Serum Powder Flash (AKA Blackjack)

[ErkBek]

Meadbert has been really pushing Serum Powder in Flash for few weeks now, but everyone including myself just kept telling him why it shouldn't work in theory. I quit my job last week so I've had at lot of free time again so I decided to give it a try. It took 50 games of goldfishing to get to a list I really liked, at that point I goldfished 100 games with that list. Here's the list I decided on and it's goldfish results

Serum Powder Flash

Lands 13
1 Island
2 Trop
3 Usea
7 Fetches

Accel 6
1 Lotus
1 Sapphire
1 Jet
1 Emerald
1 Petal
1 ESG
(4 green pacts)

Combo 17
4 Poision Slivers
2 Heart Sliver
4 Green Pact
3 Hulk
4 Flash

Protection 9
4 FoW
4 Blue Pact
1 Chain of Vapor

Draw/Tutor 15
4 Serum Powder
4 Brainstorm
3 Merchant Scroll
1 DT
1 VT
1 MT
1 Ancestral

-----------------100 goldfish game results------------------------------

Turn 1 kill w/o protection
18

Turn 1 kill w/ at least 1 protection spell
14

Turn 2 kill w/o protection
13

Turn 2 kill w/ 1 protection spell
22

Turn 2 kill w/ 2 or more protection spells
10

Turn 3 kill w/o protection
4

Turn 3 kill w/ 1 protection spell
4

Turn 3 kill w/ 2 or more protection spells
5

Turn 4 kill / Fizzle
10

Serum Powder was used a total of 39 times over the course of the 100 games.

The reason powder works so well in here is Flash can easily win on a 6 card hand. Powder in your opening 7 cards allows you to see up to 3, 6 card hands before having to mull down to a scary 5 card hand. The only unrestricted card I'd rather see in my opening 7 than Powder is Flash.

Powdering away slivers came up about occassionally, but definitely not enough to remove powder entirely. Adding the 2nd Heart Sliver really helped this problem.

There are 3 games I noted: 1) Mull to 3 and kill on turn 3. 2) One of the Turn 4/Fizzles was a turn 4 kill with 4 counter backup. 3) One game was a turn 3 kill with counter backup via sliver beats. INSANE!

I don't know what the sideboard should look like for this list. I'd probably start with 3 Reverant Silence, 3 Massacre, and 4 Leylines.

The scary conclusion: this deck killed by turn 2 over 75% of the time. 50% of those kills had at least 1 counter backup. I'm not sure yet if this build is better than powder-less flash, but I think it should be given some consideration.

[cophos]

Including a card like Serum Powder but cutting an integral combo piece (Hulk, to a certain extent Scroll) at the same time seems outright counterproductive. In order to maximize Powders effect I´d strongly suggest fitting in the last Protean Hulk. (Scroll)

Since the list´s already extremely tight, the addition of Serum Powder in general must be thought through very carefully. The deck´s fairly consistent already. (Because of the waste number of available tutors.) Cutting actual combo parts or tutors in favor of a card, whose purpose is to enable getting these cards in your starting hand, seems simply wrong in my opinion.

Don´t get me wrong, I highly appreciate your work on the deck. Especially because Serum Powder seems to fit into the Flash archetype perfectly. I´m looking forward to reading your take on the exposed issue.

(I also think that Serum Powder would be a evident addition to the deck regarding a possible restriction of Merchant Scroll.)

[Purple Hat]

EDIT: my math did suck.  you gain ~10% chance to have hulk in either your opening hand or in your powder mulligan by taking out a hulk and adding 4 powder.

[TopSecret]

Have you considered adding in Street Wraiths
to up the OMG I'M CHEATING THE 60 CARD RULE factor?

Did you consider Imperial Seal?

You could always add in a Research//Developement
to Merchant Scroll for when you remove your win conditions.

Congratulations!
If this deck becomes popular,
I will probably quit Vintage
and start playing craps.

[ErkBek]

Including a card like Serum Powder but cutting an integral combo piece (Hulk, to a certain extent Scroll) at the same time seems outright counterproductive. In order to maximize Powders effect I´d strongly suggest fitting in the last Protean Hulk. (Scroll)

Question for you. If you could run a 50 card build of flash, would you? Of course, but you'd have to start to cut back on something. I'm positive Scroll and Hulk would be the first 2 cards you'd start to cut back on.

EDIT: my math did suck.  you gain ~6.5% chance to have hulk in either your opening hand or in your powder mulligan by taking out a hulk and adding 4 powder.

I ran the numbers too and found that 5.2% of the time you are more likely to have at least 1 green pact / hulk in hand, however resulting from the addition of another hulk there will be 1.22% higher chance that you'll see either triple or quadruple pact / hulks. I've go to think that I'd mull at least half of those hands because of that alone, so it really would be about a 4.5% difference. I can tell you though, I'd much rather be looking for a hulk than a flash though.

