Value in Hearthstone

this started as a post i told pat id write for why you can't just call lava burst bad based on it being less efficient than fireball, and why it would have more value if lightning bolt didn't exist, but now its become more of a "How do you play Hearthstone and how do you assess card value"

so the way i approach hearthstone is just a pretty general
Q: how do i kill my opponent before they kill me

and then each card in your deck is part of an answer. when your opponent plays a card, theyre asking you a new question -- one that's part of the main question, but a more specific part of it. then when you play a card, you're trying to answer a question. ill sometimes refer to a card your opponent plays as a question and a card you play as an answer

very often, your card asks its own questions back to the opponent. this contributes, directly or indirectly, to your own answer to "how do i kill them first"
sometimes you dont have any choice but playing two cards as a single answer to one question but -- depending on the question, especially if its the "win" question, and the circumstances of those specific answers -- that can be totally fine. and sometimes your card answers multiple questions

better yet, cards (questions / answers) don't ever truly leave the game. even when they seem fully answered, a question still cost a certain amount of resources (card-draws + mana) to ask and a certain amount to answer. even when things seem to have completely zeroed out, you're still out that one answer this game and they are too.

a card's value is different as you change context
value compared to all other cards
a general baseline to estimate if this answer will leave you ahead or behind as compared to
1) what it will be used on
2) what else fills a similar role (provides a similar answer)
you can't take much from this value as it really doesn't mean much. how could it -- it barely has any context in a game that's literally all about context. what you can take though is a general idea

value within your deck (summarized)
1) how "useful" of an answer will this card be within your deck. this part alone could go on for awhile, but super simplified: does it increase the value of your other answers / has its own value increased; does it satisfy some meta read on the current problem questions being asked; does it provide a "useful" answer that none of your other cards do (ie flamestrike in a mage deck without any other aoe). increasing the value of your other answers is complex though -- it doesn't just have to be like shield + watcher, it could be like frost bolt so you dont have to burn some better removal down the line on undertaker.
2) how good is this answer, the way it will be used in your deck, for the questions asked by other peoples decks. this is part a vanilla comparison, stat point for stat point how does this card line up compared to what it will be answering. because youre looking at the card as you want to use it, there's a lot more strategy / context involved in this value assessment than in the global one

i made a miracle deck a few days ago and i quickly stopped running assassinate because it rarely ended up being my most efficient answer, and it felt too slow/mana heavy for the rest of the deck. not that id argue assassinate was mana efficient to begin with

value within a certain game
1) how good is this answer for the questions specific to this game. this value becomes much more pronounced with your skill as a player. a meta read into their deck comp may encourage you to save some answers depending on what you expect to see.
2) how good this answer is compared to your other answers. whenever your opponent asks you a question your immediate options for answering it are defined by your hand -- even if your answer is to ignore it. having two of one card in hand is rarely good because the same answer is rarely usable for equal value soon after using the first
likewise, the value of an answer is increased by your immediate limitations. you cant forget that the value of a card, at this point, is only based on how well it can help you win the game

We have different approaches
Just answering your opponent will almost never get you ahead, you need to play your own questions for him to answer
For me, the Q is this: how do i kill my opponent in the most efficient possible way
Forcing them to answer you is one of the best ways to do that, but its not the only way

Lets look at miracle rogue as its a really good example
It has answers for pretty much everything, but it also asks almost unanswerable questions by using conceal
Conceal itself doesnt answer anything, its only a question
Conceal on auction asks the question of "can you stop me from drawing my deck without targeting my guy"
Conceal on a big vancleef or on a few minions 1 of which has cold blood that gives lethal asks "do you have taunt"
And the deck can answer the taunt if they manage to put it down with sap

Answers dont win games, they even things out and/or give you time to ask a question, questions are what win
If you draw nothing but answers you will eventually lose and thats true for all decks (healadin could get away with nothing but answers, but it wouldnt be easy)

Cards that both ask a question and answer one, like fire elemental for example are generally the best cards in the game, thats why sylvanas, rag, cairne, black knight, and other legendaries are so good and used so much
The value of those types of cards are inherently higher than others because of it

