Not quite hypothetical question.

[Wordslinger]

mods, feel free to move this if you need to.

Situation is this:

You are thinking of entering a five round GPT. Your deck has a ~10 win percentage against any variant of combo, but a 85 percent win percentage against every deck in the rest of the field.

Would you feel comfortable bringing this deck to that GPT? Say these numbers can't be adjusted by card choice. Would you invest in a separate deck?

would like to hear the community's opinions on this one.

[wonderPreaux]

mods, feel free to move this if you need to.

Situation is this:

You are thinking of entering a five round GPT. Your deck has a ~10 win percentage against any variant of combo, but a 85 percent win percentage against every deck in the rest of the field.

Would you feel comfortable bringing this deck to that GPT? Say these numbers can't be adjusted by card choice. Would you invest in a separate deck?

would like to hear the community's opinions on this one.

Your EV of winning a match would be (X*.85) + (Y*.10) where X is the percentage of noncombo and Y is the percentage of combo. The question "do i invest in a separate deck" would thus be a function of:
1: how much combo you expect
2: what the EV of the other deck is in comparison

this is assuming you are equally able to proficiently use both decks. of course, 5 rounds means smaller sample size, and ofc, more variance.

[thecrav]

Is this GPT local? If so, you certainly know the local metagame better than we do. If only a few people are likely to bring combo, this could be a solid gambit. Choose to roll over an unlikely match in order to shore up all the likely matches.

You may find some similar posts on Vintage forums about Dredge. In 100% proxy tournaments where price doesn't matter, people are unlikely to bring the perceived budget option, Dredge. On MTGOv4 until recently, people were unlikely to play dredge due to the Graveyard UI. I frequently chose to not pack Graveyard-specific hate online knowing that I was quite unlikely to face it.

[Aggro_zombies]

If you're trying to T8/T4 a five round GPT, you realistically probably only need to win 3 rounds, then ID one round, then try to win or draw the fifth round. If you're 85% against "not combo" you can budget the fifth round as your auto-loss to combo and then just hope your 3-1 opponent isn't also a combo deck or is willing to ID with you in round 5.

Still kind of a dicey plan though since one bad pairing means you're probably out of contention. If you could scout the event beforehand and figure out the relative proportions of combo and not-combo, that would allow you to make a better decision about running this deck.

[Wordslinger]

yeah this actually happened during the treasure cruise era. Playing Enchantress, i crushed Landstill in round one, UR delver twice, then promptly lost to elves and bad mulls. Properly built, the deck is a wrecking ball to fair decks.
it just loses to combo way to hard. the legacy scene here is really developed, people generally have cards for whatever so the meta is really hard to predict, any one person could be on 3 or 4 different decks. i just wanted to know
if it was worth it to still play a deck that gives up percentages to that much of the meta if it's win percentages were that good against the rest of it to compensate.

[Megadeus]

Doesn't enchantress have like a million ways to beat combo? Between White Leyline, Ghostly prisons, elephant grass, and all sorts of other shit, I feel like you should be able to configure the deck to wreck any meta as long as you build it properly.

[btm10]

Doesn't enchantress have like a million ways to beat combo? Between White Leyline, Ghostly prisons, elephant grass, and all sorts of other shit, I feel like you should be able to configure the deck to wreck any meta as long as you build it properly.

The Omnitell matchup is nearly unwinnable, and if you get paired against a Storm deck you basically have to hope they're forced to go for Empty in two games. You can improve the combo matchups a bit by boarding a ton of discard (I've gone as far as 4 Duress, 4 Brain Maggot), but that's more than half of your board to go from 30/70 to 40/60 agsinst a small portion of the field.That being said, Painter is nearly unloseable, and there are actually pretty strong board plans for Elves. It's not really a wash, though. I've tried employing the above logic and it works well if get lucky. Sometimes you go up against Blade decks and BUG and smash them. Sometimes you get paired against combo five rounds in a row.