EDIT: I've got no idea how to do the math if you start including powder mulligans, or even mulligans for that matter. My numbers are based strictly on the opening 7 card hand.

Have you considered adding in Street Wraiths
to up the OMG I'M CHEATING THE 60 CARD RULE factor?

Did you consider Imperial Seal?

You could always add in a Research//Developement

I've never really been much of a fan of Street Wraith in Flash for a couple of reasons. The main is that you want to have as many absolutes in your hand as possible when making mulligan decisions. When having 6 cards plus street wraith, I'm going to base my mulligan decision off 6 cards or just hope to lucksack into either the mana or combo piece you want.

Imperial Seal was in the original list, but I cut it for being terrible. It made room for the much better 2nd Heart Sliver.

R/D would be an awesome way to win....but if you're worried about mulling away poision slivers you can just run 2 Screeching Slivers (a one mana U sliver) to alleviate those problems. I don't think they're needed though. 

Congratulations!
If this deck becomes popular,
I will probably quit Vintage
and start playing craps.

Sorry dude. I just want to see Flash restricted. If it doesn't and this deck does get popular I'd have to quit again.....well, I'm currently sorta out of magic. I don't own cards but I still make decks.

[Purple Hat]

EDIT: my math did suck.  you gain ~6.5% chance to have hulk in either your opening hand or in your powder mulligan by taking out a hulk and adding 4 powder.

I ran the numbers too and found that 5.2% of the time you are more likely to have at least 1 green pact / hulk in hand, however resulting from the addition of another hulk there will be 1.22% higher chance that you'll see either triple or quadruple pact / hulks. I've go to think that I'd mull at least half of those hands because of that alone, so it really would be about a 4.5% difference. I can tell you though, I'd much rather be looking for a hulk than a flash though.

EDIT: I've got no idea how to do the math if you start including powder mulligans, or even mulligans for that matter. My numbers are based strictly on the opening 7 card hand.

I think I got this right:  chance to get hulk in your powder hand or the 7 cards you powder into= (1-chance you don't draw powder)(1-chance you don't draw hulk in the remaining 6 pulls out of a 59 card deck)(chance you draw hulk out of your remaining 53 cards assuming it wasn't in the first 7) + (chance you draw hulk out of cards that aren't powder in a 60 card deck)(chance you don't draw powder in a 60 card deck)

that should cover all possible hands if you assume you'll keep any hand with at least 1 hulk in it and pitch any hand that doesn't contain hulk.  this is clearly untrue, but it makes the math somewhat manageable. It turns out to be something like 76% or 77% for a 7 of I think where as the chances of drawing an 8 of are around 66%.

[LordHomerCat]

As if Flash wasn't already starting-hand-dependent.  Thanks Eric ><

Man, those are some pretty frightening goldfish numbers.  I mean, 75% of the time on turn 2?  This deck isn't even hard to play either, which makes that pretty scary.  I can also understand not wanting street wraith here, because it doesn't go well with powder (making you mull more hands due to uncertainty where you have wraith instead of a possible piece you are missing) and straight up drawing cards doesn't seem that strong in Flash compared to tutoring for the right ones.  I know a lot of people have stated they wanted 2 ESG, did it ever come up that you wanted to be able to tutor for a second one (like maybe using it to cast scroll turn 1 but not having a second land for flash turn 2 with an extra G pact or something) ?

[ErkBek]

I know a lot of people have stated they wanted 2 ESG, did it ever come up that you wanted to be able to tutor for a second one (like maybe using it to cast scroll turn 1 but not having a second land for flash turn 2 with an extra G pact or something) ?

I don't understand why people want 2 ESG's. That play there is suicide. Your green pact kills you on your upkeep.

One thing that came up a couple times in the 150 games was casting flash, Pacting for the ESG, and then casting a poision sliver with the ESG. In that respect another ESG may be useful, but I really doubt that situation would come up very often.

[TopSecret]

Warning: There is a good chance that my math is really inaccurate and/or atrocious!
             Please point out any stupid mistakes if found!

The number of combinations of n distinct objects taken r at a time is given:

C(n, r) = n! / r!(n - r)!

Initial Hand:

n = 60
r = 7

60! / 7!( 60 - 7)!

60! / 7! (53)!

60 * 59 * 58 * 57 * 56* 55* 54 / 7 * 6 * 5 *4 * 3 * 2

1946482876800 / 5040

386206920 possible combinations of an initial 7 card hand from a 60 card deck.