Value compared to all other cards
You can directly compare cards and it does have context
There are vanilla cards that all have different mana costs that set a baseline, people call this the "vanilla test"

Mana	Card	Total Stats
0	Wisp	2
1	Murloc Raider	3
2	Bloodfen raptor/Rivercroc	5
3	Spider Tank	7
4	Yeti	9
5	Feugen*	11
6	Ogre	13
7	Wargolem	14**

Spoiler:

You can use this as a baseline for value, those are the minimum stats you should get for a minion that costs that much, if it matches that it passes the vanilla test, if it doesnt it fails and is generally bad
For example, the total stats of a 3 mana minion should be, at minimum 7, any less that that is low value, more is more value and you can compare cards, even class cards of different classes against eachother using this
Minion ability will change that value, so a card like dancing swords has 8 total stats and should be high value, but the ability to let your opponent draw a card that early in the game is too bad to outweigh its stat total in most cases and it also just dies to a 4 mana card like yeti so its low value
Harvest golem falls below the stat line at first, but ends up being above and its ability makes it hard to deal with so its really high value for 3 mana
Because of that you can say harvest golem is better than dancing swords

You can take this even farther and get the value of game mechanics
Novice engineer is a wisp with a card draw ability and costs 2 mana, so a 1/1 is already establish as cost 0, so engineers cost comes entirely from drawing a card immediately, so you can say that the ability to draw a card is worth 2 mana
Loot horder is a murloc raider with a delayed card draw, so delayed card draw is worth 1 mana
Windfury harpy is worth 6 mana and is a yeti with windfury, so windfury is worth 2 mana (as seen in the actual windfury spell)

It all matches up like this with some exceptions
Those exceptions are either really crappy cards that fall way above what they should cost or amazing cards that cost very little for a lot
Some of those bad exceptions are actually good, like windspeaker is getting windfury at a 1 mana discount, but windfury in general isnt great so windspeaker isnt good, but if windfury was good (and it could be in a future meta game, probably not though) windspeaker would be a good card

So you can compare minions and game mechanics with context by using those baselines

Spells also have a table like that, though its not as clear cut
Vanilla in this case means cards that have no downside/extra cost/health gain on use/card draw on use

Mana	Card	Damage
0	Moonfire	1
1	Arcane shot/holy smite	2
2	Frostbolt/wrath	3
3	Kill Command	3 (5)
4	Fireball	6
5	Explosive Shot	5 (+2)(+2)
6	Avenging Wrath	1*8
7	N/A	N/A
8	N/A	N/A
9	N/A	N/A
10	Pyroblast	10

So lava burst's total cost is 5 with overload and falls below the vanilla test, and the only pay 3 now thing isnt helping enough to make it worth that failure, in fact it could even set you back because if you use it on turn 3 you only have 2 mana the next turn instead of 4
So you can say that its worse then explosive shot and worse than fireball in terms of value

If we look at starfire though, deal 5 dmg for 6 mana and draw a card, immediate card draw is established as costing 2 mana, so 4 mana for 5 dmg fails the vanilla test by 1, but the card draw is good enough to make up for its failure making stafire an average card when compared to fireball


Value within your deck
I agree with you here, thats why i said you could put 1 lava burst in your deck depending on the deck
The structure of the deck does make some cards more valuable than its vanilla test would indicate, conceal in miracle rogue again, perfect example, normally that card sucks, but in miracle its value skyrockets

But like you said about assassinate, theres not much that will make that cards value any higher than the vanilla test (it falls way behind poly or hex which also silences the thing it gets rid of)
Lava burst is similar, your deck would need to be spell focused and have auctioneer to make lavaburst worth it, you didnt tell me what type of deck you were making, i assumed standard shaman, and lava burst is almost as bad in standard shaman as assassinate is in mircale

You were saying hypothetically if lightning bolt didnt exist lava burst value would go up, i was saying no it wouldnt, because it still fails the vanilla test, it would get used more, sure, but not because its better, its just the only direct spell shaman would have that ignores taunt that isnt complete crap like frostshock
Shaman would be significantly worse overall if that was the case

Value within a certain game
This type of value cant be used as a measurement of any kind

