Unlocking Legacy – Flash

Grand Prix Columbus - May 19-20, 2007! Countdown to Grand Prix: Columbus! Kevin Binswanger dispels some of the hype and myth about Flash, and tries to talk strategy and metagaming for Grand Prix: Columbus. Decklist? Discussion of policy for errata’ing and banning? It’s all inside.

Grand Prix Columbus - May 19-20, 2007!

"Nearly all men can stand adversity, but if you want to test a man’s character, give him power."
Abraham Lincoln

Important Fact #1: Flash will be legal for the Grand Prix
Despite two weeks worth of Grand Prix Trials being conducted with the errata and near continuous discussion of the topic in #mtgjudge (Sorry guys and girls!), people still are not certain. So let me go ahead and clear it up. Flash has been in Oracle for two weeks with the new wording, and Oracle is definitive. In other words, short of an emergency ban (which won’t happen, more on that in a bit), Flash will be legal.

Before I get into the meat of this discussion, there are two important points:

1) Read Aaron Forsythe‘s "Power-Level-Errata-B-Gone"
2) The singular form of "errata" is "erratum"

With that, said, onward!

So what does the erratum mean, and why did they do it? Updates to the Oracle database can only happen when a new set is released; each time Internet celebrity, superstar and potential AI Yawgatog compiles a listing of changes to the errata and releases them online. You can find the changes from Planar Chaos to Future Sight here. Note that only the changes are Official, and the groups he uses are only for convenience sake. In this case, the change to Flash comes about in a slow of changes Yawgatog aptly titles "Changes involving cards in unrevealed zones.” These changes all involve making choices instead of searches in hidden zones. You are allowed to fail a search with a condition on it in a hidden zone, but choices cannot be failed if there is a legal target. This series of errata clears up the problem where a judge must be called to verify that there are no legal targets for Burning Wish, for example. All the errata here deal make choosing or putting cards into play with a specific characteristic; Didgeridoo gets the "You may put a Minotaur" errata treatment that brings it in line with Aether Vial’s printed text.

Flash’s errata does two different things. Number one definitely makes sense, and no one seems to be arguing against it. It makes putting the creature into play a choice instead of a required action; otherwise resolving Flash with no creatures in play would require a judge call to verify. (Optionally revealing your hand is the other way to resolve this situation). That particular change leaves the card functionally almost identical, if slightly better, but it makes the rules surrounding the card much nicer. No it is part two of the card’s errata that seems to be the sticking point for most people. Previously the card only put the creature into play if you paid the mana cost less {2}; this was a power level errata to prevent a dangerous combo with Academy Rector.

Alright, so they removed power-level erratum from Flash. Why? Scott Johns, in his "Taking You to the Beginning, At Last" offers this summary:

Currently Wizards has two main systems for dealing with problem cards, either banning/restricting or issuing errata. Generally speaking, the intention has been to use banning and restricting for cards that are too powerful, and to issue errata to cards that either have mistakes (like the original version of Impulse) or need to be updated because of things like rules changes (such as Wall cards gaining the defender ability when that rule was changed).

The issue is that there is a small subset of cards that proved too powerful when printed but were dealt with through errata rather than banning or restricting. Examples of this would be Great Whale or Basalt Monolith, each of which received "power-level errata" to prevent degenerate combos, thus avoiding having to restrict or ban the cards in question. So, in the interest of using each system what it was intended for, with Monday’s Oracle update R&D is removing a bunch of errata that deals with power level.

In other words, current policy is that errata is only intended to be used to fix cards that do not work as printed under the rules (Waylay) and that banning and restricted are intended for cards that are above the power level (Skullclamp). Older cards with power-level erratum are being fixed when they are found. Mark Gottlieb was inevitably going to remove the power-level errata at some point; most people just criticize the timing of the action (especially since it was not immediately banned). So why now? My argument is that Flash was already receiving erratum, and the rules team decided to go ahead and remove the power-level erratum from the card, since that is current policy.

