Hello there! We are conducting a survey to better understand the user experience in making a first edit. If you have ever made an edit on Gamepedia, please fill out the survey. Thank you!

Talk:Advanced rulebook

From Hearthstone Wiki
Jump to: navigation, search
For previous discussions, see Talk:Advanced rulebook/Archive01 (archived 2015-06-03)

Rewrite project[edit source]

Following discussion with various editors, I've created Hearthstone_Wiki:Advanced rulebook project. The aim of the project is to rework the current article into something more accessible and readable. As this is a fairly involved process, with several interested editors, it made sense to host the effort on its own page, rather than simply experimenting with the live version. All editors are welcome to contribute to the project there. -- Taohinton (talk) 22:12, 3 May 2015 (UTC)

Don't worry, I only stopped working on it because I was sleeping :) I still have plenty more I want to write. --180.235.156.10 22:32, 3 May 2015 (UTC)
Completion

The project has now been completed and the draft transplanted onto this page. Discussion related to the project can still be found on the project page, but new talk topics should be posted on this page. -- Taohinton (talk) 04:56, 28 May 2015 (UTC)

Glossary[edit source]

In the "Force play" entry of the glossary it is stated that "Notable due to a bug that makes the resultant minion's order of play the oldest, not the newest, but only when considering simultaneous Deaths": wasn't this fixed?

Late On Summon Phase[edit source]

Hello everyone! I have doubts about that statement "Late On Summon Phase happens after applying auras" is true. The only prove i saw is: "Taohinton 2015-02-02, summoned Timber Wolf with a Starving Buzzard on board, the Buzzard clearly gained the Wolf's Attack buff prior to triggering its card draw effect". But that doesn't look convincing. So here what i suppose to test: Setup - You have Buzzard and Champion in play. Buzzard is pinged to 2 health. You have a lot of ganged up Flame Leviathans in deck. Play Raptor. Buzzard reacts and draws a Flame Leviathan. Raptor stays alive? — Preceding unsigned comment added by CodeComposer (talkcontribs) 08:16, 1 June 2015‎

Hello, CodeComposer, and welcome to the wiki! To make conversations on talk pages easier to follow, please sign all the edits you make. (You can do this by pressing the 'signature' button in the editor pane, or by typing two - then four ~ with no spaces.)
The reference provided for the timing of the first application of auras upon summoning a minion is https://www.youtube.com/watch?v=bs4BncUyOFo , which shows the sequence 'Questing Adventurer triggers, then the Aura is gained, then One-eyed Cheat triggers', which corroborates Taohinton's video-less claim. However, arguments based on what can be seen alone are not 100% foolproof, as can be seen when things like the Frothing Berserker vs Lightbomb bug happen: https://www.youtube.com/watch?v=7narpEF6bUY So thinking of tests for claims that have a gameplay side-effect are worthwhile.
The problem with your test is that we already know that auras and enchantments are recalculated, following the conclusion of an outermost Phase, before death processing happens. This can be seen in, for example, https://www.youtube.com/watch?v=lm1t1FU-ftc - by the time the phase Mal'Ganis is summoned in is over, the Voidwalker is mortally wounded, yet the recalculation FIRST of auras means that the Voidwalker is brought back to positive health before deaths are checked for. Your test has a similar problem - regardless of if the Stormwind Champion's aura starts applying for the first time before or after the "Late On Summon Phase", the aura WILL be applied before death processing, and the Raptor survives.
In addition, there is circumstantial evidence that the entire start of playing a minion is one big Phase with no death processing and a forced aura/enchantment update inside of it, or a similar construct. For example, in https://www.youtube.com/watch?v=H1758zVywbg the minion is played from the hand with the 'pending destroy' flag already set, so it should be destroyed whenever a Phase concludes, but we see it survive until after the Starving Buzzard reacts and draws one card, and die just before its Battlecry goes off. The only reason why I haven't already used this video to rewrite this section of the Advanced Rulebook is because it relies on a bug to demonstrate it, and I'm reluctant to do that. However, Illidan Stormrage can also be used to kill a played minion before its Battlecry goes off, such as in https://youtu.be/TI4hwVe7Z1s , so if you did a similar video but also threw in a Starving Buzzard/Undertaker/One-eyed-Cheat, you could show with no bugs whether death processing happens between the "On Play Phase" and "Late On Summon Phase", which I WOULD happily use.
Let me know if you have any additional thoughts! --Patashu (talk) 09:35, 1 June 2015 (UTC)

Turn limit/89 turns[edit source]

As the new video contradicts the older video reference, has the game been changed, or was the previous information simply incorrect/interpreted incorrectly? If the former we should document it in the rule changes section. Also, presumably this is the same in all game modes? -- Taohinton (talk) 17:06, 9 June 2015 (UTC)

As far as I can tell, the turn limit was 89 turns since Jan 2014 and it was just incorrectly interpreted. See this video: https://www.youtube.com/watch?v=KRp6RtvYMwM#t=1m10s In this video they deliberately avoided draw mechanics to try and get as far into fatigue as possible. Player 1 goes first AND last, and takes 19 fatigue damage on his last turn. He starts with 3 cards in hand and draws a card for his first 27 turns, then takes fatigue for the next 19 turns, indicating 46 turns. The most likely explanation is that he drew a card once by mistake. If the turn limit was closer to 100 turns, obviously he would reach at least 23 fatigue before the game ended - note that no mechanics put cards back INTO your deck before GvG, so you couldn't get a higher turncount without also a higher fatigue back then. Just like 'battlecry occurs before the minion hits the board', saying 'the game lasts 50/100 turns' was an outdated, incorrect piece of misinformation that was never correct.

However, the change in turn time from 90 to 75 IS a real change that occurred during GvG, at least according to Xinhuan.

If you want to discuss rule changes that have occurred through the game, here are some suggestions:
  • Mortally wounded minions, pre-GvG, didn't block summoning effects
  • Various ways to get more than 7 minions at once that have been patched out (Shadow Madness ending, Redemption + Harvest Golem, etc) (probably related to the above!)
  • Pre-Naxx Secrets used to activate during your turn
  • Pre-GvG turns lasted 90 seconds, now they last 75 seconds
  • Warsong Commander's move from Late On Summon Phase to Secret Activation Phase (Was this GvG or Naxx?) to After Summon Phase (BRM) (I guess this isn't a rule so much as a specific minion being moved around though)
  • The period after Naxx where KT could resurrect minions that were mortally wounded/pending destroy during the End of Turn Phase
  • The period after Naxx where Knife Jugglers could target mortally wounded minions (I assume that other things like Ragnaros would have also done it during this period)
  • The period until a patch after GvG where Cogmaster, Old Murk-Eye and Grimscale Oracle's auras had a Priority going before enchantments, and thus could be hidden by Aldor Peacekeeper/Humility/Reversing switch

(Obviously some of these are more like bug fixes than rule changes, but I'll let you decide on that) --Patashu (talk) 23:10, 9 June 2015 (UTC)

Thanks. I agree most of these are more like bug fixes, and I really wouldn't want to list those kinds of changes to the rules, just major ones to the overall intended ruleset, like the turn timer or AoE healing. The secrets one is already there, and I've added the 90-75 one (although pre-beta not GvG, as shown by your ref). -- Taohinton (talk) 12:11, 22 June 2015 (UTC)

About Graveyard[edit source]

Not sure if it's the right place to put questions...Don't know how to use wiki correctly XD I think how Kel'Thuzad and Resurrect interact with a Druid of Fang that die before the battlecry Phase is important. With this we can know more about how Graveyard works.

— Preceding unsigned comment added by Willyk (talkcontribs) 08:48, June 13, 2015‎

Hi WillyK! This is the right place to put questions, and new information that you can't figure out how to integrate yet. For future reference, please sign your edits to a talk page by using the 'signature' button in the editor panel (it's the one with a blue pencil, or if you can't find it, type two - then four ~ with no spaces in between)
That's a good idea for a test. I've added it to my list of tests I want to see happen here: http://hearthstone.gamepedia.com/User_talk:Patashu#Self-Transforming_Battlecry_vs_Resurrect If you're willing to help out, join #hearthsim on http://webchat.freenode.net/ and say hi! --Patashu (talk) 00:09, 14 June 2015 (UTC)

Mismatched citation[edit source]

Rule M5 has a footnote (as I write this, number 197) which leads to a video that is mana-related but not order-of-play mana-related or Summoning Portal-related. I don't want to just yank it since I assume whoever put it in meant it to go somewhere, or perhaps even had another video they meant to link there. - jerodast (talk) 06:10, 12 July 2015 (UTC)

Thanks for the heads up! I moved it to be a reference for Rule M4. The two parts of Rule M5 are proven by references in the examples (there's one for Millhouse vs Loatheb and one for Summoning Portal vs Shadowstep). --Patashu (talk) 12:28, 12 July 2015 (UTC)

Heal ALL characters[edit source]

The recent edit states that "ALL"-type AoE ignoring mortally wounded characters affects healing as well as damage, but is this safe to say? I can't actually think of a triggered heal that does this. Certainly, it is safe to say that Healing Totem does NOT ignore mortally wounded, but the opposite seems like assumption. - jerodast (talk) 00:13, 12 August 2015 (UTC)

Good question. You forgot one, and I don't blame you for forgetting - Holy Nova. The facts are that Holy Nova does all its damage (then on-damage triggers) then all its healing, and I've seen multiple reports that if you or one of your minions dies as a consequence of the damage (possible with Eye for an Eye, Axe Flinger and with Imp Gang Boss + Knife Juggler) that character doesn't receive the healing. I plan on making my next Hearthstone Science video about documenting this, and then I'll add it as a reference. --Patashu (talk) 08:36, 12 August 2015 (UTC)
Haaa, sneaky. Oh Hearthstone, when will "simultaneous" actually mean "simultaneous" :P I guess I'll have to add 2-part spells to my list of triggered effects now (although it's more of a pseudo-triggered effect). - jerodast (talk) 11:31, 12 August 2015 (UTC)

Warsong Commander[edit source]

I'm in the process of updating the various elements of the wiki in response to the Warsong Commander nerf, but I'll leave it up to the regular editors of this page to decide whether to remove related examples or substitute other cards, etc. -- Taohinton (talk) 23:33, 20 October 2015 (UTC)

Hi Taohinton. jleclanche/Adys had the suggestion that examples that no longer apply due to card or rule changes should be moved to a separate page (could even be a page in user space - the idea is just to preserve them in some way). What do you think? --Patashu (talk) 01:06, 21 October 2015 (UTC)
I can't really see the value of adding this to the main wiki. Subsequent changes may render them wholly unintelligible, and the rulebook's already pretty compendious. A user page is a fine idea though, especially for research/past notes. I would suggest using a subpage like User:Patashu/Sandbox. -- Taohinton (talk) 01:17, 6 November 2015 (UTC)

