The amount of snowball potential with an unanswered Shadowjeweler Hanar is mind bogglingly degenerate.

Quote from Wurstfett >>

Why is everyone here talking about combo, control and aggro... but not about (true) midrange? Is this archetype really so dead at the moment as it looks like?

 There are actually quite a few very strong midrange decks atm: even shaman, midrange shaman, spell hunter, and deathrattle hunter.  These are just in the tier 1-2 range too.  Not all of them are the classic 'play on curve' midrange maybe you're used to, but do fit the description of midrange.

Quote from SlydE >>

Quote from Czhorat>>

Quote from SlydE>>

That is not so impressive tbh, they have a long way to go, but they should get there!

It puzzles me that nobody has set up bots to grind against eachother to optimize decklists and plays yet. At least for tournaments with known decklists, that should be very useful and not too difficult. Pro chess players use computers all the time, so it should not be considered cheating. Computers should be able to completely nail the draws and rng elements of the game much better than humans, and never do serious misplays.

 That's a likely next step; deep learning algorithms should be able to maximize results. I suspect that part of the problem is that the speed of games is limited by animation time; you can have two computers play chess against eachother as fast as they can process. Two AIs squaring off in Hearthstone probably can't play the thousands of games they'd need to effectively learn without spending a prohibitive amount of time on it.

The games would be played out in a simulator ofc, just like chess computers don't need an actual chess board and chess pieces. The math behind hearthstone is pretty manageable, even stuff like "summon a random minion" and "craft a beast" has a limited number of outcomes that can be calculated. I don't even think deep learning would be needed for learning to play the game close to perfectly, I would be more interrested in how the computers build decks!

 This might be incredibly useful for game balance purposes, creating very deep evolving metas perhaps (think SSBM, only intentional) or even npc logic.  Unfortunately, using this learning system against itself could solve it equally as well.  Maybe the future of a game like Hearthstone is just single player, so people don't have the incentive to use tool/machine assisted systems?

In short, single players games could take on the robust challenge of what multiplayer now serve for players.  Clearly the end game is a realistic sim, but now I'm getting ahead of myself...

Quote from Bee >>

Quote from mur72 >>

Quote from Bee >>

it's not possible for a game like Hearthstone

 Ofcource it is possible. RNG in Hearthstone ofcource gives results from left and right, but in the long run with millions of games played, the AI can be victorius against even the best. Especially when the decks are known. After beating the best at GO, which was thought to be impossible ten years ago, ofcource machine intellect will beat humans in whatever game in the long run. Hearthstone first, then MtG and other more complex games with randomness.

It's impossible to always make the best play because you don't know what your opponent has in his hand. For example; you need to play your minions and you can either play around defile or hellfire, not both. You have to make a choice and in the end it all comes down to RNG.

That's why making a computer that ''can't'' be beaten in a game like Hearthstone is impossible. Also keep in mind that it'd be near impossible to code it against every possible archetype.

 Some of this is true.  It's not impossible to make the 'best' theoretical play, because the best play is simply the play with the best odds.  If you have mountains of data and thousands+ of years of training data (as in the example with OpenAI's dota 2), you can actually theoretically find the most effective strategies and which play is technically correct in every scenario.  It is impossible to know the opponent's hand, certainly.  And no matter how perfectly the system plays, it will be a victim to bad luck like everyone else.

'Coding against every possible archetype' is basically what's happening - in that a neural network is an algorithm, but taking in thousands or millions of years of training data will accommodate nearly every deck permutation - within a very small margin of error.

Again, if a large organization like OpenAI was working on Hearthstone, I'm fairly certain this game would have been solved by now.  It's very basic compared to Dota 2. 

Quote from Somm >>

I beat legend rank players playing on my phone while watching Netflix so It is not impressive if a bot con do It.

  There is a big difference between a bot in the classical sense (has access to in game data and an extremely narrow skillset) and unsupervised learning using generalized machine learning algorithms (neural nets, vision, controlling a keyboard and mouse even?).

The final straw in the camel's back will be when these systems can experiment and optimize their own decks and gather hundreds of years of training data in a single day or even hours.  That's when the HS meta is solved on day one moments after release...

Still, not too far off most likely... which is a bit scary to me.

Quote from Czhorat >>

It and that Hearthstone can join chess and Go in the panties pantheon of games in which the best humans can't beat the best computers.

