What Decks are 50-50 against the 'Field'?

[carefulmug]

I'm a long time Canadian Threshold player and have always appreciated the deck's capacity to be (roughly) 50-50 against anything across the table.

I recently got into RG Lands but have become disinterested considering its abysmal OmniShow matchup.

I'm wondering what decks other than Can Thresh are out there that might not have spectacular matchups against everything, but no terrible ones against anything, either.

[apple713]

I'm a long time Canadian Threshold player and have always appreciated the deck's capacity to be (roughly) 50-50 against anything across the table.

I recently got into RG Lands but have become disinterested considering its abysmal OmniShow matchup.

I'm wondering what decks other than Can Thresh are out there that might not have spectacular matchups against everything, but no terrible ones against anything, either.

miracles. People who play miracles win because they are better players than their opponents (generally speaking). The deck has pretty even matchups but shines in the hands of a good player.

[Lormador]

Miracles has some pretty tough matchups, but it might be close to the effect you're talking about (50/50 against the field). I still don't exactly look forward to playing against Infect, 12Post, MUD, Lands, or DnT when I play Miracles.

Then again, RUG has some tricky matchups as well... I don't think there's any getting away from it. Deck advantage and disadvantage remains a central pillar of the game.

The various Blade decks might possibly be close to 50/50 against the field, particularly Esper.

[phazonmutant]

Miracles has some pretty tough matchups, but it might be close to the effect you're talking about (50/50 against the field). I still don't exactly look forward to playing against Infect, 12Post, MUD, Lands, or DnT when I play Miracles.

Then again, RUG has some tricky matchups as well... I don't think there's any getting away from it. Deck advantage and disadvantage remains a central pillar of the game.

The various Blade decks might possibly be close to 50/50 against the field, particularly Esper.

More like 45% against the field.

I think Storm is powerful enough that it doesn't have any truly atrocious matchups. Some are tricky, but not even Miracles is hugely lopsided.

[Megadeus]

I think the only match up where I feel at a real disadvantage with storm is reanimator.

[Bahra]

I think D&T is close to 50/50 if not favoured against most things. Except for Elves and turn 1 combo decks it's pretty much even or favoured against everything.

And I am of the opinion that D&T with a red splash makes it more even against the field than before. Making the elves match up and a few others much closer. But if you're new to D&T definitely start with just the mono white version.

[Lemnear]

Isn't the answer as easy as looking at the DtB section and realize that these decks are there mainly because they perform excellent and/or are widely played to gather all the points required to remain in that section?

[apple713]

More like 45% against the field.

I think Storm is powerful enough that it doesn't have any truly atrocious matchups. Some are tricky, but not even Miracles is hugely lopsided.

i wouldn't recommend a combo deck for what he's asking. most combo decks lose to themselves. Furthermore, unless something has drastically changed, sneak and show is the best combo deck in the format. D&T is maybe its worst matchup... and blood moon/pyroclasm rock it.

Recently omni has been performing better but it's less consistent than sneak. Sneak is underrepresented and doesn't show up as high on tier lists because its among the most expensive decks in the format comparatively.

[Lemnear]

i wouldn't recommend a combo deck for what he's asking. most combo decks lose to themselves. Furthermore, unless something has drastically changed, sneak and show is the best combo deck in the format. D&T is maybe its worst matchup... and blood moon/pyroclasm rock it.

Recently omni has been performing better but it's less consistent than sneak. Sneak is underrepresented and doesn't show up as high on tier lists because its among the most expensive decks in the format comparatively.

Oh, did he ask for an Autopilot.dec which a monkey can Top8 a GP with? Not recommending a deck because it's not pickup & win level overpowered is a joke as calling SneakShow the best deck in the format with all it's clunky draws and an additional color compared to OmniTell and the weakness to Karakas and Containment Priest

Edit:
I mean drawing two mana, Sneak Attack, Emrakul, Griselbrand, FoW and Preordain is common Junk SneakShow wields you as a possible starting 7. Dumb, linear, clunky

[TheArchitect]

Miracles is more like 75-25 against the field and then has like 3 even matchups and 3 bad ones.

Even RUG delver has some horrendous matchups. Maverick, nic fit and pox are particularly bad. Esper stoneblade is pretty 50/50 against most things. I am not sure about UWR stoneblade, but I imagine it is similair. Besides elves, D&T is pretty even vs the field.

I would not recommend a combo deck as their matchups tend to be more polarized.

[nevilshute]

Oh, did he ask for an Autopilot.dec which a monkey can Top8 a GP with? Not recommending a deck because it's not pickup & win level overpowered is a joke as calling SneakShow the best deck in the format with all it's clunky draws and an additional color compared to OmniTell and the weakness to Karakas and Containment Priest

Edit:
I mean drawing two mana, Sneak Attack, Emrakul, Griselbrand, FoW and Preordain is common Junk SneakShow wields you as a possible starting 7. Dumb, linear, clunky

There's that good-spirited, friendly and over-bearing Lemnear we have all come to love :smile:

I will agree with the sentiment of your post anyway, which is to say that I don't see how Sneak and Show is more consistent than Omni in the post-DTT world. I also don't think Sneak and Show has ever been "the best combo deck in the format". But in truth I shy away from any of these "best deck" hypes.

In truth, I think - if not Canadian - then some other Delver variant is probably still the closest you are going to get at "close to 50/50 across the board". But even then you risk running into lopsided matchups. It's the name of the game I guess.

If we're talking truly lopsidedness then I think what makes most sense is to look at each deck and then ask if they have any attrocious matchups among the other tier 1-1½ decks? Miracles doesn't, it's closest truly horrible matchup is 12-post which is tier 2. Infect might not be great but it's nowhere near as bad as 12-post. Storm is also pretty set in this way of looking at things. Storm folds to MUD but MUD is tier 2, if perhaps, a bit closer to 1½ than 12-post. Storm also has a bad time against Reanimator, but again, that deck is not tier 1-1½ (might be 1½ I guess). D&T on the other, has a torrid time against Elves which in my opinion makes a little less alround reliable, but it's still an excellent choice. Canadian and BUG-delver both seem around 45/55 at worst against the tier 1-1½ strategies.

[apple713]

Miracles is more like 75-25 against the field and then has like 3 even matchups and 3 bad ones.


if any deck was actually 75/25 vs the field it would warp the meta to where you are either playing that deck or a deck built specifically to beat it. Since that is not the case you are clearly wrong...


There's that good-spirited, friendly and over-bearing Lemnear we have all come to love :smile:

I will agree with the sentiment of your post anyway, which is to say that I don't see how Sneak and Show is more consistent than Omni in the post-DTT world. I also don't think Sneak and Show has ever been "the best combo deck in the format". But in truth I shy away from any of these "best deck" hypes.



I'll admit that I haven't played omni post DTT but I'm having a hard time believing that a card that only has a use after turn 3 fixed the extremely clunky draws and consistency issues to bring the deck to be better than sneak. 2 card combo is almost certainly better than a 3 card combo.

Considering that containment priest only made the poor matches worse it didn't have that detrimental of an effect i don't believe.

[FoolofaTook]

Has Miracles been out of the DTB section for more than an odd month here or there over the last year plus?

[apple713]

Has Miracles been out of the DTB section for more than an odd month here or there over the last year plus?

doesn't really matter, if it was truly the case where it was so well positioned, everyone would be playing it or against it. The meta would be 12 post v miracles v the rest of the meta...