So, what's the probability of drawing a Hulk or Green Pact in the opening 7?
Assuming 3 Hulks and 4 Green Pacts, they are 7 of the same card for these purposes.
How many combinations of such a hand are there?
Assuming only drawing 1 Hulk/Pact:

60 - the 7 hulk/pacts = 53 cards

So, for all the combinations of cards non-Hulk/Pact...

n= 53
r = 6

53! / 6! (53 - 6)!

53! / 6! (47!)

53 * 52 * 51 * 50 *49 * 48 / 6 * 5* 4 *3 * 2

16579385600 / 720

22957480 possible combinations of 53 cards.
However, since you're assuming only one card'll be a Hulk/Pact,
just multiply the result by 7.

22957480 * 7 = 160702360

So, the probability of drawing exactly 1 Hulk/Pact in the opening hand is:

160702360 / 386206920 = 0.4161043

hmmm.
So, if you wanted to start calculating the Serum Powder probabilities
it'd get stupid pretty fast, because after calculating the initial probability of
drawing a Powder, you'd have to take into account a whole new hand.
This is where I wish I had a graphing calculator.

For example:

n = 56
r = 7

So, then once the result is gotten with that,
it could be multiplied by 4, for just the probability of drawing just one Powder,
or you could include the possibility of drawing any number of Powders
by multiplying the result by 4!.
Choosing 4! would make it complicated though, because there'd be
24 possible scenarios instead of just 4.

You could then start calculating the probability of drawing whatever with the new hands.
Assuming that you used Serum Powder on the first 7
the sample size (n) would be 7 smaller, on all accounts, with one less Powder possible.
However, you'd then have to start calculating the probabilities of having removed
different numbers of Hulks/Pacts and/or Powders if you wanted serious accuracy.
This would create multiple different scenarios of assumptions,
needing multiple similar calculations to find an accurate probability
of what you'd want to know about.

Heck, you could even calculate the approximate goldfish rate of a deck
if you compiled all possible combinations that would yield the desired kind of kill
and then divided it by all the possible combinations drawn from that point.
That would take a serious amount of calculations though,
and/or a ridiculous program that could do it for you.

[Methuselahn]

Have you considered adding in Street Wraiths?

I've never really been much of a fan of Street Wraith in Flash for a couple of reasons. The main is that you want to have as many absolutes in your hand as possible when making mulligan decisions. When having 6 cards plus street wraith, I'm going to base my mulligan decision off 6 cards or just hope to lucksack into either the mana or combo piece you want.

Your Street Wraith argument in light of the fact that you run Powders confuses me.  Wraith and Powder are similar in certain ways.  I don't think one would notice it while golfishing alone though. 

There are times where you draw a nutty hand that also has a Serum Powder, a hand that you don't want to mulligan.  You could easily have acceleration, land, Flash, Hulk, Pact, and a spare Powder.  Naturally, you wouldn't want to toss this hand.  Then you start the game.  Your opponent drops a Leyline and suddenly the Serum Powder becomes a non-absolute that's far worse than Wraith.  I suppose, in the end, this is the ultimate question.  Is a better goldfish rate stronger than having a slower deck with more flexibility?

These goldfishing numbers are impressive, but let's not forget that you have to play against another Vintage deck.  Already an inconsistant deck, you thrust yourself further towards catastrophic failure if the Powder Flash deck doesn't get the nuts or your opponent goes first and has any kind of proactive disruption.   Adding Serum Powder does make the deck more like Black Jack, but unlike the poker game, this is still a turn based game that swings in different directions constantly.  Goldifishing doesn't address this at all and Serum Powder doesn't help you after you say "keep."

I immediately wonder what happens to the win % numbers when you start factoring things that are harder to deal with when you can't rely on Misd/Pact/Reb because of things like 2sphere, duress, slice and dice, Leylines, or any other hate that your opponent casts on their turn with backup.  Suddenly, the fact that you went all in with a less flexible, powder-based deck to be able to cast Flash sooner looks like a bad idea.  With Powder, you have to half guess/hope that your opponent cannot do anything relevant during their first turn because you have committed to a deck that relies more on throwing away an otherwise flexible strategy (non-powder flash, relatively speaking anyway) for a faster clock. 

Powder is even more dangerous here than in other Powder decks because not only are you trying to find a two card combo, the deck is already fairly inconsistant as-is.  Back when Trinisphere was not restricted, many on Reflection were working with aggro decks and a common conclusion came to be that Powders were really good, and only good, for finding 1 card comboes.  In our case, Trinisphere.  The evidence here was proven by Eric Miller's "the Riddler" (a Trini deck with Mask-Nought)  Nowadays, you have Ichorid, a Serum Powder deck that abuses a different one card combo, Bazaar of Baghdad.  I'm not sure SP can ever be as strong here as it is in a one card combo deck.