 

https://thenextweb.com/artificial-intelligence/2018/08/15/researchers-teach-hearthstone-bot-to-dominate-legend-rank-players/

 

Obligatory XKCD https://xkcd.com/1002/

 

 

 The fact that OpenAI has a good, but not total grasp, on Dota 2 is probably telling of where we're at with AI in regards to multiplayer gaming.  Dota 2 is a much better test of skill than a game like Hearthstone, which is probably why it was undertaken by this venture first.  It's probably another few years before anyone (including OpenAI), attempts to make this technology more universal and other companies start getting involved.

I really have no doubts that Hearthstone is already a solvable game by these types of systems.  The digital CCG multiplayer games are no further than a few years away from atleast having a company like Blizzard collecting data  in this way for game balance - and perhaps sophisticated bots playing perfectly (which would be the end of a game like Hearthstone with its matchmaking currently).

EDIT: just read the article - probably faster than my time estimates then haha

Quote from DiamondDM13 >>

Quote from SnipaElite >>

 You actually live in some kind of weird fantasy land - argue constantly by cherry picking, entirely avoid pain points in your arguments, constantly use anecdotal evidence to support your claims, wander off topic, and just slop on lazy arguments ad nauseam in each post.  You very rarely get upvotes, which isn't a suprise, given how weak your logic generally is.

Nothing in your latest post describes me for instance.  What a lazy argument!

I didn't avoid anything, I made quite a long post replying to your previous post, I don't know if you bother to read it, but picking my post directed to a different user which, being a parallel conversation, obviously deviates from the conversation we were having, might result in that post not containing the reply meant for you or our conversation...

 

 You'd have to pay me good money to point out all the flaws in that vast bird-walk of a post.  You have heart, but your logic meanders so much it's painful and dull to read.  The discussion wasn't about skill, so you wasted your time and created another filler post. 

Quote from DiamondDM13 >>

Quote from SlydE >>

Not to mention, Shudderwock decks were mid tier decks before the Cubelock nerfs.

It is curious that all while the OTK decks get all the hate, the old fast snowball decks are the ones which win the most games... If Control decks can just kill all of those, an occational loss OTKs should not hurt as much.

 Well, that is kind of because people want to win all games, against all playstyles, regardless of natural weaknesses of their own playstyle. People don't want to lose, and that is natural, no one likes to, but some people can't accept that they will always be unfavoured at one point or the other and have an extremely hard matchup they will probably not win. I just did a few games with a friend, I was playing Token Druid against his Control Warlock, he asked me to play that matchup. I won a game and lost the other two, when I'm expected to lose almost every single game. Obviously doesn't feel good to lose, but reality is, with an average draw, the class of both decks will always favour the Control Warlock, and if both play correctly, that is even more certain. Player decisions should influence the outcome of the game, but if we both play correctly, he will win.

I think people just don't want to accept that they can play a game correctly in an unfavourable matchup, and lose regardless. That should be a given, since if the opponent always plays correctly, they are favoured to win. It should only be frustrating when you lose playing properly and the opponent playing incorrectly, not when both play correctly.

 You actually live in some kind of weird fantasy land - argue constantly by cherry picking, entirely avoid pain points in your arguments, constantly use anecdotal evidence to support your claims, wander off topic, and just slop on lazy arguments ad nauseam in each post.  You very rarely get upvotes, which isn't a suprise, given how weak your logic generally is.

Nothing in your latest post describes me for instance.  What a lazy argument!

Quote from danklesminky >>

Quote from Infirc >>

i'm honestly at a loss when people say "control deck" and then complain about decks having a big win condition, seriously can someon update me in wtf control means for this community?

A control deck is just a deck that's mostly reactive instead of mostly proactive.

 Yeah, I mean in reality cards like Shudderwock have really blurred the line.  After WW, this was basically a powerful OTK combo in a fairly strong control shell.  There isn't much space for a control deck without a strong win condition now - which makes them more like hybrids.

Quote from DiamondDM13 >>

Quote from SnipaElite >>

 There's a lot to unpack here, so I'm just going to address the 90/10 10/90 problem. 