[nevilshute]

if any deck was actually 75/25 vs the field it would warp the meta to where you are either playing that deck or a deck built specifically to beat it. Since that is not the case you are clearly wrong...




I'll admit that I haven't played omni post DTT but I'm having a hard time believing that a card that only has a use after turn 3 fixed the extremely clunky draws and consistency issues to bring the deck to be better than sneak. 2 card combo is almost certainly better than a 3 card combo.

Considering that containment priest only made the poor matches worse it didn't have that detrimental of an effect i don't believe.

Not to derail the thread too much, but DTT helps with consistency on two fronts. Yes, if the game goes to turn 3 and beyond then a value DTT can help you get to the point where you can combo off. But what seems to be even more important (I haven't played the deck, so this is my impression from across the table) is that DTT also functions as pseudo combo piece once Omniscience is in play. So yes, the deck is still theoretically a 3-card combo, but in reality it feels more like a 2 card combo because all you need is show and tell + omniscience + one more action spell, be that a cunning wish, an emrakul or a DTT.

[apple713]

Not to derail the thread too much, but DTT helps with consistency on two fronts. Yes, if the game goes to turn 3 and beyond then a value DTT can help you get to the point where you can combo off. But what seems to be even more important (I haven't played the deck, so this is my impression from across the table) is that DTT also functions as pseudo combo piece once Omniscience is in play. So yes, the deck is still theoretically a 3-card combo, but in reality it feels more like a 2 card combo because all you need is show and tell + omniscience + one more action spell, be that a cunning wish, an emrakul or a DTT.

still you are looking for 2 cards that can at most be 4 of's AND a third piece that is usually 9 cards (DTT cunning emrakul). In Sneak you are looking for two cards that have 8 copies each. Statistically sneak has every advantage.

[nevilshute]

still you are looking for 2 cards that can at most be 4 of's AND a third piece that is usually 9 cards (DTT cunning emrakul). In Sneak you are looking for two cards that have 8 copies. Statistically sneak has every advantage.

By the same token, in sneak and show you are running 16 cards that are quite interdependent across the 4 types. Griselbrand doesn't combo with Griselbrand or Emrakul etc. The fact that you have been able to replace Enter the Infinite with DTT you are running fewer cards that aren't bricks outside of the combo but that still help you combo. I mean don't get me wrong, I understand what you are saying and I won't argue statistics, but am just trying to flesh out what I consider to be the advantages of the Omni tell deck in terms of consistency.

[Kanti]

if any deck was actually 75/25 vs the field it would warp the meta to where you are either playing that deck or a deck built specifically to beat it. Since that is not the case you are clearly wrong...




I'll admit that I haven't played omni post DTT but I'm having a hard time believing that a card that only has a use after turn 3 fixed the extremely clunky draws and consistency issues to bring the deck to be better than sneak. 2 card combo is almost certainly better than a 3 card combo.

Considering that containment priest only made the poor matches worse it didn't have that detrimental of an effect i don't believe.

But it's not an A+B+C card combo, its more of an A+B+X, where X just needs to be any cantrip. It's literally SnT>Omni>Ponder>ohheylookDTT>DTT>Ponder>Preordain>CWish>win. Lines like that are so common that it would be dumb to think otherwise. And yes, DTT did shoot it's consistency through the roof, and also aids with this whole 2 card combo thing. The only thing Sneak and Show is more consistent at is getting more dead draws.

So I'm looking for 2 cards, then I'm looking to chain cast my deck until I find a CWish. Chaining my deck isn't too hard when it consists of 12 cantrips and 4 DTT. So then instead of needing a 1/8 card as you say (Cwish or DTT) it's more like a 1/16 card (DTT, CWish, Ponder, Preordain) and that's not even counting G.Probe.

[mike1987]

Miracles do have some bad matchups like 12 post, landstill.

The deck that i personally feel has 50-50 across the board are the stoneblade decks, esper perhaps.

[FoolofaTook]

Miracles do have some bad matchups like 12 post, landstill.

The deck that i personally feel has 50-50 across the board are the stoneblade decks, esper perhaps.

Esper Stoneblade has a really iffy matchup against Burn. Other than that it works as a 50/50 list. My local meta has 4 Burn players in the 20 or so regulars and so Esper Stoneblade has declined here.

[Lemnear]

if any deck was actually 75/25 vs the field it would warp the meta to where you are either playing that deck or a deck built specifically to beat it. Since that is not the case you are clearly wrong...

In my books, a deck is highly warping the metagame if people are "forced" to dedicate 4+ SB slots to that particular deck alone. These days you either dodge the low manacost spells which get stuck in counterbalance or "have to" run Decay.


I'll admit that I haven't played omni post DTT but I'm having a hard time believing that a card that only has a use after turn 3 fixed the extremely clunky draws and consistency issues to bring the deck to be better than sneak. 2 card combo is almost certainly better than a 3 card combo.

Considering that containment priest only made the poor matches worse it didn't have that detrimental of an effect i don't believe.

Well, it's pretty unlikely to cast SneakAttack + hasty creature by turn 2/3 in SneakShow anyway, isn't it? You also reduced the number of clunky combo parts from a whooping 16 (and the fact that Griselbrand does nothing if paired with anotjer creature in your hand) to 4 S&T and 4 Omniscience plus an Emrakul which is about half the number SneakShow has to run and you can even spare a color. You don't suffer from clunky draws as DTT is easily castable unlike the former staple Enter The Infinite even w/o Omniscience in play and ergo does double duty for this deck as it digs for S&T/Omni pre-combo and for the kill with Omni on the field. You can even play Probes for additional information/protection/redundancy and quicker filling of your yard.

In general I find it really counterproductive to make statements about the current state of OmniTell without experience and suggesting SneakShow in the process

Edit:

There's that good-spirited, friendly and over-bearing Lemnear we have all come to love :smile:

I will agree with the sentiment of your post anyway, which is to say that I don't see how Sneak and Show is more consistent than Omni in the post-DTT world.

I love you too, pal xD

[iamajellydonut]

In my books, a deck is highly warping the metagame if people are "forced" to dedicate 4+ SB slots to that particular deck alone.

Outside of ANT, I can't think of a single deck that dedicates excessive amounts of cards in their sideboard for their Miracles match-up. Even Maverick just kind of rolls with the punches. I run Abrupt Decay because it's an amazing fucking card. I run Wear/Tear because it's an amazing fucking card. I run Surgical Extraction because it's an amazing fucking card. I run Red Elemental Blast because everything else is blue too. I run Krosan Grip because it's a shitty but generally necessary card. I run Pithing Needle because it's fucking amazing.

Who's actually that afraid of the big bad Miracles? It's not that good.

[Lemnear]

Outside of ANT, I can't think of a single deck that dedicates excessive amounts of cards in their sideboard for their Miracles match-up. Even Maverick just kind of rolls with the punches. I run Abrupt Decay because it's an amazing fucking card. I run Wear/Tear because it's an amazing fucking card. I run Surgical Extraction because it's an amazing fucking card. I run Red Elemental Blast because everything else is blue too. I run Krosan Grip because it's a shitty but generally necessary card. I run Pithing Needle because it's fucking amazing.

Who's actually that afraid of the big bad Miracles? It's not that good.