Event Queue[edit source]

The description of the event queue is completely confusing. A queue is a representation of a BFS, and is FIFO. A stack is a representation of a DFS and is LIFO. Since the events are executed Last-In-First-Out, and similar to a Depth-First-Search, why is all the terminology on the page discussing queues? It's a stack, just like Magic the Gathering. This should be described in those terms. — Preceding unsigned comment added by 98.248.47.229 (talkcontribs) 22:47, 24 October 2015‎

I use Queue when I mean Queue. If multiple Events or Triggers are raised simultaneously, they Queue in the same order they were played in, and then we resolve recursively in the same order they queued in. There is no equivalent to the MtG stack. --Patashu (talk) 09:42, 25 October 2015 (UTC)
I disagree. To be very clear, computationally it's exactly the same. The main difference is that instead of storing 1 queue per depth of the event stack (and storing the priority of each minion that was alive at the time the event began processing), you just store all the triggered events in the stack itself. To process an event, we navigate the play order backwards, pushing each event onto the stack until we reach the oldest card, which is the final push. Now the oldest card is on the top of the stack, and executed. Any events that trigger off of it are executed before any of the other events, so they too are then pushed onto the top of the stack. You are describing a stack, a depth first search, a last-in-first-out. Hopefully it's clear that these two definitions are equivalent. So why does it matter to use one or the other? A stack has advantages such as being a single data structure to store all the events. Instead of using recursion, a simple loop until empty will process all the events in the correct order. — Preceding unsigned comment added by 98.248.47.229 (talkcontribs) 01:50, 26 October 2015‎
I thought about this on the train ride home. To be clear, the Stack and the Queue model both have the same predictive power and describe the same things - so we should pick the model that has better pedagogical value.
To that end, I think it is more appropriate to use the Queue model for the following reasons:
1) When describing the game mechanic, what is important is clearly explaining that events and triggers are both processed in age order (which developers have been quoted to say) and this is best explained as a Queue rather than that of adding to a Stack in reverse order so that it can be processed backwards.
2) There is no literal 'Stack' area of the playfield in Hearthstone. In MtG, even a newbie can see that people play spells in response to other spells, then resolve in reverse order, in the Stack Zone. It's easy to extrapolate this to adding other objects like copies of spells and triggered abilities after learning a few rules. However, in Hearthstone there is no Stack - you can never see it with your eyes, it is never talked about by the devs, or even mentioned in Power.log. However, it is possible to remember global play order - and a Queue naturally builds itself from the oldest Entity, adding newer Entities to the end in play order.
3) In general, in MtG, objects that are resolving on the stack cannot be interrupted (you can arbitrarily nest replacement effects but in general, the way triggers can create triggers can create events is more complicated in Hearthstone), in contrast to how it works in Hearthstone. In addition, spells and effects can be 2-step in Hearthstone - like Blizzard that allows on-damage triggers to run (spawning new minions in the case of Imp Gang Boss) before THEN freezing all enemy minions, and Light of the Naaru that allows Shadowboxers to trigger before deciding if its target is damaged or not.
4) While I don't like that it exists in its current form, the Greedy Queue of the Combat Preparation Phase is hard to model as a Stack: http://hearthstone.gamepedia.com/Advanced_rulebook#Combat Note in particular that triggers that are added to the Greedy Queue are added at the end - the equivalent in a Stack would be to add them to the bottom of a Stack, which doesn't make sense.
5) There is strong evidence to suggest that recursion, and limits of recursion, matter in Hearthstone. Auchenai Soulpriest + Mistress of Pain recurses only 61 times, but the game handles Queuing hundreds of simultaneous Acolyte of Pain triggers, which would overflow a stack that is merely ~60 entries deep! Draw effects that are interrupted by on-draw triggers re-play their animation afterwards. Knife Juggler nested inside of himself does NOT trigger, but Acolyte of Pain nested inside of himself does trigger but on an as-of-yet undescribed delay ( http://hearthstone.gamepedia.com/User_talk:Patashu#Trigger_Re-entrancy_Exploration ). These kinds of effects would be extremely difficult to include in a Stack model.
6) Delayed Event Queues (described in the video on the Summon Resolution Step: https://www.youtube.com/watch?v=QC8_jNOvdEs ) are a natural extension of a Queue model, defining Queues that are emptied (resolved one by one) at a later timing. In a Stack model, you still need Delayed Event Queues (you would dump them onto the Stack to empty them) - so you have to introduce the concept of a Queue anyway! (I think MtG has the equivalent to what we call a 'Delayed Event Queue' though, because you have to dump triggered abilities onto the stack whenever a player gains priority, so they had to be somewhere in the mean time. So this isn't an extremely strong point, but I'm including it for completion's sake.)
7) A very large number of videos and documentation use the Queue model. They would become outdated if we swapped to the equivalent Stack model, so it would have to be MUCH simpler to be worth swapping.
I am open to hearing any further thoughts you have. To be clear, these are not entirely my own views - Xinhuan shares them too, and I imagine Adys does as well (since his Hearthstone Simulator, Fireplace, is implemented with a queue/recursion model not a stack model). --Patashu (talk) 08:22, 26 October 2015 (UTC)
Thanks for the well thought out response. I think I must acquiesce, but with comments. I think the issue with greedy combat queue is simplified by assuming there is one additional step to combat, the "Declare Target" phase. Similarly, the "Play Spell on Dragonkin Sorcerer" phase is probably more realistically a "Declare Target" phase for spells, where Spellbinder is acting exactly like Redemption by forcing itself to trigger last.
In terms of limits to recursion, my understanding is Bouncing Blade also has a limit. This is probably more likely due each turn having a time limit, and wanting to prevent infinite loops or other odd behavior. Remember a stack is just RAM like anything else, it can store the exact same number of events as a series of queues could. The queue implementation relies on recursion, which by its nature means it's using the program's "call stack". This is probably what you are thinking of when you're thinking about the depth limit of stacks. Since the stack I'm suggesting is in memory, it's only limited by memory, the same way the Queue's are. The main difference is there is no recursion, so paradoxically, the stack implementation isn't using the "call stack", but the Queue implementation is. Therefore if anything would be limited by the stack depth, it's actually the queue implementation.
However, I think I must agree that updating all the documentation/videos would be a large undertaking for something that at the end of the day is trying to describe the same behavior. The queue model predicts the behavior correctly, so an update isn't necessary. Instead maybe effort is better put into simply rewriting sections of the advanced rule book to be more clear and to remove interactions which are outdated (e.g. there is a reference to sylvanis/sylvanis interaction which depends on order of play. My understanding is it actually depends on which attacks and which defends). I think for the programming minded of us, at most it's worth adding a footnote somewhere. — Preceding unsigned comment added by 98.248.47.229 (talkcontribs) 15:51, 26 October 2015‎
This was a productive, thoughtful conversation, so I'm happy to have it :)
To clarify a few things:
1) "I think the issue with greedy combat queue is simplified by assuming there is one additional step to combat, the "Declare Target" phase.". It's not that simple. When you have an Ogre Brute and an Explosive Trap, regardless of order of play/Dominant Player Bug, if the Ogre Brute redirects from minion to hero or hero to minion, Explosive Trap will trigger at some point. There's no Phase where Ogre Brute triggers and a different Phase where Explosive Trap triggers - they can trigger in either order depending on the setup.
2) "Similarly, the "Play Spell on Dragonkin Sorcerer" phase is probably more realistically a "Declare Target" phase for spells, where Spellbinder is acting exactly like Redemption by forcing itself to trigger last." I'm lazy, but this is something that needs to be investigated. We figured out that 'Spellbender always triggers after Dragonkin Sorcerer' BEFORE we found out about the Dominant Player Bug. They might really be able to trigger in either order. Here's the full list of stuff that needs to be tested about that: http://pastebin.com/BPz6ZkY9
3) RE recursion: My point by bringing up the 61 limit on Auchenai Soulpriest + Mistress of Pain triggers, and also bringing up that having 160 acolyte of pain triggers intermingled with damage AoEs and flame leviathan triggers on draws, is that in a stack implementation, the former should take up ~60 stack entries, whereas the latter should take up ~160 stack entries. If the stack is limited to 60 stack entries (because Auchenai Soulpriest + Mistress of Pain stops), then how can it hold 160 acolyte of pain trigger stack entries? But in a Queue/Recursion model, this is natural (one would posit that Queues can only be nested up to ~60 deep, for example). So you're exactly correct when you say the Queue implementation is going to be limited by the stack depth - because that's exactly what seems to happen! (And Bouncing Blade's hard limit isn't related to recursion, it's a hard coded limit in the spell, because each damage event resolves before the next one begins - no kind of recursive structure is getting deeper and deeper.)
4) (e.g. there is a reference to sylvanis/sylvanis interaction which depends on order of play. My understanding is it actually depends on which attacks and which defends). You're mistaken. Sylvanas and Sylvanas simultaneously dying still depends on order of play. What you might be thinking of is simultaneous damage of combat - regardless of order of play, the defender's damage event is resolved first. This should be mentioned in the Combat section of the advanced rulebook.
All this said, though, I am always open to feedback on how the advanced rulebook can be made easier to read and more accurate. It's a bit of a mess because this iteration of the advanced rulebook was being written at the same time that I was doing the research and learning necessary FOR it to be written, so there's still some relics in it of outdated thinking. So feel free to leave comments here whenever you like. The more constructive/specific the feedback, the easier it is for me to use, as well. And if you want to talk about this stuff in real time, feel free to join #hearthsim on http://webchat.freenode.net and say hi :) --Patashu (talk) 22:12, 26 October 2015 (UTC)

Enrage mechanics[edit source]

Enchantments give the following example: "Examples: You play an Amani Berserker and Enrage it, giving it 5 Attack. You then play Humility on it, giving it 1 Attack. You then heal and Enrage it a second time - the new Enrage is at the end of order of play, going after the Humility effect and it now has 4 Attack." Since the 4.0.0 change to Enrage priority I guess enrage should get it's own (sub-)point.

I'll change the Advanced rulebook section on Enrage if and only if it is not fixed in the upcoming patch this week. --Patashu (talk) 22:36, 2 December 2015 (UTC)

Ordering (I am writing this pseudo-rulebook-section for my own benefit in understanding the topic, not necessarily for adding to the rulebook as it may turn out to be redundant or extremely esoteric)[edit source]

There are three categories of game mechanics in Hearthstone which raise the need for things to be resolved or processed in a particular order. When multiple triggers or multiple Events are to be resolved in the same Queue, there must be some criteria in place to order their resolution. When multiple adjustments are to be continuously applied to a single Tag, a similar criteria must exist to order their processing. Note that triggers, Events, and Tag-sets never compete with each other's order; only multiple occurrences of a single category can conflict. The following concepts are related to ordering criteria:

Dominance - Whether a trigger is owned by the Dominant Player or the Secondary Player.

