Oh, I'm not offended. We're just chatting, here.Originally Posted by andy
This is a seductive argument because it makes sense, logically, but it simply doesn't work like that in reality. If the above quote were true, there are a number of things we should be able to take from this.When you're just trying to find the T1 goldfish rate, the deck absolutely is linear. In any situation in which you are deciding between multiple lines of play, exactly one of those lines of play has the highest chance of resulting in a T1 win, and that is the correct play.
First, you are wasting your time goldfishing hands to work out the T1 win rate experimentally, when you could instead just work out the T1 win rate mathematically - it's all just statistics, right? Make a list of all the hands that win on T1 and compute the probability that you will draw a winning hand in your opening grip. Far more accurate than that experimental stuff; luck and pilot error have been taken out of the equation.
Second, we can use mathematics as a guide to building the perfect T1 deck. Obviously, decks with options make poor T1 decks, right? Every time you have to make a choice, you risk making the wrong choice. Look at my Bayou v Mox conundrum in my last post: if the odds are (say) 60-40 rituals or Culling then I've just cut my chances of a T1 win by 40% (because even though I'll always make the statistically correct play, my deck will unhelpfully feed me the "wrong" card 40% of the time).
This is probably why Belcher is so strong.
Human beings are simply not capable of actually playing like this. Chessmasters, who represent the top .00001% (estimated) of the population, think a dozen moves ahead with 32 pieces and no unknowns. Storm combo requires players to think a dozen or more cards ahead with 53 unknowns in their own deck, and another seven unknowns in their opponent's hand, and a further 53 unknowns in their opponent's deck.But the correct play doesn't depend on what you actually drew in any instance of this game. The correct play depends only on the probabilities that you will draw various cards. One of these plays is a higher percentage play than the other, even if we haven't yet identified which play is best.
NOBODY can actually play storm simply as a game of percentages.
This supports my initial point, then. If I lost, based on my choice, all goldfishing will do is tell us about my (or your) T1 percentages, not the deck's upper limit. Another pilot, making different choices, will get different results.No, since you actually fizzled. Although the proper way to determine which is the correct play is take that starting hand, shuffle the rest of the deck, and play 30+ games making the same choice every time. Then play 30+ games making the other choice every time. Then see which choice is the higher percentage play. (Note: if the two plays are very close in percentage, it may take significantly more games, even hundreds, to determine which is better.)
You can't have it both ways. Either the actual result takes precedence (as you argued above), or it does not. If you believe strongly in making the statistically correct play, you need to count the correct play as a "win", regardless of what the deck actually did. And that being the case, you don't actually need to goldfish - just crunch some numbers.The "wrong play in hindsight" is a particularly common fallacy. To give an example that makes it more clear why this reasoning is wrong: suppose you have Belcher in play with 1 land left in your 50-card deck. You have BBBB GG floating and Witness in hand, with Manamorphose in the graveyard, nothing else in hand, and no other unused mana on the table. Do you go for the Belcher activation, or do you Witness -> Manamorphose hoping to topdeck the perfect card to continue? The probability that your Belcher activation does at least 20 damage is 60%. Assuming all D4s are left in your deck, your chance of topdecking a D4 is 16% (so your chance of winning with the Witness line of play is strictly less than 16%). If you activate Belcher, seeing a D4 on top of your deck and your 1 land 10 cards down, that doesn't make your line of play "wrong in hindsight". It was obviously the higher percentage play, and you should make the same choice every time you find yourself in this position.
And for what it's worth, I consider the play that wins to be the correct play, regardless of the odds. Identifying the winning play from the merely probable is what separates the master pilots from the merely competent. I hope to join their ranks one day.
Eh. Colour me foolish, but I've played draw-style storm long enough to believe that a good pilot takes one part stats and one part experience before adding one part instinct, and seasoning with intuition for the perfect game.You shouldn't make your play decisions based on an assumption of what the deck will give you. You should make your play decisions knowing the chances of drawing what you need to continue your spell chain.
Edit: Do keep in mind that when people report X% T1 wins, they are probably reporting actual wins, which will involve games where they made the "wrong" choices, but "guessed" correctly. If I am right, and some form of luck or intuition does play a role in piloting storm, then no amount of testing or algorithms can duplicate the results of a particular pilot - win percentages would be highly individual, up to a theoretical maximum.
The difficulty, again, is the unknowns. Let's suppose you have BBB floating, and an LED in play. In your hand you have IT and a D4. We will keep things simple by saying that you can only cast the D4 and crack the LED, or IT for Slithermuse, cracking the LED. Essentially, D4 with B floating, or D7 with nothing floating.This in itself isn't enough to reject the 80% on-the-draw claim. We also need to know that the difference in T1 win rate (53.3% vs. 80%) isn't due to other factors (such as deck design or incorrect plays on my part). That's why I post my hands, in the hopes that other pilots claiming significantly higher T1 win rates can identify any play errors which might be affecting my results.
Trouble is that after you chose, you will never know what cards you would have seen if you had chosen differently. It's hard to identify "play errors" of that nature.
@Kanti: 0/ It's good to see another Solidarity player here. Personally, I've found PSI to be a really good Solidarity replacement, as it functions on very similar principles, but faster.
Like, 3 turns faster ^^. And who'd have thought? We do play storm the same Silent :D
A very interesting discussion broke out here. Actually two discussions, one about Slithermuse and one about the philosophy of the deck.
When I came to a point where I thought I could play P.S.I. at least decently I started keeping track of all my goldfishs. However those aren't only real goldfishs, I like to roll a dice for a turn 1 Cabal Therapy/Duress/Inquisition of Kozilek, Force of Will/Daze/Spell Pierce or turn 2 Hatebears. I still do simple goldfishs, alternately on the play and on the draw, and I can look up the T1 percentage when I am at home. But I think this will be a not very telling statistic. Imho this deck is so good not because you can win on turn 1 very often, but because you can safely win on turn 1, 2 or sometimes even 3 very often. A 80 % T1 rate is still bad when the other 1/5 of your games are total fizzles. At least I see it that way.
Discard is played very rarely in the current meta so I don't see why we should press for T1 kills every time. Plus, like Vacrix said, for a high T1 ration you need to take mulligans very aggressively. Even though this deck DOES mulligan very well, drawing only 5 cards can be a pain in the ass.
For the math-part: I play my games mostly on instinct. 30 seconds are often just not enough to do stochastic calculations. I do however, 5-6 times per month take my time during goldfishing and do the math on almost any decision that the deck wants from me. The instinct I mentioned has to be developed like that I find, so you can later recall the percentages when you have a identical or similar situation in a tournament.
It was already mentioned and I agree: Humans can't always find the best line of play in storm decks. At least not in 30 seconds time.
On the Slithermuse-subject: I'd never cut her. Unfortunately I did not keep track of games, where IGG would have been better than Slithermuse, but it did not happen often. Slithermuse can also be played during D4 chains, she's not only in this deck to be searched by IT.
EtW is a bad replacement. This deck wants to kill on turn 1 or 2 and all cards in the IT-toolbox should help you do that.
PS: I forgot: Hillarious picture^^.
Last edited by Summerrain; 07-05-2012 at 01:06 PM.
I too love me a slithermuse. But I just had a thought... Maybe its better 1 of on the board vs hatebears on the play (or something like that...). Also dont forget - he's an out to leyline of sanctity ^^ (comon now, we've all been there XD)
Empty the Warrens vs. Slithermuse
I've been pretty happy with Empty the Warrens. Outside of hands with Pact, ie. mostly shorter spell chains, Empty the Warrens is much better. Its great past the first turn and on the draw. Slithermuse is only good when you are on the play, which is half the time. Empty the Warrens is good on the play and the draw, can still be cast naturally from the hand but with more interactions with Chrome Mox (PiF, Cantor, Pact--> Cantor if you have LED to pay for Pact). It saves you +1 sideboard space as well since you want to board EtW in as at least a 1'of anyway. I haven't come to any conclusions yet because I have limited experience at the moment without Slithermuse (which I've been running for years).
On turn 1 percentages
I guess I'll have to do a bunch of playtesting and record the results. Honestly I find this discussion rather trivial though. We know the deck is fast. It can win before bears come down consistently post-board once you have X copies of Empty the Warrens in the maindeck, though the higher X is, the better your chances. I think that fine tuning the build for speed, boarding, and/or resilience in real life situations is more important for getting the deck somewhere rather than testing its reputation for 60%. 60%, by the way, refers to a seasoned pilot. I pilot around 60% pretty consistently but I've been playing and learning this deck since I first read about it in 2007. Its hard to pick up the deck and be a seasoned pilot within less than 1000's of games. As multiple people mentioned, there are hands with multiple business that are rather ambigious and based on scouting, predicting matchups, or playing it safe vs. balls to the wall, different pilots are going to play it differently. This also takes into account a known metagame vs. an unknown metagame. Frankly, in a known metagame, I play much differently than in an unknown metagame. In my metagame, I'll keep a no-business Mox x2 hand if I know that all the control players showed up. If I see aggro players, that hand is a mulligan unless its game 1 and I know they don't have disruption. A pilot will have a difficult time playing this deck to maximum efficiency without knowing the general metagame and how they expect an opponent to board really well. Granted, we always have the surprise factor with our extensive post-board options; I don't know of another deck that can board in the full 15 for its weakest matchup. However, a pilot cannot expect to ride a good sideboard plan and the dark horse factor on the journey to 'seasoned pilot.'