Having run Serum Powder/mulligan calculations ages ago, I came to the conclusion that no matter how the numbers pan out, they are just not practical.   Even if you get the math spot on with Powder calculations and goldfishing rates, you have to consider that these numbers are based on far, far more games than you will actually see in a single tournament.  IMO, to adequately rely on these numbers and actively choose a Powder deck, you need to enter a meta with very little in the way of hate decks.  This doesn't seem realistic to me.  Mathmatically, you could even get better results in a tournament than expected, but still go home with no prize simply because games 1 and 3 against match X didn't go in your favor, the kid dropped, and you got screwed out of tiebrakers, or what have you.  There are too many random and non-goldfishing, interactive plays happening to outweigh the benefits here, it seems.

That all said, I'm glad to see that you've posted all of this.

Sorry dude. I just want to see Flash restricted. If it doesn't and this deck does get popular I'd have to quit again.....well, I'm currently sorta out of magic. I don't own cards but I still make decks.

[jab]Wait a minute.  The evidence is clear, you didn't actually retire! 

You care about the meta.
You still build decks. (which requires FAR more time investment than just borrowing cards and showing up to a tourney, btw.)
You still post here in mtg strategy related threads.
You've attended scg events following your own retirement thread.

This is similar to the crack head down the street who claims that he has come clean because he only smokes half as much now! hehe.  [/jab]

[LordHomerCat]

I know a lot of people have stated they wanted 2 ESG, did it ever come up that you wanted to be able to tutor for a second one (like maybe using it to cast scroll turn 1 but not having a second land for flash turn 2 with an extra G pact or something) ?

I don't understand why people want 2 ESG's. That play there is suicide. Your green pact kills you on your upkeep.

One thing that came up a couple times in the 150 games was casting flash, Pacting for the ESG, and then casting a poision sliver with the ESG. In that respect another ESG may be useful, but I really doubt that situation would come up very often.

I was thinking more of the situation where you draw yours naturally and have to use it to cast scroll or something, and then only have 1 mana source next turn along with an extra G pact which can no longer get mana (since you used it already).  I haven't really played with flash at all, but its definitely a situation that could occur, although I'm not sure if it would be worthwhile to plan for that. 

[ErkBek]

Your Street Wraith argument in light of the fact that you run Powders confuses me.  Wraith and Powder are similar in certain ways.  I don't think one would notice it while golfishing alone though. 

There are times where you draw a nutty hand that also has a Serum Powder, a hand that you don't want to mulligan.  You could easily have acceleration, land, Flash, Hulk, Pact, and a spare Powder.  Naturally, you wouldn't want to toss this hand.  Then you start the game.  Your opponent drops a Leyline and suddenly the Serum Powder becomes a non-absolute that's far worse than Wraith.  I suppose, in the end, this is the ultimate question.  Is a better goldfish rate stronger than having a slower deck with more flexibility?

Street Wraith would give you more flexibility in that hand, but chances are if this is game 1, you're screwed either way unless you get savagely lucky. Let's say you've got a very typical hand of 2 Land, 1 Mox, Merchant Scroll, Brainstorm, Sliver, and SW/Powder. I'm thinking Powder is looking a whole lot better than SW here. Powder draws you 7 cards, SW draws you 1. In fact, SW just complicated my mulligan significantly. If SW was a Flash of Hulk you'd be keeping that hand. Powder that shit away and draw 7 new cards.

I immediately wonder what happens to the win % numbers when you start factoring things that are harder to deal with when you can't rely on Misd/Pact/Reb because of things like 2sphere, duress, slice and dice, Leylines, or any other hate that your opponent casts on their turn with backup. Suddenly, the fact that you went all in with a less flexible, powder-based deck to be able to cast Flash sooner looks like a bad idea.  With Powder, you have to half guess/hope that your opponent cannot do anything relevant during their first turn because you have committed to a deck that relies more on throwing away an otherwise flexible strategy (non-powder flash, relatively speaking anyway) for a faster clock

I'd be interested too. Leyline is the biggest problem of the bunch, but only 1 upper tier deck plays that maindeck (goblins). I'm pretty sure most builds of flash are going to have a very rough game 1 vs. Leyline though. I'm not too sure where you ever got the idea that flash was a flexible deck game 1 because it's not, with or without serum powders. You could actually make the arguement that this build is more flexible than past builds because your opening hands tend to me stronger, but that comes at the cost of worse slightly topdecks/Brainstorms and 1 less merchant scroll. I think this build will have a larger problem with Leyline game 1, but about the same if not slightly better game vs. other problematic cards that cost 2-3 mana purely because of its increased speed.

