Repairing Game state (Specific Example)

[M.Solymossy]

My question is based on a vintage gamestate I observed the other day at an event.  The 'judge' who wasn't even a judge, ruled it, and I was informed today that it was wrong.  Here is what went down.

It's the top 8, Game 3. Player B has a goblin welder in play and he has MindSlaver in the graveyard.  Player A contemplates for about 5 minutes about welder's resolution, and then on his upkeep contemplated  for 5 minutes before casting vampiric tutor.  Player A casts vampiric tutor on upkeep and Player B  casts Shadow of doubt.  Playerr A casts Pyroblast.  Player B Misdirect's it. Player A then manadrain's Misdirect.  Player B uses force of Will.  Player A force back.  Now Player B activates his goblin welder, bringing in Ruby, and pyroblast's back.  It's important to note that player Bhad 1 card in hand (Yawgmoths Will) and there are >8 people watching from behind.  So player A goes to draw, and draw Brainstorm.  He then cast yawgmoths will.

THIS is where we have the problem, because  player B  had JUST cast the welder, and Because of the incredible thought and time in this game, both players just overthought the situation.  IF vampiric tutor resolved at the end of the chain, Player B would have gotten tendrils of agony, and easily won the game on storm count.   Player B argues for about 5 minutes, before it is pointed out by player B that he WOULD HAVE slaved player A if he had a turn, which is why player A was  trying to combo him out.  They go over his mana, and he realizes that Player A was right, but he still argues because it's an irreversable game state.  The judge rules that Player B is right, and that nothing can be reversed. 

The problem I see is that after shadow of doubt was cast, Player A used a fetchland, then realized that Shadow of doubt was in effect until end of turn, and Player B forced him to shuffle his library.  The only problem is the draw step and the shuffle.  But since the deck was randomized anyways due to the resolution/shuffle of Vampiric Tutor, this shouldn't be a problem to repair, correct?

Can the judge take the word of the 8 or more spectators, most of which didn't even know the players, as to what cards were drawn?  Can the gamestate be reversed?   The game was won or lost on that turn.

[Dante]

Can the judge take the word of the 8 or more spectators, most of which didn't even know the players, as to what cards were drawn?   

Any sanctioned judge who relied on "spectators" to try and repair the game state like that should be stripped of any judgeship immediately.

[M.Solymossy]

So this bars the question of, what the correct ruling should have been.

[Gort32]

Put it this way: the word "spectator" is mentioned 9 times in the DCI floor rules (http://www.wizards.com/dci/downloads/DCI_UTR_2jan06_EN.txt).  Most of these basically read "Spectators aren't allowed to do squat" (talk with players, take notes, etc) with the following exception:

Spectators and members of the press who believe they have observed rules violations should inform a judge, but must not interfere with the match.

There is no clarification to this rule.  It doesn't say anything about a judge taking a spectator's word for anything - they are spectators, not players, and have no responsibility within the game.

There was a decision made after all of this took effect but before it was caught - Brainstorm was cast.  After this point it simply isn't possible to fully restore the gamestate.  As such, the judge should just hand out warnings and/or game losses as necessary depending on the REL (worse against Player B, but with at least a warning for Player A for not keeping track of game state).

Later, if Player A fetches while under a Shadow of Doubt, he is not allowed to search but must still shuffle. 

Sadly, an argument of "I would have done . . ." has no meaning in the rules either...

(disclaimer: I am not a registered judge but have been fairly and unfairly slapped by many real judges)

[M.Solymossy]

Brainstorm was drawn for turn, but never cast.

[Anusien]

A judge is not going to not listen to the word of spectators if they have relevant information, but it's up to the judge to determine the steps of the gamestate for themselves.  Now, let's look at everything that happened legally.  As far as I can tell based on your explanation, the only game actions took after the activation of Goblin Welder were Player B casting Pyroblast, Player A drawing a card, and Player A casting Yawgmoth's Will.  If both players can correctly identify the card drawn for the turn (and their story lines up with the spectator's great, but I'm going to take the word of both players more strictly since we know they're paying more attention), then we'll back up to before Welder was activated (putting the card back but not hsuffling, bringing Player B's fetchland back and shuffling and granting him +1 life), undo the Weld, and resolve the stack.