Priority - A special preference hard-coded into specific cards, affecting the ordering of their triggers, Events, and Tag-sets.

Order of play - The relative timestamp value of an Entity, often used when ordering triggers, Events, and Tag-sets.

 Rule S1a:  The sole criteria used to order trigger resolution is the Queue order; the order in which the triggers were queued.  The criteria used to order trigger queuing is dominance, then priority, then order of play, evaluated at queue-time.
 Rule S1b:  The criteria used to order Event resolution is priority, then order of play, evaluated at [creation-time or trigger-time?].
 Rule S1c:  The criteria used to order Tag-sets is priority, then order of play, evaluated at every Aura Update (Health/Attack).  [Add notes about Aura/Enchantment priority and temporarily out of order Tag-sets.]


This means that the Dominant Player's triggers will queue before the Secondary Player's, but if the owner of both triggers is the same player (equal dominance), triggers with higher priority will queue first, and if dominance and priority are the same, order of play will be the determining factor. The difference between these criteria is that the Dominant Player's triggers queue AND resolve before the Secondary Player's triggers even queue, while priority and order of play merely determine order within a single Queue.

/*Work in progress, this needs to be restructured*/

QUEUING (use side of Queue, then priority, then order of play)
Inherent in play minion/weapon/secret trigger
Side of Queue: Minion/weapon/secret controller at queue time
Order of play: Minion/weapon/secret timestamp
Inherent removed from play minion/weapon trigger (Inherent Deathrattle)
Side of Queue:
Order of play:
Enchantment-granted in play minion/weapon trigger
Side of Queue: Enchantment controller at queue time
Order of play: Enchantment timestamp
Enchantment-granted removed from play minion/weapon trigger (Added Deathrattle)
Side of Queue:
Order of play:
Enchantment-granted-Enchantment-granted...in play minion/weapon trigger
Side of Queue:
Order of play:
Enchantment-granted-Enchantment-granted...removed from play minion/weapon trigger
Side of Queue:
Order of play:
APPLYING (Dominant Player Bug is not relevant; use priority then order of play)
Enchantment-granted in play Aura
Order of play:
Aura-granted in play Tag-set/Enchantment /*should test whether aura needs to be in play or just Enchantment*/
Order of play: Aura-emitting Minion/weapon/hero/player timestamp
Aura-granted-Aura-granted...in play Tag-set/Enchantment
Order of play:

Brainstorming (suggestions/thoughts/questions)[edit source]

"Death is the only Event that is not resolved in the middle of a Phase." What about Summon Events (including Play Events)?--DicePower (talk)

Death Events and Summon Events - this is correct. I tend to leave talk about Summon Events/Summon Resolution Step out of the main Advanced rulebook though, because it complicates things. (I leave out talk about the Dominant Player Bug and the Split Queue for a similar reason, even though it is also technically acting all the time.) --Patashu (talk) 10:01, 18 January 2016 (UTC)

"Phase: A Phase begins when multiple Events are raised simultaneously, or when multiple triggers respond simultaneously to a common Event, or as part of a Sequence. (Phases can be nested, but only the outermost Phase ending is meaningful.) When the outermost Phase ends: Run Aura Update (Health/Attack), Death Creation Step, Aura Update (Other). Then if any Deaths occurred, a Death Phase starts to resolve their consequences (and this Phase can have its own followup Death Phase for new Deaths and so on). This continues until there are no new Deaths - then the Phase is finally considered resolved and then the normal next Phase begins." Does this mean that a Phase begins every time a Queue is formed, even a nested one? I have been only been using the term to mean named divisions of Sequences or for Death Phases.--DicePower (talk)

So the thing is that when you summon a minion, an After Summon Phase is resolved for it inside of whatever Phase you happened to be in. In addition, Forced Death Phases are a thing. So attempts to come up with a model where there is no such thing as nested Phases inevitably fail. So yes, you can have nested Phases. --Patashu (talk) 10:01, 18 January 2016 (UTC)
Indeed, but my question is whether EVERY Queue begins a Phase, as the Phase definition suggests. For example, if inside the Spell Text Phase of Whirlwind multiple Damage Events are raised simultaneously, do you mean for this to begin a nested Phase to resolve the Event Queue? The definition seems to suggest so, but I've never seen you use the term Phase this way. The same goes for multiple triggers responding to a common Event.--DicePower (talk) 05:45, 20 January 2016 (UTC)
Because a Phase ending inside of a Phase does not have any side-effects, it is purely a question of definition - if it is clearer to talk about inner Phases starting every time a Queue is populated, or if it is just unnecessary baggage. --Patashu (talk) 04:20, 27 January 2016 (UTC)

Perhaps it would be helpful to define "Player Action" at the very beginning of the guide, as it's a concept that's easy to grasp, and can be used to define "Sequence" in terms of something everyone understands. Then move onto defining "Phase" in terms of Sequences, etc.--DicePower (talk)

Good idea, done. --Patashu (talk) 10:01, 18 January 2016 (UTC)

"Trigger" has quite a few meanings and can be confusing when first reading the rulebook (i.e., "triggers, including Deathrattles and on death triggers, trigger when triggered by Events").--DicePower (talk)

Hmm, Trigger and Event are important to define. I better do that. --Patashu (talk) 10:01, 18 January 2016 (UTC)

The definition of "resolving" seems possibly insufficient. There's more to resolving Mad Bomber's Battlecry than recursively resolving each Acolyte of Pain he hits. The resolution also includes the act of dealing damage to random minions.--DicePower (talk)

I think people know how Mad Bomber's Battlecry works, and realize that when I say to resolve the Battlecry at this time, the WHOLE thing is happening (including all selecting of targets, creating Damage Events, resolving Damage Events, etc). But if you can think of a concise way to word it, let me know. --Patashu (talk) 10:01, 18 January 2016 (UTC)
Are you sure you mean to include selecting targets? I was under the impression that resolving an effect began at trigger-time. Is this false? I do agree that it's fairly clear what you mean in the definition. It's one of those things where I only questioned my understanding after re-reading it many times.--DicePower (talk) 05:45, 20 January 2016 (UTC)
When I talk about a player targeting something, this is 'conscious targeting', involves a red arrow, checks target validity, and is done and remembered before the played card enters play and is paid for. When I talk about an effect targeting something, it's part of the spell/battlecry/trigger effect, happens at that time, and since it is not 'conscious targeting' it doesn't get stopped by effects like Stealth and Elusive, but must obey any requirements it states on the card text (like Lightwell must target a friendly undamaged minion, and Enter the Coliseum must target minions with the highest Attack on their respective sides of the Battlefield). Otherwise Ogre Brute could not target an enemy minion in Stealth! --Patashu (talk) 04:20, 27 January 2016 (UTC)

"Play" and "Summon" are often implied to be mutually exclusive terms. Hearthstone uses the terms differently; playing is a type of Summoning. It might be better to fix our terminology to adhere to the game's terminology, where it actually exists. This would probably mean something like using "Summon" interchangeably with "enter into play," and "Summon by effect" to specify that the minion is not played.--DicePower (talk)

Can you give an example of where in the Advanced rulebook this could lead to confusion? (I'm not against it, just want to know what I have to deal with.) --Patashu (talk) 10:01, 18 January 2016 (UTC)
Hmm, I actually can't find too many examples at the moment, apart from implications drawn from Sequence names. One example is in the Summon Resolution Step section. "Whenever a minion enters play (whether due to being played or summoned), a 'Summon Event' is created, but not resolved." Like defining "resolution," I don't think this is a huge deal, as people will understand. I just wanted to put these thoughts out there in case there is ever a reason to rework vocabulary.--DicePower (talk) 05:45, 20 January 2016 (UTC)

Similarly, it seems to me the term "target" should be reserved for cards that require the player to select an Entity when played (with the possible exception of the phrase "attack target"), since Hearthstone does not consider Vaporize or Knife Juggler to target in any sense.--DicePower (talk)

This is a good idea, but it's too natural to call what Mad Bomber or Knife Juggler does 'targeting'. Maybe instead we could come up with a new word or phrase that means 'consciously targeted by a player', and consistently use that? --Patashu (talk) 10:01, 18 January 2016 (UTC)
I'm not sure if we actually need the distinction to get points across. My reasoning for suggesting an adjustment is so that, i.e., if someone reads Faerie Dragon's text, and then reads the Advanced Rulebook, "target" means the same thing in both contexts. I have seen players actually get confused as to why Vaporize killed their Elusive minion.--DicePower (talk) 05:45, 20 January 2016 (UTC)
If you come up with good terminology to use that reads well, let me know. --Patashu (talk) 04:20, 27 January 2016 (UTC)

Is Grim Patron's trigger condition really an exception to Rule 5? I can't see anything in the rule that would allow one to predict the Patron outcome one way or the other.--DicePower (talk)

Yeah, this doesn't make sense. I think it used to make sense to have it here, but no longer does. I'll take it out, since the example is used elsewhere already and better. --Patashu (talk) 10:01, 18 January 2016 (UTC)

"A similar setup using Cone of Cold instead will show that the Cone of Cold hits and freezes nothing, since its target, now in the Graveyard, has no adjacent minions on the playing field." This explanation may not be correct, since it does freeze Battlefield Zone minions if it's in the hand when Cone of Cold resolves.--DicePower (talk)

I think it's OK to use, since the most common way to encounter this (target of CoC dies to VT+KJ) correctly causes nothing to freeze. I've seen it happen multiple times on video, not once has it randomly frozen something. The hand case is a bit weird, but convoluted to the point where people won't see it, and it might even be a bug that something in play is frozen. --Patashu (talk) 10:01, 18 January 2016 (UTC)

"The Battlecry still occurs even if the played minion Dies first. If the Battlecry uses the played minion as a target, then it affects its position in the Graveyard, not the position it had on the board, because it has already left play." I don't think this is even possible, is it?  :P--DicePower (talk)

I think the example is intended to indicate that Defender of Argus won't taunt any minions in play, since it checks what minions are adjacent to the position in the graveyard. Can it be worded more clearly? --Patashu (talk) 10:01, 18 January 2016 (UTC)
See, you even confused me with "target," haha!--DicePower (talk) 05:45, 20 January 2016 (UTC)

What do you think about changing Rule H2 to something like "However, when a minion's maximum Health is reduced, its current Health is only reduced if, and to the extent that, it exceeds the new maximum" and/or adding an example to show, i.e., a 4 Health (5 max) minion losing a Mal'Ganis Aura?--DicePower (talk) 07:29, 21 January 2016 (UTC)

I think it's OK, there are examples very close below already. --Patashu (talk) 04:37, 27 January 2016 (UTC)