...and Maverick has been successful to make a comeback since Miracles is a thing?

[iamajellydonut]

...and Maverick has been successful to make a comeback since Miracles is a thing?

That's not the point, and Maverick is shit for a lot more reasons than just "Miracles exists".

[apple713]

But it's not an A+B+C card combo, its more of an A+B+X, where X just needs to be any cantrip. It's literally SnT>Omni>Ponder>ohheylookDTT>DTT>Ponder>Preordain>CWish>win. Lines like that are so common that it would be dumb to think otherwise. And yes, DTT did shoot it's consistency through the roof, and also aids with this whole 2 card combo thing. The only thing Sneak and Show is more consistent at is getting more dead draws.

So I'm looking for 2 cards, then I'm looking to chain cast my deck until I find a CWish. Chaining my deck isn't too hard when it consists of 12 cantrips and 4 DTT. So then instead of needing a 1/8 card as you say (Cwish or DTT) it's more like a 1/16 card (DTT, CWish, Ponder, Preordain) and that's not even counting G.Probe.

even if you call omni a 2 card combo finding 2 cards that are at best 4 of's is still worse than finding 2 cards that are 8 of's. EVEN going as far to designate DTT as a combo piece, finding 2x 6 of's is still harder than finding 2x 8 of's

[rufus]

...
I'm wondering what decks other than Can Thresh are out there that might not have spectacular matchups against everything, but no terrible ones against anything, either.

There are decks like Dragon Stompy that seem to care more about their own draw than the opponent's.

[Lemnear]

even if you call omni a 2 card combo finding 2 cards that are at best 4 of's is still worse than finding 2 cards that are 8 of's. EVEN going as far to designate DTT as a combo piece, finding 2x 6 of's is still harder than finding 2x 8 of's

with more cantrips total and the use of DTT to look 7 cards deep and cherrypick, I doubt the chance to find both combo pieces by turn 3/4 is any lower than running SneakShows 8/8 split, but that's not the topic. Evading clunky draws/hands of several Creatures and dodging opposing hate is. We could go into details and compare how efficient a Sneaky Emrakul is compared to a S&t'd Omniscience for closing out games

[Star|Scream]

even if you call omni a 2 card combo finding 2 cards that are at best 4 of's is still worse than finding 2 cards that are 8 of's. EVEN going as far to designate DTT as a combo piece, finding 2x 6 of's is still harder than finding 2x 8 of's

I don't get it. Why do you keep arguing when you've admitted to not even playing with OR against the deck? It's putting up enough numbers to be a DTB.

[BTY]

i wouldn't recommend a combo deck for what he's asking. most combo decks lose to themselves. Furthermore, unless something has drastically changed, sneak and show is the best combo deck in the format. D&T is maybe its worst matchup... and blood moon/pyroclasm rock it.

Recently omni has been performing better but it's less consistent than sneak. Sneak is underrepresented and doesn't show up as high on tier lists because its among the most expensive decks in the format comparatively.

That is not always true, especially with the cantrip density of certain combo decks like storm variants and omnitell. Those decks both have a high density of cantrips, a lot of/enough basics+fetches so you don't lose to wasteland much (with the exception of some versions of TES) and the combo is relatively compact. For storm, you just usually need mana and some form of business card, and omnitell you need the S&T, omni and usually just cantrips and DTT to find the kill. I have been playing both a lot lately and they are both very consistent and have very few bad match ups, the worst probably being Reanimator and MUD sometimes for storm and Other show and tell decks/griselbrand decks with Force of Will for omnitell, MUD can also be kind of tough too but all of those decks aren't very common at all. The biggest thing holding storm back is that it can be tough to see winning lines sometimes or how to adequately play against discard or something but all in all I think it could be a perfectly reasonable choice.

[Richard Cheese]

Just pick up another flavor of Delver. More creatures around? Try Patriot. More combo? Give BUG a spin.

[wizard_of_gore]

- Various GBx midrange builds (Jund, Rock, BUG and similar)
- Deadguy ale, maverick, death and taxes
- Various Ux stoneblades
- Tempo Delver decks
- Miracles
- Grixis midrange/control

[Admiral_Arzar]

Every deck has bad matchups. The key is to find decks that are good against most of the field - i.e. Miracles is 50/50 or better against basically all tier one decks, but suffers against Tier 2 or fringe decks like MUD, Cloudpost, Infect, and Stax. Delver has issues with Veteran Explorer.dec, anything mono-black, etc. However, these decks are going to be strong against most of what you will see, which is the reason they see so much play.

[Megadeus]

I think the closest you will get to a 50/50 deck is a Stoneblade Variant. The counterspells give you a shot against combo decks and the stoneforge package plus removal and other bombs like Jace give you game against the fair decks.

[mishima_kazuya]

The key is just knowing your deck better and playing better than your opponent.
Turns your not so good match-ups, into 50/50 match-ups.

[Varal]

Are you talking about 50/50 for an average player playing against an average player or for an expert player playing against another expert player? This can make a big difference since not all decks/matchups have the same tolerance for mistakes.

[slave]

In my books, a deck is highly warping the metagame if people are "forced" to dedicate 4+ SB slots to that particular deck alone.
Dredge did exactly this for quite some time IME. Hate these days is so varied, and in many cases main deck material (DRS etc.), that I don't think this is the case for the majority of players.

I think the strongest deck is Miracles, it's a beast in the hands of a good pilot.
But out of curiosity, what could you possibly play to hate them out in the first place?

[FoolofaTook]

Dredge did exactly this for quite some time IME. Hate these days is so varied, and in many cases main deck material (DRS etc.), that I don't think this is the case for the majority of players.

I think the strongest deck is Miracles, it's a beast in the hands of a good pilot.
But out of curiosity, what could you possibly play to hate them out in the first place?

Vendilion Clique and Stifle main list.

[mishima_kazuya]

BG/x decks like Shardless BUG or Jund are good against Miracles. The combination of Tarmogoyf, card advantage from Ancestral or P.Fire+Liliana and Abrupt Decay for Counterbalance is great. The caveat is that you need to pray Miracles doesn't assemble a board of basic lands and topdeck Entreat the Angels for sad times. :cry:

[carefulmug]

Thank you for your responses!


The key is just knowing your deck better and playing better than your opponent.
Turns your not so good match-ups, into 50/50 match-ups.

I agree with your philosophy, only my goal is to take 50/50 match-ups and render them wins, not to take 30-70 match ups and render them even.


Are you talking about 50/50 for an average player playing against an average player or for an expert player playing against another expert player? This can make a big difference since not all decks/matchups have the same tolerance for mistakes.

I'm considering these values in a vacuum by ignoring mind games and bluffs but anticipating ideal execution on both sides. However, I can see how this may create for more gray area than is helpful.


Nonetheless, t sounds like the consensus is Delver, Storm, DnT, Miracles, S&T, and, to a lesser extent, Elves and Blade.

[Zupponn]

There are decks like Dragon Stompy that seem to care more about their own draw than the opponent's.
Yeah, the deck is definitely at the least 50/50 against the field, but sometimes 30/70 against itself.

[Raystar]

This

A very elitist assumption....

[Lemnear]

A very elitist assumption....

I agree. Assuming you are a supreme player just because you play the deck with the best anti-combo/creature MB is a joke.