The card received erratum, but will it be legal? The answer is yes. There are two ways it could become illegal for the tournament: emergency banning and re-errata. The card will not be re-errata’ed for two reasons. This would be power-level errata which is no longer policy; re-errata’ing the card is simply not an option for Wizards of the Coast under current policy. Also, Oracle is essentially read-only until new sets are released, so the earliest Flash could receive erratum would be the release of the next set, 10th Edition, in the middle of summer (sometime around July 13).

Will the card receive emergency banning? Not likely at all. Only one card has received emergency banning in the history of Magic: Memory Jar. Randy Buehler explains in "Extended Thoughts."

If you compare this situation to the current situation in Extended, you can see that there are two crucial differences: 1) The problem spanned all Constructed formats, 2) There was a huge buildup of angst about the format before the emergency ban candidate was even printed. I do not expect the DCI to issue any emergency bannings in the future unless both of these conditions are true — there would have to be a preexisting problem (that included both public outcry and previous bannings) and the problem would need to span all formats and audiences. Furthermore, there was a real changing of the guard inside R&D as a reaction to the problems of the Urza block. If those of us who left the Pro Tour to join Wizards over the last four years do our jobs correctly, these conditions will never come up again.

Replace "Extended" with "Legacy" and his words make perfect sense to the situation in question. Now the people on the other side of this issue, people like Chris Coppola (Machinus), would argue that Flash is a unique situation. I answer with the following excerpt from the same article:

I fully expect us to have to ban cards again in the future (read my "Banning: Good or Bad?" article from February for an argument about why this is actually a good thing). However, when a mistake does happen, that doesn’t mean an emergency ban is called for — I actually think it’s okay if a Pro Tour is dominated by a ban-worthy card. In fact, our general policy is to use the pros as guinea pigs – we let them try to break a format and then if they succeed we can fix it before the next qualifier season. Note that we care more about qualifier seasons that the Pro Tours because so many more people play in them (plus breaking a Constructed format is a pretty skillful thing to do, so rewarding the pros who do it most successfully seems reasonable).

Yes Columbus will not be a Pro Tour and the people working on Legacy are not, on the whole, professional Magic players. But judging from the Grand Prix Trial results (more on that later), Flash in its current incarnations is not broken beyond control. This suggests that scoring a Grand Prix Trial victory with Flash requires either deckbuilding skill, playskill, or more likely both. The analogy holds. Moreover if they were going to emergency ban it, they would have done it already. The longer they wait to do so, the worse the GP metagame becomes. The only proper time to ban it would have been with the errata; they have had at least enough time to consider Flash without the power-level erratum as we have, so they must be fine with it at least until June 1.

Finally in Wednesday’s "Ask Wizards," Aaron Forsythe answered this question:

Q: Flash and Protean Hulk now make a combo that’s one heck of an elephant in the Legacy room, and I fear for GP–Columbus. Are there any plans to fix this at the last second before the Grand Prix?
-Kevin, Iowa USA

Despite the same name, this question was not asked by me. The response is quoted here in full:

We will be sticking to the normal Banned & Restricted list update schedule; as a rule, we do not ban cards at other times. You can read Randy Buehler old article Extended Thoughts for more insights into the B&R policy and our avoidance of "emergency bans."

For those who aren’t aware, Flash was given power-level errata in 2000. We removed that power-level errata, as is our policy, when it was found during the most recent Oracle update. We will not be reissuing errata for the card Flash or any other cards changed during the Future Sight Oracle update.

All the attention paid to the recent functionality change of Flash has caused us to reevaluate how we disseminate such information, however, so we are working on a much more public and visible method of highlighting Oracle and Comprehensive Rules changes going forward.

I maintain that while the timing may be disadvantageous, the DCI is not to blame. People are condemning Wizards for un-errata’ing Flash nearly a month before the Grand Prix. There are two lines of response to this argument.

The first response defends the DCI by basically pointing out that there is nothing they can or should do to respond. The DCI and the Rules Team are distinct teams in responsibilities if not people. The Rules Team should be ignoring power level in fixing cards to the way they should work according to the rules; this is the point of removing power-level errata. The DCI can only do two things when it comes to disruptive cards: ban or restrict it when the B/R cycle is upon us, and emergency ban or restrict it. I’ve already argued that they are unlikely to emergency ban it, so there is literally nothing they can do. Anti-Wizards folks like to say that un-errata’ing Flash was a horrible move at this time because it is so powerful, but the analysis already presented is that basically they should un-errata the card as soon as possible because it should always have worked that way. After all if they take how powerful the card is into consideration when deciding whether to remove power-level errata, that kind of defeats the purpose.