"You have 9 cards in hand, Baron Rivendare and either The Skeleton Knight or Anub'arak on the field. When the minion dies, it bounces itself to your hand on the first Deathrattle resolution, then attempts to move from your hand to your hand, sees that your hand is full and discards itself from your hand." I think this example no longer applies, but I'm not 100% certain.--DicePower (talk) 07:29, 21 January 2016 (UTC)

Thanks, fixed in 4.0 so I removed it. --Patashu (talk) 04:37, 27 January 2016 (UTC)

In the "Zones" section, the term "zone" seems to refer to an area unique to one player in some context, and an area belonging to both players in other contexts.--DicePower (talk) 07:29, 21 January 2016 (UTC)

Hmm, you're right - there's really two distinct concepts here. Moving between zones (played, force play from hand, force play from deck, mad scientist DR, dying/equipped over/weapon destruction/hero destruction (-> graveyard), bounce to hand, transform/recycle (-> setaside), malorne DR/entomb (-> deck), tracking/discover (deck -> setaside, setaside -> hand or setaside -> graveyard), Lord Jaraxxus Battlecry/replacing your Hero (lots of -> setaside)) and changing controllers (mind control and all friends). Should they be separated? --Patashu (talk) 04:37, 27 January 2016 (UTC)

"Your current mana is capped at 10 but has no lower limit. Negative current mana is displayed as 0. (Note that current Overloaded Mana Crystals do not reduce the maximum of 10.)" I'm not sure what the parenthetic part means. By "maximum," do you mean maximum Mana, or current Mana cap? Is its purpose to explain that if you start your turn with 1/10 Mana (9 Overloaded), you can Innervate and play a 3 cost card?--DicePower (talk) 07:29, 21 January 2016 (UTC)

Edited and added a clarification. --Patashu (talk) 04:37, 27 January 2016 (UTC)

Should Rule DH8 say "proposed damage," or is "reported damage" different?--DicePower (talk) 07:29, 21 January 2016 (UTC)

I guess it's fine to change it to 'proposed'. --Patashu (talk) 04:37, 27 January 2016 (UTC)

Is there a reason to keep cards in hand as part of the definition of "In play"? Also, Sabotage is listed in the definition of "Forced Death Phase." I was going to remove it but figured it must have been added for a reason.--DicePower (talk) 07:29, 21 January 2016 (UTC)

Cleaned up the definition of Entity and in play a bit. Removed Sabotage. --Patashu (talk) 04:37, 27 January 2016 (UTC)

There are two different ruled named "Rule DP2."--DicePower (talk) 22:45, 22 January 2016 (UTC)

It seems like the second, isolated Rule DP2 is mostly restating Rule DP1 using different vocabulary, and they seem amenable to being merged.--DicePower (talk) 22:45, 22 January 2016 (UTC)

Wow, I have no idea how this happened :S Lol. Cleaned up a bit. --Patashu (talk) 04:37, 27 January 2016 (UTC)

"(Spellbender triggering will not stop later triggers in this Phase, similar to Counterspell and the On Play Phase.)" Is the meaning of this supposed to be that later triggers in the same Phase will resolve as if, and/or evaluate their trigger conditions as if, the old target was still the current target? Aside from a trigger condition that no longer passes, I don't see how Spellbender would stop anything, even in later phases, as it doesn't remove the subject from play.--DicePower (talk) 20:38, 23 January 2016 (UTC)

I mean that if you cast a spell on your friendly Dragonkin Sorcerer but Spellbender triggers first, Dragonkin Sorcerer still triggers even though its queuing condition is no longer true. Reworded to make it clearer. --Patashu (talk) 04:37, 27 January 2016 (UTC)

"Shadow Madness also remembers which side it 'should go back to' even if stolen by Sylvanas Windrunner in the mean time." What if this happens before Shadow Madness's Spell Text Phase?--DicePower (talk) 19:07, 26 January 2016 (UTC)

Then you get what happens with Djinni of Zephyrs + Shadow Madness - the minion is stolen from the side it's already on to the side it's already stolen on. Then later when SM's End of Turn Phase trigger runs, the minion will be stolen to your side yet again. --Patashu (talk) 04:37, 27 January 2016 (UTC)

The full list of Auras that are known to update in Aura Update (Other) (excluding Solo Adventure-only content) is as follows: Baron Rivendare, Auchenai Soulpriest, Brann Bronzebeard, Mal'Ganis's Immune effect, Prophet Velen." Are Warsong Commander and Wee Spellstopper other examples?--DicePower (talk)

This is an excellent question, and something that deserve an Esoteric rulebook entry. Basically, Aura Update (Health/Attack) is actually Aura Update (Enchantment), and Aura Update (Other) is actually Aura Update (Tag). This sounds esoteric because it is - basically, Auras like Tundra Rhino, Crush and Aviana create Enchantments and attach them to Entities they effect, whereas Auras like Mal'Ganis (immune Effect) and Prophet Velen do not create and attach Enchantments - they just directly update values of tags while they're alive. This difference in mechanism (directly setting a tag vs creating and attaching Enchantments that themselves set tags) also has a difference in timing - whether they update during Aura Update (Enchantment) or Aura Update (Tag). So not only is the current list for Aura Update (Tag) comprehensive, EVERYTHING that isn't in Aura Update (Tag) is in Aura Update (Enchantment).
The reason why I haven't already done a find-replace on this terminology is because Aura Update (Health/Attack) is a LOT clearer to laymen about what it is. As in, if I say Aura Update (Enchantment) I have to explain a while layer of mechanism and what goes in it and what doesn't. Yet, the only time where Aura Update (Health/Attack) having a different timing from Aura Update (Other) matters currently is with Health Auras like Stormwind Champion vs simultaneous AoE damage. So if I'm ambiguous about stuff like Mana Cost auras, Tundra Rhino, etc's timing, it doesn't lead people to make wrong predictions, because there are no predictions to make right or wrong. That's why information about this should strictly be in the Esoteric rulebook for now. --Patashu (talk) 05:33, 19 January 2016 (UTC)
Very interesting, I guess for now it's just easiest to memorize the Tag list, since it's not obvious that Mal'Ganis's Immunity is a Tag rather than an Enchantment.--DicePower (talk) 06:23, 20 January 2016 (UTC)

Warsong Commander is a conditional Aura we've hypothesized to likely wait until Auras are updated to apply to new minions. Do Flametongue Totem and Dire Wolf Alpha wait until the next Aura Update (Health/Attack) to remove the Attack boost from one minion and apply it to another after a minion enters/leaves play, or are they simply turned on and off during updates, with appropriate minion selection applying instantly?--DicePower (talk)

...Usually. I've written about it a bit here. More testing required: http://hearthstone.gamepedia.com/User_talk:Patashu#confusing_aura_thing (For example, I think one side-effect is that the extra update might make a dead Stormwind Champion's Aura immediately turn off, changing the outcome of a Stormwind Champion + AoE simultaneous death to losing Aura then taking damage, instead of taking damage then losing Aura.) --Patashu (talk) 05:33, 19 January 2016 (UTC)

Has anyone actually tested adding a Dragon minion to the hand before the Battlecry Phase of playing Blackwing Corruptor or Rend Blackhand, after failing to select a target for the Battlecry?--DicePower (talk)

No, but what would it do? You didn't select a target, is one going to be spontaneously selected for you? --Patashu (talk) 05:33, 19 January 2016 (UTC)
Crash Hearthstone perhaps, or bug something?--DicePower (talk) 06:23, 20 January 2016 (UTC)

"Faceless Manipulator and Druid of the Claw will still transform even if they are removed from play by the Illidan combo. Mirror Entity consequently copies the post-transformed state from the Graveyard. However, if the post-transformed minion has a Deathrattle, it will not trigger, because the minion already died before the transformation happened." Does this contradict Rule 10? Applying the same rationale behind Sacred Trial VS Dr. Boom would predict an outcome where the Deathrattle queues and resolves.--DicePower (talk)

I do not understand the question. --Patashu (talk) 05:33, 19 January 2016 (UTC)
Given Rule 10, it seems that just like we would predict that Sacred Trial will trigger if all conditions are met at trigger-time (using no memory of whether or not they were true prior to the current Queue), we would also predict that if Faceless Manipulator targets a minion with a Deathrattle, is sent to the Graveyard, and then Transforms, the Deathrattle WOULD queue and resolve. The tough part is thinking of a way to Transform Faceless in the Graveyard BEFORE its Death Event is resolved.--DicePower (talk) 06:23, 20 January 2016 (UTC)
Deathrattles only trigger at a very specific time - the moment the minion dies and its Death Event is resolved. If you make a copy of it, such as with Faceless Manipulator - tough, too late, that timing window already passed. If you're trying to get something to happen like 'a minion dies, then transforms into a deathrattle minion or gains an added DR, THEN its Death Event is resolved', good luck with that, I have no idea what it would do at all! (Though I still want to test killing a minion with Cabal then stealing it in the Graveyard - That's testable, and I have NO IDEA what it will do.) --Patashu (talk) 04:42, 27 January 2016 (UTC)

Do attack damage and counterattack damage really occur sequentially? They appear to be simultaneous, and we know from the Tavern Brawl snowball test that Events aren't always resolved in the order they're created, so I'm not sure I see a reason to treat them as any more sequential than i.e., multiple minions being damaged by Whirlwind.--DicePower (talk)

I re-worded it slightly to make it more clear that the damage is dealt simultaneously before being resolved, just that the ordering is not order of play, but (attack, counterattack). (matters if you attack with Grim Patron into Imp Gang Boss while Knife Juggler is in play for example) If it can be made even better let me know how. --Patashu (talk) 05:33, 19 January 2016 (UTC)
Looks great to me.--DicePower (talk) 20:41, 20 January 2016 (UTC)

If an attacker or defender are removed from play, at what timing do they stop being "the attacker" or "the defender," respectively, for the purposes of evaluating queuing conditions and trigger conditions?--DicePower (talk)

Never. Actually, I seem to have left this information out. It's implicit, but better to make it explicit. --Patashu (talk) 05:33, 19 January 2016 (UTC)

I noticed that Mogor the Ogre does not have a queuing condition or trigger condition listed specifying that that attacker must still be in play. Is this intentional? If so, if a defender changes while the attacker is not in play, will another Proposed Attack Event be resolved after the current one?--DicePower (talk)

It probably does have one. It probably also requires that the attacker is not mortally wounded, and that there is a non mortally wounded second enemy target to redirect to. However, I don't have evidence confirming or denying this, so I can't put it into the Advanced rulebook yet. What would you suggest I do? --Patashu (talk) 05:33, 19 January 2016 (UTC)
It would be interesting to know, as the outcome is relevant to the explanation of the Combat Preparation Phase. We discussed a similar question on the old model of the Greedy Queue though, and you pointed out that many tests would be needed to establish the "aborts" result.--DicePower (talk) 20:41, 20 January 2016 (UTC)