1.) If you have 10/90 you have 90/10 matchups.  Neither of these outcomes are healthy for the game.  While it should technically be possible to construct a deck with these odds, the power curve should not favor them.  If you have a 10% chance of winning a game, it is almost entirely out of your control to win by definition - unless you're playing an incompetent opponent (knowing you I have to clarify this point).  The odds speak for themselves, this kind of matchup devolves into a 'which deck did my opponent roll?'.  Changing the odds to a 70/30 and 80/20 is totally anecdotal.  Try playing a control warrior against mecha'thun priest and tell me how you improved your odds.  Ofc, the same principles apply in the inverse situation - 90/10.

2.)  The community likes 50+ % matchups instead of 90/10.  If someone actually wants this, they aren't thinking about the health of the game.  If the power curve prefers decks that are more balanced (but still slightly unfavored against their respective archetype), there is a greater chance that player decisions matter.  Especially when we consider the control vs OTK dynamic, where the OTK just draws their whole deck in 9/10 turns and wins.

I get the feeling I'm arguing against someone playing devil's advocate - it feels disingenuous.

 1.) Yes, it is mostly out of your control to win those matchups. Except for when your opponent fucks up, or when you someone find the one path that can win you the game. What 90/10 matchups tell you is that the strategy of these two decks heavily favours one of them when facing each other, but that is entirely fine. Due to the numerous possibilities when building a deck, this kind of situation will always happen and it is HEALTHY that it does happen, even if some people don't like it.

This situation only happens because the game allows both of the strategies to exist and that is excellent, a card game should allow as many strategies as possible, that is why people play card games, the numerous amounts of possible combinations and strategies you can play with the hundreds or thousands of cards that are printed.

Obviously, when players play two extreme strategies, their matchups will lead to extremely results. If you make a deck which goes full all in on a specific strategies, a specific win condition, any deck which can withstand that win condition will hard counter you, and any deck whose strategy is naturally weak to that win condition, will be hard countered by you.

You cannot build a perfect, unbeatable deck. You have to make choices while building a deck. Sometimes, like in the case of OTK decks or Fatigue decks, two very extreme win conditions, you will give up a lot of your matchups just to win against what you hard counter. It is part of balance. Decks with extreme win conditions tend to always have extreme matchups, this is healthy, regardless of whether or not you particularly enjoy that, it is healthy and balanced. It is part of having a very wide card pool and a very limited amount of slots to pick cards for.

2.) I'm not sure what you mean by the community. Hearthpwn? Hearthstone's playerbase? There's 70+ million registered accounts for this game, some of which, like you, don't enjoy hard counter matchups. But do not confuse that with balance. The existence of hard counter matchups doesn't cause problems for the game's health or balance. They exist because people have the option to play extreme win condition strategies and they enjoy doing so. Believe, if people didn't like these strategies, if they didn't want them and the hard counter matchup dynamics that come with them, people wouldn't be playing them. So it is not the community that wants 50+ matchups, it's some that want that, and some that don't particularly feel the need for it.

This is a strategy card game, ideally, the sole two factors determining games would be your strategy and your technical play. Strategy being the deck you prepared, and technical play being how well you piloted the deck and according to the dynamic of the matchup you are facing.

Obviously, you also have card draw order, which is a random element which decides games. It is part of most card games and we accept and deal with it, otherwise we wouldn't be playing a card game.

Finally we have the random effects, this is not present in all card games, but Hearthstone is HEAVILY focused on these. These determine far more games than they should be allowed to. Still if we continue to play the game despite knowing this, we accept that.

What determines how influential player decision is doesn't change because of the hard counter matchups. It limits it, but you have far greater problems in the random aspects above. And 50+ Decks are not anymore balanced than extreme win condition decks. They are both balanced regardless of whether one can have extremely good and bad matchups.

You will not find any 50+ Deck that involves a minimally specific win condition. Any deck whose win condition diverges from the basic Midrange strategy will be a deck that involves matchups with winrates going even to the 70/30s. That is because the strategy for Midrange doesn't tend to change much no matter what you are facing, and because the Midrange strategy doesn't tend to be particularly weak or strong against anything (Sure, it can be particularly weak to Aggro and strong against Control based on it's constitution, but you get the idea).

I'm not playing Devil's Advocate, I simply speak to you from a different perspective. I have personal preferences, but I also have the ability to look past those. I dislike Aggro strategies, I dislike Midrange strategies even more, but I'm not going to be posting about how unfair the Aggro vs Combo dynamic is just because I happen to not like Aggro and like Combo. I understand the importance of both, I understand the need for multiple strategies to exist, I know how important that is specially for a game like Hearthstone trying to appeal to as wide an audience as possible. I also understand that the Rock-Paper-Scissors dynamic is perfectly natural in a balanced game, regardless of how frustrating it is when playing Ladder.