[kombatkiwi]

I'm not a fan of the "I'm such a savant for playing Control/Gifts/Doomsday/etc" statements but assuming that the data posted in the BR thread is correct then he's kind of right

only a 51% win percentage for Miracles overall


SCG Legacy Results for Major Archetypes (DC - Providence)
January to now

MUD: 38-22-2, (63%)
RG Lands: 69-44-8, (61%)
Food Chain: 18-12-2, (60%)
BUG Delver: 138-97-10, (59%)
Infect: 60-44-2, (58%)
BURG Delver: 21-16, (57%)
UWR Stoneblade: 41-31-6, (57%)
Junk: 14-11, (56%)
Lands: 15-12-2, (56%)
Shardless BUG: 106-84-9, (56%)
Painter: 31-26-2, (54%)
RUG Delver: 99-86-7, (54%)
Grixis Delver: 12-11-3, (52%)
Miracles: 158-153-33, (51%)
DeadguyAle: 11-11-2, (50%)
Dredge: 58-59-2, (50%)
Elves: 95-95-4, (50%)
Grixis Control: 23-23-5, (50%)
Patriot: 64-65-5, (50%)
Storm: 114-116-7, (50%)
Twelvepost: 37-37-5, (50%)
Goblins: 19-20, (49%)
Sneak and Show: 94-97-5, (49%)
Death and Taxes: 123-131-14, (48%)
Maverick: 52-56-5, (48%)
Omnitell: 43-48-3, (47%)
Reanimator: 69-80-5, (46%)
Burn: 43-52-1, (45%)
Jund: 27-33-1, (45%)
Deathblade: 44-55-8, (44%)
UW Stoneblade: 14-18-3, (44%)
U/R Delver: 13-18, (42%)
Merfolk: 21-31-1, (40%)
Esper Stoneblade: 12-19-1, (39%)
Tin Fins: 7-12, (37%)
Nic Fit: 9-22-3, (29%)
Enchantress: 10-27-3, (27%)
High Tide: 7-22, (24%)

Edit
I meant to imply that skill is clearly important if it's overall 50% against the field but undoubtedly a DTB.
The statement that all Miracles wins are due to the miracle pilot being a better MTG player than their opponent is obviously hyperbole.

[Raystar]

I'm not a fan of the "I'm such a savant for playing Control/Gifts/Doomsday/etc" statements but assuming that the data posted in the BR thread is correct then he's kind of right

only a 51% win percentage for Miracles overall



Edit
I meant to imply that skill is clearly important if it's overall 50% against the field but undoubtedly a DTB.
The statement that all Miracles wins are due to the miracle pilot being a better MTG player than their opponent is obviously hyperbole.

I believe there are other effects being discounted:

Miracles is probably the strongest deck in the meta: more people assemble it and bring it to tournament, this means that the percentage of "non dedicated" players is probably a bit higher than with other decks, possibly leading to a lower win percentage.
Saturation: when the field gets saturated by an archetype you start seeing many mirror matches. The net result is that the winning percentage will probably level to an even state.


The points above are, of course, totally based on my biased view of the state of competitive Legacy. I don't possess data to back it apart from tournament reports where people are starting to play against Miracles more than 50% of the time.

[Quasim0ff]

Those results are pretty interested - I mean, three stoneblade variants with sub-44%, as well as Omnitell with below 50%.


*Worth noting is also, regarding Miracles: Miracles is a deck with folds to very few decks, but have a shot at beating pretty much every thing else. That's something good players will play, which also makes for it becoming the best deck. Simply due to the level of overall play from the (best) pilots.

[apple713]

I believe there are other effects being discounted:

Miracles is probably the strongest deck in the meta: more people assemble it and bring it to tournament, this means that the percentage of "non dedicated" players is probably a bit higher than with other decks, possibly leading to a lower win percentage.
Saturation: when the field gets saturated by an archetype you start seeing many mirror matches. The net result is that the winning percentage will probably level to an even state.


The points above are, of course, totally based on my biased view of the state of competitive Legacy. I don't possess data to back it apart from tournament reports where people are starting to play against Miracles more than 50% of the time.

just looking at the food chain numbers its easy to see that when a small number of dedicated players plays a deck it performs better. there is no way that food chain is 60% vs the meta. its solid but those numbers have factors like you have mentioned that have been left out and need to be taken into account.

In short, i do believe skill accounts for a lot of a decks potential. The number for miracles is probably representative of an average player playing the deck.

[Gunseng]

BG/x decks like Shardless BUG or Jund are good against Miracles. The combination of Tarmogoyf, card advantage from Ancestral or P.Fire+Liliana and Abrupt Decay for Counterbalance is great. The caveat is that you need to pray Miracles doesn't assemble a board of basic lands and topdeck Entreat the Angels for sad times. :cry:
And by good you mean "have a more or less equal matchup"? I play a lot of BUG Delver and I don't see myself favoured in any way against Miracles. You have to chip away their life points with a single threat at a time and hope that they don't topdeck a win condition which you can't counter. This is not fun, especially because they are a lot better in the late game than you are.

On topic: In my opinion Miracles is by far the strongest deck in the meta. It is either favoured or equal to all tier 1 decks and only at a disadvantage against random tier 2 decks which are played far less often. If it were more fun to play, I would pick it up in a heartbeat. I even consider it oppressive in the current meta.

[Quasim0ff]

And by good you mean "have a more or less equal matchup"? I play a lot of BUG and I don't see myself favoured in any way against Miracles. You have to chip away their life points with a single threat at a time and hope that they don't topdeck a win condition which you can't counter. This is not fun, especially because they are a lot better in the late game than you are.

On topic: In my opinion Miracles is by far the strongest deck in the meta. It is either favoured or equal to all tier 1 decks and only at a disadvantage against random tier 2 decks which are played far less often. If it were more fun to play, I would pick it up in a heartbeat. I even consider it oppressive in the current meta.
What BUG decks do you play? Delver?

[Gunseng]

Yep, BUG Delver. Are there any other ones? :)

[Quasim0ff]

Yep, BUG Delver. Are there any other ones? :)

Shardless :)?
That deck has a very good matchup with miracles. Do you play Hymn or Stifle BUG?

[Gunseng]

My list is (almost) identical to Bob Huang's (http://www.channelfireball.com/videos/channel-bob-huang-legacy-bug-delver/). It is a Hymn Build and I have always played Hymn in the past. I have not played against too many miracles with this list, so this version may be better than my other lists in the past. Still, I wonder how you define a "very good" match up versus Miracles? 60% in BUG Delver's favor? Personally, I would not agree with this assessment.

[FoolofaTook]

A very elitist assumption....

It's the truth though. Miracles is a monster in the hands of a very good player. There are some people I sit down opposite when they are playing Miracles and I know I'll beat them with whatever I happen to be playing on the day. It's a pile very unforgiving of unforced errors and it makes people think two and three turns ahead all the time. It makes them balance the use of fetches more than any other pile. It puts people under more time pressure than any other pile.

You're right that Miracles is played by more unqualified players than any other list in the meta right now. That's because it has a fairly high bar to meet before you are really qualified to play it.

[Quasim0ff]

My list is (almost) identical to Bob Huang's (http://www.channelfireball.com/videos/channel-bob-huang-legacy-bug-delver/). It is a Hymn Build and I have always played Hymn in the past. I have not played against too many miracles with this list, so this version may be better than my other lists in the past. Still, I wonder how you define a "very good" match up versus Miracles? 60% in BUG Delver's favor? Personally, I would not agree with this assessment.