So in short, I will do a large play testing session (after practicing first of course) and share the results. I assure you it won't ever be less than 50% and never has been in my testing. Often I've been disappointed that I haven't been able to create a build that consistently tops 60% though. Best percentages I've ever seen in a larger sample size, and I've mentioned this before, was DireLemming's 50 games testing of a Burning Wish list; 70% turn 1 kills. I don't think that can necessarily be replicated but anything above 50% is ludicrous speed for Legacy.
Burnt Offering
Last time I looked into this card I was sorely disappointed. It required multiple maindeck Skyshroud Cutters and having a forest in play is a big qualifier for a ritual.
However, it occured to me that I had not explored some previously outdated tech. Vine Dryad is Cutter 2.0 and was replaced by Odious Trow a while back. Both cost 4, both have alternative casting costs and both produce 4 total mana, any combination of red and black mana. This opens up the possibility of Burning Wish and Past in Flames in the maindeck. I'm having a lot of trouble with a build at the moment but playing 61 cards just to see if it even works, I'm consistently goldfishing with it as a live card. Cutter is fucking awesome either way; I might have to start maindecking this guy. Its nice to know you can convert Pact directly into a tall man without having to pay mana, meaning you have more floating and it makes longer spell chains easier. Given that Pact makes longer spell chains more risky, for example, if you are chasing business or mana to continue, having more floating is extremely helpful.
I have reached no conclusions about Burnt Offering yet though. I'm currently playing 2 (1 copy being the 61st card). Access to +2 Culling the Weak is fucking savage. The deck list is tight though. Burnt Offering will be significantly better as well after we board against aggro; if we are sitting on the right cards, Burnt Offering is a black ritual that can add red mana. This is huge. If cast naturally without Pact, we have +2 red sources. Otherwise, we can potentially pay for Pact with a LED that we did not use to cast a natural EtW (I've done this plenty of times).
Also.. this never occured to me until now; Allosaurus Rider might be a pretty beast Pact target if we can configure the deck to play Burnt Offering. One black add 7? I think I have a boner. All you need in hand is Pact + ESG/Land Grant/tallman, + IMS, + 1 business. No idea if this is good or not but it might create some insanely explosive turn 1 plays.
Luck is a residue of design.
I'm an aspiring Psychedelic Trance musician. Please feel free to enjoy my sense of life:
http://soundcloud.com/vacrix
Expect me or die. I play SI.
In a real match situation, I would need to know more (e.g., round, current record, opponent's deck if I had scouted it, etc.). For the purposes of maximizing the T1 goldfish rate, I see two main options:
1) Petal, Ritual, Contract, hope to continue.
2) Mox (imprinting LG), LED, Petal, Ritual, Contract, sacing LED in response for BBB.
Given my testing so far, I think option (2) is the higher percentage play, although I don't have a large enough sample size for this scenario to be sure.
I don't yet know all of the relevant percentages, but yes, I try to do this. At the very least, the percentages for successful Belcher activations are fairly easy, and I would imagine any good Belcher pilot knows that the probability of a successful Belcher activation with 50-ish cards left in the deck and 1 Taiga remaining in the deck is approximately 80%. I don't understand the attitude that we cannot or should not try to know the important probabilities relevant to SI.
Yes, and, because it makes sense logically, I think the argument is correct.
If I knew enough of the relevant probabilities to do this calculation, I would. There are basically no current Legacy combo decks that are simple enough to be analyzed in this manner. So I'll have to settle for goldfishing.
With a draw-based combo deck, it's not the case that a starting hand is either a winner or a loser on its own. Starting hands each have some probability of winning and losing. Since we don't know the probabilities for all starting hands, this approach is not likely to work.
If there is some game situation in which you have exactly two options, Play A and Play B, such that Play A has a 60% chance of leading to a win and Play B a 40% chance, then you should choose Play A every time you find yourself in that situation. If you don't have any ability to distinguish between the situations where Play A would win and where Play B would win, then you simply cannot win more than 60% of the time when you find yourself in this position. This is not a failure of your play decisions. This is just how probability works.
Of course this method of play is possible. But I'm not even arguing that we should do this entirely, because the amount of effort it would take to generate the necessary data is too large. Still, we can try to identify clear opportunities to raise our T1 win rate. We don't have to give up entirely because perfection seems unattainable.
I wasn't trying to have it both ways. In the situation I described, I would count the Belcher activation play as a loss. But if I were put into that situation a hundred times, I would expect the Belcher activation play to win 60% of the time, whereas I would expect the Witness line of play to win far less than 16% of the time. This makes the Belcher activation clearly the correct play, even if it doesn't always work out.
This is not completely correct. Consider the Belcher/Witness lines of play from the previous example. No matter how good the pilot, they cannot have knowledge of where their 1 land is in the deck. We're talking about situations where we cannot know which play will actually pan out (because they depend on cards drawn). The best we can hope to know is which play is more likely to pan out. This is true even of the most skilled pilots.
Even if this were the case, the law of large numbers says that this can't happen forever. If somebody is reporting 60%+ T1 win rates for large sample sizes of games, then either it's actually possible with the deck or they're misreporting. I would guess the latter. I would guess that most people reporting certain T1 percentages are simply not keeping careful data. It's more likely that people are reporting from memory; or combining on-the-play with on-the-draw data; or including "virtual wins" (e.g., first turn Belcher with no activation, or first turn Empty the Warrens); etc.
Perhaps I'm misreading your example, but I think you mean the choice is between D4 with BBB floating or D7 with 0 floating. I'm still not sure which of these is correct, and this decision comes up fairly frequently. But one of them is a higher percentage play, and it's up to us to determine which one it is. (Part of this may depend on other features of the game state. For example, I currently believe the choice between these two lines of play depends heavily on whether or not a land has already been played. But as I said, I don't have enough sample hands yet to know which play is correct.)
If you feel the discussion is trivial, and Silent Requiem, The Spanish Tunnel King, kicks_422, and Summerrain seem to think we can't (or shouldn't) use data to determine a correct play, perhaps I should stop wasting screen space with goldifsh sample hands. But honestly, I'm surprised to find such an adamant defense of play-by-instinct and such a strong deference to the luck of the draw in a group of combo players.
But it's not really one black mana to add 7. You pay two green cards for Allosaurus Rider. So in the end, you've used four cards and one black mana to get 7 mana in a combination of black or red. +6 mana for the cost of 4 cards is just +1.5 mana per card. You already get a better investment from Dark Ritual and Culling the Weak on Arbor/Skyshroud Cutter.
Please don't stop doing here what you, in this post, refer to as "waste of screen space". You are doing a great job and you are asking questions that I think are relevant and healthy for this deck. I really think you got me wrong on the context. When I say "play-by-instinct" I mean "play-by-instinct-backed-up-by-memory-of-calculations". Imho the goal is to do mathematical sound goldfishs, remember as many percentages of as many situations as you can and then, on a tournament, decide in between the 30 seconds rule what your best line of play is. Out of memory from your previous calculations.If you feel the discussion is trivial, and Silent Requiem, The Spanish Tunnel King, kicks_422, and Summerrain seem to think we can't (or shouldn't) use data to determine a correct play, perhaps I should stop wasting screen space with goldifsh sample hands. But honestly, I'm surprised to find such an adamant defense of play-by-instinct and such a strong deference to the luck of the draw in a group of combo players.
I feel a bit sad now, because your approach is a very fine, scientific one and I simply don't have enough time to use this approach every time I practice the deck. However what I wanted to say in the last post and I'll say it again now, long story short:
You don't have enough time on a tournament for a really acurate stochastic calculations -> You have to decide on instinct/intuition (whatever you call it) and this "instinct" better be lined by serious and scientific goldfishing/testing/brainstorming.
That's the way decisions in this deck are done, in my opinion. Of course you sometimes stick to the wrong decision and after the tournament, when you replay it, notice there we're better line of plays out there.
I want to apologize here, because I promised to post my turn 1 percentages. I'm not at home yet, so I can't look them up. Sorry. I'll post them tomorrow.
I haven't defended not getting data on this on that basis. Keep in mind this deck has been around for years. I think our time is better spent discussing things that aren't already known. I'll do a solid 50 games testing and post my turn 1 percentages. Since I've been playing this deck longer than most people who post here frequently I figure I will be able to show you roughly 60% turn 1 kills as seasoned pilot.
Testing has shown Allosauraus riders to be mediocre to bad; I've had some explosive hands with it and then other times I just wished it was something else. I'm still going to explore the Burnt Offering variations though; I think it could be promising as a 1 or 2 of.
EDIT:
Summerrain is right though. You have memorize plays and percentages, do basic calculations to discover the chances of what options you are chasing with a D4, gamble when you think its apporpriate, and slow play if you have enough perpetuals to play around Spell Pierce, Daze, etc, as well as play around Wasteland (relevant sometimes). Intuition and Instinct are important skills to develop with this deck. Its not a surefire win when you go off with D4s like it is with Ad Nauseam, Doomsday, or EtW; you have to have confidence in your build, know it well, and mulligan according to matchups if you can scout (which is yet another skill).