A part of the community, the one you are part of, doesn't seem to understand this. I know it can be frustrating to enter a game and know from the start that you have a very slim chance of winning that matchup, no one is stating the contrary, but wanting Blizzard to remove this is completely different. If Blizzard were to remove all these extreme win conditions from the game and make the only strategies possible in the game be limited to bland Midrange decks which can attain an average winrate across the board, do you even imagine how many people would leave the game?

Believe it or not, there are a ton of people that want to play more interesting strategies than simply Midrange. Hell, if you like to play Control, which is likely considering your dislike for OTKs, you would leave the game as well. Control, as a strategy, doesn't really have an average winrate across the board, it has very strong and weak matchups even removing Combo from the picture. You can easily be on the 70/30s or 80/20s against Aggro decks and the opposite against Midrange decks. Do you also consider this an unhealthy, unbalanced problem? No, it is just what happens when your strategy diverges from the most basic standard strategy the game can have. The more you deviate from the basic strategy of the game, the more extreme matchups become.

You're a guy who prefers quantity over quality.  None of your arguments are very compelling, and you aren't even able to follow your own points through.  For instance, in your wall of text you attack me for using the word 'community', when that was, in fact, a paraphrase of your previous point.  It's like you're asking yourself what you meant?  Reflect on that - and choose your points more carefully.  More isn't better and doesn't improve your stance.

It is clear based on our discussion here that you're the close minded player, and yet you assumed I was.  I actually enjoy all the archetypes equally (except midrange more than aggro).  I love control, I love the more nuanced OTK decks, and I love midrange.  As an example, I have hated midrange for instance if the game plan is just to play on curve, and for that meta will maybe prefer one of the other two.  Tons of strategies can exist in my paradigm, and they would flourish more than using your vague notion of what can exist.  I think greedy strategies are certainly necessary to exist and naturally exist in pretty much every game - they are what fundamentally drive the more balanced decks to include the cards they do.

I hate losing to turn 5 aggro as much as anyone, but it needs to be fairly strong to maintain a healthy ecosystem.  Certainly, the best metas in HS that I can remember are true to this.

The ideal power balance is probably something like the attached crude graph I quickly drew up.

Quote from DiamondDM13 >>

Quote from SnipaElite >>

 Absolutely, interaction has a lot of different successful applications - including tempo as you mention.  As long as it exists as a viable counter strategy, I find that acceptable.  This is almost always highly class limited in HS though, depending on the expansion.

I play w/e classes I find most interesting in an expansion, and would ideally consider all of them interesting.   If a class feels extremely limited by these oppressive strategies (like Jade or Odd Pally), then it's a less interesting meta.  I generally play at Legend or near Legend rank, so at this level I have to be playing highly optimized classes/lists.  In some cases, you're lucky to have 1-2 more balanced strategies.

Greedy decks rarely if ever appeal to me, because their ability to adapt to a matchup is just so limited - so any meta that feels too slanted in this way I'll just sit out.  Grinding out a greedy meta has to be the most mind-numbing experience to everyone right?  This is why I can't imagine how you can refute my earlier points - skill/decision-making rarely comes into play if a deck has a ton of 10/90% win matchups. 

 Well, sometimes yes, it is limited by class, but they can always do something, there tends to always be a slight chance no matter what class, it can be just extremely hard to identify what is that very low % line of play that gives you the win. Also, regarding how I beat the Shudderwock Shamans, I just froze their boards in the pre-Shudderwock turns which meant they either were forced to Overload themselves killing their board, or they had to risk not getting Shudderwock back because the Grumble would send Shudderwock back as the last minion. I just had to push them to the point they had to actually to Shudderwock (which is very hard as a deck like Freeze Mage).

Well, playing at Legend Rank you can easily play whatever gimmicks you want, specially if you are not grinding for HCT Points. It's the Rank I've played the most crappy gimmicks at in the entire ladder. Same cannot be said about 4 to 1 since there, you don't survive in that manner.

Well, but don't forget that OTK decks are also greedy. OTK decks tend to be the greediest of Combo decks and pay for it with little to no flexibility. Take Infinite Burn Mage and normal Freeze Mage. They are similar in game plan, but their win condition is different and that difference makes one very versatile while the other is as far as it gets from it.