No, Shardless BUG is favored towards the Shardless deck. That deck has so many angles of attack, as well as better CA than miracles. It can play a very strong control-setup, with planeswalkers, 2 for ones (or 1 for 4s with shardless > AV etc).

BUG Delver is pretty close to 50/50. That depends, highly, on playerskill as well as who goes first.

[Raystar]

It's the truth though. Miracles is a monster in the hands of a very good player. There are some people I sit down opposite when they are playing Miracles and I know I'll beat them with whatever I happen to be playing on the day. It's a pile very unforgiving of unforced errors and it makes people think two and three turns ahead all the time. It makes them balance the use of fetches more than any other pile. It puts people under more time pressure than any other pile.

You're right that Miracles is played by more unqualified players than any other list in the meta right now. That's because it has a fairly high bar to meet before you are really qualified to play it.

That' a different point of view than the one I commented upon, I agree to this. What was said was

"...miracles. People who play miracles win because they are better players than their opponents (generally speaking). The deck has pretty even matchups but shines in the hands of a good player..."

The above implies that Miracles appears as a better deck because people that play it are better players. This I don't agree: Miracles is a better deck, full stop. It is such a better deck that the community is starting to wonder if it is warping the meta around itself.

Of course it is a difficult deck to play and good players get the best out of it but the same applies to ANT, TES, Canadian, BUG, Shardless BUG and I don't hear the same statements about them.

[Quasim0ff]

That' a different point of view than the one I commented upon, I agree to this. What was said was

"...miracles. People who play miracles win because they are better players than their opponents (generally speaking). The deck has pretty even matchups but shines in the hands of a good player..."

The above implies that Miracles appears as a better deck because people that play it are better players. This I don't agree: Miracles is a better deck, full stop. It is such a better deck that the community is starting to wonder if it is warping the meta around itself.

Of course it is a difficult deck to play and good players get the best out of it but the same applies to ANT, TES, Canadian, BUG, Shardless BUG and I don't hear the same statements about them.

I do agree with the argumentation that, Miracles isn't* way less forgiving than any other legacy deck. There are very few decks (besides something like Sneak and Show) where you can just try to jam it, and then go off again and again, in case you don't get there.

However, it is MORE unforgiving, in my estimation, compared to something like Delver/Shardless, and likely on the same page as Storm, where you can really fuck it up for you, in case you miss something. I think the legends version is much less forgiving, compared to the ponder version, due to the amount of cards - thus ressources - you have available.

[plowshares]

I really like that every thread on here has become "Is Miracles oppressive". I do think that Miracles is one of the most unforgiving decks out there. I regularly playtest with several accomplished miracles (and legacy in general) players, and it can be a real struggle to have a respectable win percentage when I'm playing Canadian, which I believe has a 50-50 shot against the deck. However, at SCG opens and the like, it is very easy for me to beat worse Miracles players; they manage their resources poorly and I am able to capitalize on that. Miracles might be a more powerful weapon than something like Delver, Shardless, or DnT, but it is much more difficult to wield.

[Quasim0ff]

I really like that every thread on here has become "Is Miracles oppressive". I do think that Miracles is one of the most unforgiving decks out there. I regularly playtest with several accomplished miracles (and legacy in general) players, and it can be a real struggle to have a respectable win percentage when I'm playing Canadian, which I believe has a 50-50 shot against the deck. However, at SCG opens and the like, it is very easy for me to beat worse Miracles players; they manage their resources poorly and I am able to capitalize on that. Miracles might be a more powerful weapon than something like Delver, Shardless, or DnT, but it is much more difficult to wield.
Agree, it gets pretty boring, when every new thread ends with "how to smash the miracles foe?"

[FoolofaTook]

That' a different point of view than the one I commented upon, I agree to this. What was said was

"...miracles. People who play miracles win because they are better players than their opponents (generally speaking). The deck has pretty even matchups but shines in the hands of a good player..."

The above implies that Miracles appears as a better deck because people that play it are better players. This I don't agree: Miracles is a better deck, full stop. It is such a better deck that the community is starting to wonder if it is warping the meta around itself.

Of course it is a difficult deck to play and good players get the best out of it but the same applies to ANT, TES, Canadian, BUG, Shardless BUG and I don't hear the same statements about them.

Miracles is different than most of the lists above in that it really doesn't have bad matchups in the commonly played lists. The closest are the card advantage lists in the BGx realm and even against them Miracles can win playing hellbent for several turns. Top plus cantrips is just very strong at finding what you need when you need it.

Each of the lists you listed has a commonly played list or two that is basically a 30/70 matchup where it's not likely to win the match going in. Miracles doesn't have that in the commonly played lists at the moment.

[spirit of the wretch]

Miracles doesn't have that in the commonly played lists at the moment.

The OmniShow MU does seem pretty hard from my (admittedly limited) testing.

[LOLWut]

I believe there are other effects being discounted:
[...]

Saturation: when the field gets saturated by an archetype you start seeing many mirror matches. The net result is that the winning percentage will probably level to an even state.
[...]


here are the records in non-mirror matches and match win percentages excluding draws for all archetypes with at least 20 matches against known decks

[Chatto]

So now what? Can we conclude that it doesn't matter which deck you play, just play the deck you are the most familiar with and jam some anti-Miracles tech somewhere in your 75?

That would be a pretty sad conclusion...

[Megadeus]

It's sad that the conclusion is that you can play whatever you want?

[maharis]

Miracles is a better deck, full stop. It is such a better deck that the community is starting to wonder if it is warping the meta around itself.

Of course it is a difficult deck to play and good players get the best out of it but the same applies to ANT, TES, Canadian, BUG, Shardless BUG and I don't hear the same statements about them.

Agree. In fact Lemnear and I don't agree on much but he is 100% spot on here:


Assuming you are a supreme player just because you play the deck with the best anti-combo/creature MB is a joke.

I have lost to clods on Miracles before because of one blind flip on Counterbalance. I have drawn six cards off a Sylvan Library and not had a way to beat Terminus and Entreat off the top.

I've also beaten the deck by playing smart myself, and getting a little bit lucky. However I've always felt that the onus is on me to beat them and not the other way around. The deck counters your spells for free, AND it plays a full suite of countermagic besides. It maxes out on the best boardwipe effect (tucking vs. destroying is very relevant), and can win in one turn, AND it plays Jace which enables/supplements all of the above or just wins by itself. It can clique/venser lock you if they feel like it, plays Snapcaster Mage and Dig Through Time, can adopt Monastery Mentor... It's full of actual powerful effects.

As draw-go decks go historically, I would say the nature of the Miracle mechanic and the raw power of Counterbalance even without top make it much more forgiving of errors than control decks of the past. As card selection has gotten better, the gulf has only grown.

[Chatto]

It's sad that the conclusion is that you can play whatever you want?

It's sad you have to dedicate slots to a certain deck, but that's Legacy of course.

[Quasim0ff]

The OmniShow MU does seem pretty hard from my (admittedly limited) testing.
Omni crushes Miracles with a God-Draw.

Otherwise it's a grindfest, unless they have access to Boseiju.

[Megadeus]

It's sad you have to dedicate slots to a certain deck, but that's Legacy of course.