Luck is a residue of design.
I'm an aspiring Psychedelic Trance musician. Please feel free to enjoy my sense of life:
http://soundcloud.com/vacrix
Expect me or die. I play SI.
On a personal note, playing the deck with such a heavy weight put on percentages and probabilities makes it not fun anymore.
The Source: Your Source for "The Source: Your Source for..." cliche.
You don't have to take it to that extreme. I for one am for testing the shit out of the deck in a purely goldfish sense as it allows you to take whatever you learned from those sessions into a real game.
Like we know it's a bad idea to Belch someone with 2 lands left in the deck (unless you are running Taiga, though I'd like to know the percentages on that). We know that because of tesitng.
About the whole right or wrong decision thing, I gotta go with andy. You can not say you did not make the best decision because you have hindsight.
Thanks. The "waste of screen space" that I was referring to was not my posts asking these questions. It was massive sample hand dumps, such as this one:
Aside: Why is there no spoiler/hide text functionality on these boards?
A few stats:
- 30 sample hands, all on-the-play
- 10 T1 wins (33.3% T1 win rate)
- 15 fizzles (50% fizzle rate. Note: fizzle just means I started a spell chain which didn't eventually win.)
- Every single time (7 times) my spell chain started with a D4 floating 0 mana, I fizzled. From now on in my goldfishing, I'll be mulliganing 7-card (and probably 6-card) hands on-the-play which can only play a D4 floating 0 mana.
- 4 times the spell chain starts with a D7 floating 0 mana. These were split between wins and fizzles.
- When opening with a D4 floating mana (14 times), there were 6 wins and 8 losses. Opening with a D7 floating mana only occurred 3 times (2 wins and 1 loss).
- I mulliganed at least once in 12 games (40%), compared to 19 games (63%) with my first set of 30 hands. Although the T1 rate didn't go up that much, the mulligan rate fell significantly.
And now the games:
1. (T1) LG, Mox, ESG, MM, LED, Muse, Contract: Keep.
T1: LG (1) -> Bayou (1, B). Mox (Muse) (2, BU). ESG (2, BUG). MM (3, BBB) -> ESG (3, BBBG). LED (4, BBBG). Contract, sac LED (5, BBBG).
D4: CRit, Bargain, Arbor, Tendrils.
CRit (6, BBBBG). Bargain (7, BG).
D4: DRit, Petal, Mox, Cantor.
DRit (8, BBBG). Petal (9, BBBG). Tendrils for 20.
2. 2xLG, Mox, Pact, ESG, CRit, Contract: Keep.
T1: LG (1) -> Bayou (1, B). ESG (1, BG). CRit (2, BBB). Contract (3).
D4: Petal, LED, Contract, Belcher.
(I don't see that I can continue here. I can't get to 4 mana for the Belcher. I can't get BBB for the Contract. I can't do anything with a Witness floating BB. Easy T2.)
3. Arbor, Petal, LED, MM, Tutor, Bargain, Tendrils: Mull. (Good chance of T2, no chance of T1.)
LG, Mox, Pact, 2xCRit, Tutor: Mull. (All I see is Tutor floating BB, not enough for T1.)
2xMox, ESG, CtW, LED: Mull.
Petal, 2xDRit, CtW: Mull.
Mox, LED, MM: Fail.
4. (T1) LG, Petal, Pact, Cantor, 2xCRit, Bargain: Keep.
T1: LG (1) -> Bayou (1, B). Petal, sac (2, BB). CRit (3, BBB). Bargain (4).
D4: Mox, CRit, LG, Tutor.
Pact (5) -> ESG (5, G). Mox (LG) (6, GG). Cantor, sac (7, BG). CRit (8, BBB). CRit (9, BBBBBB). Tutor (10, BBBB) -> Tendrils for 22.
5. (T1) LG, Mox, 2xPact, Cantor, CRit, Contract: Keep.
T1: LG (1) -> Bayou (1, B). Pact (2, B) -> ESG (2, BG). CRit (3, BBB). Contract (4).
D4: DRit, Contract, Bargain, Tendrils.
Mox (Contract) (5, B). DRit (6, BBB). Bargain (7).
D4: Mox, DRit, LED, Contract.
Mox (Contract) (8, B). DRit (9, BBB). Pact (10, BBB) -> ESG (10, BBBG). Tendrils for 22.
6. Arbor, Pact, CtW, LED, Witness, Cantor, Contract: Keep.
T1: Arbor. Pact (1) -> ESG (1, G). Cantor, sac (2, B). CtW (Arbor) (3, BBBB). LED (4, BBBB). Contract, sac LED (5, BBBB).
D4: LG, Pact, LED, Tendrils.
(I don't think I can continue here. Witness is in the yard, so my only castable Pact target is Trow, which will leave me unable to cast Tendrils.)
Fizzle.
7. (T1) LG, ESG, Pact, DRit, CRit, LED, Bargain: Keep.
T1: LG (1) -> Bayou (1, B). ESG (1, BG). Pact (2, BG) -> Cantor (3, B). DRit (4, BBB). CRit (5, BBBB). LED (6, BBBB). Bargain, sac LED (6, BBBB).
D4: Petal, CtW, Contract, Tutor.
Contract (7, B).
D4: LG, CtW, Trow, Tutor.
CtW (Cantor) (8, BBBB). Trow (9, BBB). CtW (Trow) (10, BBBBBB). Tutor (11, BBBB) -> Petal. 2xPetal, sac (13, BBBBBB). LG (14, BBBBBB) -> nothing. Tutor (15, BBBB) -> Tendrils for 32.
8. (T1) Arbor, LG, 2xPact, CtW, LED, Muse: Mull. (Probably an easy T2, but I don't see how I can cast that Muse this turn.)
LG, 2xPetal, ESG, LED, Contract: Keep.
T1: LG (1) -> Bayou (1, B). ESG (1, BG). 2xPetal, sac (3, BBBG). LED (4, BBBG). Contract, sac LED (5, BBBG).
D4: DRit, CRit, Pact, Tendrils.
Pact (6, BBBG) -> Cantor (7, BBB). DRit (8, BBBBB). CRit (9, BBBBBBBB). Tendrils for 20.
9. LG, ESG, Pact, Cantor, Tutor, Bargain, Contract: Mull. (Can't get to BBB to D4.)
LG, LED, Cantor, Trow, Muse, Contract: Mull.
Pact, CtW, Witness, Tutor, Contract: Mull.
Mox, Petal, DRit, Contract: Keep.
T1: Petal, sac (1, B). DRit (2, BBB). Contract (3).
D4: LG, ESG, CRit, Contract.
LG (4) -> Bayou (4, B). ESG (4, BG). CRit (5, BBB). Contract (6).
D4: 2xPact, CtW, Tendrils.
(Already played Bayou this turn. Can't get to 4 for Tendrils.)
Fizzle.
10. Arbor, LG, ESG, Petal, CRit, LED, MM: Mull.
LG, Mox, Petal, Trow, LED, Tutor: Keep.
T1: LG (1) -> Bayou (1, B). Mox (Trow) (2, BB). Petal, sac (3, BBU). LED (4, BBU). Tutor, sac LED (5, BBBU) -> evoke Muse (6).
D7: Mox, ESG, MM, Contract, Bargain, Tutor, Belcher.
Mox (Contract) (7, B). ESG (7, BG). MM (8, BB) -> CtW.
Fizzle.
11. (T1) Petal, ESG, CRit, LED, 2xContract, Tutor: Keep.
T1: Petal, sac (1, B). ESG (1, BG). CRit (2, BBB). LED (3, BBB). Tutor, sac LED (4, BUUU) -> Muse (5).
D7: Arbor, LG, Mox, Pact, DRit, 2xTendrils.
Arbor (5). LG (6) -> Bayou. Mox (Tendrils) (7, B). Pact (8, B) -> ESG (8, BG). DRit (9, BBBG). Tendrils for 20.
12. Bayou, Mox, CRit, Trow, Contract, Tutor, Muse: Keep.
T1: Bayou (B). Mox (Trow) (1, BB). CRit (2, BBB). Contract (3).
D4: DRit, LED, Bargain, Belcher.
Fizzle.
13. ESG, CRit, 2xLED, Trow, Witness, Tutor: Mull.
Arbor, Cantor, MM, Tutor, 2xTendrils: Mull.
ESG, CtW, LED, Muse, Tendrils: Mull.
Mox, DRit, LED, Cantor: Mull.
Petal, ESG, Cantor: Fail.
14. Petal, 2xPact, Contract, 2xTutor, Belcher: Mull.
Bayou, LG, CtW, CRit, LED, Bargain: Mull. (Probably an easy T2, but not a T1.)
Petal, ESG, 2xCRit, Pact: Mull.
ESG, Tutor, 2xContract: Mull.
Mox, Pact, MM: Fail.
15. LG, 2xMox, ESG, Tutor, Bargain, Belcher: Mull. (T1 Belcher with 3 PMSs is great, but no chance of T1.)
Arbor, 2xLG, CtW, DRit, Belcher: Mull. (Another great T2 hand.)