Freeze Mage can adapt it's strategy more easily to it's matchup, it can go for high burn plans, it can go for exhaustion, it can even go for minion tempo in some cases and matchups, it is quite a flexible deck when you think of it, and if that doesn't seem real, trust, I've played that deck for thousands and thousands of games, while most games you win by the normal game plan, a great chunk will be won by different, adapted game plans. Infinite Burn Mage however, it is extremely rare the occasions in which you actually manage to beat the opponent without the 5 piece combo. The deck is not flexible, it's entire game plan is tailored to execute that game plan and if that plan is not possible, you will almost always lose.

Still, when you think about it, Infinite Burn is a much, MUCH stronger win condition than the normal Freeze win condition. One deals infinite damage solely limited by the turn timer, the other deals limited burn or minion damage. Infinite Burn is a much greedier plan, it is much more all in, not flexible, but it has a practically guaranteed win when executed.

Greedy Metas simply mean there is less Aggro punishing these Greedy Combo decks and when that happens, they can easily punish Control, and the more greedy the Control decks get, the easier it is to punish them. Still, we just started this expansion, and what you tend to see most at the start is very greedy stuff, and very strong aggro stuff. People either wanna play with the new fun strategies, or they want to take advantage of those playing these fun strategies.

And decision making does come into play even in 90/10 matchups. It's not just a coincidence that Control Warrior is one of my favourite Freeze Mage matchups. It is the hardest matchup, the most hard to beat, which makes it a fun challenge to try and complete, one I completed far more than one would expect. Generally, Skill/Decision making tends to help turn those 90/10 matchups in your favour. Sure, you don't make them positive, but you won't be losing 90/10, you'll get them to 80/20 or 70/30, which is already very good.

Matchup winrates come from how both strategies interact. The reason Freeze Mage is fucked by Control Warriors is because the Mage decides to mostly give up on all minion damage, all those constant damage sources, and solely uses Burn, which is a single use source. The moment Warrior has more Armour than the Burn it wins. The more compromises you make with your deck, the most likely you are to have very difficult matchups. That is not a design flaw, that is just how balance works.

If you have 90/10 matchups, you also tend to have 10/90 or close matchups in your favour. If you take the Control Warrior, the Warrior has the 90/10 against the Freeze Mage, but it also loses that badly the moment you play something like Jade Druid (pre-Geist). You had an extremely good matchup and an extremely bad matchup.

It seems sometimes people just don't think about this, yes, they are getting destroyed in 10/90 matchups, but there are also 90/10 matchups in their favour that they are destroying in. Neither of these are problem, the one real problem in the community, is that people want to have decks with 50+ winrates against everything and that is a complete balance problem.

Most of the time, decks with average winrates across the board tend to be across the board, they don't tend to have extremely good or extremely bad matchups. The decks with extreme matchups, they tend to have them for both sides, extremely good and extremely bad.

 There's a lot to unpack here, so I'm just going to address the 90/10 10/90 problem. 

1.) If you have 10/90 you have 90/10 matchups.  Neither of these outcomes are healthy for the game.  While it should technically be possible to construct a deck with these odds, the power curve should not favor them.  If you have a 10% chance of winning a game, it is almost entirely out of your control to win by definition - unless you're playing an incompetent opponent (knowing you I have to clarify this point).  The odds speak for themselves, this kind of matchup devolves into a 'which deck did my opponent roll?'.  Changing the odds to a 70/30 and 80/20 is totally anecdotal.  Try playing a control warrior against mecha'thun priest and tell me how you improved your odds.  Ofc, the same principles apply in the inverse situation - 90/10.

2.)  The community likes 50+ % matchups instead of 90/10.  If someone actually wants this, they aren't thinking about the health of the game.  If the power curve prefers decks that are more balanced (but still slightly unfavored against their respective archetype), there is a greater chance that player decisions matter.  Especially when we consider the control vs OTK dynamic, where the OTK just draws their whole deck in 9/10 turns and wins.

I get the feeling I'm arguing against someone playing devil's advocate - it feels disingenuous.

Quote from DiamondDM13 >>

Quote from SnipaElite >>

 'Most card games do it' doesn't mean it's a bible to follow.   I listed two examples of OTK decks that have a more conditional yet were considered okay decks in their day.  Interaction in a game makes it more enjoyable to intelligent players at the very least.  Yes, you want lots of strategies, but not at the expense of in-game decision making.   MTG was terrible about this for example.