Sure... But the cards you board for miracles aren't stupid super narrow shit like random life gain that you would have to board versus burn. Against Miracles there are all sorts of cards that are fine against them. Needle is a very solid effect that has plenty of random overlap. Councils judgement, decay, KGrip all have overlap into hurting other decks. Sure, you need to take into account that the deck exists while putting your board together, but the ways you can attack the deck are varied and not too narrow for the most part unlike the cards you need to board against burn (life gain, leyline), or storm (Rule of Law, Canonist, Leyline Again I suppose).

[twndomn]

Decks themselves don't make you 50-50 against the 'Field', the familiarity of the format (or local meta) allows you to play any decks and be 50-50 against the field.

[Barook]

It's sad that the conclusion is that you can play whatever you want?
Makes me wonder why not more people play MUD, since it doesn't have the same cost barrier like Lands.

[Megadeus]

Makes me wonder why not more people play MUD, since it doesn't have the same cost barrier like Lands.

Probably because they just don't own the cards. When you are breaking into the format the only deck that you can go to from MUD is other stompy things like Dragon or Sea Drake or go into Tezzerator (which just the cost of the additions like USeas, Forces, Jaces are probably triple what the whole MUD deck cost).

[Quasim0ff]

It's sad you have to dedicate slots to a certain deck, but that's Legacy of course.

How is that, in any way, sad?

That's magic 101. Some decks play in a different axis than you do, so you need to optimize your deck to beat that. This is how it works in standard, modern and (even more so!) vintage.

5 cards is too few(!) to have a reasonable shot at beating dredge - something often being about 20% of the field to a non-proxy event! You need approximately the same to beat MUD consistently - Usually the top-played archetype (all mono-brown decks together). That leaves, with 6 to each dredge and MUD 3 cards for the blue matchup. Imagine that scenario in legacy.

Needle is excellent vs miracles, null rod is so too - Both are good cards, which are fine against several other top deck - Death and Taxes as well as Storm are both cards fine against. Needle is also excellent vs. planeswalkers.

Krosan Grip, Decay etc. are good vs most of the UWx stoneforge decks.

Sure, it might look like a lot of your sideboard is geared towards beating miracles. That's not the case - The cards overlap alot.

[Zombie]

Why would you board Needle against Storm?

[Megadeus]

Why would you board Needle against Storm?

Preemptive Stone Rain on their fetches???? Duh

[Kanti]

It's mostly when you're deck has literally nothing else to do, and you're trying to side out some chaff that like StP.

[Namida]

It's mostly when you're deck has literally nothing else to do, and you're trying to side out some chaff that like StP.

At least an opponent using Swords to Plowshares on their own creatures to gain life is something marginally worth being concerned about, which is more than I can say about Pithing Needle naming basically any card in the Storm deck. Also, every time I have had an opponent resolve a Pithing Needle against me when I was playing Storm, it named LED.

[Megadeus]

At least an opponent using Swords to Plowshares on their own creatures to gain life is something marginally worth being concerned about, which is more than I can say about Pithing Needle naming basically any card in the Storm deck. Also, every time I have had an opponent resolve a Pithing Needle against me when I was playing Storm, it named LED.

And I am sure that all of those times they were extremely confident with their play and thought it was a great one too.

[btm10]

And I am sure that all of those times they were extremely confident with their play and thought it was a great one too.

Phyrexian Revoker namimg LED, on the other hand, has won me games against TES.

[Megadeus]

I have won games where my opponent was unable to get hell bent thanks to Revoker. Such a fun card. It was so sweet in Deadguy

[btm10]

I miss running him. Too bad the meta is incredibly hostile to one toughness dudes right now.

[nevilshute]

As an anecdote I can add, that I was once playing at my LGS for the weekly legacy tournament - so pretty casual atmosphere - and I was up against another regular. I was on ANT and he was on some delver list I think. It's game 2 and I'm on the play I go turn one Swamp --> Duress. He lets it resolve and shows me a hand containing among other things a Pithing Needle. I blink twice at seeing it, as he's a pretty experienced player. He then remarks something like "needle does nothing against you, right?" and we both chuckle a bit, but inside I am looking at the 2x polluted deltas and 2 or 3 cantrips in my hand which are my only lands and thinking... fuck.

I could choose to take the needle, but he had other relevant cards. In the end I took a flusterstorm or something and just prayed he wouldn't decide to take the line that would devastate me.

[Crimhead]

I'm a long time Canadian Threshold player and have always appreciated the deck's capacity to be (roughly) 50-50 against anything across the table.

I recently got into RG Lands but have become disinterested considering its abysmal OmniShow matchup.

I'm wondering what decks other than Can Thresh are out there that might not have spectacular matchups against everything, but no terrible ones against anything, either.
I guess you didn't play Lands very much, or else you'd realise it's almost an auto-win against your Thresh deck!

Thresh runs just 18 lands; none of them basic. If Lands doesn't drown you with Port, Ghost Quarter/Wasteland, and Tabernacle, your 11 creatures cant withstand PF, Maze, EE, and in a pinch Chasm. Canadian Thresh is a good deck, but it's paper-mache against Lands.


Makes me wonder why not more people play MUD, since it doesn't have the same cost barrier like Lands.MUD has a lot of feel bad losses which can fool the head.

If You play Miracles or Tempo, and you get bellow average draws, the game will often at least appear to have been close. Also, you'll have made so many decisions (often involving shuffles), you know things would have been different had you played just a little differently. It's easy to convince yourself a lot more of your losses would have been wins if only you'd done x, even though many of those games would still have been loses, or you did in fact make the best play but got unlucky.

With MUD, on the other hand, if you get sub-par draws and lose you know you got screwed by variance. And most people won't notice that many of those losses they would have lost even with an average hand, because their opponent drew exceptionally well.

Variance and perception mess with people, and many don't understand statistics. People see that MUD has a 58% win-rate, but they say it's inconsistent, or it losses to itself. that win-rate accounts for all those times MUD "loses to itself"! All losses are calculated. Winning 58% of its matches in the current meta makes it more consistent than any deck which wins fewer - it wins more consistently.

People also look at top8s too much. They don't realise that an archetype's probability of top8ing is solely a function of it's win rate and the percent of the field it occupies. Nothing else is a factor (except pilot skill). Even pros fall victims to misunderstanding probability. Post secondary maths majors can get confused by probability more so than by other branches of maths.

Pros who don't own the cards they need have to consider the investment vs just how much more they expect earn in a year. Just cause you are a pro doesn't make you rich. Hundreds of dollars is a lot to anyone - even a small business operator considering an increase in overhead.

Play-style is a thing, too. People tend to prefer aggro/control, control (no prison), and combo. This matters even to a pro! Would you accept a promotion if it made the work far less present and increase in salary was only moderate? Pros who play for money are probably smart enough to get a good job. They choose MTG because they like it. If they are going to shell out big money upfront and switch from a deck they love to a deck they loathe, the are going to want to see a substantial increase in the decks positioning.

[Varal]

I guess you didn't play Lands very much, or else you'd realise it's almost an auto-win against your Thresh deck!

Thresh runs just 18 lands; none of them basic. If Lands doesn't drown you with Port, Ghost Quarter/Wasteland, and Tabernacle, your 11 creatures cant withstand PF, Maze, EE, and in a pinch Chasm. Canadian Thresh is a good deck, but it's paper-mache against Lands.