CRit, LED, MM, Witness, Bargain: Mull.
Petal, DRit, CtW, Bargain: Keep.
T1: Petal, sac (1, B). DRit (2, BBB). Bargain (3).
D4: DRit, Contract, Witness, Muse.
Fizzle.
16. Arbor, Petal, CRit, Pact, Witness, Contract, Tutor: Keep.
T1: Petal, sac (1, B). Pact (2, B) -> ESG (2, BG). CRit (3, BBB). Contract (4).
D4: Mox, LG, ESG, Tendrils.
Fizzle.
17. (T1) Mox, Petal, Pact, CtW, LED, MM, Tutor: Keep.
T1: Pact (1) -> Arbor. Petal, sac (2, B). CtW (Arbor) (3, BBBB). Mox (MM) (4, BBBBG). LED (5, BBBBG). Tutor, sac LED (6, BBBUUU) -> evoke Muse (7, BB).
D7: Mox, 2xLG, Petal, ESG, LED, Tendrils.
Petal, sac (8, BBB). Mox (ESG) (9, BBBG). Tendrils for 20.
18. (T1) LG, DRit, CRit, LED, MM, Contract, Tendrils: Keep.
T1: LG (1) -> Bayou (1, B). DRit (2, BBB). CRit (3, BBBB). LED (4, BBBB). Contract, sac LED (5, BBBB).
D4: CRit, CtW, Bargain, Muse. (7 in graveyard.)
CRit (6, BBBBBBB). Bargain (7, BBBB).
D4: Petal, CtW, LED, Tendrils.
Petal (8, BBBB). LED (9, BBBB). Tendrils for 20.
19. LG, ESG, 2xDRit, CRit, 2xLED: Mull.
Mox, ESG, CtW, DRit, CRit, Bargain: Keep.
T1: Mox (CtW) (1, B). ESG (1, BG). CRit (2, BBB). DRit (3, BBBBB). Bargain (4, BB).
D4: Arbor, Petal, 2xTutor.
Fizzle.
20. Mox, DRit, CtW, LED, 2xBargain, Tutor: Keep.
T1: Mox (CtW) (1, B). LED (2, B). DRit (3, BBB). Tutor, sac LED (4, BUUU) -> Muse (5).
D7: Arbor, LG, Mox, 2xESG, CRit, Contract.
Fizzle.
(Note: I checked to see whether Bargain instead of Tutor -> Muse would have been better. That would have fizzled as well: Arbor, Cantor, ESG, LG.)
21. Pact, DRit, 3xCRit, LED, Tutor: Mull.
Mox, DRit, Pact, ESG, CtW, Tendrils: Mull.
Petal, 2xDRit, LED, Bargain: Keep.
T1: Petal, sac (1, B). 2xDRit (3, BBBBB). LED (4, BBBBB). Bargain, sac LED (5, BBBBB).
D4: Mox, DRit, CRit, MM.
Fizzle.
22. Mox, CtW, 2xDRit, Pact, MM, Tutor: Keep.
T1: Pact (1) -> ESG (1, G). Mox (CtW) (2, BG). MM (3, BB) -> Petal, sac (4, BBU). 2xDRit (6, BBBBBBU). Tutor (7, BBBBU) -> Muse (8, B).
D7: 2xMox, 2xPetal, LG, ESG, Tutor.
Mox (ESG) (9, BG). LG (10, BG) -> Bayou (10, BBG). Mox (nothing) (11, BBG). 2xPetal, sac (13, BBBBG). Tutor (14, BBB) -> Bargain (15).
D4: Trow, Contract, Bargain, Tendrils.
Fizzle.
23. Pact, 2xDRit, CRit, LED, Trow, Tutor: Mull.
Bayou, LG, CtW, 2xLED, Tendrils: Mull.
Mox, CtW, DRit, Pact, Bargain: Keep.
T1: Pact (1) -> Trow. Mox (Trow) (2, B). DRit (3, BBB). Bargain (4).
D4: Mox, Pact, Tutor, Bargain.
Pact (5) -> Arbor. Mox (Tutor) (6, B). CtW (Arbor) (7, BBBB). Bargain (8, B).
D4: Bayou, Pact, Petal, Tutor.
Fizzle.
(Note: higher percentage play from the outset would have been Pact -> Arbor, Mox (DRit), CtW (Arbor), Bargain floating B. Wouldn't have helped here, though.)
24. LG, Pact, ESG, Petal, 2xLED, Bargain: Keep.
T1: LG (1) -> Bayou (1, B). Petal, sac (2, BB). ESG (2, BBG). Pact (3, BBG) -> Cantor, sac (4, BBB). 2xLED (6, BBB). Bargain, sac 1 LED (7, BBB).
(Question: how to best utilize the LEDs here?)
D4: DRit, MM, Witness, Contract.
DRit (8, BBBBB). Contract, sac LED (9, BBBBB).
D4: CRit, LED, Contract, Bargain. (11 in graveyard.)
CRit (10, BBBBBBBB). Contract (11, BBBBB).
D4: DRit, LG, CtW, Bargain.
DRit (12, BBBBBBB). Bargain (13, BBBB, 1 life).
D4: LG, DRit, 2xPact.
Fizzle. (Witness is in the graveyard already.)
25. Mox, ESG, CRit, Witness, Contract, Bargain, Tendrils: Keep.
T1: Mox (Contract) (1, B). ESG (1, BG). CRit (2, BBB). Bargain (3).
D4: LG, Pact, Bargain, Tutor.
Fizzle.
26. (T1) 2xMox, Pact, LED, 2xBargain, Tutor: Keep.
T1: Pact (1) -> ESG (1, G). 2xMox (2xBargain) (3, BBG). LED (4, BBG). Tutor, sac LED (5, BUUU) -> Muse (6).
D7: Mox, Petal, CtW, Pact, 2xCRit, LED.
Mox (CtW) (7, B). Petal, sac (8, BB). LED (9, BB). CRit (10, BBB). CRit (11, B), response Pact (12, B), response sac LED (12, BGGG, threshold). Pact resolves -> Witness. CRit resolves (12, BBBBBBGGG). Witness (13, BBBBBB) -> Tutor (14, BBBB) -> Tendrils for 30.
27. (T1) 2xPetal, DRit, CRit, LED, Tutor, Bargain: Keep.
T1: 2xPetal, sac (2, BU). DRit (3, BBBU). CRit (4, BBBBU). LED (5, BBBBU). Tutor, sac LED (6, BBBBBU) -> evoke Muse (7, BB).
D7: Arbor, Pact, MM, DRit, CRit, Contract, Bargain. (8 cards in graveyard.)
DRit (8, BBBB). CRit (9, BBBBBBB). Bargain (10, BBBB).
D4: Mox, Pact, CtW, Tendrils.
Tendrils for 22.
28. LG, 2xDRit, 2xLED, 2xTendrils: Mull.
LG, 2xPetal, 2xLED, Tendrils: Mull.
LG, Pact, CtW, CRit, Tutor: Mull.
CRit, CtW, Tutor, Muse: Mull.
Pact, LED, Tutor: Fail.
29. LG, 2xESG, Bargain, 2xTutor, Belcher: Mull.
LG, ESG, CRit, LED, Trow, Tendrils: Mull.
CtW, Tutor, Contract, 2xBargain: Mull.
Mox, LG, Petal, Bargain: Mull.
LG, Tutor, Tendrils: Fail.
30. Bayou, LG, Pact, CtW, CRit Contract, Bargain: Keep.
T1: Bayou (B). Pact (1, B) -> ESG (1, BG). CRit (2, BBB). Bargain (3).
D4: ESG, LED, Tutor, Tendrils.
Fizzle.
And now for the on-the-draw hands:
- 30 hands, all on-the-draw
- I counted Slithermuse as a draw-to-5 in these games rather than draw-to-7. This seems like a more consistently representative choice than draw-to-6.
- 16 T1 wins (53.3% T1 win rate)
- 12 fizzles (40% fizzle rate. Note: fizzle just means I started a spell chain which didn't eventually win.)
- I mulliganed at least once in 13 games (43.3%).
- There were 4 hands (6, 14, 23, 30) which I kept needing my draw card in order to start going off. I tried to only do this when I had an extremely high number of outs (e.g., in game 30, where I basically needed any colored spell for a Chrome Mox imprint). Of these 4 games, game 23 was a loss and the rest were wins.
- In game 15, you can see the calculated win rate is approximately 42%. At the time, I had 6 out of 14 T1 wins (42%), so I figured a mulligan would lower my chances. Now that I've recorded a 53.3% T1 rate over a larger sample size, I would probably mulligan this hand.
And now the hands:
1. (T1) Petal, DRit, 2xMM, Contract, Tutor, Tendrils: Keep.
T1: Draw Pact. Petal, sac (1, B). Pat (2, B) -> ESG (2, BG). MM (3, BG, draw Mox). MM (4, BB, draw LG). DRit (5, BBBB). Contract (6, B).
D4: LG, Petal, DRit, Belcher.
LG (7, B) -> Bayou (7, BB). Petal (8, BB). DRit (9, BBBB). Tendrils for 20.