Now what if both player's can't agree on which card is drawn?  Then I'd talk to the spectators.  I'd probably take a few of them aside and ask them individually which card was drawn that turn, both to make sure they don't give away any information (don't know how much of this was done publically) and to see how certain they are away from the group.  If all the spectators agree and at least one of the players agree, I'm probably going to go with that story and put that card back then resolve Brainstorm.  If I can't find any sort of consensus on which card it is, I'm going to randomly put one back as part of backing up so as to restore the proper game state.

[Komatteru]

I think using spectators to corroborate facts from players is ok, but if there is a disagreement between the two, then you have to go with what the players' say, as they were the ones actually involved in the game.  Asking a spectator "Player A says he drew Brainstorm for the turn, did you see it?" or "Did you see the card Player A drew for the turn?" for comparison to what Player A says is one thing, but saying, "OK, I have dilemma, and I want you to tell me what happened, and I'll fix it according to that." is something completely different.

Another thing is that the judge should have told Player B to shut up while he was sorting out the game state.  If Player B was arguing for an irreversable game state--that was largely created because of his action--then that's grounds for suspicion.  It's like the defense putting a witness on the stand, when they know full well that the witness is going to say something that has been ruled inadmissable, then creating a motion for a mistrial.  Essentially, players are not allowed to present an argument to a judge before he's made his ruling first, because that can bias judges and short them the opportunity to conduct a thorough and fair investigation.  In fact, a player making repeated arguments to a judge over how to rule or what penalty to apply is grounds for an unsportsmanlike misconduct warning or game loss (if the badgering is really severe and unacceptable).

I'm going to randomly put one back as part of backing up so as to restore the proper game state.

That could actually be worse. You need to be really careful how you use that because if VT was cast with a certain card in hand, and you just changed the cards in hand, then maybe that player wouldn't want to cast VT anymore, for instance.

[Anusien]

I'm aware that putting a random card

I'm going to randomly put one back as part of backing up so as to restore the proper game state.

That could actually be worse. You need to be really careful how you use that because if VT was cast with a certain card in hand, and you just changed the cards in hand, then maybe that player wouldn't want to cast VT anymore, for instance.

According to The Penalty Guidelines, under "Drawing Extra Cards" has this to say:

Correcting the problem always involves putting the extra card (if known, or a random card from the player's hand, if not known) on top of the deck. This ensures that a player does not receive an advantage from a free shuffle (which he or she would receive if the card were shuffled back into the deck) and that the game will be impacted as little as possible.

While I'm aware that the real problem is a Procedural Error -- Major, part of resolving that game state (if possible) is correcting for an extra card draw.  The judges don't have a responsibility to pander to the players in this circumstance, especially since there's a high possibility of the player drawing Yawgmoth's Will and trying to make it seem like he drew that card instead.  In order to keep the player from being harmed here, he should pay more attention to the board state; a broken game state is the fault of BOTH players.

I'm aware that randomly putting a card is not the ideal situation, but I'm not going to have any sympathy.  If we can't determine which card was the one drawn that turn, we either have a choice to end the current game or try and repair the game state.  In this circumstance I feel the game state is repairable, so I'm going to do so.

[Komatteru]

If we can't determine which card was the one drawn that turn, we either have a choice to end the current game or try and repair the game state.  In this circumstance I feel the game state is repairable, so I'm going to do so.

However, if you decide to put a random card back with VT on the stack, you don't repair the game state.  You make it worse by creating a completely different one similar to the original one.  A hand with Brainstorm vs. one with Yawgmoth's Will is completely different for the purposes of resolving Vampiric Tutor.  You've actually further disrepaired the game state, as you probably changed the decision the player was going to make by changing that card in hand. If you can't determine the card to go back, in this case, you'd have to leave it how it is.

Putting back a random card is a good solution when the card is going to be drawn for sure.  With VT, the card you put back will NOT be drawn, as VT will shuffle the library and send the gamestate into a state of further disrepair.