Rev Silence and Massacre are both nuts off the board for playing through the hate. Also, if Slice and Dice gets popular there is a Sliver that gives all Slivers +0/+1 that only costs 1 mana that you could board.

[jab]Wait a minute.  The evidence is clear, you didn't actually retire!  Razz

You care about the meta.
You still build decks. (which requires FAR more time investment than just borrowing cards and showing up to a tourney, btw.)
You still post here in mtg strategy related threads.
You've attended scg events following your own retirement thread.

This is similar to the crack head down the street who claims that he has come clean because he only smokes half as much now! hehe.  [/jab]

There's a reason they call it the cardboard crack. But really, when I moved to SC wanted to just stop playing, I wanted to focus on other things in life. A streak of bad luck changed my plans considerably. To get away from things and see some familar faces, I decided to go to SCG Roanoke. Through that decision I met a magic player in the area that was looking for a roommate and not asking for much for rent, considering that I was living in a hotel at the moment it was an easy decision. Most of his friends play magic so I guess found my way back into it.

[nataz]

Let's say you've got a very typical hand of 2 Land, 1 Mox, Merchant Scroll, Brainstorm, Sliver, and SW/Powder.

You mulligan that?

Between BS and Scroll -> Recall, you have the ability to see 7 cards (8 if its wraith) by your next turn. Land, mox -> scroll for recall. Draw for next turn, Brainstorm, fetch (assumed), recall. Thats pretty much insane.

Also, its not terrible to go land -> BS t1. Considering that you run on color moxen its most likely that the mox is either black or blue (i'd run another ESG over Green right now i think), opens up more options depending on what you see. Even at worst case where you BS on their endstep (if you are worried about duress say), at least you can pull a shuffle effect off the merchent scroll next turn, or better yet get a free shuffle off your land if its a fetch. Either way, you should get three new cards off your recall.

My point is to me, that hand isn't that bad at all, why would you risk a worse 7? I tested the powder version in about 2 dozen games, and realized that in a lot of hands serum powder was just dead weight. Sure I could risk another 7, but with the hand above, you would not have a deck with only 3 brainstorms, 3 scrolls, 16 mana sources. I'm not sure its really worth it.   

[ErkBek]

Let's say you've got a very typical hand of 2 Land, 1 Mox, Merchant Scroll, Brainstorm, Sliver, and SW/Powder.

You mulligan that?

Between BS and Scroll -> Recall, you have the ability to see 7 cards (8 if its wraith) by your next turn. Land, mox -> scroll for recall. Draw for next turn, Brainstorm, fetch (assumed), recall. Thats pretty much insane.

So you're goal with this hand is to get a turn 3 kill with at least 1 counter backup. A turn 2 would be nice though.

You've got 2 lines of play (let's assume that's a emerald or jet)

option 1
Turn 1) Cycle SW + BS (hit Green Pact or Hulk in 4 cards - 44% chance)
  Turn 2a) Scroll for Flash
  Turn 3a) Cast Flash
        -The odds that you'll hit at least 1 counter in the 5 cards you see are about 50% (that's not counting hitting another blue card too for FoW). So all and all you've got a 22% chance of having a turn 3 kill with counter backup with this route. A turn 3 kill with 1 counter backup is very unimpressive considering the deck averages a turn 2 with counterbackup.

Turn 1) Cycle SW + BS (hit Flash - 27% chance)
  Turn 2b) Scroll for Mystical and mystical for green pact
  Turn 3c) Cast Flash
         -So you saw 4 cards that game 1 of which was flash. The odds of hitting Pact or FoW + blue card are 32%. So the odds this really working out is 8.5%.

Turn 1) Cycle SW + BS and you hit something really broken that doesn't include flash or hulk, let's say a 10% chance

****This line of play gives you a win percentage around 40% of the time assuming all you need is 1 counter backup to kill on turn 3. I wouldn't expect this line of play to be remotely effective in a tournament. I'd totally rather have Serum Powder and take 7 new cards.

option 2
Turn 1) Cycle SW and scroll for ancestral
  Turn 2a) Cast Ancestral and cast flash (hit flash and hulk in the 5 new cards = 14%)
  Turn 2b) Cast Ancestral and BS (hit both flash and hulk = 34%    and pact/FoW = 24%)
  Turn 3b) Cast Flash
I'll be generous and say there is a 5% chance you hit the nuts some other way via brokeness. 

****This line of play still only reaches your goal about 43% of the time.

I've done my best with the numbers, I think they are right. Either way though, if you look at the lines of play there is a ~60% chance you're going to lose. It's funny, but with this deck its safer to mulligan these so called "safe" hands. Notice in option 2, you blew your first 2 turns looking for about the same amount of cards you'd get with serum powder on turn 0.

Conclusion: This hand sucks.