2. Arbor, LG, DRit, CRit, CtW, Petal, Mox: Mull.
Bayou, 2xLG, Petal, ESG, LED: Mull.
Mox, LG, 2xESG, Trow: Mull.
Pact, CtW, LED, Bargain: Keep.
T1: Draw MM. Fail.
3. (T1) 2xLG, Pact, DRit, Cantor, Contract, Bargain: Keep.
T1: Draw Mox. LG (1) -> Bayou (1, B). DRit (2, BBB). Contract (3).
D4: Arbor, CtW, Tutor, LED.
LED (4). Pact (5) -> ESG (5, G). Cantor (6). Mox (Bargain) (7, B). CtW (Cantor) (8, BBBB). Tutor, sac LED (9, BBBBB). Tendrils for 20.
4. ESG, 2xCtW, 2xDRit, Contract, Tutor: Mull. (about 68% to NOT draw one of my 17 outs)
Mox, Petal, ESG, CtW, LED, Tutor: Keep.
T1: Draw Tendrils. Mox (CtW) (1, B). ESG (1, BG). Petal, sac (2, BGU). LED (3, BGU). Tutor, sac LED (4, BBBU) -> evoke Muse (5).
D5: DRit, CRit, LED, Cantor, Contract.
Fizzle.
5. Cantor, DRit, LED, Tutor, Contract, Belcher, Tendrils: Mull. (About 40% to hit an IMS.)
Mox, DRit, ESG, LED, Bargain, Tendrils: Keep.
T1: Draw Mox. LED (1). Mox (Tendrils) (2, B). DRit (3, BBB). Mox (ESG) (4, BBBG). Bargain, sac LED (5, BBBG).
D4: Bayou, LED, CtW, Pact.
Bayou (5, BBBBG). LED (6, BBBBG). Pact, sac LED (7, BBBBGGGG) -> Witness (8, BBBBG) -> Bargain (9, BG).
D4: Mox, Petal, DRit, CRit.
Fizzle.
6. (T1) Arbor, Pact, 2xDRit, LED, MM, Contract: Mull. (17 outs).
LG, Petal, MM, Tutor, Trow, Contract: Keep. (?)
T1: Draw DRit. LG (1) -> Bayou (1, B). Petal, sac (2, BG). MM (3, BB, draw ESG). DRit (4, BBBB). Contract (5, B).
D4: Pact, CRit, CtW, Cantor. (5 in graveyard)
Pact (6, B) -> ESG (6, BG). ESG (6, BGG). Cantor (7, BG). Trow (8, B). CtW (Trow) (9, BBBB, threshold). CRit (10, BBBBBBB, hellbent). Tutor (11, BBBBB) -> Tendrils for 24.
7. (T1) Bayou, ESG, DRit, CRit, LED, MM, Cantor: Mull.
2xLG, ESG, 2xDRit, Contract: Keep.
T1: Draw Witness. LG (1) -> Bayou (1, B). 2xDRit (3, BBBBB). Contract (4, BB).
D4: DRit, CRit, LG, MM. (4 cards in graveyard.)
LG (5, BB) -> Arbor. DRit (6, BBBB). ESG (6, BBBBG). MM (7, BBBGG, threshold) -> Petal (8, BBBGG). CRit (9, BBBBBBGG). Witness (10, BBBBB) -> Contract (11, BB).
D4: CRit, CtW, Tutor, Bargain.
CtW (Witness) (12, BBBBB). Bargain (13, BB).
D4: Pact, Belcher, 2xContract.
CRit (14, BBBBB). Pact (15, BBBBB) -> ESG (15, BBBBBG). Sac Petal (15, BBBBBBG). Belcher, activate with 0 lands left (38 damage).
8. Mox, Petal, CRit, CtW, LED, 2xBargain: Keep.
T1: Draw Tutor. LED (1). Mox (CtW) (2, B). Petal, sac (3, BB). CRit (4, BBB). Bargain, sac LED (5, BBB).
D4: 2xPact, CRit, Trow.
Fizzle.
9. LG, Mox, 2xCtW, CRit, MM, Contract: Keep.
T1: Draw Petal. LG (1) -> Arbor (1). Mox (CtW) (2, B). CtW (Arbor) (3, BBBB). Contract (4, B).
D4: LG, 2xMox, LED.
Petal, sac (5, BG). MM (6, BB) -> draw LG.
Fizzle.
10. (T1) Petal, Pact, ESG, CtW, Bargain, Tutor, Tendrils: Keep.
T1: Draw Petal. Pact (1) -> Arbor. Petal, sac (2, B). CtW (Arbor) (3, BBBB). Bargain (4, B).
D4: DRit, LED, Pact, Contract.
ESG (5, BG). Tutor (6) -> DRit. Petal, sac (7, B). 2xDRit (9, BBBBB). Tendrils for 20.
11. Mox, Pact, ESG, 2xCtW, Witness, Tutor: Mull.
ESG, CtW, Tutor, Bargain, 2xContract: Mull.
CtW, CRit, 2xContract, Tendrils: Mull.
Arbor, Petal, ESG, Tutor: Mull.
DRit, LED, Tendrils: Mull.
Petal, LED: Fail.
12. ESG, Pact, DRit, LED, Cantor, Bargain, Tendrils: Keep.
T1: Draw MM. ESG (G). Pact (1, G) -> ESG (1, GG). MM (2, BG, draw Petal). Petal (3, BG). LED (4, BG). Cantor (5, B). DRit (6, BBB). Bargain, sac LED (7, BBB).
D4: Bayou, LG, ESG, MM.
ESG (7, BBBG). MM (8, BBBB, draw Petal).
Fizzle.
13. (T1) Arbor, LG, Petal, CtW, Muse, Tutor, Belcher: Mull.
LG, 2xPact, LED, 2xBargain: Mull.
Pact, ESG, Petal, DRit, Bargain: Keep.
T1: Draw Muse. Petal, sac (1, B). DRit (2, BBB). Bargain (3).
D4: LG, 2xCtW, Contract.
LG (4) -> Bayou (4, B). ESG (4, BG). Pact (5, BG) -> Trow (6, B). CtW (Trow) (7, BBBB). Contract (8, B).
D4: Mox, CRit, 2xBargain. (7 in graveyard.)
Mox (Contract) (9, BB). CRit (10, BBBBB). Bargain (11, BB).
D4: Arbor, CtW, DRit, Tendrils.
DRit (12, BBBB). Tendrils for 26.
14. (T1) 2xCtW, DRit, Trow, Bargain, Tutor, Tendrils: Mull.
2xPact, CRit, CtW, LED, Tutor: Keep. (27 outs)
T1: Draw Petal. 2xPact (2) -> Arbor, ESG (2, G). Petal, sac (3, BG). CtW (Arbor) (4, BBBBG). LED (5, BBBBG). CRit (6, BBBBBG). Tutor, sac LED (7, BBBBBBB) -> Belcher (8, BBB), activate for 27 damage.
15. (T1) Mox, ESG, 2xDRit, LED, Witness, Belcher: Keep. (Belcher activation with 2 lands left only wins about 37% of the time, but I also have 10 outs to increase my odds to 62%. EV = 42% win rate with this hand.)
T1: Draw Contract. Mox (DRit) (B). DRit (BBB). ESG (BBBG). Belcher. LED, activate Belcher with 2 lands left for 24 damage.
16. (T1) 2xPact, 2xBargain, 2xTutor, Belcher: Mull.
Pact, 2xPetal, LED, Contract, Tutor: Keep.
T1: Draw Mox. Pact (1) -> ESG (1, G). Mox (Contract) (2, BG). 2xPetal, sac both (4, BBGU). LED (5, BBGU). Tutor, sac LED (6, BBBBU) -> evoke Muse (7, B).
D5: Arbor, Pact, DRit, CtW, LED.
Arbor (7, B). CtW (8, BBBB). DRit (9, BBBBBB). LED (10, BBBBBB). Pact, sac LED (11, BBBBBBGGG) -> Witness (12, BBBBBB) -> Tutor (13, BBBB) -> Tendrils for 28.
17. (T1) CRit, LED, Cantor, Tutor, Contract, Muse, Belcher: Mull.
2xPact, DRit, Tutor, 2xBargain: Keep.
T1: Draw DRit. Pact (1) -> ESG (1, G). Pact (2, G) -> Cantor, sac (3, B). 2xDRit (5, BBBBB). Bargain (6, BB).
D4: Pact, CtW, DRit, CRit.
Pact (7, BB) -> Arbor. CtW (Arbor) (8, BBBBB). DRit (9, BBBBBBB). Bargain (10, BBBB).
D4: Petal, LED, DRit, Tendrils.
Tendrils for 22.
18. 2xMox, ESG, CtW, Trow, Bargain, Contract: Keep.
T1: Draw CRit. ESG (G). Trow (1). Mox (Contract) (2, B). CtW (Trow) (3, BBBB). Bargain (4, B).
D4: Bayou, LG, Bargain, Belcher.
Bayou (4, BB). (Obvious correct play here is to LG -> Arbor, Mox (Bargain), CRit, Belcher with 3 permanent mana sources. But chasing the T1:)
CRit (5, BBB). Bargain (6).