For example, yesterday while playing against slayer, i had a 4/5, slayer had a 2/4 on board and played a 5/5 onto that board
So on my turn i knew the trade he would make would be 5/5 into my 4/5, even if i killed the 2/4 with my 4/5, so hed be left with a 5/1 and w/e else he played on his turn and all i would have was what i could play on my turn right now, i was priest with no way to deal 1 damage and i had no minions that could deal with the 5/1 without being insanely costly (like losing a 5/5) and no aoe that could incidentally kill the 5/1, i didnt have death to kill the 5/5 either, so the only line of play i could see was mindcontrol on the 5/5, even though i knew he had rag/sylvanas/ysera etc in his deck (he was also low on hp so i figured the game wouldnt last long enough for him to get ysera or sylvanas going)

Basically i used mindcontrol as shadow word death knowing that it was bad value because the 5/1 that would be left would get too much value if i let him make the logical trade
It helped me to win, but doesnt change the fact that i got bad value out of it

Its like that for any card that you have to use in an inefficient way during a game

The only time this works is when you get more value out of a card, like frost bolt on an azure drake with fireblast to finish it, 4 mana to kill a 5 mana thing, and you only spent a 2 mana spell card to do it
But even in that case you only getting slightly more value out of an already valuable card

Frostbolt on undertaker is bad value, 2 mana on a 1 mana card, and it protects whatever follow up they have from that frostbolt
It seems like good value, because undertaker can snowball out of control, but when i play my priest deck, i love when people burn frostbolt/wrath/lightning bolt on my undertaker because it means i can play out my cultist or blademaster (without healing it) and know its 95% safe

The way i play it is to try to answer them with my worst possible answer, so against undertaker, i could frostbolt it right now, or i could play a 2/3 or 3/2 and let that deal with it (with fireblast backup potentially), the odds of him being able to buff it up to insane levels on turn 2 or 3 is pretty low, if all i have is frost bolt id still wait for my next draw before using it, its possible ill draw into something that can deal with it, or its possible he cant buff it yet, either way it will still be in range of my frost bolt next turn, most likely
The only time i wouldnt wait is a play like undertaker coin leper/chow, that implies he has a turn 2 followup which will screw me if i dont deal with it now

But the cards value never changes, you just use it in a low value or extra value situation
Frostbolt on undertaker is low value
Frostbolt on azure drake is extra value
Frostbolt's base value is still killing a 2 or 3 mana minion with 3 hp

Even playing bgh against zoo, knowing they wont have any targets for bgh to hit (unless something insane happens with 2x argus/clerics and doomgaurd) is low value since bgh base value is killing a minion and being a 4/2 body on the board, so it lost half its value, sure it will help you win by playing it like that and should be played, but its worse than other 3 mana minions that couldve been there in that situation

You never responded to this seren, also i updated the table with spider tank, now that theres a proper 3 mana vanilla minion (7 is what it shouldve been to fit the pattern)

lol im probably not going to <3
two things, both enough on their own: a) i dont enjoy hearthstone as much anymore b) im not lookin for another argument when we're already on different pages. idk what the saying is for fool me three times is but i think im in the red on this front

i liked reading the parts of yours that werent (as well as some of the parts that were) commentary on mine. if your view on value changes you should post how and why cuz ill still be interested. maybe someday if i make a hearthstone bot ill respond with some of the tests i try

Its different for a bot, bots dont create decks so value within a certain game is all that matters to them and i think if you could program a bot to follow what either of us said about ingame value it would perform really well

Unless you mean make a bot to create an arena deck or something to play arena for you infinitely as long as it kept earning enough gold to keep going
Youd need 2 separate systems for it to play really well (like a human), 1 for the deck building part that combines what i was saying about value compared to all other cards and value compared to other cards in the deck as the deck is made and 1 for actually playing the deck which would be based on value within a certain game


When i said value within a certain game cant be used as a measurement of any kind i only meant that because it keeps changing on a game by game basis theres no way to use it as a measurement, i still think its important to recognize value within a certain game though
If i didnt recognize it as being important i wouldnt have used mind control the way i did in that story i told