What is the definition of a defender "dying indirectly"? It seems both the bugged case and the non-bugged case of cancelling combat involve at least one Death Phase following the Preparation Phase.--DicePower (talk)

Dying directly: One Death Phase. Dying indirectly: 2+ Death Phases. I'll make this explicit. --Patashu (talk) 05:33, 19 January 2016 (UTC)

Does the brief placement of Enchantments later in order of play than Health/Attack Auras lead to any significant outcome that wouldn't have resulted if such Auras were instantly moved to the end of the list?--DicePower (talk)

Yup! Watch the very reference I link there for one example. (If you're not sure, I mean https://www.youtube.com/watch?v=Fmfa9NFThsM&t=2m51s .) --Patashu (talk) 05:33, 19 January 2016 (UTC)
I have, but I'm not sure I see the relevance. Wouldn't Dire Wolf Alpha and Grimscale Oracle end up with 2 Health even if Equality was immediately placed earlier in order of play than Stormwind Champion, instead of the inverse being true until initial Aura Updates?--DicePower (talk) 20:41, 20 January 2016 (UTC)
No - because if Equality was immediately placed in its final order (before the Stormwind Champion's Aura), then Hearthstone will recalculate the minion's HEALTH as 2 - and it was also 2 previously, so no change will be made to current Health. It's the very fact that when we apply the Equality enchantment we recalculate HEALTH and find it to be 1, then later update auras, recalculate HEALTH and find it to be 2, that causes the heal. --Patashu (talk) 04:42, 27 January 2016 (UTC)

What are the conditions under which a minion with an irresolvable Battlecry can or cannot not be played? Or is Cabal Shadow Priest's restriction unique?--DicePower (talk)

I think Cabal is the only one. --Patashu (talk) 05:33, 19 January 2016 (UTC)

Is the Pending Destroy tag cleared when a minion is returned to the hand (I was under the impression it remained until the minion entered play again)?--DicePower (talk)

As of Patch 3.2 when this was fixed, the tag is now cleared when the minion is replayed, but remains set until that time. Short edit to make this explicit. --Patashu (talk) 05:33, 19 January 2016 (UTC)

In rule DC2, what is the difference between triggering when a card is drawn, and resolving the card draw? Isn't the former part of the latter?--DicePower (talk)

Think of there being a 'Card Draw' block. It begins with the card being drawn from the deck and placed in your hand, then you queue and resolve on-draw triggers, then the card's on-draw effect is resolved, and now the card draw is resolved. Is there a way I can word it to be clearer? --Patashu (talk) 05:33, 19 January 2016 (UTC)
I changed "Then the card draw itself is resolved" to "When this is complete, the card draw has been resolved" as I think it's a bit less ambiguous. Do these sound interchangeable? I was misunderstand the old wording as saying "the next step in the process is to resolve the card draw."--DicePower (talk) 20:41, 20 January 2016 (UTC)
Sounds fine to me. --Patashu (talk) 04:42, 27 January 2016 (UTC)

"Aura update: Refers to both Aura Update (Health/Attack) and Aura Update (Other) happening. (Sometimes in Hearthstone these can happen without an intervening Death Creation Step, such as between the two Battlecries when Brann Bronzebeard is in effect.)" Does this mean that Brann also runs Aura Update (Other), in addition to Aura Update (Health/Attack)?--DicePower (talk)

It's not tested yet. One way to test it would be to transform Mal'Ganis with Recombobulator while Brann is out, and to see if the Immune effect leaves between Battlecries, or not until the end of the Battlecry Phase. (But I suspect it will leave immediately.) --Patashu (talk) 05:33, 19 January 2016 (UTC)

If the Battlecry is a Discover Effect, do Brann's Aura Update(s) occur during the Battlecry Phase, the Unnamed Step, neither, or both?--DicePower (talk)

Well, Brann is emitting an Aura so long as he's in your control, and said Aura updates (turns on or off for both players) every Aura Update (Other). And it just so happens that for Discover Battlecries, you need to control Brann at the end of the Sequence, instead of just before the Battlecry Phase as usual. --Patashu (talk) 05:33, 19 January 2016 (UTC)
I was actually referring to the additional Aura Update(s) (Health/Attack being the known one) created by Brann, not his Aura turning on and off. However, like we discussed, it's becoming clear that a question like this is one tiny piece of a larger puzzle that we might need a more systematic way of keeping straight.--DicePower (talk) 20:41, 20 January 2016 (UTC)

Does using the bug for Immunizing Gorehowl followed by the opponent playing Bloodsail Corsair (with Brann's Aura active) allow the weapon to survive with -1 durability?--DicePower (talk)

Probably! I'm interested in having this tested, as it would be quite amusing. --Patashu (talk) 05:33, 19 January 2016 (UTC)

Is there an Aura Update of either type, or a (Special) Summon Resolution Step, before/after Instant Weapon Destruction?--DicePower (talk)

Dunno. But further findings seem to indicate that there is some kind of 'Weapon Update Step' that occurs just as any outermost Phase ends, where Hearthstone registers that your Weapon has entered play, changed or left play and performs the appropriate interactions. See https://github.com/HearthSim/hs-bugs/issues/303 for more information. So we could try and figure out the timing of this 'Weapon Update Step'... --Patashu (talk) 05:33, 19 January 2016 (UTC)
Highly bizarre!--DicePower (talk) 20:41, 20 January 2016 (UTC)

If Faceless Manipulator or Druid of the Claw is played, and Commanding Shout resolves before the Battlecry Phase, does the subject lose the Commanding Shout Enchantment when Transforming, and have to wait until the Special Summon Resolution Step to regain it?--DicePower (talk)

Yup! See here: https://www.youtube.com/watch?v=QC8_jNOvdEs&t=12m54s It's actually really cool, so I'll include it as an example in the rulebook. --Patashu (talk) 05:33, 19 January 2016 (UTC)

"Both players play Murloc Tidecallers and then a Faceless Manipulator is played targeting the Murloc Tidecaller. We know that the Faceless Manipulator does not transform until its Battlecry, that Murloc Tidecaller reacts before the Battlecry and does not participate in the "Summon Resolution Step", but the following occurs: We see the Faceless Manipulator instantly transform into a Murloc Tidecaller, play the Murloc Tidecaller sound clip (despite how Faceless Manipulator does not normally do this), then all Murloc Tidecallers buff in reaction, then the Battlecry 'flash' goes off and the Faceless Manipulator copies the post-buffed state." Does any of this occur before Commanding Shout resolves during the initial Summon Resolution Step, or is this bug at least neatly contained in the Battlecry Phase?--DicePower (talk)

I have no idea. This needs to be re-explored now that we have a more modern understanding of mechanics. It probably all occurs inside of the Battlecry Phase - when I was investigating this bug initially, I don't think I ever thought to use an On Play Phase trigger or Summon Resolution Step trigger. --Patashu (talk) 05:33, 19 January 2016 (UTC)

Does Rule DP2 apply if control shifts from Player A to Player A? For example, if Shadow Madness's Enchantment on a doubly-stolen minion triggers (grants it Charge, etc.), but doesn't actually move the minion, does Power Overwhelming still abort?--DicePower (talk)

Wow, good question. I have no idea. It probably does too, because right now steal from Play to same Play has all its side-effects, such as summoning sickness and full zone checking. --Patashu (talk) 05:33, 19 January 2016 (UTC)

I think order of play should be defined in more detail; perhaps it even deserves its own subsection in Advanced Mechanics 101. The only introduction currently offered (without mining other sections of the rulebook to construct a model) is...

"Order of play means the order the Entities each Event/trigger is associated with entered play, from oldest to newest. Minions, Heroes, Weapons, Secrets, attached Enchantments and added Deathrattles are all Entities, and all exist in the same order of play list together."

This doesn't provide any insight into how different categories of Entities are compared to each other. For example, there are at least four ways to compare the timestamps of Entity A which has an inherent effect, and Entity B which creates an Enchantment on Entity C.

1) Compare the instant Entity A entered play with the instant Entity B entered play,

2) compare the instant Entity A entered play with the instant Entity B put the Enchantment into play,

3) compare the instant Entity A entered play with the instant Entity C entered play, or

4) some inconsistent combination of 1-3, depending on the Entities/Enchantments in question.

"But in addition to this, if you play Aviana after Emperor Thaurissan discounts your cards your minions will cost 1, but if Emperor Thaurissan triggers after Aviana enters play, Emperor Thaurissan's effect will be processed second and your minions will cost 0" provides a defeater for "1," and seems to render "3" incoherent (unless some secondary timestamp system is used to track i.e., "order of entering hand," similarly to how Death Events used to reference non-standard timestamps). "You play Ragnaros the Firelord and cast Shadow Madness on an enemy minion. They are queued in this order, and thus Ragnaros the Firelord damages an enemy character, THEN the stolen minion is returned." This appears to decisively rule out "3," leaving us with "2," and "4." However, the Repair Bots VS Shadow Madness example is at odds with "2," because the Shadow Madness Enchantment entered play before Repair Bot #2 entered play ("Your opponent has a Repair Bot (which triggers to heal during the End of Turn Phase). You cast Shadow Madness on it, then Faceless Manipulator on the copy. During the End of Turn Phase, we see all triggers end in order of play - Repair Bot #1's trigger, Repair Bot #1's Shadow Madness ends, Repair Bot #2's trigger, Repair Bot #2's Shadow Madness ends."). This leaves us with "4."

Substituting effects granted by Enchantments (i.e., Explorer's Hat) in place of inherent effects is an addition layer of complexity that merits exploration and explanation.