D4: ESG, Petal, 2xLED.
Fizzle.
19. Mox, Petal, Pact, CtW, CRit, LED, Trow: Mull.
LG, Trow, Tutor, Bargain, Belcher, Tendrils: Mull.
2xLG, CRit, LED, MM: Mull.
CtW, LED, MM, Muse: Mull.
CtW, LED, Tutor: Mull.
DRit, Contract: Keep.
T1: Draw Petal. Petal, sac (1, B). DRit (2, BBB). Contract (3).
D4: Bayou, ESG, Bargain, Tendrils.
Fizzle.
20. ESG, Petal, DRit, Trow, Tutor, 2xContract: Keep.
T1: Draw Muse. Petal, sac (1, B). DRit (2, BBB). Contract (3).
D4: CRit, LED, MM, Tendrils.
Fizzle.
21. (T1) Mox, 2xPetal, ESG, LED, 2xBargain: Keep.
T1: Draw Tutor. ESG (G). Mox (Bargain) (1, BG). 2xPetal, sac both (3, BBGU). LED (4, BBGU). Tutor, sac LED (5, BBBBU) -> evoke Muse (6, B).
D5: LG, Mox, MM, ESG, Bargain.
LG (7, B) -> Bayou (7, BB). ESG (7, BBG). MM (8, BBB, draw DRit). DRit (9, BBBBB). Bargain (10, BB).
D4: LG, ESG, CRit, Tutor.
Mox (LG) (11, BBG). ESG (11, BBGG). CRit (12, BBBBBBG). Tutor (13, BBBBB) -> Tendrils for 28.
22. (T1) LG, Pact, CtW, 2xCRit, 2xBargain: Keep.
T1: Draw DRit. (Note: safer line of play is LG -> Bayou, DRit, CRit, Bargain, saving the Pact for afterwards.)
LG (1) -> Bayou (1, B). DRit (2, BBB). Pact (2, BBB) -> Trow (3, BB). CtW (Trow) (4, BBBBB). Bargain (5, BB).
D4: Mox, Petal, Contract, Tutor. (6 in graveyard.)
Petal, sac (6, BBB, threshold). CRit (7, BBBBBB). Bargain (8, BBB).
D4: ESG, Pact, CtW, Tendrils.
CRit (9, BBBBBB). Tendrils for 20.
23. Pact, DRit, CRit, CtW, LED, Contract, Bargain: Mull. (21 outs)
Petal, LED, Trow, 2xMM, Bargain: Keep. (24 outs)
T1: Draw LG. LG (1) -> Bayou (1, B). Petal, sac (2, BG). LED (3, BG). MM (4, BG, draw CtW). MM (5, BG, draw Contract). Trow (6, B). CtW (Trow) (7, BBBB). Contract, sac LED (8, BBBB).
D4: LG, Tutor, 2xBargain.
Bargain (9, B).
D4: ESG, CtW, Contract, Tendrils.
Fizzle.
24. (T1) LG, Pact, 2xPetal, CRit, MM, Bargain: Keep.
T1: Draw Petal. 3xPetal, sac 2 (3, BB). CRit (4, BBB). Bargain (5).
D4: Mox, CRit, CtW, Tutor.
LG (6) -> Arbor. Sac Petal (6, B). CtW (Arbor) (7, BBBB, threshold). Pact (8, BBBB) -> ESG (8, BBBBG). MM (9, BBBBB, draw DRit). DRit (10, BBBBBBB). CRit (11, BBBBBBBBBB). Mox (nothing) (12, BBBBBBBBBB). Tutor (13, BBBBBBBB) -> Tendrils for 28.
25. (T1) Petal, 2xDRit, Muse, Tutor, Contract, Belcher: Keep.
T1: Draw Mox. Mox (Tutor) (1, B). Petal (2, B). 2xDRit (4, BBBBB). Contract (5, BB).
D4: ESG, LED, MM, Contract.
LED (6, BB). ESG (6, BBG). MM (7, BBB, draw Bayou). Bayou (8, BBBB).
(Note: Belcher activation with 1 land left has a 57% chance of winning here. Contract and Muse will probably see enough cards to have a higher percentage.)
Contract (9, B).
D4: DRit, CtW, Pact, Bargain.
DRit (10, BBB). Pact (11, BBB) -> Trow (12, BB). CtW (Trow) (13, BBBBB). Bargain (14, BB).
D4: CRit, Witness, Contract, Tendrils.
CRit (15, BBBBB). Tendrils for 32.
26. 2xPetal, ESG, CtW, DRit, CRit, Contract: Keep.
T1: Draw Bargain. 2xPetal, sac 1 (2, B). DRit (3, BBB). Bargain (4).
D4: ESG, LED, Tutor, Belcher.
2xESG (4, GG). Sac Petal (4, BGG). LED (4, BGG). CRit (5, BBBG). Tutor, sac LED (6, BBUUU) -> evoke Muse (7, B).
D5: Arbor, LG, Mox, CRit, MM.
Mox (LG) (8, BG). MM (9, BG, draw CtW).
Fizzle.
27. Petal, Pact, DRit, 2xCRit, Trow, Bargain: Keep.
T1: Draw LG. LG (1) -> Bayou (1, B). DRit (2, BBB). Bargain (3).
D4: Mox, 2xESG, MM.
2xESG (3, GG). MM (4, BB, draw Contract). Petal, sac (3, BBG). Pact (4, BBG) -> Cantor, sac (5, BBU, threshold). CRit (6, BBBBBU). Contract (7, BBU).
D4: Mox, Petal, Muse, Belcher.
(After 2 D4s, Belcher with 1 land left is a coin flip. Muse is safer.)
CRit (8, BBBBBU). Petal (9, BBBBBU). Mox (Trow) (10, BBBBBU). Mox (nothing) (11, BBBBBU). Evoke Muse (12, BB).
D5: (1 in hand.) ESG, CtW, MM, Contract.
Tap Mox (12, BBG). MM (13, BBB, draw LED). ESG (13, BBBG). LED (14, BBBG). Contract, sac LED (15, BBBG).
D4: LG, 2xPact, LED.
LED (16, BBBG). Pact (17, BBBG) -> ESG (17, BBBGG). LG (18, BBBGG) -> Arbor. Pact, sac LED (19, BBBBBBGG) -> Witness (20, BBBBB) -> Muse. Sac Petal (20, BBBBBU). Evoke Muse (21, BB).
D5: 2xPetal, Mox, 2xTutor.
2xPetal, sac both (23, BBBB). Mox (Tutor) (24, BBBBB). Tutor (25, BBB) -> Contract (26).
D4: DRit, Tutor, Contract, Tendrils.
Fizzle.
28. (T1) Mox, CRit, LED, Trow, MM, Contract, Tendrils: Mull.
LG, Petal, 2xDRit, CRit, Muse: Keep.
T1: Draw ESG. ESG (G). LG (1, G) -> Bayou (1, BG). Petal, sac (2, BGU). 2xDRit (4, BBBBBGU). CRit (5, BBBBBBGU). Evoke Muse (6, BBBB).
D5: LG, DRit, LED, Belcher, Tendrils.
LG (7, BBBB) -> Arbor. Belcher (8). LED, sac (9, BBB). Activate Belcher with 0 land left for 46.
29. Arbor, LG, Mox, ESG, DRit, Bargain, Belcher: Keep. (Approx. 43 outs to make this a likely T1 hand.)
T1: Draw DRit. (Can't play and activate Belcher; have to try the D4)
Mox (DRit) (1, B). DRit (2, BBB). Bargain (3).
D4: LG, Pact, MM, DRit.
ESG (3, G). Pact (4, G) -> ESG (4, GG). MM (5, BU, draw Mox).
(This was a very safe T2 hand, with Belcher, 0 lands left in deck, and 3+ permanent mana sources.)
Fizzle.
30. (T1) 2xMox, CRit, CtW, LED, Bargain, Belcher: Keep. (40+ outs to play a D4 floating at least BBB.)
T1: Draw CtW. 2xMox (2xCtW) (2, BB). LED (3, BB). CRit (4, BBB). Bargain, sac LED (5, BBB).
D4: LG, ESG, LED, Contract.
LG (6, BBB) -> Bayou (6, BBBB). ESG (6, BBBBG). LED (7, BBBBG). Contract, sac LED (8, BBBBG).
D4: ESG, DRit, LED, Tutor.
LED (9, BBBBG). ESG (9, BBBBGG). DRit (10, BBBBBBGG). Tutor (11, BBBBBG) -> Tendrils for 24.
I also sat down and played out 30 games looking for T1 wins on the play. I did not record the games in as much detail as andy, but I did record the results.
Of 30 games (on the play) I won 16, for a win percentage of 53.3%. This included four EtW wins, for which I required 10+ goblin tokens. In three of those cases, I had 14+ goblins. Obviously I could not play Pact as part of an EtW chain.
Because I'm testing EtW over Slithermuse, I took each EtW win and then kept playing them out as if I had cast Slithermuse instead. Of the four EtW games, Slithermuse would have won two instantly, and fizzled twice. I'm not sure that's better odds than hoping that my opponent can't deal with my goblins.