MUD has a lot of feel bad losses which can fool the head.

If You play Miracles or Tempo, and you get bellow average draws, the game will often at least appear to have been close. Also, you'll have made so many decisions (often involving shuffles), you know things would have been different had you played just a little differently. It's easy to convince yourself a lot more of your losses would have been wins if only you'd done x, even though many of those games would still have been loses, or you did in fact make the best play but got unlucky.

With MUD, on the other hand, if you get sub-par draws and lose you know you got screwed by variance. And most people won't notice that many of those losses they would have lost even with an average hand, because their opponent drew exceptionally well.

Variance and perception mess with people, and many don't understand statistics. People see that MUD has a 58% win-rate, but they say it's inconsistent, or it losses to itself. that win-rate accounts for all those times MUD "loses to itself"! All losses are calculated. Winning 58% of its matches in the current meta makes it more consistent than any deck which wins fewer - it wins more consistently.

People also look at top8s too much. They don't realise that an archetype's probability of top8ing is solely a function of it's win rate and the percent of the field it occupies. Nothing else is a factor (except pilot skill). Even pros fall victims to misunderstanding probability. Post secondary maths majors can get confused by probability more so than by other branches of maths.

The only problem with this is that playskill is the most important factor in determining victory especially in low level tournaments like SCG Open or Grand Prix. It might be possible that skills is less important at the Pro Tour because everyone is closer to optimal but I'm still not sure about it.

Winning percentage doesn't say everything, the less a deck is played the more likely it will stray away from 50% either up or down. There's some additional complexities, a deck ability to win in the 3-0 and 0-3 brackets don't have the same value. I've no idea how much of an effect it has on the results but the best 3 percentage decks might not be the best decks in the format.

Playing the game theoretical best deck might not even be the best deck to win an event because you're not playing against abstract "rational" opponents always making the best play but flawed "irrational" opponents making bad plays. A higher variance deck might have less opportunity to profits from opponent mistakes than a lower variance lower average deck. It's similar to make a theoretical "bad" play in poker to abuse an opponent flaw in its play. The goal isn't to optimize your expected value against theoretical "rational" opponents but real "irrational" opponents.

[Crimhead]

The only problem with this is that playskill is the most important factor in determining victory especially in low level tournaments like SCG Open or Grand Prix. It might be possible that skills is less important at the Pro Tour because everyone is closer to optimal but I'm still not sure about it.AFAIK, there is no pro tour for Legacy anymore.

But it sounds like you want to ignore stats altogether! What we have is an uncontrollable variable. You can't test a deck without somebody piloting it, and you can't test a player without giving them a deck. The best we can do is to track the results of a deck over multiple tournaments played by multiple people. Probably all decks have good pilots and bad pilots, so the more data we have the less skewed it will be by pilot skill.

If you don't look at a deck's win-rate, what possible data do you have to suggest that one deck is stronger than another? I hope you realise that looking at top8s also is troubled by variable pilot skill! Top8 data has all the uncontrolled variables that win-rate data has, plus even more uncontrolled variables!


Winning percentage doesn't say everything, the less a deck is played the more likely it will stray away from 50% either up or down.The stats in question (http://www.reddit.com/r/MTGLegacy/comments/37lfon/scg_legacy_results_for_major_archetypes_dc/) ignore mirror matches, so is there s no reason ubiquitous decks will be close to a 50/50 win-rate. All we can say is that the more data we have on an individual deck, the more accurately the stats reflect it's actual strength.

Interestingly, if your deck has a (non-mirror) win rate >50%, you're better of with fewer people also playing that deck (mirror matches bring down your average). Only if your deck is <50% (non-mirror) should you want other people playing the same!


There's some additional complexities, a deck ability to win in the 3-0 and 0-3 brackets don't have the same valueThis is an illusion. If you are 3-0 and I am 0-3, presumably I'll now be having easier matches than you. But the flip side is I need to win more of my (easy) remaining matches that you have to win in order to match your record.

You are suggesting that is is easier to win a higher per cent of your matches if you lose round one than if you win! I hope you can see the absurdity of this idea. If two players go 8-2, will you suggest one of them has more impressive results based on which rounds they won or lost? Cause that's what you are saying!


Playing the game theoretical best deck might not even be the best deck to win an event because you're not playing against abstract "rational" opponents always making the best play but flawed "irrational" opponents making bad plays. A higher variance deck might have less opportunity to profits from opponent mistakes than a lower variance lower average deck. It's similar to make a theoretical "bad" play in poker to abuse an opponent flaw in its play. The goal isn't to optimize your expected value against theoretical "rational" opponents but real "irrational" opponents. I don't think there is a relationship between variance of a deck and opportunities that deck offers to exploit mistakes. I'm not seeing that. But let's give you the benefit of a doubt on this one...

Suppose certain decks are better helped by having weak weak players than other decks. If we look at Legacy stats, surely those decks will have wins for this very reason, and those wins will be recorded and contribute to that decks calculated win-rate. In other words, that phenomenon is being accounted for already!

TLDR - You seem to have an idea that there are factors which contribute to a decks probability of winning a match, which somehow will be lost when we observe how much that deck actually wins! It baffles me. If any factor affects a decks ability to win - that deck will win more or less accordingly and this will be reflected in it's win-rate. Conversely, anything that doesn't affect a decks win-rate is not a factor as to the decks probability to win any given match.

I'll ask again, if you are going to dismiss win-rates as relevant data, how can we meaningfully conclude anything about what decks are stronger than others in the meta? Experience/anecdotal evidence? Voodoo/Ouija?

[Varal]

AFAIK, there is no pro tour for Legacy anymore.?

True, I was talking about skills of the players. I've played in GP against opponents pawning games due to missed triggers and likewise I've pawned games in this way.


But it sounds like you want to ignore stats altogether! What we have is an uncontrollable variable. You can't test a deck without somebody piloting it, and you can't test a player without giving them a deck. The best we can do is to track the results of a deck over multiple tournaments played by multiple people. Probably all decks have good pilots and bad pilots, so the more data we have the less skewed it will be by pilot skill.

This will give you a good estimate of a deck results played by a random player against a random opponent but this shouldn't be what you're looking for. You're not a random player. You might be a bad player or a good player but definitely not a random player. It should be noted that being good or bad isn't a single statistics. It depends on the deck you're playing. Player A might be awesome with Lands while bad with Storm while player B will be the reverse.


If you don't look at a deck's win-rate, what possible data do you have to suggest that one deck is stronger than another? I hope you realise that looking at top8s also is troubled by variable pilot skill! Top8 data has all the uncontrolled variables that win-rate data has, plus even more uncontrolled variables!

I'm not saying that win rate isn't a valuable tool. I'm just saying it's flawed and not the only information you should use to select a deck.


The stats in question (http://www.reddit.com/r/MTGLegacy/comments/37lfon/scg_legacy_results_for_major_archetypes_dc/) ignore mirror matches, so is there s no reason ubiquitous decks will be close to a 50/50 win-rate. All we can say is that the more data we have on an individual deck, the more accurately the stats reflect it's actual strength.

Magic is a zero-sum game so your initial estimate of a deck winning rate should be 50%. The less a deck is played, the higher the variance on its winning ratio will be.