[Moxlotus]

Anusien--it doesn't appear that you actually repaired the game state--you just changed it.  Not only did you not repair it--you have a 50/50 chance of making it worse.

Why can't you just ask the spectators?  I know the rules don't say you have to--but they don't say you can't.  Would you not do it for the sake of not doing it?  If they all agree, then why wouldn't you take their word?

[Komatteru]

You can't always rely on spectators since they weren't involved in the game directly--they could be biased for one side or the other, and that you can't determine.  You know whom the players each think is right...unless your players are daffy...

You might be able to use spectators for corroboration, but not as the sole source of info.  You still have to take it with a grain of salt.

[Toad]

I am sorry, but after reading all that twice, I still do not understand what is the problem we are trying to adress here. Could someone enlighten me?

[Mr. Nightmare]

Player B uses his turn to cast Welder, with a Slaver and a Ruby in the yard.  Welder is Summoning Sick.

Ends his turn.

On Player A's upkeep, he Vamps.  B responds with Shadow of Doubt.  A counterwar ensues, during which Player B uses the freshly cast Welder to bring in Ruby and uses Ruby to REB to win the counterwar.  Player A is left with YawgWill in hand.  Vamp did not resolve, so without a shuffle A draws his card, which is Brainstorm. He casts Will, they then both realize what happened (a sick Welder was used to win a counterbattle).  Issue:  Winning the battle over Vamp gets Player A Tendrils for the win.  Losing it means Slaver is active next turn and he probably loses.

Is this game state repairable, and does spectator assistance (from behind Player A) make any difference?


Hope that helps.

[President Skroob]

I don't understand why it is difficult to repair this gamestate...

Is the bone of contention the fact we can't be sure whether he drew a Brainstorm without asking spectators? I'm sure the player would say that he drew Brainstorm and the spectators would agree.

The Ruby goes back in the graveyard, the item welded out for it comes back in, Welder is untapped, REB is back in hand, Brainstorm is put back on top of the library, but the player doesn't have an advantage from knowing the Brainstorm is there because Vampiric resolves immediately without the REB to end the counterwar and the library is shuffled upon its resolution. Sure it looks like a lot of steps, but it all seems very simple and obvious.

Am I missing something?

All the facts seem to be here for us.

[diopter]

I think the point of contention is which card was drawn for the turn: Brainstorm, or Yawgmoth's Will? Player A knows, Player B probably doesn't, and the spectators behind Player A may have seen. The question is, if we can't come to a conclusion as to which card was drawn by consulting the players, is it OK to talk to the spectators? Or just put back a random card, as Anusien suggested? And putting back a random card, may or may not change th gamestate. If Yawgmoth's Will was the card drawn for the turn, and the gamestate reversal leaves Player A with Yawg Will in his hand instead, he's just won the game.

[M.Solymossy]

Player B had even said "it was assumed [player A] had will".  I'm assuming that is why vampiric tutor was shadow of doubt'ed.

[Toad]

Oh, Welder has summon sickness, I didnt realize that 

Repairing the game state is a big no no. It is always messy and leads to complications all the time. Assuming that the player didnt use his Welder knowing it had summon sickness (which would be cheating), I would simply keep the game in that state. Both players are responsible for the game state and the legality of actions. It is too late to rewind the Welder activation, and since both players agreed on resolving the Welder ability, we will keep it as it is.

Basically, do nothing to the game state, give a Warning for the player using the Welder (PE Major), a Caution for his opponent (failure to notice a problem with the game state).

But basically, never ever start rewinding the game over multiple phases.

[Moxlotus]

But basically, never ever start rewinding the game over multiple phases.

So I take it you disagreed when the judges at some PT did it using a video camera?

I just find it interesting that all the different judges are coming to different conclusions.

[Anusien]

By the way, I'm not a judge.  But to clarify things:

Anusien--it doesn't appear that you actually repaired the game state--you just changed it.  Not only did you not repair it--you have a 50/50 chance of making it worse.

Why can't you just ask the spectators?  I know the rules don't say you have to--but they don't say you can't.  Would you not do it for the sake of not doing it?  If they all agree, then why wouldn't you take their word?