The second line of response is more sophisticated, but I feel that it is also stronger. You either have to assume that the Rules Team did or did not know how powerful Flash would be when they removed the erratum. If they did not know how powerful the card would be they were simply doing their job removing power-level errata. After all no one condemned Mark Gottlieb for removing the errata from Karmic Guide (ironically part of a combo with Flash) or Cloud of Faeries. If they did know how powerful Flash would be then you can point to the change to the Two-Headed Giant starting life total change as an example of a successful format upheaval right before a Grand Prix.

The other note I should make here is that Future Sight will NOT be legal for Grand Prix: Columbus. Future Sight becomes legal on the 20th of May, and the Grand Prix begins on the 19th of May. Thus Columbus will be the last Legacy tournament to not feature Future Sight. If you remember, this is similar to US Nationals when Skullclamp’s banning went into effect on the 20th, but the tournament started on the 18th.

Important Fact #2: You Will Get a Turn

People have been suggesting ludicrous statistics with the Flash deck, so I decided to get to the bottom of the most-hyped one, the so-called Turn 0 Win Percentage. With a ridiculous hand, it is possible to start the game on the draw and win with Flash on your opponent’s first upkeep, before they have a chance to play spells. It requires having Gemstone Caverns, Flash, Protean Hulk, and either Elvish Spirit Guide or Simian Spirit Guide and one other card. Some of the most egregious offenders claimed as much as a 25% chance for this to happen. Thanks to the wonderful folks in Efnet’s #scg, I calculated the following percentages. I built a spreadsheet to do most of the heavy lifting; you can find it online here.

Pre-Future Sight
The chance of winning on turn 0 with a 60 card deck and 7 cards: 0.02320994144796760000 or 2.32%. If you are willing to mulligan down to 5 in search of the combo (the smallest possible hand since Caverns requires an extra card), you can get the percentage all the way up to 0.03897086665250050 or 3.89%.

Post-Future Sight
The addition of Summoner’s Pact changes things, since Pact can stand in for either an Elvish Spirit Guide or a Protean Hulk. The chance of winning on turn 0 with a 60 card deck and 7 cards: 0.06488575606050760000 or 6.48%. If you are willing to mulligan down to 5 in search of the combo (the smallest possible hand since Caverns requires an extra card), you can get the percentage all the way up to 0.12050513521525100000 or 12.05%.

In the spreadsheet you can tweak the numbers on all the combo pieces as well as starting hand size and library size (adjust library size to 56 to see how little Street Wraith affects things). If you want to make any other adjustments, the password is "test.” Adding in the percentage of having a pitch counter in hand becomes somewhat intractable; adding Pact of Negation makes the entire problem grow in complexity by an order of magnitude and Force of Will’s blue card makes the entire problem somewhat ridiculous to calculate. The point of this side note was merely to reassure you that you will get a turn. This also suggests weighting your deck in favor of consistency rather than speed.

That 12% figure post-FS may be somewhat alarming, but I would actually point to the opposite statistic; if your opponent mulligans down to 5 cards and misses the combo, there is every chance that they lose the game. While the combo may be powerful and fast, you are more than five times more likely to have Force of Will in hand than for your opponent to have a turn 0 kill. The existence of lack of a turn 0 kill is not what makes Flash respectable. For this reason I do not even consider Gemstone Caverns worthwhile in my build of Flash.