Auras seem a bit simpler. The only thing needed to be determined is whether the timestamp for an Aura reflects the moment the emitting minion entered play, or the moment the Aura began to emit. To easily check this, just play Kel'Thuzad then Emperor Thaurissan, during a turn in which friendly Naga Sea Witch(s) and/or Aviana(s) is killed. During the End of Turn Phase, Emperor Thaurissan's Enchantments will be created after the emitting minion's Summon, but before the application of its Aura.--DicePower (talk)

Ok, here's a longer description of play order. If you like it, I can put it into the Advanced rulebook.
  • The Timestamp of a Hero, Minion, Secret or Weapon is the last time it entered play (either by being played, force played or summoned.)
  • The Timestamp of an Enchantment (created by a spell, Battlecry or triggered effect) (including added Deathrattles and added Enchantments' triggers) is the time it was generated.
  • The Timestamp of a trigger or Aura that is inherently part of an Entity (such as Aviana's aura, Shadow Madness's trigger to revert in the End of Turn Phase or Sword of Justice's trigger) is the same as that Entity's timestamp. (Note that while Auras technically make and attach Enchantments to other Entities, the timestamp of those added Enchantments is still the timestamp of the Aura that created them, no matter what.)
  • Whenever it is ambiguous how two Events, two Triggers or two Enchantments/Auras should be ordered (they have identical priority and are created or Queued (whichever is appropriate) simultaneously), they are sorted by play order, from oldest to newest.
Unanswered questions:
  • What is the Timestamp of something in your hand? For example, if you have Aviana and any giant (such as Mountain Giant), under what circumstances does Aviana apply first?
  • As you mention with your Aura example, it's technically not known if the timestamp of an Aura is the same as when the emitting Entity entered play, or if it's the timestamp of the first time it emitted. I expect it to be the former but we could test it to make sure.
--Patashu (talk) 09:40, 18 January 2016 (UTC)
This is very good info, and may be useful to include. I especially like the note about Auras creating Enchantments (perhaps such a note could be added to the Auras section as well). The term "timestamp" no longer appears in the rulebook, since Force Play has been done away with, so it's just a matter of deciding if it's worthwhile to introduce an additional vocabulary word into the mix (it is one we use commonly in live mechanics discussion, so perhaps it's worth the addition). I'm not sure it goes the whole distance to establish an order of play model though, since I can reword my example using timestamp vocabulary as you defined it in the first three bullet points, without the uncertainty dissappearing. I'll also invent a few other terms to try to state the problem more clearly.
Instantiated trigger: A trigger that constitutes one element of a trigger Queue. Comes into existence when the Queue is created, and is destroyed after the Queue resolves. Examples: "The first (instantiated) trigger aborted, but the second one resolved successfully, mortally wounding all enemy minions." "There were too many (instantiated) triggers in the Queue to keep straight."
Base trigger: A trigger that has the potential to be instantiated. When the timing and queuing conditions of an in play base trigger are met, the base trigger gives birth to an instantiated trigger. Does not come into or go out of existence. Examples: "I crafted a Cult Master because its (base) trigger is so good." "Casting Arcane Missiles against Grim Patron was a poor idea; when hit twice, its (base) trigger created two (instantiated) triggers, helping the opponent win.
NOTE: Events are always instantiated (technically you could talk about how many base Events are present in a base trigger, but it's awkward and probably not relevant here).
Parent: The Entity that causes a trigger, Event, Enchantment, or Aura to come into play. Examples: "The Aura was turned off shortly after its parent minion was removed from play." "The Enchantment's parent, Shattered Sun Cleric, was later stolen by Mind Control Tech." "The Damage Event's parent, Whirlwind, damaged two other minions as well." "Soul of the Forest parents Enchantments which each grant a base trigger. This base trigger parents an instantiated trigger when the Enchanted minion dies. However, Soul of the Forest is not a direct parent of the instantiated trigger, so is not required to be in play at queue time."
Bearer: This term is used once in the rulebook: "An on-Death triggered effect triggers for a Death so long as its bearer is not dead yet and thus still in play." I assume a good definition is something like "The character a base trigger and/or Enchantment belongs to." Instantiated triggers and Events would not have bearers. Examples: "Deathlord is a favorable bearer of Inner Fire's Enchantment." "That Battlecry can be powerful, but it's bearer's stats are rather unimpressive." "The direct set Enchantment's bearer was Aviana, while its parent was Crazed Alchemist."
When a base trigger is inherent, the parent and bearer are the the same, and no Enchantment is causally present, so we know that when evaluating the relative order of an instantiated trigger whose parent base trigger is inherent to ITS parent/bearer, the relevant timestamp is the parent/bearer's.
So if Entity A is such a parent/bearer, the possibilities become:
1) Compare Entity A's timestamp with the Enchantment's parent's timestamp.
2) Compare Entity A's timestamp with the Enchantment's timestamp.
3) Compare Entity A's timestamp with the Enchantment's bearer's timestamp.
4) Case by case.
The additional layer of complexity I was trying to get at can be simplified to the question, "If an Enchantment is removed from play before its instantiated trigger queues (i.e., Explorer's Hat), is the answer to the above question modified due to possible information loss (of who its parent or bearer was, or its own timestamp) when the Enchantment was moved to the Removed From Game Zone?"
So I think maybe the difficulty still exists when using the language of timestamps (just as when you reworded the Aura question). The possibility for multiple interpretations seems to come in when applying "Whenever it is ambiguous how two Events, two Triggers or two Enchantments/Auras should be ordered (they have identical priority and are created or Queued (whichever is appropriate) simultaneously), they are sorted by play order, from oldest to newest," since it's not always clear (and it seems not always consistent) which timestamps are compared in order to establish that play order.--DicePower (talk) 20:26, 18 January 2016 (UTC)
The timestamp of an added Enchantment (such as Explorer's Hat) doesn't lose its timestamp when it leaves play. (Remember, this is not MtG, Entities don't ever become new Entities/cease to exist due to changing zones.) That's how Deathrattles are able to queue in order of play with Secrets/on-death triggered effects, despite the DR no longer being in play at the moment this occurs.
Right, it's the same Entity after being removed from play. I was thinking more along the lines of parent, bearer, or Entity timestamp (turns out from your below answer that Entity timestamp is the relevant factor) information loss (or information ignorance), analogously to your Dominant Player Bug on Deathrattles explanation. "It looks like the Ancestral Spirit has the controller of it's caster, but gets detached and removed from play when the death processing itself happens. Afterwards, the minion it was attached TO triggers (as opposed to the Ancestral Spirit itself triggering) and does the Ancestral Spirit effect, so it's in the order of play of the minion's controller, not the caster's."--DicePower (talk) 19:50, 20 January 2016 (UTC)
So, let's look at your question.
If Entity A (which inherently has a trigger, maybe it's Ragnaros) and Entity C's Enchantment (maybe it's Shadow Madness) queue simultaneously, they are ordered by the following timestamps: When A entered play, and when C's Enchantment entered play (and was attached to C) respectively.
The answer is the same in other cases - for example, if Entity A is Aviana, and Entity C's Enchantment is an ET tick, if the ET tick entered play before Aviana C costs 1, but if the ET tick entered play after Aviana C costs 0. The answer is also the same if you throw in added DRs, Secrets, etc.
Basically, only Entities have timestamps. If you ask the question 'what is the timestamp of X trigger/enchantment/aura?' then look at the Entity MOST CLOSELY ASSOCIATED WITH X, and use that Entity's timestamp. If an Enchantment is attached to an Entity, anything that Enchantment does (like emit an aura, change Health/Attack/Mana Cost or have a triggered effect/add a Deathrattle) uses the timestamp of the Enchantment, because that is the Entity most closely associated with that effect.
But the curveball is with Auras - do you use the timestamp of the Aura emitter, or the timestamp of when the Enchantment was attached during Aura Updates? (I think we should do tests with Aviana and Shadowfiend, Aviana and Giants, etc. to make sure we're 100% confident about this.) So that's the open question, and answering that will tie everything together neatly. --Patashu (talk) 04:50, 19 January 2016 (UTC)
Excellent! This is very clear, and it seems this mechanic is consistent after all (I was probably misinterpreting the Repair Bot example, assuming Repair Bot #2's Shadow Madness's timestamp would be a copy of Repair Bot #1's Shadow Madness's timestamp, like a Tag, rather than being allocated its own unique timestamp).--DicePower (talk)

Player A (Dominant Player) plays Explorer's Hat on a Clockwork Gnome controlled by Player B (Secondary Player). When Clockwork Gnome dies during a Death Creation Step or through Full Zone Instant Removal, the Explorer's Hat Enchantment is moved to the Removed From Game Zone. When Clockwork Gnome's Deathrattles are queued during the next Death Phase, is the expired Enchantment still controlled by Player A, evoking the Dominant Player Bug, or will the Deathrattles queue in order of play?--DicePower (talk)

Good question. It turns out that Deathrattle triggers using the minion's controller, not the added Deathrattle's controller, when they differ. This is very obscure. See here for more information: http://hearthstone.gamepedia.com/User_talk:Patashu#Dominant_Player_bug_on_Deathrattles.3F
Sweet, this is actually really valuable in widening the scope of order of play exploration.--DicePower (talk) 20:41, 20 January 2016 (UTC)

Do both Aura Update (Health/Attack) and Aura Update (Other) occur during an On Play Phase with Overload?--DicePower (talk)

Untested, but I guess that Aura Update (Other) is likely to also happen. What's a good way to test? Summoning Stone? --Patashu (talk) 05:33, 19 January 2016 (UTC)

How does Feign Death work with multiple Deathrattles that Summon minions? Can the new minions end up being Summoned in positions not adjacent to their respective Deathrattle minions?--DicePower (talk)

It looks like Feign Death always puts its DR summoned minions on the far right, regardless of what the DR is. --Patashu (talk) 05:33, 19 January 2016 (UTC)

If a minion dies through Full Zone Instant Removal, is its board position recorded at its time of death, or in the following Death Creation Step alongside any other minion killed the orthodox way?--DicePower (talk)

I'm guessing when the Death Event was created (so the moment it left play). Untested though. --Patashu (talk) 05:33, 19 January 2016 (UTC)

Player A controls Knife Juggler (3/1 with Commanding Shout Enchantment) and Voidcaller. Player B controls a stolen Knife Juggler (3/1 with Commanding Shout Enchantment), Dreadsteed, and Vaproize. Voidcaller dies to Vaporize, Summoning Dreadsteed. What happens?--DicePower (talk)

You're trying to do something clever and it won't work. Eventually a knife will hit a KJ or a Hero (minions at 1 Health with Commanding Shout aren't ignored by random negative effects, ONLY by Bouncing Blade) and the chain reaction will stop. --Patashu (talk) 05:33, 19 January 2016 (UTC)
I forgot that Hearthstone has heroes.--DicePower (talk) 20:41, 20 January 2016 (UTC)

Minor nitpick/question. In the example given under Rule 7, "You have a Duplicate put into play, a Sludge Belcher injured to 4 Health and Dark Cultist, played in that order. Your opponent casts Flamestrike.... If the minions were played the other way around [making Dark Cultist the oldest minion], the buff [from Dark Cultist] would have no target at time of activation and the Slime would remain at 1/2." Wouldn't the Dark Cultist's deathrattle queue first and buff the Sludge Belcher out of range of Flamestrike (Belcher first taking 4 damage, being reduced to 0 health and marked as mortally wounded, then receiving the +3 health buff from Dark Cultist)?--Pestevan (talk) 16:17, 18 October 2016 (UTC)

No, because Hearthstone does not check for deaths in the middle of a Phase (in this case, the Spell Text Phase). The entire spell 'Flamestrike' must resolve before any of the damaged minions can die, and all the minions will die simultaneously in the Death Creation Step before on-death triggers queue and resolve in the following Death Phase. --Patashu (talk) 23:06, 18 October 2016 (UTC)

Questions requiring testing[edit source]

QUESTION: Even though there is nothing in Hearthstone that queues during the After Summon Phase if the subject of the Sequence has been removed from play, does the empty Phase still exists and allow post Phase steps to resolve?