Shudderwock and Topsy Priest are pretty good examples to talk about.  Shudderwock is clearly an offender and I agree with your points.  Priest has had a deck like Topsy priest for some time now (inner fire/divine spirit shenanigans).  Topsy, in my opinion, is a bit too far - but is kept in check currently by its performance (to my knowledge anyway).  A deck that can literally win through charging boars and clearing (theoretically) any amount of taunts from an empty board position is a dangerous design.  The older priest decks have been somewhat strong and allowed more universal counter play.  Skill vs luck is a fine discussion to have after you have considered all the interactions with a deck/strategy first.

You seem to show up a lot when I post... have to say, your logic isn't ever totally on point.  But you have heart! 

 Intelligent players know that interaction is not solely the ability to directly disrupt the opponent by the use of a single card. Interaction is any action you can take to affect something, in this case that something would be the strategy of your opponent.

A year ago, we were faced with an annoying issue in the game called Jade Idol. It allowed a very strong strategy which didn't have direct means of disruption until Skulking Geist was printed. That doesn't mean there were no interactions that intelligent players could take against it, it just meant there were no interactions the less experience players could take against it.

You could still play Control and beat Jade Idols, you just had to know how to heavily out tempo the Druid with the limited tempo tools you had as a Control deck. It was hard, very hard, but that is what the experienced players could do, that non experienced ones could not.

The first day of Witchwood, I faced a bunch of Shudderwock Shamans as an adapted Freeze Mage. I didn't have any single direct disruption tool like Dirty Rat available to me and because of Healing Rains, I simply couldn't just Burn down the Shamans. Still, I didn't lose those games. Now I ask you, how did I beat them? Can you guess how I disrupted their game plan without having Dirty Rats?

Well, the Priest InnerFire decks were Combo decks but not OTK decks. They required minions from the previous turn to survive to be able to attack. Regardless, it is perfectly fine to have a deck which can do what Topsy can. It requires the player to be very proficient from a Technical Play level to be able to execute the Combo turns effectively and win the game.

The Win Condition of a deck should only be as powerful as the amount of Technical Play required to pull of the winning Combo. That is why things like Patron were slightly overpowered but still fine. These decks are actually the best kind of decks, they are decks that perform REALLY badly in the hands of bad players, yet perform extremely well on the hands of great players. This is how the game should be as a whole.

The actual design flaw in Hearthstone is not the power of Combo decks, it's the power of other decks. It is terrible to see how high the winrates of decks like Odd Paladin are when you take into consideration the amount of Errors the players are allowed to do while piloting them and still get away with the win. Obviously, in an ideal scenario the game would punish every players for their technical fails, but the fact that there exist decks out there which are in such a high power level that they still perform so strongly even while piloted incorrectly, that is where they need to look at when it comes to design flaws.

So yes, Skill and Luck is a discussion for when you take interaction into consideration, the problem is, not all players actually know how interaction works and more importantly, most players don't know the interactions past the basic direct disruption mechanics, as in, use X card to disrupt strategy Y.

Oh and about posting a lot when you post, I'm sorry, I don't tend to memorise names and such all that much, I just post based on the text written in the post itself.

 Absolutely, interaction has a lot of different successful applications - including tempo as you mention.  As long as it exists as a viable counter strategy, I find that acceptable.  This is almost always highly class limited in HS though, depending on the expansion.

I play w/e classes I find most interesting in an expansion, and would ideally consider all of them interesting.   If a class feels extremely limited by these oppressive strategies (like Jade or Odd Pally), then it's a less interesting meta.  I generally play at Legend or near Legend rank, so at this level I have to be playing highly optimized classes/lists.  In some cases, you're lucky to have 1-2 more balanced strategies.

Greedy decks rarely if ever appeal to me, because their ability to adapt to a matchup is just so limited - so any meta that feels too slanted in this way I'll just sit out.  Grinding out a greedy meta has to be the most mind-numbing experience to everyone right?  This is why I can't imagine how you can refute my earlier points - skill/decision-making rarely comes into play if a deck has a ton of 10/90% 90/10% win matchups.