[nataz]

Option 1: draw 4 cards with SW + Brainstorm:

The point is, you aren't just looking for hulk/pact and the numbers are more complicated then you make out. What if you see a recall or a lotus off the brainstorm? Perhaps a demonic or a vampiric tutor instead. Considering you are going to have 3 mana on turn 2, its not terrible impossible that you hit one of your other fast mana sources (4 Pact -> ESG, ESG itself, Lotus, Petal, Second and third mox = 9 possible hits), for 4 mana on turn 2 which would enable a scroll -> flash a full turn earlier.

Not to mention bluffs like scroll into -> flash with mana up but no pact/hulk (to bluff the win against a better hand to buy time). the great thing about running out an early flash is they have to counter it, often with a FOW, which puts them down 2 cards in hand to my 1. As long as you are careful with pacts, you can run out multiple theats and still have a decent chance to come back with a good top deck/brainstorm. 

I've done my best with the numbers, I think they are right. Either way though, if you look at the lines of play there is a ~60% chance you're going to lose.

Perhaps part of the disconnect stems from goldfishing v. actual playtesting? I've won plenty of games against very good players after turn 2. Sure, its not the prefered way to win, but beating down with slivers, or a hulk, or baiting out flashes does win more then its fair share of games. 

To be fair, for me the argument is not really serum powder v. SW, its serum powder at all. Looking at your list, you've had to make room for 4 powder, 1 extra heart sliver, and you are thinking about adding more slivers to make it more consistant. All of this is to up the kill rate to a stable 75% w/out counter back up by turn 2. While that is impressive, its not like list w/out powder are that much slower. What was Mons getting, something like mid high 60's mid 70's with no powder for a t2 goldfish? That % gap comes at the cost of space.

That space can be filled with things that help in real games. Things like an extra set of counters (Daze, Misdirection) or perhaps maindeck Duress. It comes at the cost of extra chain of vapors, and the loss of extra tutoring power (imperial seal). It comes at the cost of less pure card draw (think cantrips like opt/serum visions, or even street wraith). Quite frankly, I'm not convinced the 5%-15% (and I think I'm being generous here) is worth the loss in real world stability.

look, I'll be the first to admit I could be wrong, very wrong in fact. I'm not wedded to the idea of powderless flash. In fact, if you look back at the SCG thread, I was one of the first to respond with comments. I like the idea, and I did give it at least 2 dozen test games over the past few months. It just never seemed promising enough to really persue when I was simply wining games with hands similar to those discribed above, and powder would have been better as anything else.

Well, intrigued by the new deck and the claims made here, I used this build for 100 goldfish-games, 50 each on the play and draw:

   3 Underground Sea
        1 Flooded Strand
        4 Polluted Delta
        1 Island

   1 Black Lotus
        1 Lotus Petal
        1 Mana Crypt
        1 Mox Emerald
        1 Mox Jet
        1 Mox Pearl
        1 Mox Ruby
        1 Mox Sapphire
        1 Elvish Spirit Guide

        1 Misdirection
        4 Pact of Negation
        4 Force of Will

        1 Chain of Vapor

        1 Demonic Tutor
        1 Vampiric Tutor
        1 Mystical Tutor
        1 Imperial Seal
        4 Merchant Scroll
        4 Summoner's Pact

        3 Street Wraith
        1 Ancestral Recall
        4 Brainstorm

        1 Carrion Feeder
        1 Karmic Guide
        1 Body Snatcher
        1 Kiki-Jiki, Mirror Breaker

        4 Flash
        4 Protean Hulk


I've tested some PitchLong with Pacts in place of MisD and Wraiths, and PoN impressed me there a lot. Same for Wraith, especially with topdeck tutors.
So when I realized Duress was slowing me down a real lot, I replaced it.
As for Wraith, it's just the perfect deck for it. You're searching to combo-pieces and only a little mana. 56 cards would be great. Sadly, I haven't found anything I'd  really like to cut (I might actually want go to only two Wraiths with another land).
It is probably better to actually create a well balanced 56 than anything else, But so far I'm not sure how. I just know it's not a mana that has to go.

