I've not played much French EDH. I'm wondering, in a tournament setting, how fast does my non-control deck need to be? Specifically, how fast does it need to be able to kill. Obviously, I don't need to blindly go for the kill from the get go each game, but I'm just trying to get an idea.
I can build the deck to be more all in:
Can kill turn 3.
Kills by turn 4 - 25% of the time.
Kills by turn 5 - 65% of the time.
Kills by turn 6 - 8% of the time.
Kills turn 7+ - very rarely.
Or less all in:
Can kill turn 4.
Kills by turn 5 - 48% of the time.
Kills by turn 6 - 48% of the time.
Kills turn 7+ - very rarely.
NB. These are stylized numbers based on somewhere from 50-100 goldfishes per deck.
NB. These numbers assume that I will go for the kill rather than disrupt my opponent with the cards in my hand.
I got sick of transforming into a midrange deck for G2/G3, no one is playing graveyard hate locally.
I just choose the best 6-8 pieces of removal for G2/G3 against aggro.
Against control I bring in Sin Collector and Abrupt Decay if I anticipate graveyard hate.
Against mid-range I bring in Abrupt Decay if I anticipate graveyard hate.
Against Slaughter Games I bring in Slaughter Games and Sin Collector, as I'm completely dead to Slaughter Games naming AoGR.
I know this sounds crazy, but what happens if we drop the green entirely? Yes, the ramp goes out the window. It also limits the deck to 4 proper Enchantresses.
However, it allows for a dedicated second color. Going RW allows for 3 things main that could help. Blood Moon, Assemble the Legion, and Seal of Fire are the big ones that I'm seeing just off the top of my head. It creates a win condition that doesn't create more card draws, in addition to shutting off the manabases for a LOT of decks, and providing removal.
You can run a 3rd color easily here with fetches, just tone down the green, don't need to cut it.
Imagine your favorite pro football teaming having to play a middle school football team. Do you think you pro team is going to find enjoyable to wreck the snot out of those kids?
So this was a deck i brewed up using birthing pod and gifts to find combos like Melira + persist, Kiki-Jiki + Pestermite, and body double combo.
I've been tooling around with it and I'd love to see what other people have to say about it.
Help is appreciated, Thanks in advance!:)
Maybe. The difference just isn't statistically significant because the sample size is only 10, at least for Kiki Pod.
I think you'd be better of using confidence intervals rather than p-values. This is better at communicating the information to non-stats people. The other thing I'd recommend is to bootstrap your residuals, as this will go some way to counteracting the small sample size.
Maybe it underperformed at this competition, or it has a higher learning curve than other decks in modern. But if you go to the Modern Bannings thread, several users have been collecting data on MTGO; it seems that whilst it's a relatively small part of the meta there (~12% of decks), Pod's win rate is consistently higher (~20% of 4-0 lists). Note that this includes all dailies (since Wizards only posts 1/3 of all dailies nowadays, or something to that effect).
Not to mention the deck made it to the finals of a second consecutive GP on its own strength, and did happen to win the match. Maybe Pod pilots are better than other pilots, but I think that's pretty unlikely. It's win rate online suggests that it is the best deck, or close to it.
Fair enough. I can get behind that data supporting your claims.
A deck can be the best deck without being totally dominant. You're right that Pod's results indicate that it has a relatively even match up with much of the field, but that alone is probably enough to constitute being the 'best deck'. The fact that this is also the second GP to be taken out by Pod indicates that it's probably not a fluke (irrespective of the misplay in game 3 of the finals).
I'm sorry, I gave exaggerated numbers that were misleading - implying that I meant total dominance; they were there to account for the relatively small sample sizes.
However, a deck that is 50% (i.e., AT CHANCE) against the field can not be the best deck. EDIT: it could be tied for equal best, but that would make the argument you are advocating less compelling anyway.
Also, Pod managed to maintain roughly 25% of the meta all the way through to the finals (4 in the top 16, 2 in the top 8). It's not unbeatable, but it's certainly looking like it's the best deck right now.
Your logic is flawed.
This simply suggests the deck has reasonable match ups against the field.
If a deck was the best deck, starting at 20% of the field, you would expect it to be at least 35% (+) of the top 16.
If a deck was awful, starting at 20% of the field, you would expect it to be at most 5% (-) of the top 16.
I realize this. But, are there no rules of thumb for when I need to either have established control of the game or be about to win by?
I can build the deck to be more all in:
Can kill turn 3.
Kills by turn 4 - 25% of the time.
Kills by turn 5 - 65% of the time.
Kills by turn 6 - 8% of the time.
Kills turn 7+ - very rarely.
Or less all in:
Can kill turn 4.
Kills by turn 5 - 48% of the time.
Kills by turn 6 - 48% of the time.
Kills turn 7+ - very rarely.
NB. These are stylized numbers based on somewhere from 50-100 goldfishes per deck.
NB. These numbers assume that I will go for the kill rather than disrupt my opponent with the cards in my hand.
FYI:
My deck:
It is so consistent.
Spam/Trolling infraction issued
-:pops:
My sideboard is just very geared toward aggro now.
3 Pillar of Flame
2 Drown in Filth
2 Rolling Temblor
2 Slaughter Games
3 Sin Collector
I got sick of transforming into a midrange deck for G2/G3, no one is playing graveyard hate locally.
I just choose the best 6-8 pieces of removal for G2/G3 against aggro.
Against control I bring in Sin Collector and Abrupt Decay if I anticipate graveyard hate.
Against mid-range I bring in Abrupt Decay if I anticipate graveyard hate.
Against Slaughter Games I bring in Slaughter Games and Sin Collector, as I'm completely dead to Slaughter Games naming AoGR.
You can run a 3rd color easily here with fetches, just tone down the green, don't need to cut it.
What I'm thinking is:
4 Heartless Summoning
3 Enduring Renewal
Other Spells (Seventeen)
3 Idyllic Tutor
1 Tezzeret, Agent of Bolas
3 Muddle the Mixture
3 Path to Exile
2 Dismember
3 Mana Leak
1 Spell Pierce
1 Mana Tithe
1 Wurmcoil Engine
2 Sphinx Summoner
4 Faerie Mechanist
4 Perilous Myr
1 Myr Servitor
1 Bottle Gnomes
Yes.
Why don't you try over here:
http://forums.mtgsalvation.com/forumdisplay.php?f=638
I think you'd be better of using confidence intervals rather than p-values. This is better at communicating the information to non-stats people. The other thing I'd recommend is to bootstrap your residuals, as this will go some way to counteracting the small sample size.
Fair enough. I can get behind that data supporting your claims.
I'm sorry, I gave exaggerated numbers that were misleading - implying that I meant total dominance; they were there to account for the relatively small sample sizes.
However, a deck that is 50% (i.e., AT CHANCE) against the field can not be the best deck. EDIT: it could be tied for equal best, but that would make the argument you are advocating less compelling anyway.
Your logic is flawed.
This simply suggests the deck has reasonable match ups against the field.
If a deck was the best deck, starting at 20% of the field, you would expect it to be at least 35% (+) of the top 16.
If a deck was awful, starting at 20% of the field, you would expect it to be at most 5% (-) of the top 16.