[Tylert]

The Omnitell matchup is nearly unwinnable, and if you get paired against a Storm deck you basically have to hope they're forced to go for Empty in two games. You can improve the combo matchups a bit by boarding a ton of discard (I've gone as far as 4 Duress, 4 Brain Maggot), but that's more than half of your board to go from 30/70 to 40/60 agsinst a small portion of the field.That being said, Painter is nearly unloseable, and there are actually pretty strong board plans for Elves. It's not really a wash, though. I've tried employing the above logic and it works well if get lucky. Sometimes you go up against Blade decks and BUG and smash them. Sometimes you get paired against combo five rounds in a row.

Rule of law? Isn't it great vs omnitell? it comes online when your opponent casts show and tell and your opponent can only cast one spell a turn...
Paired with elephant grass / Oblivion ring for emrakul?

[Megadeus]

Rule of law? Isn't it great vs omnitell? it comes online when your opponent casts show and tell and your opponent can only cast one spell a turn...
Paired with elephant grass / Oblivion ring for emrakul?

This. Dredge seems almost unloseable as Elephant Grass stops them cold and you get RIP. Plus if your deck is so strong against any non combo deck for the most part in game 1, I don't see why having 8-10 cards to bring in for combo MUs is so bad. A rule of law backed by a sterling grove or something seems very difficult for omni/storm to beat. I guess Sneak and Show doesn't care as much, but you still have access to tools to beat them I would imagine.

[btm10]

This. Dredge seems almost unloseable as Elephant Grass stops them cold and you get RIP. Plus if your deck is so strong against any non combo deck for the most part in game 1, I don't see why having 8-10 cards to bring in for combo MUs is so bad. A rule of law backed by a sterling grove or something seems very difficult for omni/storm to beat. I guess Sneak and Show doesn't care as much, but you still have access to tools to beat them I would imagine.

Dredge is basically a bye for Enchantress, though I generally don't consider it combo as much as its own crazy thing. The issue with the Enchantment-based disruption package is that it's very vulnerable to bounce until Grove is online, and that takes time. I've tried boarding in 8 discard and 5 permanent-based answers to combo and the fact that the permanent based disruption doesn't hard lock them (and the non-Leylines don't even come down until turn 2 at the earliest) is a problem, as is not playing Brainstorm to find the hate. The best way to say all of this might be that Enchantress has excellent fair matchups, but you can only pick 1-2 combo matchups to fix postboard, and are punished severely dor choosing incorrectly.

I also think that 85/15 overstates how good some of your matchups are - it's not like you don't need your other sideboard slots at all. If Wordslinger is going up against a room full of Blade decks and one guy on ANT then I'd play Enchantress in a heartbeat because you have one guy who you have to prepare for and you know exactly what he's got, and your other matchups really are 85/15 good. But stuff like RUG and BUG Delvers are closer to even and suddenly you want sideboard cards for them.


Rule of law? Isn't it great vs omnitell? it comes online when your opponent casts show and tell and your opponent can only cast one spell a turn...
Paired with elephant grass / Oblivion ring for emrakul?


ORing is bad against Omni of they know how to set up, and Rule of Law is only good before they cast Show and Tell since it slows you both down and you aren't really online until turn 3-4 anyway. If you can get multiple pieces of disruption down it's tractable, but they need to be specific and they need to be protected by Grove. Sneak is considerably easier to beat because of Grass and the fact that they usually don't have access to Cunning Wish in response to ORing/Banishing Light's etb trigger.

[Wordslinger]

no i definitely meant 85/15 postboard vs fair decks. Also a problem is that as soon as you play forest-wild growth they know the coast is clear. the main issue is this: Against combo with countermagic, your earliest piece of interaction
comes in on turn two and gets countered. Turn three dead. Against combo without it, they generally have discard. land, growth or guile, pass. Thoughtseize you? better hope you have more ways to interact in your top three.

What are we hoping for against storm combo, really? leyline, land, land, grove, three other cards? to just have some sort of chance in the matchup? keep in mind we still have to kill them and we're giving them basically infinite time to set
up . We have no pressure and our clock is downright diseased. What's to stop them from just boarding into decays and chains, casting ad naus from twenty(cause lets be honest here, there's no way they aren't at a high life total) and just... eot ad naus, kill your stuff, untap, kill you.