Now, after much BlaBla, here are my results (remember, goldfishing, not playtesting, no real opponent)

Oh yeah,
tX = turn X kill
tXbY = turn X kill with Y active counters

for back up.

on the play 50 games

t1:   =  6  12% 12%
t1:   =  2
t1b1: =  3
t1b2: =  1

t2:   = 24 48% 60%
t2:   =  5
t2b1: = 14
t2b2: =  5

t3:   = 10 20% 80%
t3:   =  2
t3b1: =  3
t3b2: =  5

t4:   =  4 8% 88%
t4:1  =  1
t41b: =  1
T4b2: =  1
t4b3: =  1

later =6 12% 100%

on the draw 50 games

t1:   = 16 32% 32%
t1:   =  3
t1b1: =  9
t1b2: =  4

t2:   = 29 58% 90%
t2:   =  9
t2b1: = 11
t2b2: =  9

t3:   =  3 6% 96%
t3:
t3b1:
t3b2: =  1
t3b3: =  2

t4:   =  0 0% 96%
t4:
t4b1:
t4b2:
t4b3:

later =  2 4% 100%


combined 100 games:

t1:   =  22  22% 22%
t1:   =  5
t1b1: =  12
t1b2: =  5

t2:   = 53 53% 75%
t2:   = 14
t2b1: = 25
t2b2: = 14

t3:   = 13 13% 88%
t3:   =  2
t3b1: =  3
t3b2: =  6
t3b3: =  2

t4:   =  4 4% 92%
t4:1  =  1
t41b: =  1
T4b2: =  1
t4b3: =  1

later =  8 8% 100%


Some interesting tidbits from these results here:

3/4th of all goldfish games end on turn 2 (not the claimed 90%, but not bad)

kills turn 1 through FoW: 17% kills till turn 2 with 1+ backup on the play: 46% (this deck should directly beat single Disruption nearly half the time on the play, just by goldfishing!)

THe deck has continuously more protected than unprotected kills, every single turn.

What makes this deck scary to me is that you can hardly always try to mulligan into double FoW, so if they have it, they have it. And that's 46% of the time. No other combo-deck I've seen so far  Most importantly, many kills without protection could have been the turn after with 1 or even 2 extra disruption-pieces.

In this context, a figure to think about:

kills till turn 2 on the draw against FoW + turn 2 drain: 44%
Even if you get double counter up on turn2, you still loose nearly half of these

games, only 16% of them on turn 1, where you only have one counter.


Fizzles (typically mulligan to oblivion):

8%
Seems kind of reasonable for combo, in my experience.


Now, looking at the goldfishing statistics for speed, this deck isn't something very special. GrimLong and Belcher both provide similar stats. The huge difference lies in the fact that the deck almost always has at least one backup-counter (in 73 of these 100 games to be exact). This is new.If you had an active FOW, you could usually expect to live to reach the next M.Scroll or Mana Drain. Here you can't. Well, make of this what you want, hope it further's the debate. I'll have to find 4 Leyline of the Voids, somewhere, for post Future Sight...

[ErkBek]

Nataz, how would you play the hand I posted? Apparently you'd play it differently since you basically blew off my entire post. Scrolling up Flash as a bait spell? Come on. I also addressed the Brainstorming into brokenness that doesn’t include flash or hulk with a rough estimate of 10%. Really, you need to hit something like Lotus and something or Vamp + Mox.

How are you planning not to get crushed by anything your opponents do (race you, duress, mindcensor, etc). I'm very aware that flash can slow roll hands, I have played the deck in real games, but I'd like to bring your attention to something very relevant from Steve's last article on the GAT - flash matchup

Both decks are hyper-aggressive. What I’ve noticed is that if the GAT player can survive the first two turns, however, its chances of winning reach over 80%. What’s more, if the GAT player can survive the Flash player’s third turn, the GAT player is probably over a 90% favorite. Keep that in mind.

I think vs. bomberman and the mirror things only get worse as you hit turn 3 and 4.

That space can be filled with things that help in real games. Things like an extra set of counters (Daze, Misdirection) or perhaps maindeck Duress. It comes at the cost of extra chain of vapors, and the loss of extra tutoring power (imperial seal). It comes at the cost of less pure card draw (think cantrips like opt/serum visions, or even street wraith). Quite frankly, I'm not convinced the 5%-15% (and I think I'm being generous here) is worth the loss in real world stability.

You're being generous? It appears to me to be 15%. No way should you count those goldfish games on the draw. Also, I wouldn’t count those turn 4 kills as much of anything, because you’ve given your opponent too much time to either kill you, get a disruptive perm on the board, duress you multiple times, or destroy your manabase. You talk about these other protection cards (duress, daze, MisD, and extra chains) that are only going to slow down those goldfish numbers (which are from are from a 9 counter, 1 chain deck).

My list put the game away by the end of turn 3 90% of the time, opposed to 80% of the time. Also, my turn 1 kill percentage is 8% higher.  Both are pretty significant differences in my eyes.

I really don’t think my list is much more unstable than other list. Really you pretty much lose to a resolved mindcensor or Chalice @ 2 either way. I think you’re better off running a faster list to beat these 2 cards (not to mention you ignore crypt and extirpate). Leyline and meddling mage are better dealt with a slightly greater tutor suite in you list (+1 Scroll and +1 Iseal).