Important Fact #3: There is Strategy in this Article
What makes this Grand Prix interesting to me is the timing. With another month or two of testing it is possible that someone could come up with the best version of Flash, but right now the deck is still essentially in its infancy. It was barely more than a week ago when the UG build with Disciple of the Vault and Worldly Tutor was the supposed best deck in the format (before losing to Affinity 2-0 in a GPT in Roanoke, Virginia). Now the Kiki-Jiki kill has been discovered and we can shave more slots off the kill, and players are splashing black instead of green. So there is no best build of the deck. Maybe it turns out that the actual best build is UGW for Meddling Mage on hate, Living Wish for tutoring, Trinket Mage for Top and Counterbalance, and a reasonable beater backup plan. Or maybe the best build is a more refined UB version with more cards everyone forgot. With the information we have available right now, here is the build of the deck I find more powerful:

5 Island
3 Flooded Strand
4 Polluted Delta
1 Swamp
2 Underground Sea
1 Bayou
2 Tropical Island
1 Karmic Guide
1 Kiki-Jiki, Mirror Breaker
1 Carrion Feeder
1 Body Snatcher
4 Protean Hulk
1 Benevolent Bodyguard
4 Duress
4 Lim-Dul’s Vault
4 Flash
4 Merchant Scroll
4 Force of Will
4 Brainstorm
4 Lotus Petal
4 Mystical Tutor
1 Chain of Vapor

The Benevolent Bodyguards come from Pat Maeder and the rest of the innovations are basically Team Reflection letting me into their testing. Most of the final work comes from Matt Hargis’s tournament report here on The Mana Drain. What makes the deck challenging is how many open questions there are. What kind of hate are people going to run? It is theoretically popular (judging from online conversations) to run Threshold with Predict to counter opposing Mystical Tutors. If that plan is a reality at Grand Prix: Columbus than Mystical Tutor, a card that I dislike personally, becomes the wrong call. But maybe the kind of decks that run Predict lose to the rest of the decks that also aim to beat Flash. Or maybe once you enter the tournament with three byes, you are just facing the mirror match and need the Mystical Tutor to draw more Duresses.

The green lands in the maindeck are only for sideboarded Pernicious Deeds against the aggro-control decks. The problem is that you have to weaken your manabase significantly to allow for Pernicious Deed. Before I made that addition to the maindeck I had one Swamp and one or two Underground Seas as non-Island, non-fetchland lands. Adding the four or so Wasteland-ready lands potentially weakens yourself in matchups like BW, BR and Goblins. Now you should roll over Goblins anyway, but BW (Pikula) and BR (Red Death) are not auto-win matchups by any stretch of the imagination. Part of the reason those matchups are as good as they are is because Wasteland is essentially dead, and you can live with Sinkhole; if all of a sudden they can Wasteland half your lands things become problematic; nearly a quarter of your lands are now vulnerable to Wasteland and sometimes those sit in your opening hand.

The best card in the deck is actually Lim-Dul’s Vault. Rich Shay suggested the card off-handedly to me in testing, but I picked it up and ran with it in testing; it and not Duress was initially the card that prompted me to go UB in testing. Lim-Dul’s Vault singlehandedly makes the B/x Disruption matchups favorable, because it leaves you virtually immune to hand disruption as well as tutors up the cards you need. There are a few relevant things to consider. You can put the cards back in any order, so LDV becomes twice as powerful with a Brainstorm in hand or in the top 5 cards of your library. This card not only acts like Worldly Tutor and gets you Protean Hulk, but it is not unusual to be able to use it to tutor both Flash and Protean Hulk to your hand. With Brainstorm and adequate mana, it is possible to cast LDV, and then untap, Brainstorm and get both pieces of the combo and win. With Brainstorm, remember that the only card that will not go to your hand from the top five is the fifth card, so any combo pieces go there. LDV is also the reason I avoid Gemstone Caverns in the deck; I want to be able to cast it as soon as possible and a low number of off-color sources already makes this difficult.

The other intriguing thing about Flash is that once you put Protean Hulk to the graveyard, there are two radically different combos you can win with. The Disciple of the Vault kill requires at least 9 slots, and I would not run it without at least 4 Disciple and 7 X-drops so that you can have one of either piece in hand and still be able to win. This combination does not require an attack step and it is immune to targeted removal, Pithing Needle and Tormod’s Crypt. The Kiki-Jiki, Mirror Breaker requires at least 3 slots (but is normally run with 5), but it is vulnerable to all the above mentioned removal. If enough of the combo pieces are stranded in hand or in the graveyard, it can be difficult to win; specifically a Hymn to Tourach hitting the Carrion Feeder means that it requires two Flashes to win instead of just one. Aside from the extra slots, the Kiki-Jiki kill has two advantages: you are virtually or completely immune to effects like Angel’s Grace and Gilded Light, and you have a respectable backup plan. If somehow you cannot win with the Disciple of the Vault plan, you are reduced to putting four 1/1 clerics on the table and turning them sideways. By contrast in one game I misplayed against Rich Shay on the Kiki-Jiki combo and allowed him to cast Flash with multiple Pithing Needles down on wrong target. He showed me my mistake by Flashing Protean Hulk, fetching Karmic Guide, and bringing back Protean Hulk. I died in three quick swings.