There was one game that did suggest a Slithermuse play but required Pact, so I couldn't use EtW. Instead, I just tutored for Cruel Bargain, which I was able to cast with one mana floating. It was enough.
There were three lost games I might have won if I had been playing Eternal Witness (Summoner's Pact & LED plus plenty of mana).
Ill Gotten Gains was broken; PiF would rarely have been as good. As the game goes on, of course, PiF becomes more powerful, but for winning on turn 1 IGG is simply better.
Five of my losses came from mulling to fewer than three cards. The rest were fizzles due to poor D4s. I say poor, but many of these losses would have put me in a very strong position to win next turn.
I haven't tested being on the draw yet, but I imagine that, like andy, my percentages would improve with the added card.
For reference, here's my list:
4 Cruel Bargain
4 Infernal Contract
4 Infernal Tutor
4 Goblin Charbelcher
1 Tendrils of Agony
1 Ill-Gotten Gains
1 Empty the Warrens
1 Odious Trow
1 Wild Cantor
1 Mox Diamond
4 Elvish Spirit Guide
4 Summoner's Pact
4 Culling the Weak
4 Dark Ritual
4 Cabal Ritual
4 Lion's Eye Diamond
4 Lotus Petal
4 Chrome Mox
4 Land Grant
1 Bayou
1 Dryad Arbor
30% T1 is incredibly low, andy. I looked over the games briefly and you ought to try mulliganing more. Also, 50% fizzle rate? I really think you should try to first play the deck normally, ie. learn pass the turn plays, set up the turn 2, gain inevitability turn 1 with Belcher or EtW, etc. and then move on to pushing the turn 1 because it doesn't look like you are very good at it ATM and it probably just means you need to be more familiar with the deck.
I'm really more interested in working on the sideboard. We've tried to optimize the turn 1 rate but the deck didn't convert well into the post-board plan. The deck is so fast that optimizing the turn 1 rate really only matters if you have to deal with a combo infested metagame. Otherwise, the deck can comfortably win with 8 cards on the draw basically every game. Initially I was a bit surprised at 50% on the draw; I get at least that on the play. But then I remembered that you just started playing the deck. For a new comer your percentages are acceptable. I'm going to do 50 to 100 hands today with Skyshroud Cutter and EtW in my list and see how that goes. I'll post percentages and hands.
Also, Skyshroud Cutter is really good. Being able to play both Land Grant and Pact in the same turn and get both a 0cc creature and a manasource has been extremely versatile. I've been running it over Odious Trow at the moment. I haven't had to fetch him for Chrome Mox in a long time while Cutter I find almost every other game. It smooths goldfishing out with +1 extra mana. Granted this is in a Burnt Offering list that I'm trying to configure, not sure if 1 or 2 copies is best, but even when Burnt Offering is irrelevant, this guy is really good.
Luck is a residue of design.
I'm an aspiring Psychedelic Trance musician. Please feel free to enjoy my sense of life:
http://soundcloud.com/vacrix
Expect me or die. I play SI.
Running EtW over Slithermuse would have given you a "win" (10+ goblins) here, assuming you are happy with a play like that.10. Arbor, LG, ESG, Petal, CRit, LED, MM: Mull.
LG, Mox, Petal, Trow, LED, Tutor: Keep.
T1: LG (1) -> Bayou (1, B). Mox (Trow) (2, BB). Petal, sac (3, BBU). LED (4, BBU). Tutor, sac LED (5, BBBU) -> evoke Muse (6).
D7: Mox, ESG, MM, Contract, Bargain, Tutor, Belcher.
Mox (Contract) (7, B). ESG (7, BG). MM (8, BB) -> CtW.
Fizzle.
I would have mulled that (if going for a T1 win), personally. Slithermuse in hand is essentially a mull to 6 already, and going into a D4 with zero floating is rarely going to get you there.12. Bayou, Mox, CRit, Trow, Contract, Tutor, Muse: Keep.
T1: Bayou (B). Mox (Trow) (1, BB). CRit (2, BBB). Contract (3).
D4: DRit, LED, Bargain, Belcher.
Fizzle.
Again, EtW over Slithermuse would have given you a "win". Interesting to see how often Slithermuse just doesn't get there (I'm thinking of my own testing, too).20. Mox, DRit, CtW, LED, 2xBargain, Tutor: Keep.
T1: Mox (CtW) (1, B). LED (2, B). DRit (3, BBB). Tutor, sac LED (4, BUUU) -> Muse (5).
D7: Arbor, LG, Mox, 2xESG, CRit, Contract.
Fizzle.
(Note: I checked to see whether Bargain instead of Tutor -> Muse would have been better. That would have fizzled as well: Arbor, Cantor, ESG, LG.)
As you already had plenty of black floating, I would have cracked LED for green in your list (red in mine), as this opened up Eternal Witness plays and MM. Would the extra draw have turned this into a win? Who knows, but it's a line of play you shut off by floating only black.21. Pact, DRit, 3xCRit, LED, Tutor: Mull.
Mox, DRit, Pact, ESG, CtW, Tendrils: Mull.
Petal, 2xDRit, LED, Bargain: Keep.
T1: Petal, sac (1, B). 2xDRit (3, BBBBB). LED (4, BBBBB). Bargain, sac LED (5, BBBBB).
D4: Mox, DRit, CRit, MM.
Fizzle.
For balance I note that EtW was not available here, due to Pact (not that Slithermuse got you there anyway). Tutor for PiF would have been auto win, but IGG would have been one mana short.22. Mox, CtW, 2xDRit, Pact, MM, Tutor: Keep.
T1: Pact (1) -> ESG (1, G). Mox (CtW) (2, BG). MM (3, BB) -> Petal, sac (4, BBU). 2xDRit (6, BBBBBBU). Tutor (7, BBBBU) -> Muse (8, B).
D7: 2xMox, 2xPetal, LG, ESG, Tutor.
Mox (ESG) (9, BG). LG (10, BG) -> Bayou (10, BBG). Mox (nothing) (11, BBG). 2xPetal, sac (13, BBBBG). Tutor (14, BBB) -> Bargain (15).
D4: Trow, Contract, Bargain, Tendrils.
Fizzle.
While I can't say it would have been a win, I think you list really benefits from floating heavy amounts of green, both for Witness and Manamorphose. Cracking the second LED for GGG would have opened up several different lines of play.24. LG, Pact, ESG, Petal, 2xLED, Bargain: Keep.
T1: LG (1) -> Bayou (1, B). Petal, sac (2, BB). ESG (2, BBG). Pact (3, BBG) -> Cantor, sac (4, BBB). 2xLED (6, BBB). Bargain, sac 1 LED (7, BBB).
(Question: how to best utilize the LEDs here?)
D4: DRit, MM, Witness, Contract.
DRit (8, BBBBB). Contract, sac LED (9, BBBBB).
D4: CRit, LED, Contract, Bargain. (11 in graveyard.)
CRit (10, BBBBBBBB). Contract (11, BBBBB).
D4: DRit, LG, CtW, Bargain.
DRit (12, BBBBBBB). Bargain (13, BBBB, 1 life).
D4: LG, DRit, 2xPact.
Fizzle. (Witness is in the graveyard already.)
If hunting for a first turn win, I would have mulled this. Witness and Tendrils are dead, so you effectively have a mull to five here. I hate going into a D4 with no mana floating, so I'd chance it by going to six.25. Mox, ESG, CRit, Witness, Contract, Bargain, Tendrils: Keep.
T1: Mox (Contract) (1, B). ESG (1, BG). CRit (2, BBB). Bargain (3).
D4: LG, Pact, Bargain, Tutor.
Fizzle.
This is why I run 4 Belcher/1 Tendrils. If either of those Tendrils in the first hand were Belcher, you had a realistic shot at a T1 win PLUS and immediate, uncounterable second attempt on turn 2.28. LG, 2xDRit, 2xLED, 2xTendrils: Mull.
LG, 2xPetal, 2xLED, Tendrils: Mull.
LG, Pact, CtW, CRit, Tutor: Mull.
CRit, CtW, Tutor, Muse: Mull.
Pact, LED, Tutor: Fail.
Last edited by Silent Requiem; 07-07-2012 at 02:35 PM.
I agree. At this point, I don't think I'm frequently just missing higher percentage lines of play. The low T1 rate is likely down to the particular deck list I'm using or the mulligan rules I'm using (I'm leaning toward the latter).
This suggestion has been made several times. In a match setting, I'm perfectly comfortable passing the turn when necessary, playing grinding games, etc. But of course, passing the turn will never, ever improve your T1 percentage.
I need the Arbor to sac to Culling the Weak. Otherwise, I have no way to start the spell chain.
I'm still in the process of tweaking my set of mulligan rules, but within each set of 30 hands, I use a consistent set of mulligan rules. For each of these sets of 30, one of the rules was to keep if I could start a D4 spell chain. I'm going to change that for the next sets of 30, since my data so far shows a D4 floating nothing to be insufficient nearly always.
Note that the Bargain before cracking LED uses up BBB, so if I don't get BBB (instead of GGG) with LED, I can't play a drawn D4. With 8 D4s in the deck compared to 2 Manamorphose + 1 Witness, I thought getting black was the higher percentage play.