I said to ask the spectators, however the spectators are almost always paying less attention to the game state than the players.  If both players agree on which card, go with that one, otherwise check with the spectators, and if that isn't conclusive than do it randomly.  It's possible the game state will be one that is strategically worse for a player than the ideal one, but the Penalty Guidelines say to repair game states that way if applicable.  Plus, it's not like he's a hapless victim; maintaining the game state is the responsibility of both players.

[Dante]

I had a similar situation happen to me at one of the SCG Chicago's and every judge involved including the head judge said that they couldn't take spectators opinions on anything as binding or relevant (the issue in question was which card I drew that should have been revealed to Uba Mask).

[Komatteru]

Unfortunately, that was over a year ago, and the DCI changes its policies almost on a weekly basis.  I am not sure as to current DCI philosophy on this.

[Anusien]

I discusses this on #mtgjudge and while there isn't 100% consensus, the overwhelming opinion says that if you know 100% what the card is that was drawn that turn, you can consider repairing the gamestate, but if you don't, you probably don't want to try to repair the gamestate.  Take that as you will.

[Komatteru]

Asking a Level 2, I got the following:

-Judges can ask spectators as to what they saw.  Players cannot, but judges can.

-A judge doesn't have to repair the game state.  You may if you have near certainty that you can do so correctly, but if the players managed to get the game to that state, then that's their problem. (Agrees with Anusien posted)

-If the (head) judge suspects the other player of cheating, i.e., is insisting that the game state is unrepairable, now wishing to get away with the play he made  because he knowingly did so with the intent to create such a game state, then DQ without prize is an appropriate action to take.

[Jacob Orlove]

Keep in mind that it's quite possible for all the spectators to be the friends of one of the players, which is not so great when you're going to them for information.

[Anusien]

Keep in mind that it's quite possible for all the spectators to be the friends of one of the players, which is not so great when you're going to them for information.

True, the damage is already done, but that's why I'd try to interview the spectators seperately from each other and the players.  That and to avoid giving away any information about cards in hidden zones.

[Godder]

Keep in mind that it's quite possible for all the spectators to be the friends of one of the players, which is not so great when you're going to them for information.

Also, if the spectators are other players in the tournament, they have a vested interest in the result.

[Apollyon]

As a side note: if you are a spectator and lie to a judge in order to help a friend if the judge asks you questions in something like this, you can be DQed for lying to the judge. You don't have to be playing in the match for your actions to merit a DQ.

As for repairing gamestates, in general, I try to not repair them. It leads to all kinds of headaches of "Well, one player might gain an advantage." and other kinds of messy "fixes".

Both players are responsible for maintaining the accuracy of the gamestate. Both players missed it, so assuming that it's not cheating, warn and move on. Again, investigation is key.

[Komatteru]

As a side note: if you are a spectator and lie to a judge in order to help a friend if the judge asks you questions in something like this, you can be DQed for lying to the judge. You don't have to be playing in the match for your actions to merit a DQ.

Such action could also warrant suspension from the DCI, even if you are not actually playing in the tournament.

[Khahan]

Doesn't the judge have the duty to conduct a thorough and full investigation?  To me, that includes asking the spectators.

Now, what weight do you give the answers you get?  That is something that is up the judgement of the judge.  You may come to the conclusion that the spectators are all over the place with their stories and you can't rely on the info.

You could realize that 3/5 of the spectators are teammates of one of the players...and could be giving biased information.

One of the spectators could be a good friend of yours that you know you can rely on. Does that mean the judge just blindly goes w/ what his friend says? No. But maybe it carries a bit more weight than the statement of a players teammate.  Maybe not.

No matter what, the judge needs to weigh all the information he gets to the extent he feels appropriate and then make his decision.  The floor rules do state that spectators can be interviewed. But the rules do not state what kind of credence should be given to those answers.

[Godder]

In chess, the physical evidence comes first (including any Arbiters who happened to be witnesses, although that normally makes the investigation easy), followed by the word of the players, then non-playing spectators, and playing spectators last. I'd suggest that Judges use that list, as well as any warnings and anything else that has been noted about the players related to their past conduct in similar situations (it may move them down the list, for example).