The Disciple kill is nominally faster, and many people still cling to the Disciple of the Vault kill for whatever reason. For that reason I would not feel safe relying on something that only effects one build of the combination: not Gilded Light nor Swords to Plowshares. The hate that is effective are things that hurt all control decks: opposing counters, Duress, strong discard in the form of Hymn to Tourach. What amazes me the most about the deck is how potentially good things like Disrupt and Spell Snare are on both sides of the matchup, since now the Flash decks and the anti-Flash decks run a lot of two casting cost spells and almost no creatures.

So what do I turn to in order to beat Flash? Well, I’m not necessarily sure that I do. Since Flash is probably the best deck in Legacy (not that this is a bad thing; there always has to be a best deck), like in my last article I would question playing anything else. Sure, you could get the right pairings, play your hate or metagame deck and conceivably Top 8 or win, but it’s unlikely. That same amount of effort will probably pay off more for you if you just concentrate on testing and making the best Flash deck and player you can. Still in testing, the best strategies people have found were either Blue-based aggro-control decks (like U/B/W HanniFish), black-based disruption decks, combo decks, and U/B control decks. Other people have discussed the other options, but I wanted to give special attention to the notion of combo decks. On a scale from The Epic Storm combo to Reset High Tide, Flash is not ridiculously fast. The cards that make it able to beat the hate, like dropping the acceleration and running Lim-Dul’s Vault, make it a slower and more consistent deck. Decks like Iggy Pop or The Epic Storm don’t have to make that same tradeoff. Much in the same way that Flash continually threatens to go off the moment Threshold taps out, The Epic Storm continually threatens to just go off through hate versus Flash. "Sure, counter my Xantid Swarm", the TES player says. "Meanwhile you can’t go off because you are down two cards." In any sort of fair fight, the player with draw 7s is in a better position to win than the player that just has Brainstorm. Flash’s advantage in the matchup is that it can just win at instant speed with less setup, but TES is much more consistent at going off turn 1-2.

Important Fact #4: The Sky is Not Falling
A lot of what follows in this fourth and final section is me rambling about tournament results and possibly banning criteria. If you do not care about my thoughts regarding whether Flash is ban-worthy, you won’t miss any more strategy by pretending the article ends here. With that said…

The Mana Drain Forum Goer Cisco Fernandez compiled all the GPT results available onto this page,

Week Zero: Flash was errata’ed with the pre-release of Future Sight. No tournaments were held during this time. However, I did discover the amazing interaction between Flash and Karmic Guide. Cast Flash and put Karmic Guide into play, then do not pay the cost. It is put into the graveyard, and then when Flash finishes resolving, its Comes into Play trigger is put on the stack. You are now free to bring back Karmic Guide… with its own ability.

Week 1: To my knowledge, there is only one tournament where Flash was played; the GPT in Roanoke, Virginia. One player brought the U/G version of Flash with Disciple of the Vault. He went 2-1 (losing to Affinity) and scooped the U/W/R Pyroclasm player into the Top 4. The Flash player received packs for Top 8 but did not make the Top 4 elimination round cutoff.

Week 2: For all extents and purposes, the cat is out of the bag. We know of six major tournaments within this time. Hulk wins half of them, and places 11 people into the Top 8. This is twice the number of players as Threshold. I hear some gasps in the audience, but I am not actually concerned by this. The Goblins and Threshold numbers were actually incredibly similar to this prior to Grand Prix: Philadelphia. The rest of the decks just got phenomenally better.

