stating my degree. I'd rather let the evidence speak for itself, it is a much more powerful tool to convince people.
1
ngames = 10000000;
wongame = zeros(1,ngames);
axehit = 0;
for i = 1:ngames
a=4;
b=4;
c=4;
axehit=0;
while a>0^b>0^c>0
hit = ceil(rand(1)*3+0);
if hit == 1
axehit = axehit+1;
a=a-1;
elseif hit == 2
axehit = axehit+1;
b=b-1;
elseif hit == 3
c=c-1;
end
end
if axehit > 1
wongame(i)=1;
end
end
winpercentage = sum(wongame)/ngames*100
3
You guys are underestimating the 'skill level' of these bots, they have reached legend in the past. I used to play against them (it was mostly the shaman bot) quite a bit before the big bot ban wave and I can tell you that they can be a tough opponent.
Read the thread below if you're interested in the topic (it's a bit long)
http://www.hearthpwn.com/forums/hearthstone-general/general-discussion/19510-the-botting-problem-in-hearthstone-discussion
2
So I just tried to kill 10 million faerie dragons with arcane missiles and managed to kill 4,998,271, that's 49.983% Although I'm quite positive the 3/7 crowd will still not be convinced...
For those interested, I used Matlab:
close all;clear all;clc;
ntimes = 10000000;
nhits = 3;
health = 2;
for j=1:ntimes
h=health;
for i=1:nhits
if h >0
h=h-round(rand(1));
end
end
health_vec(j) = h;
end
number_killed=sum(health_vec==0)
probability = number_killed/ntimes
0
Hi Guys,
I think GvG opens up some possibilities for a mech druid. The new mech cards combined with Innervate will make some crazy opening plays viable, especially when combined with Savage Roar.
The main drawback of this idea seems obvious, the deck will run out of steam too quickly. Some card draw is really necessary or this deck will be dead in its tracks before it can deal lethal. Ancient of Lore is of course premium card draw which is available to druids but this may not be enough.
The idea is to add Gadgetzan Auctioneer to draw cards of spare parts, even after the nerf this may be viable. The chances of getting to turn 6+ seem quite well seeing the aggressive openings which are possible.
I also added one Force of Nature for the good old Force of Nature + Savage Roar combo. Rationale is that it's likely that the opponent is on 14 health or less on turn 9 due to your aggressive opening. In addition you'll do some heavy drawing in turns 6-8 so drawing into the combo by turn 9 is a possibility.
So without further ado, here's my first attempt.
I don't know whether this concept is at all viable and I'm pretty sure the deck itself can still be fine tuned in many positions.
Main goal of this topic is to have some fun with a mechdruid and to see if this concept can be tweaked to be competitive. What do you guys think? Any feedback or suggestions is always welcome :)