[ErkBek]

Nataz im'd me which was cool, so we got our differences worked out. We also discussed the possibility of running the kiki kill over the flash kill but run 2 of each of the creatures for the kill (and 0 body snatcher). This way we'd still have 6 dudes, but we could potentially somewhat alleviate the RFG'g problem. It's debatable which kill is better in the current metagame.......but something tells me that vintage players love the sliver kill a little too much.

[Implacable]

The Sliver kill is clearly the objectively better one; it uses 1 more slot, but dodges a huge amount of frequently seen hate, and thus flips several matchups (Bomberman, Slaver) into the hugely favorable column.  However, with the rise of Goblins, I am considering placing the Kiki kill in the sideboard; While Fanatic stops both kills, blockers (the general strategy of Goblins against Flash) do not stop infinite Karmic Guides.

[Harlequin]

So I'm not that familiar with the deck list, but It appears that you have cut:

1 Hulk
1 Scroll
1 mana
1 other

for the extra 4 Serum powders.  Now let's reconsider this hand: 2 Land, 1 Mox, Merchant Scroll, Brainstorm, Sliver, and Powder.

Something that appears to be overlooked is that if the powder is not increaseing the winning chances of this had by +25% then it's not worth it.   "Why?" you might ask... If you kept the original list, then you had a 25% chance to have that Powder be a Hulk. 


From the opening post you stated that you used powder 39 times.  I think a key consideration you missed was how often could you have replaced the powder with a Hulk or Merchant scroll (or mana) to make the hand keepable, and how quickly that hand could have won.  Now granted it would have to be a very large %, because after you say: "yes this hand would be keepable with powder -> scroll" you have to remember that you only have a 25% chance to actually make good on that.  But still if you say that 30 out of 39 of those hands would have been keepable with a hulk/scroll/mana in place of powder, then 9 out of 30 of those hands would have been better without powder.    Now you can compare those 9 hands to powerding down 39 times, and see how they compare.  If in those powders the results are much better than 9 out of 39 - then powder is worth it.

Now I'm not saying that this is a reason why powder shouldn't be used - I'm just saying that its a better way to evaluate making a small change like this to an astablished decklist.

[Sephiroth]

Hi kobefan,

i like the deck very much, it seems really powerful (and also the name is pretty good  ) and serum powder helps a lot! 

But , i would to ask you a couple of questions:

- using 8 pacts (4 green and 4 blue) and a lot of cards cc 0 , you dont risk to suffer a lot more that "standard" flash COTVs ?

- did you have some problems with a combo that passes from attack phase like slivers one or you never had this kind of problem ?

- have you tested agains gifts x1 decks ? Do you cobnsider it a good match up for us ?

Thanks for helping and sorry for my bad english.

[ErkBek]

Something that appears to be overlooked is that if the powder is not increaseing the winning chances of this had by +25% then it's not worth it.   "Why?" you might ask... If you kept the original list, then you had a 25% chance to have that Powder be a Hulk. 


From the opening post you stated that you used powder 39 times.  I think a key consideration you missed was how often could you have replaced the powder with a Hulk or Merchant scroll (or mana) to make the hand keepable, and how quickly that hand could have won.  Now granted it would have to be a very large %, because after you say: "yes this hand would be keepable with powder -> scroll" you have to remember that you only have a 25% chance to actually make good on that.  But still if you say that 30 out of 39 of those hands would have been keepable with a hulk/scroll/mana in place of powder, then 9 out of 30 of those hands would have been better without powder.    Now you can compare those 9 hands to powerding down 39 times, and see how they compare.  If in those powders the results are much better than 9 out of 39 - then powder is worth it.

I did this for about 35 games for my initial list, but doing this for 100 games is very tedious and hard to keep track of and to go and explain. My notes for the initial 35 games can be found on the Serum Powder flash thread on SCG if you are interested. As for what I cut from my initial list that probably wasn't optimal

-1 Mox Ruby
-1 Mana Crypt
-1 Iseal
-2 MisD
-1 Merchant Scroll
-1 Hulk / Meta slot
+1 Fetch
+1 Heart Sliver
+4 Powder

- using 8 pacts (4 green and 4 blue) and a lot of cards cc 0 , you dont risk to suffer a lot more that "standard" flash COTVs ?

Most Flash decks have adopted running at least 6-8 pacts. I think you have to take this risk with this build.

- did you have some problems with a combo that passes from attack phase like slivers one or you never had this kind of problem ?

Not really, but then again, these are goldfishing results, not actual games. If you are worried about this you could switch to the kiki kill with 2x of all your dudes.

- have you tested agains gifts x1 decks ? Do you cobnsider it a good match up for us ?

Nope, but I've got to think it'll be like the GAT matchup, only better.