TEST: Player A controls Illidan Stormrage and Knife Juggler. Player B controls Unstable Ghoul (1/1), Mirror Entity, and Starving Buzzard. Player A plays Angry Chicken. During the Flame of Azzinoth's nested After Summon Phase, Knife Juggler lands on Unstable Ghoul, ultimately killing Unstable Ghoul and Angry Chicken. During the outermost After Play Phase, Mirror Entity Summons a copy of Angry Chicken, but there is no Summon Resolution Step following this Phase, so the Summon Event cannot yet be resolved. Since the subject of the outermost Sequence has been removed from play, the outermost After Summon Phase is said to be "cancelled entirely," but this would imply that post-Phase steps, including the Special Summon Resolution Step, are also skipped. The outcome of the test depends on whether or not Starving Buzzard queues and resolves for the copied Angry Chicken following the would-be After Summon Phase.--DicePower (talk)


QUESTION: If a trigger with a queuing condition is controlled by the Secondary Player, is the condition evaluated when the Secondary Player's side of the Queue begins to form, or when the entire Split Queue begins to form?

TEST: Player A (Dominant Player) controls Mekgineer Thermaplugg and Knife Juggler. Player B (Secondary Player) controls Acolyte of Pain and Chillmaw, has no Dragon minions in hand, but has a Dragon minion at the top of the deck (perhaps it is the last card in the deck). Chillmaw dies, and during the Dominant Player's side of its Death Event Queue, Mekgineer Thermaplugg Summons a Leper Gnome, triggering Knife Juggler, which lands on and triggers Acolyte of Pain, which adds a Dragon to Player B's hand. The outcome of the test depends on whether or not Chillmaw queues for Player B.--DicePower (talk)


QUESTION: How is an Immune weapon with 1 or less durability affected by Bloodsail Corsair?

TEST: Player A controls Gorehowl and Wisp. Player B controls Misdirection and Explosive Trap. Player A attacks Player B's hero. Misdirection and Explosive Trap trigger, killing Wisp and cancelling combat in the following DCS/Death Phase, and allowing Gorehowl to remain Immune. On the following turn, Player B plays multiple copies of Bloodsail Corsair, or Bloodsail Corsair with Brann Bronzebeard's Aura active. The outcome of the test depends on the resulting durability of Gorehowl.--DicePower (talk) 19:09, 26 January 2016 (UTC)


QUESTION: If a minion with a queued trigger is stolen before it triggers, but the minion's source and destination are both a single player's Battlefield Zone, does the trigger still abort?

TEST: Player A (Secondary Player) controls Goblin Sapper and Sylvanas Windrunner. Player B (Dominant Player) casts Shadow Madness, targeting Goblin Sapper, and then targets it with Power Overwhelming. Sylvanas Windrunner dies, granting to Player A control of Goblin Sapper, Shadow Madness's Enchantment, and Power Overwhelming's Enchantment. During the End of Turn Phase, both Enchantments queue for the Secondary Player, but Shadow Madness resolves (pseudo-stealing Goblin Sapper) before Power Overwhelming can trigger. The outcome of the test depends on whether or not Power Overwhelming triggers.--DicePower (talk) 19:09, 26 January 2016 (UTC)


QUESTION: If a minion with an Enchantment-granted queued trigger is stolen before it triggers, but the Enchantment's (and not the minion's) source and destination are both a single player's Battlefield Zone, does the trigger still abort?

TEST: Player A (Secondary Player) controls Sylvanas Windrunner (5/3), Lightwell, Auchenai Soulpriest, Target Dummy, and Spider Tank. Sylvanas Windrunner is Player A's only damaged character. Player B (Dominant Player) controls Spellslinger. Player A plays Mimiron's Head, and then casts Corruption, targeting Spellslinger. Both players pass. During the Start of Turn Phase, Lightwell, Mimiron's Head, and Corruption queue for the Secondary Player. Lightwell mortally wounds Sylvanas Windrunner. In Mimiron's Head's Forced DCS/Death Phase, Sylvanas Windrunner is killed, and steals Spellslinger, but Corruption's Enchantment is only pseudo-stolen. The outcome of the test depends on whether or not Corruption triggers.

NOTES: Dominance may not matter in the above test, since I doubt Corruption shares Power Overwhelming's ability to re-queue for the Secondary Player after aborting once.--DicePower (talk) 21:43, 30 January 2016 (UTC)


QUESTION: If a queued trigger's associated Entity's timestamp is updated before it triggers because the Entity was removed from and re-entered play, does it still trigger, and if so, is the order of the Queue updated to reflect the new timestamp?

TEST: Need a Demon with a start of turn trigger to test this.--DicePower (talk) 05:58, 4 February 2016 (UTC)

[10:34] <PattuX> i can only think of: pagle (1 minion) into flame leviathan, mortally wounding ambusher + new deathwing (2, 3), mimiron's head triggers, creating a forced death phase (+3 mechs = 6), scaled nightmare returned to hand and put into play again (7). cant have any additional start of turn triggers out, boardwise, to see what happens to the order. but at least we

[10:34] <PattuX> can tell now whether it triggers at all.

[10:37] <Patashu> instead of pagle you can use acolyte of pain + auchenai + lightwell or similar mechanism

Old Gods to the rescue! Instead of having a Nat, your opponent can have a Nat Darkfisher (and he must be the dominant player), making room for a Lightwell played after Scaled Nightmare. If Scaled Nightmare Triggers again, we are able to see whether it used a new timestam or not. --PattuX (talk) 00:14, 2 May 2016 (UTC)


QUESTION: If a created Event's associated Entity's timestamp is updated before its Event Queue begins to resolve (because the Entity was removed from and re-entered play), do triggers still queue and trigger for the Event?

TEST: Player A controls Voidcaller (3/1) and Undertaker, and has Dreadsteed (but no other unnecessary Demon minions) in hand. Player B controls Unstable Ghoul. Player A plays Anub'ar Ambusher then Voidcaller, damaging them to 5/1 and 3/2, respectively, and passes. During either player's turn, the 3/2 Voidcaller and Unstable Ghoul battle, killing Unstable Ghoul and mortally wounding both Voidcallers and Ambusher in the following DCS/Death Phase. When they die, Dreadsteed will be Summoned from, returned to (RNG please!), and re-Summoned from, Player A's hand. The outcome of this test depends on whether Undertaker triggers once or twice for Dreadsteed.

IF YES, ADDITIONAL QUESTION: If a created Event's associated Entity's timestamp is updated before its Event Queue begins to resolve (because the Entity was removed from and re-entered play), is the order of the Event Queue updated to reflect the new timestamp?

TEST: Player A controls Voidcaller (3/1), Starving Buzzard (3/4), and Unstable Ghoul. Player B has Dreadsteed (but no other unnecessary Demon minions) in hand. Player A plays Savannah Highmane, then Anub'ar Ambusher, then Voidcaller, damaging them to 6/1, 5/1, and 3/2, respectively, then ensures Flame Leviathan is within the top two cards of the deck and passes. Player casts Commanding Shout, steals (in any order) both Voidcallers (the 3/2 must be stolen via Shadow Madness or otherwise granted Charge) and Ambusher, and attacks Unstable Ghoul with the 3/2 Voidcaller, killing Unstable Ghoul and mortally wounding both Voidcallers, Highmane, and Ambusher in the following DCS/Death Phase. When they die, Dreadsteed will be Summoned from Player B's hand, two Hyena will be Summoned for Player A, and Dreadsteed will be returned to, and re-Summoned from, Player B's hand. During the following Summon Resolution Step, Starving Buzzard will draw Flame Leviathan, which will mortally wound Dreadsteed if its first Summon Event hasn't resolved yet. The outcome of this test depends on whether or not Dreadsteed is killed in the following Death Creation Step.

NOTES: At least the first test can be repeated using Unearthed Raptor to encapsulate the Deathrattles of both Voicallers and Ambusher in a single Death Event. This can only be done with one Voidcaller and Ambusher in the second test, or Highmane will lose its relative order of play. It would also be interesting to run versions of these tests where the timestamp is updated after the Event Queue begins to resolve, but before the specific Event related to the updated timestamp resolves, if it's possible to do so.--DicePower (talk) 18:15, 5 February 2016 (UTC)


QUESTION: Do non-Discover Battlecries resulting from Brann Bronzebeard's Aura trigger in the Battlecry Phase, or as high priority triggers in the following Summon Resolution Step?

TEST: Need one of several possible card templates to test this beyond visual reasoning.--DicePower (talk) 17:42, 6 February 2016 (UTC)


QUESTION: Does Brann Bronzebeard re-trigger Discover Battlecries even if Brann is not in play during the Battlecry Phase, and is never in play at the same time as the subject of the outermost Sequence?

TEST: Player A controls Starving Buzzard and Unstable Ghoul (1/2), and has Sea Reaver at the top of the deck. Player B controls Mirror Entity, Knife Juggler, and Deathlord (2/1). Player A plays Jeweled Scarab. During the initial Summon Resolution Step, Starving Buzzard draws Sea Reaver, ultimately killing Jeweled Scarab. During the outermost Battlecry Phase, Jeweled Scarab's Battlecry begins to resolve, and Player A selects a card to Discover. During the outermost After Play Phase, Mirror Entity Summons a copy of Jeweled Scarab, triggering Knife Juggler during its nested Sequence. Knife Juggler lands on Unstable Ghoul, ultimately (after the After Summon Phase) killing Unstable Ghoul and Deathlord. During the following Death Phase, Deathload Summons Brann Bronzebeard for Player A. During the Unnamed Step, Jeweled Scarab's Battlecry finishes resolving, and the selected card is added to Player A's hand. The outcome of this test depends on whether Jeweled Scarab's Battlecry now triggers again.--DicePower (talk) 17:42, 6 February 2016 (UTC)

Outdated tests[edit source]

If you find a test that no longer applies due to a card change or mechanic change, remove it from the Advanced rulebook and add it here:

http://hearthstone.gamepedia.com/User_talk:Patashu/Advanced_rulebook_outdated_tests

If you feel that an equivalent example needs to be added in its place, contact me or post on the talk page and we can figure out something to add. --Patashu (talk) 22:36, 2 December 2015 (UTC)

Removing cards from play[edit source]

I'm hoping someone can confirm my rough understanding of the destroy, discard and remove from deck effect processes. It seems the difference primarily relates to the zone in which the card was when it was removed from play. A minion is only considered to have "died" if it was in the Play Zone (hence Deathrattles not activating when discarded or overdrawn, but still being activated when attempting to return a minion to a full hand). Likewise, a minion is only considered to have been discarded when removed from the Hand Zone, hence Fel Reaver, overdraw and Tracking don't activate Tiny Knight of Evil (however cards like Burrowing Mine can still remove themselves from play when in the hand without being counted as discards).