The same comment applies to Game 24 (at the point at which I cracked the second LED). However, in this situation (open with a D4 with 2xLED available), I think I should have cracked both LEDs up front for BBBGGG. This still lets me continue with another D4, and it also gives me the opportunity to filter green mana into black mana if I need to.
I was running 4 Belcher and 1 Tendrils for my first set of 30 hands. One of the suggestions to improve the T1 goldfish rate was to move to a lower Belcher, higher Tendrils configuration, so I tried it out. I can't say how much it improved the T1 rate (the 1 higher T1 kill is not statistically significant). For a tournament, I prefer the higher Belcher count, since they're more useful against most hate and blue decks. I'm currently trying 3 Belcher, 2 Tendrils main deck in tournament play.
For my next sets of 30 (1 set on the play, 1 set on the draw), I'm considering the following changes:
- IGG over Slithermuse
- Skyshroud Cutter over Cantor (the Cantor is less useful than Trow if I'm not playing Slithermuse)
- If a 7-card hand can only play a D4 floating nothing, mulligan.
- If a 6-card hand can only play a D4 floating nothing, mulligan. (At 5 cards and below, I'm not sure.)
I'm not going to test Empty the Warrens yet, but for data purposes, I'd recommend the following:
- On the play, EtW for 10+ without a Pact or Cutter in the spell chain should be considered a win. On the draw, 12+.
- On the play, if 1 Cutter has been played, then EtW for 14+ should be considered a win. If 2 Cutters have been played, then EtW for 16+ should be considered a win. On the draw, these numbers should be 16+ and 18+, respectively.
- Any spell chain including a Pact that ends with EtW should be considered a loss. (Note: even if you have, say, land + LED in play and think you'll be able to pay for the Pact.)
I should mention, especially in regard to that last point, that I think part of what's contributing to my T1 goldfish rate being so much lower than it should is that I'm fairly stingy about what counts as a T1 win. Without playing EtW in my lists, it's pretty straightforward: I count a game as a T1 win if it literally wins on T1. I suspect that many other scenarios are being counted by other people.
For example, consider playing Belcher with 3 permanent mana sources, so you can activate T2. I do not count this as a T1 win in my goldfishing, and if a hand can do only this, it gets mulliganed. In real life, I would keep such a hand in a second, and most of the time it would translate to a win. On the other hand, I have twice lost games in local tournaments where I played a Belcher first turn after a Pact with mana available to activate in response to the Pact trigger. (In both cases, the opponent Wastelanded my land, leaving me unable to activate Belcher or pay the Pact trigger.)
I feel similarly about EtW. I haven't included it in my goldfish testing yet because, simply put, it cannot ever actually win on T1. But at the very least, to make sure we're comparing apples to apples, we should be using the same rules to determine what counts as a T1 win.
Yeah, "T1 win" can mean a lot of things. That's why I wanted to be sure to clarify what I was counting in my win percentage.I feel similarly about EtW. I haven't included it in my goldfish testing yet because, simply put, it cannot ever actually win on T1. But at the very least, to make sure we're comparing apples to apples, we should be using the same rules to determine what counts as a T1 win.
Re hand 6: I'm an idiot and I didn't read the hand properly. I've edited that out of my post.
Re hand 21: You have plenty of ways to make black mana, though. It's making green mana that's difficult. That's why I suggested floating green in your list, as it turns on Manamorphose, Eternal Witness, Wild Cantor and hardcast Land Grant.
I've also done 30 hands on the draw, now, and I won 21 of those hands for a 70% win rate. This included 5 EtW wins (12+ goblins; I thought it only right to require a higher storm count on the draw) and two "pass the turn" Belcher wins. Make of that what you will. A build that pushes Slithermuse and Tendrils ought to have a higher "true" T1 win rate, but I'm not sure it's actually a better build. Keep in mind that wherever I cast EtW, I could have cast Slithermuse, if I'd run it.
Of my nine losses, four were fizzles, and five were mulling to oblivion. I generally looked for the same hands I'd keep on the play, but did sometimes keep strong hands that only needed one more mana (NOT an initial mana source). By and large, the deck gave those to me.
Three of my four fizzles left me in a position to play Eternal Witness, if I'd been running it. There was only one occasion where I wanted to cast EtW but could not due to Pact. Slithermuse would obviously have been superior in that case.
There's no reason to waste a bunch of time goldfishing to just have a T1 win. Goldfish the deck like you would any other combo deck, with the simple rule that this one should always have a T1 or T2 win.
For that reason it is also useless to run some weird x3 Tendrils build that skimps on Belchers. I don't like running 4 Belchers in my maindeck as they clog up hands sometimes and don't imprint on Moxes, and most importantly can not make the most out of the D4 chains this deck has (which is really the only reason to run a storm kill). Still, atleast 3 is needed I've come to find. It is just too good of a card, so 3-4 is mandatory.
Any data compiled with such variables is bound to come out innacurate.
I've been a long time lurker on the Source, putting a few points here and there every so often, but mainly being a sponge and absorbing what I can.
I love combo decks, my first pet in legacy was belcher until I read stuff about PSI, and I knew that I would have to trade for enough stuff to get it together.
My first legacy deck was combo elves (which is most everything you will find in my posts for the most part), because I couldn't afford anything else, while I was assembling all the stuff for storm combo. It took me quite a while, but I have most of the stuff for much of what people would call legacy storm decks, minus blue dual lands.
Because I'm an army officer and my local legacy tournaments usually get out on Sunday at midnight to 1 am, I haven't had the opportunity to play Legacy nearly as much as I would like.
I goldfish when I can, and read everything I can about the archtype I love, which is combo.
I've made the decision that since I couldn't play combo officially until the next tournaments come around seattle, I would keep my mind and muscle memory sharp by doing the one thing that you can beyond goldfishing: teaching.
Most of you I've learned from here, so much of this will seem old hat to you, but not everyone learns best by seeing:
Some people need to see these things happen, or hear someone narrate this to them.15. (T2) 2xESG, 2xCRit, CtW, LED, Belcher: Mull.
LG, 2xPetal, CtW, DRit, Tutor: Keep.
T1: LG (1) -> Arbor. Petal (1, B). CtW (Arbor) (2, BBBB). DRit (3, BBBBBB). Petal, sac (4, BBBBBBU). Tutor (4, BBBBU) -> Slithermuse. Evoke Slithermuse (5, B).
Draw: LG, Mox, Cantor, LD, Bargain, 2xBelcher. (Can't cast Bargain or Belcher.)
Pass.
T2: Draw DRit. LG -> Bayou (B). DRit (BBB). Mox (Cantor) (BBBG). Belcher. LED, activate Belcher with 0 lands left.
So I've created a youtube channel especially for Legacy Combo, and the first deck I'm doing on it is my favorite combo deck in legacy: PSI.
I hope you will keep me honest and show me lines of play I haven't seen, I've only been goldfishing this thing for a few months after dong quite a bit of goldfishing Bryant Cook's TES.
Links will be below.
Kirby
Kirby's Komboes:
Intro to the Series
http://www.youtube.com/watch?v=gDSOVPKgvgg&feature=plcp
PSI Deck Intro
http://www.youtube.com/watch?v=OmYgLFEwWV0&feature=plcp
Hand 1:
http://www.youtube.com/watch?v=54smKsolFRk&feature=plcp
Hand 2:
http://www.youtube.com/watch?v=E1EJgQ-IzBo&feature=plcp
Hand 3:
http://www.youtube.com/watch?v=RKi36Dzs3VM&feature=plcp
Hand 4:
http://www.youtube.com/watch?v=9b1MpWA8B-g&feature=plcp
Hand 5:
http://www.youtube.com/watch?v=5f58LLxGD2E&feature=plcp
Last edited by Ellistann; 07-11-2012 at 01:13 AM. Reason: Fixing Link
The very fact flamethrowers exist means that someone, somewhere said to themselves, "You know, I want to set those people over there on fire.... but I'm just not close enough to do it."
Nice. Glad to have you on board.
I didn't get a chance to watch all the video's yet but in your introduction you mention cutting Culling the Weak for Tinderwall. If you are to play D4's then you cannot cut it. Hitting BBB consistently is impossible without it. Also, its the reason you play Pact. Without Culling, you can also cut Pact, and then you are basically just merging toward a Classic Belcher list.
EDIT:
BTW it's easier to do calculations if you think of your mana sources in net gain rather than calculating something like -2 for Crit, +5. Granted, its more relevant if you have to float mana of other colors like red for PiF or EtW, or blue for Muse, or green for Ewit (if you play it). I think of Drit as +2, Crit as +1, ThreshCrit as +3, Culling as +3. It makes the math faster. This is actually really important when it comes to analyzing a hand quickly. If you can't keep within the first 5 seconds of keeping your hand, your opponent might think you are calculating, putting you on combo.
Last edited by Vacrix; 07-10-2012 at 09:34 PM.
Luck is a residue of design.
I'm an aspiring Psychedelic Trance musician. Please feel free to enjoy my sense of life:
http://soundcloud.com/vacrix
Expect me or die. I play SI.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Reply With Quote