Quote from m4k41 >>

As Control Warrior, you have exactly 0% chance to win against Mecha'thun Priest. The few games you win are basically because the opponent messes up and overdraws his Mecha'thun, but this is an unforced mistake.

Such one sided match-ups shouldn't be in the game. Whalen is an asshole. Every class needs something to have at least 30-40% chance against anything.

I'm getting tired of these 90%-10% matchups, makes the game so frustrating.

 This is precisely the problem.  An OTK deck that doesn't allow some kind of interactivity is bad for a game because certain classes (warrior in this case) are either shut out of the game, or forced to play these ridiculous matchups.  

Ideally, it should be possible to build a deck with any class that doesn't get totally destroyed by another optimized list.  Clearly, there will always be weird rogue decks that have asymmetrical win rates, but they shouldn't be top performing decks.

Quote from robpcom3 >>

There has to be OTK decks to counter absolutely insane control decks that clear board 5, 6 times in a row of sticky, shielded and deathrattle minions. I'm playing a token druid agains a warlock and it's a joke really. He's been able to clear a full board 5 turns in a row. 

 Definitely, this is how the ecosystem is kept in check.  My points, for instance, deal with how interactive a strategy is.  If the control deck has to create a less greedy deck to deal with combo more effectively (not 50-50, but say, 40-60), then the game is healthier and doesn't feel decided at the start unless you went full greed.

Quote from DiamondDM13 >>

Quote from SnipaElite >>

In conclusion, a combo deck that wins from hand regardless of board state or how much hp/armor is probably just a terrible design.  If you need a very specific card like dirty rat to interact with a deck like this and have any chance, this probably shouldn't even be printed in the first place. 

 That is what 30% of most OTK decks do in most card games, not just Hearthstone. Hell, in some games, that is what 60% of the OTK decks do.

It is not terrible design, it is just different strategies given to players so they can win in different manner, it is what brings interest to card games, the ability to win games in so many different ways because of the insane number of combinations of cards players can use to build a deck...

Just because you don't like a strategy doesn't make it terrible design. What makes it terrible design is how it operates within the constraints of the game and it's mechanics. In Hearthstone, the OTK strategies are almost all perfectly fine, with what I would argue Shudderwock being the exception, simply because it is Random and therefore, avoids the failure margin most OTK deck have that is dedicated simply to pure fuck up from the Pilot.

You can just look at two OTK decks in current existence, Shudderwock and the Priest Topsy deck. Shudderwock operates in random order and therefore, doesn't matter what the player does, it decides on it's own if it works or not. You can play the deck correctly, have a clean board and space in hand when playing Shudderwock after maximizing the odds of it returning to hand, and RNG can still decide not to give it back, and it can do the opposite, you can fuck up the odds and the board and still get it back to hand at random.

On the other side, you have the Topsy deck, which if you attempt it, you will see that you can extremely easily fuck up the orders and get stuck with no way to win. Yet both can deal absurdly ridiculous amounts of damage, Topsy can deal even higher ones. The only difference is the player is not likely to be punished by their mistakes in one and will get fucked in the other.

I'm all for allowing different win conditions in the game, it makes it more fun. But when you give players a ridiculous win condition like Shudderwock, and don't give it room to punish player failure, I do have a problem. Yet I have no issue at all facing almost any other OTK deck in the game, I enjoy it even. Doesn't feel bad losing to someone playing correctly, it feels bad losing to someone playing badly and still winning.

 'Most card games do it' doesn't mean it's a bible to follow.   I listed two examples of OTK decks that have a more conditional yet were considered okay decks in their day.  Interaction in a game makes it more enjoyable to intelligent players at the very least.  Yes, you want lots of strategies, but not at the expense of in-game decision making.   MTG was terrible about this for example.

Shudderwock and Topsy Priest are pretty good examples to talk about.  Shudderwock is clearly an offender and I agree with your points.  Priest has had a deck like Topsy priest for some time now (inner fire/divine spirit shenanigans).  Topsy, in my opinion, is a bit too far - but is kept in check currently by its performance (to my knowledge anyway).  A deck that can literally win through charging boars and clearing (theoretically) any amount of taunts from an empty board position is a dangerous design.  The older priest decks have been somewhat strong and allowed more universal counter play.  Skill vs luck is a fine discussion to have after you have considered all the interactions with a deck/strategy first.

You seem to show up a lot when I post... have to say, your logic isn't ever totally on point.  But you have heart!