The notes in the Zones section regarding Tracking seem to imply that the reason for its quirky behaviour is that instead of drawing cards, it moves three cards to the Set Aside Zone, from where the player chooses one - which is moved directly to the hand, without being normally 'drawn', hence not triggering Chromaggus, etc - while the other two are removed from play - directly from the Set Aside Zone, hence not counting as a discard (which would be from the hand) or as a remove from deck effect (which would also be revealed to both players).

Does this sound about right? -- Taohinton (talk) 06:20, 3 March 2016 (UTC)

Sorry for the delay - for some reason I'm not getting email notifications when pages ON my watchlist are being edited. Do you have any advice to fix this?
Minions only dying if they go to the graveyard from play is correct. But there's an unusual distinction between being discarded from the hand (Discard effects; triggers Tiny Knight of Evil) and being destroyed in the hand, using Shadowflame + Violet Teacher/Anub'ar Ambusher/Knife Juggler shenanigans for example, which does NOT trigger Tiny Knight of Evil, but STILL doesn't count as a discard. This is a weird edge case; it may or may not be a bug.
The Tracking explanation is correct - since it doesn't put cards in your hand and then discard them, it doesn't trigger on discard effects, despite using the word 'discard' in its card text. --Patashu (talk) 23:17, 18 April 2016 (UTC)
Thanks for the reply. The notification problem has been discussed here. It's not clear if it's now resolved, or if it's just intermittent. If you're still getting missed notifications very recently, you might want to post a note there. -- Taohinton (talk) 02:25, 19 April 2016 (UTC)

Hero Power[edit source]

The advanced rulebook states that in the Hero Power Phase "your Hero Power takes effect then is exhausted". Visually, the Hero Power is exhausted (flipped over) before the Hero Power takes effect: is this only a visual "bug" (concerning the difference in timings caused by client-side animations in opposition to server-side processing), or should we write "your Hero Power is exhausted and takes effect" in the rulebook instead of "your Hero Power takes effect then is exhausted"?Elekim (talk) 04:49, 21 June 2016 (UTC)

I'd trust whatever power.log says more than I'd trust the visuals themselves ( https://github.com/jleclanche/fireplace/wiki/How-to-enable-logging ). --Patashu (talk) 05:12, 21 June 2016 (UTC)

Why current Mana can be negative[edit source]

How do effects that grants empty Mana Crystals (such as Wild Growth) interact with tags? Do they increase both the RESOURCES tag and the RESOURCES_USED tag in order to leave the current mana unchanged?

How do effects that destroy Mana Crystals (such as Felguard) interact with tags? Do they decrease both the RESOURCES tag and the RESOURCES_USED tag in order to leave the current mana unchanged? Elekim (talk) 09:45, 5 March 2017 (UTC)

I think so? You are welcome to play the cards yourself (such as against Innkeeper), with logging on, and check the log to see what is happening under the hood. If you discover anything new (or even the possibility of a new bug/interaction), it will be interesting to know. Instructions here: https://github.com/jleclanche/fireplace/wiki/How-to-enable-logging --Patashu (talk) 22:21, 5 March 2017 (UTC)
I run some tests, and updated the section accordingly. The only card I wasn't able to test is Kun the Forgotten King since I don't own it. I suppose it only reduces the RESOURCES_USED tag, either to 0 or to the value of the OVERLOAD_LOCKED tag (I have no idea of how it interacts with Overload...): can someone test it? Elekim (talk) 08:09, 15 March 2017 (UTC)
Don't have Kun either! I'll add a TODO. --Patashu (talk) 10:57, 15 March 2017 (UTC)

Add two table about the after play secret[edit source]

Check out more infomation in: https://www.youtube.com/watch?v=naRlE4DhRfY&feature=youtu.be

before 7.1:

Triggers Pending Destroy(Sacred Trialor Snipe+Acidmaw) Mortally wounded (Snipe or Mirror Entity + Knife Juggler) Potion of Polymorph Brann Bronzebeard + Injured Blademaster Lord Jaraxxus
Repentance Y Y N N Y(Hero Jaraxxus)
Sacred Trial Y Y N N Y(Hero Jaraxxus)
Snipe Y Y N N Y(Hero Jaraxxus)
Mirror Entity Y Y Y(Orginal minion) Y(4/7 Blademaster in Graveyard) Y(Minion Jaraxxus)
Potion of Polymorph Y Y / N N
Hidden Cache Y Y Y N N
After summon trigger Y Y N N N


After 7.1:

Triggers Pending Destroy(Sacred Trialor Snipe+Acidmaw) Mortally wounded (Snipe or Mirror Entity + Knife Juggler) Potion of Polymorph Brann Bronzebeard + Injured Blademaster Lord Jaraxxus
Repentance N Y Y(sheep) N Y(Hero Jaraxxus)
Sacred Trial Y Y Y(sheep) N Y(Hero Jaraxxus)
Snipe N N Y(sheep) N Y(Hero Jaraxxus)
Mirror Entity N N Y(sheep) N Y(Minion Jaraxxus)
Potion of Polymorph N N / N N
Hidden Cache Y Y Y N N
After summon trigger Y Y Y(sheep) N N

--10925 (talk) 02:43, 10 March 2017 (UTC)

Bolded all entries that change between the two tables. I'll make wiki edits (advanced rulebook and individual cards) when I have time. --Patashu (talk) 03:04, 10 March 2017 (UTC)
In bullet point form:
  • Repentance still triggers on mortally wounded minions, but now ignores pending destroy minions. This is a bug (it should consistently do one or the other - you can argue it consistently should because buffs effect mortally wounded/pending destroy minions, or that it consistently shouldn't because negative effects ignore mortally wounded/pending destroy minions. AFAIK there is no precedent either way, so Repentance will set the precedent).
  • Sacred Trial still triggers on mortally wounded/pending destroy minions. This is a bug (it should ignore them).
  • Potion of Polymorph now interacts as expected with all After Play and After Summon triggers (the sheep is affected by the other trigger).
  • Snipe and Potion of Polymorph now ignores pending destroy and mortally wounded minions. (It used to trigger on them.)
  • Mirror Entity now ignores mortally wounded, pending destroy and dead minions. (It used to trigger on them, and could even make copies of minions in the graveyard.)

--Patashu (talk) 03:22, 10 March 2017 (UTC)

Stepping down[edit source]

  • Just as a note, I won't be doing much (any?) work on the advanced rulebook/wiki from now on. I was already not playing or reading about the game, and with all the mechanics change in the last few patches/hotfixes I'm not operating on a recent enough model of gameplay mechanics to figure out how things work and are meant to work anymore. I am super happy with all the groundwork I laid and everything I have done for the community, though :D If Hearthsim/hs-bugs peeps want to take over (either to do incremental changes or to start a total rewrite) feel free to. It remains to be seen what the future of Hearthstone mechanics and science looks like. --Patashu (talk) 11:05, 19 June 2017 (UTC)
  • Oh yeah, something I forgot. If a section of the advanced rulebook, or a specific test, becomes outdated due to bug fixes or mechanics changes, but it's still interesting for historical reasons, you can move it to this page: http://hearthstone.gamepedia.com/User_talk:Patashu/Advanced_rulebook_outdated_tests If it's not worth preserving, feel free to delete it. (And if you think it should be better organized, or there should be a link to it from the advanced rulebook, feel free to add one.) --Patashu (talk) 23:31, 28 June 2017 (UTC)

Hearthstone mechanics have definitely become more numerous with each expansion, and hidden gameplay changes are always so hard to keep track of, it's amazing to see that the Advanced rulebook has continued for so long. It's awesome to see the work you've put into the rulebook over the years, hopefully you'll still visit the wiki every now and then. :) Aegonostic (talk) 08:46, 29 June 2017 (UTC)

Knife juggler + Volatile elemental + Razorfen hunter[edit source]

This (https://hsreplay.net/replay/Lmk6P5pFS6DJLKE33cL3UF) shows an interaction between the above three cards: Razorfens battelcry summons Boar, Boar is summoned and triggers Juggler, Juggler hits Elemental, Elemental's deathrattle triggers and hits Razorfen, Razorfen as a result never finishes hit summoning phase and Juggler does not get to do a second juggle. I think this is too advanced to include on any of the pages of the three affected cards, but it might fit somewhere on this page. However, I do not know the Advanced Rulebook well enough to figure out where to put it (or even whether it is already covered). -- BigHugger (talk) 13:26, 20 August 2017 (UTC)

It's covered in the first line of section 3.4: "When you Play a minion, the following Sequence takes place, which ends early if the minion leaves play (such as due to death)"
--PattuX (talk) 14:02, 20 August 2017 (UTC)

Upcoming Rule Changes[edit source]

Hello!

Hearthstone is having a few rule changes coming up in the next patch, as found here.

What I think is happening:

1. Effects now have a "memory" of previous board states, something that has never happened before. This could lead to quite a few differences in many interactions, some of which I believe would become very unintuitive.

2. I believe the Quad Queue is being removed, and a double queue is being added. The Dominant Player Bug, however, is not being fixed, and it is stated like it's a feature.

Another comment I found on the page: "@EightByte: The issue with Unlicensed Apothecary and evolve is getting fixed in the next update, but it's unrelated to this mechanical update."

Does this mean that the Summon Resolution Step is getting an overhaul as well?

Maxman013 (talk) 06:09, 16 October 2017 (UTC)

Damage and Healing[edit source]

Rule DH6 says that sources of damage events lose stealth. As of patch 11.0.0.23966 (April 2018), this is no longer the case -- stealth is only lost by attacking. The patch notes (link omitted because wiki filter thinks I am a spambot if I put it in) cite an example of a stealthed knife juggler not destealthing when throwing knives. I have also had a jungle panther created by wandering monster not lose stealth after killing the firefly that attacked it. 121.44.83.219 05:03, 18 June 2018 (UTC)FTWW

Question about Swiss/Swedish format[edit source]

QUESTION: In the section of this article titled "Advanced mechanics 101 (READ THIS FIRST)" there is a passage that says, "If a player concedes in a Swedish style bracket, the result is .5 - .5 in player score." Is it possible that this is actually a Swiss scoring system (not Swedish)? If it is correct as written, could some information be provided on the "Swedish style bracket" as I am unable to discover any information about such. Thanks.

In the latest official 2018 Hearthstone World Championship tournament rules pdf, info on the Swiss format can be found. Also, this wiki has info on the Swiss format. Also, Wikipedia also has a good article on the Swiss format here. I have never heard of a Swedish format, so I don't know if such a thing exists. Aegonostic (talk) 21:07, 25 October 2018 (UTC)