I'm also playing a version with four rest in peace main. No matter how you cut it this is as bad as it gets. what does it matter if you bring in eight cards if something like 40% of your deck doesn't interact with them?

[Varal]

If your deck is really 85/10, every players play optimally and only luck determines the winner, you want the deck that gives you the best chance to win a given match and that you other deck choice is 50/50. If 8 out of 15 players (excuding you) play fair deck then they're both equivalent, if more 85/10 is better, if less 50/50 is better. Assuming you know the winning percentage for every decks, you can get the probability to win a given match as the sum of (proportion of metagame * probability to win matchup).

In real life, people don't play optimally and a theoretical 40/60 matchup can be 80/20 in your favor.

[LeoCop 90]

I will play anytime a deck that crushes all the fair piles while losing to combo. That's why force of will is in the format... to get paired against combo and then lose against you XD !

[owen_meany]

Your EV of winning a match would be (X*.85) + (Y*.10) where X is the percentage of noncombo and Y is the percentage of combo. The question "do i invest in a separate deck" would thus be a function of:
1: how much combo you expect
2: what the EV of the other deck is in comparison

this is assuming you are equally able to proficiently use both decks. of course, 5 rounds means smaller sample size, and ofc, more variance.

Assuming that your hypothesis of 85% vs non-combo and 10 % vs combo is correct. Then you total win percentage will be function of the number of combo decks. If we use http://www.mtgtop8.com/ as a guideline we can see that the number of combo decks roughly equals 25 %.
This will give your deck a win percentage of: 0.25*10 % + 0.75 *85 %=66%
This would a good percentage and would put you in to the top 100 players of the SCG circuit. The question is what your acceptable win percentage based on deck.
Win percentage Combo Non-combo
77.5 % 10 % 90 %
70.0 % 20 % 80 %
62.5 % 30 % 70 %
55.0 % 40 % 60 %
47.5 % 50 % 50 %

If we use the 25 % combo as an example and use binomial distribution assuming a 5-round tournament we get
Result percentage Result Comment
12.7 % 5-0-0
19.3 % 4-0-1 ID
39.3 % 3-1-1 ID (this is not the cumulative probability)
58.5 % >=3-1-1 Equal or better than 3-1-1

Assuming that 3-1-1 is good enough to top 8 you will have about 58.5 % chance of top 8

[Wordslinger]

58.5 hmm. The problem is that the amount of combo in a given room is not entirely predictable and is itself a variable subject to some amount of variance.
59 percent isn't so much higher than something like miracles, which doesn't straight up lose to combo. hmm. slowly starting to believe my build may not be viable, even as a tier 2 option.

[owen_meany]

58.5 hmm. The problem is that the amount of combo in a given room is not entirely predictable and is itself a variable subject to some amount of variance.
59 percent isn't so much higher than something like miracles, which doesn't straight up lose to combo. hmm. slowly starting to believe my build may not be viable, even as a tier 2 option.

It is worth noting that providing that you are playing a reasonable deck the pilot effect seam more significant than the deck choice. Taking statistics from the SCG circuit in to account:
Archetype stats: http://www.reddit.com/r/MTGLegacy/comments/2wbohd/scg_legacy_results_for_major_archetypes_dc_houston/
Player stats: http://www.reddit.com/r/MTGLegacy/comments/2q7oxf/match_win_percentages_for_scg_legacy/

Looking at players like Wilson and Mann that are able to get around a 80ish win percentage for a decks that averages in the mid 50's. Or if we consider a deck like miracles where the average win percentage is sub 50 % and pilots like Duke and Lossett are averaging in the 70's. It should be noted that the archetype breakdown is for one tournament and not necessarily representative, for legacy as a whole, but it seams reasonable that the "deck-effect" is about 10% but the pilot effect could be as high as 20%. So if you feel that you can maneuver your deck to a win percentage above 65 it would require a significant amount of practice to increase your win percentage with a new "better" deck.
Just my thoughts on the matter.