A lot of people want Flash to be banned because it is simply that good. And yes, Flash is at a higher power level than several cards already on the banned list seem to be. This is making me reconsider my views on those cards; I do feel that leaving something unbanned requires less burden of proof or less fairness than to unban something. I do feel that it is a fine decision to leave questionable cards on the Banned list, simply because I do not think we have explored the entirety of the format. Maybe in another year when the format has actually progressed again; but calling for unbannings seems to say that we have given up on deck building.

The format has adjusted since the introduction of Flash, but then again turn 2 combo was on the rise and people had to adjust. The results for the current round of GPTs was linked above, and you can compare to the first set of GPTs here. There has been a significant increase in the sophistication and hybridization of decks. The first round of GPTs shows a lot of monocolored or two-colored non-interactive decks. It is not immediately clear how much of the blame to lay on Flash and how much to lay on the significant increase in popularity of Iggy and TES. If Flash were to blame, the Week One and the Week Two deck choices would be significantly different. With the exception of Flash taking Top Eight berths during week two, I do not think the results are significantly different, however the number of combo decks and more sophisticated disruption decks has increased significantly from a year ago. Last year we had FEB and Sligh, now we have Red Death, Iggy and Gamekeeper Salvagers. Basically I feel that the metagame shift that is happening to accommodate Flash was already put in motion by TES doing so well at The Mana Leak Open. It was theoretically possible for decks like RG Survival splashing black to beat Iggy, but it is very unlikely for them to beat the raw speed of Belcher or TES.

In the same vein, I do not think we should be rushing to ban Flash just because of how good it is. Yes Flash is very powerful and "invalidates" a lot of deck strategies, but based on the tournament results and the extreme popularity of things like Belcher those strategies were invalid already. Based on testing results and tournament performance, it seems very clear that Flash is beatable. For example at the Rochester GPT another fix or six players failed to make Top Eight with Flash. I think people are complaining merely because the decks they used to play cannot beat Flash. There is a radical paradigm shift in deckbuilding, but that is because the previous best deck was a monocolored beatdown deck. I think you see a similar situation happen in Standard; when the best or most popular deck changes from Dragonstorm to one of the Teferi control decks, you see the metagame shift as a result. Part of the reason Flash’s results seem so good is that everyone is used to aiming for Goblins; when the BW Confidant list you are testing with has maindeck The Tabernacle at Pendrell Vale, the results are going to look somewhat odd. It is traditionally good practice to test with stock lists that are already known, but I think Columbus is a special case only because none of the stock lists will be played as is. I think the trend you see is that the decks that cannot compete fall by the wayside, but all of the decks that you would expect to be good (the multi-colored disruption beatdown decks) continue to do well. This change was inevitable. It is actually striking that the ratio of Flash Top 8s and wins to every other deck is actually not as high as Goblins’s ratio last year.

When I raised these arguments before, the criticism came back that the only decks that are performing well are combo decks, Flash decks, and blue and black disruptive decks; supposedly that is not healthy for the format. But honestly with storm cards and things like Dark Ritual, Diminishing Returns, Infernal Tutor, Ill-Gotten Gains and Goblin Charbelcher legal, did people honestly think that Elves! was going to be a defendable metagame choice?

By the way, if you want to find out about these things as soon as possible, you should probably spend more time in IRC, channel #TheManaDrain. Thanks to Yawgatog Oracle Diffs and my procrastinating from studying, regular chatters knew about Flash the moment it was errata’ed.

Thanks goes out to all of Team Reflection for letting me join your testing process, and Josh Silvestri and Spencer Hayes for helping me with the article.

Kevin Binswanger
[email protected]


"Power-Level-Errata-B-Gone" by Aaron Forsythe
"Extended Thoughts" by Randy Buehler
"Deus Ex Errata: Debating the Changes to Time Vault" by Stephen Menendian
"Taking You to the Beginning, At Last" by Scott Johns
"Ask Wizards" by Aaron Forsythe
"Deconstructing Constructed: Jank Rares Ruin Formats: Flash" by Josh Silvestri
"Limited Lessons – Hulk Flash" by Nick Eisel

Grand Prix Columbus - May 19-20, 2007!