Interestingly, if your deck has a (non-mirror) win rate >50%, you're better of with fewer people also playing that deck (mirror matches bring down your average). Only if your deck is <50% (non-mirror) should you want other people playing the same!

Seems obvious, you want more people playing bad decks than good decks.


This is an illusion. If you are 3-0 and I am 0-3, presumably I'll now be having easier matches than you. But the flip side is I need to win more of my (easy) remaining matches that you have to win in order to match your record.

That wasn't my point, let say you have a good metagame deck. You might have a good matchup against the top decks of the format you're likely to meet at 3-0 but bad matchup against the more random, likely worst decks you'll meet in the 0-3 brackets. Magic tournaments are high variance games; it's better to go 8-0, 0-8 in two tournaments rather than 4-4, 4-4 in two tournaments. This means that a deck winning rate in the higher bracket is more important than its winning rate in the lower bracket. This is often seen in people not having sideboard slots for decks they're not likely to see after the first two rounds.



You are suggesting that is is easier to win a higher per cent of your matches if you lose round one than if you win! I hope you can see the absurdity of this idea. If two players go 8-2, will you suggest one of them has more impressive results based on which rounds they won or lost? Cause that's what you are saying!

I'm saying that in most situations you'll be more likely to win your next round if you lose the current one. I'm also saying that 2 people with the same record might now be as impressive. It's based on the structure of Magic tournaments and tiebreakers. Let say you're at a GP, you win your first 6 rounds then lose the next 3, this is more impressive than losing the first 3 rounds then winning the next 6 rounds.


I don't think there is a relationship between variance of a deck and opportunities that deck offers to exploit mistakes. I'm not seeing that. But let's give you the benefit of a doubt on this one...

Let say you play a deck with an high variance, for illustrative purpose a combo deck that always win if the opponent doesn't draw his 4-of foil card in the opening 7, you'll win a high proportion of games where you have a good draw while losing a high proportion of games where you have a bad draw. Even if your opponent plays well and has good draws, you'll win morev than him when you've good draws but you'll almost always lose when you've bad draws even if he plays badly. If you've a lower variance deck, you'll have flatter draws but when you're opponent plays badly you can turn "game theoretical loss" into win. If you can do this often enough then playing a "game theoretical losing" deck might be a winning proposition.


Suppose certain decks are better helped by having weak weak players than other decks. If we look at Legacy stats, surely those decks will have wins for this very reason, and those wins will be recorded and contribute to that decks calculated win-rate. In other words, that phenomenon is being accounted for already!

That's only if you care about random players. You should care more about you as a player and about the players/decks you're most likely to encounter while you're on a winning streak fighting for top 8 than on the players you might encounter while you're no longer competing for prizes.


TLDR - You seem to have an idea that there are factors which contribute to a decks probability of winning a match, which somehow will be lost when we observe how much that deck actually wins! It baffles me. If any factor affects a decks ability to win - that deck will win more or less accordingly and this will be reflected in it's win-rate. Conversely, anything that doesn't affect a decks win-rate is not a factor as to the decks probability to win any given match.

That's exactly what I'm saying. Give me the "best" deck, take the deck you're the best with and I'll most probably not win around 63% of matches. Why? All the "fluffy, unimportant, already accounted for, esoteric" outside factors.


I'll ask again, if you are going to dismiss win-rates as relevant data, how can we meaningfully conclude anything about what decks are stronger than others in the meta? Experience/anecdotal evidence? Voodoo/Ouija?

Combining statistics, top 8 results and experiences seems like a good start.

[Crimhead]

This will give you a good estimate of a deck results played by a random player against a random opponent but this shouldn't be what you're looking for. You're not a random player. You might be a bad player or a good player but definitely not a random player. Win-rate data doesn't reflect the performance of a single random player - it's the average results of many players. The more data we get, the better this average result matches the result of an average player. I agree it helps to know by own skill rate with various decks, and take that into account.


Magic is a zero-sum game so your initial estimate of a deck winning rate should be 50%. The less a deck is played, the higher the variance on its winning ratio will be.Because we have less data!



I'm not saying that win rate isn't a valuable tool. I'm just saying it's flawed and not the only information you should use to select a deck...
...That's exactly what I'm saying. Give me the "best" deck, take the deck you're the best with and I'll most probably not win around 63% of matches. Why? All the "fluffy, unimportant, already accounted for, esoteric" outside factors.Your own strengths are relevant. You want to look at all the best decks against the field, and pick the one you are best at playing; or look at all the decks you play well, and pick the one that is best against the field.

Also, a win-rate vs the filed is meant to reflect a deck's win rate vs each deck it might encounter, factoring in the likelihood of encountering any such given deck. The big event circuits over the past few months is not necessarily a reflection of the meta you are preparing for - not only do local metas vary, but the meta is also young, and likely still in flux. The Reddit project is thankfully providing the win-rates for individual matches. This is more valuable than just looking at the average win-rate without considering what matches are responsible for that ratio.


I'm saying that in most situations you'll be more likely to win your next round if you lose the current one. I'm also saying that 2 people with the same record might now be as impressive. It's based on the structure of Magic tournaments and tiebreakers. Let say you're at a GP, you win your first 6 rounds then lose the next 3, this is more impressive than losing the first 3 rounds then winning the next 6 rounds.I'm saying that in most situations you'll be more likely to win your next round if you lose the current one. I'm also saying that 2 people with the same record might now be as impressive. It's based on the structure of Magic tournaments and tiebreakers. Let say you're at a GP, you win your first 6 rounds then lose the next 3, this is more impressive than losing the first 3 rounds then winning the next 6 rounds.AFAIK, none of the data we have (top8, win rate, etc) distinguishes records based on order of wins and losses. This is stat we just don't have access to. As for the data we do have, there is no reason to assume some of the decks win-rates come from more top bracket matches than others'.


Let say you play a deck with an high variance, for illustrative purpose a combo deck that always win if the opponent doesn't draw his 4-of foil card in the opening 7, you'll win a high proportion of games where you have a good draw while losing a high proportion of games where you have a bad draw. Even if your opponent plays well and has good draws, you'll win more than him when you've good draws but you'll almost always lose when you've bad draws even if he plays badly.This doesn't make sense! If his deck "always win if the opponent doesn't draw his 4-of foil card in the opening 7", I have no good hands or bad hands.If he plays well and has good draws (aka, draws his "foil") I'll never win.


That wasn't my point, let say you have a good metagame deck. You might have a good matchup against the top decks of the format you're likely to meet at 3-0 but bad matchup against the more random, likely worst decks you'll meet in the 0-3 brackets...
...That's only if you care about random players. You should care more about you as a player and about the players/decks you're most likely to encounter while you're on a winning streak fighting for top 8 than on the players you might encounter while you're no longer competing for prizes.Correct me if I'm wrong, you are suggesting that decks which have more top eight finishes earn their win-rates by more top bracket matches? This is false. Say a deck is ~50/50, but is seen a lot in the top8s. For every such deck going X:0, there is another that went 0:X; for every such deck that was (X-1):1, there was another going 1:(X-1), etc.

The problem with top8 data is that it ignores completely the bad results of a deck, and takes the best results as being more representative of the deck. If I'm testing anything else, and report N success, I'm sure you'd want to know how many trials I ran!


Combining statistics, top 8 results and experiences seems like a good start.
Experience, if undocumented (in which case it becomes statistics), tends to be tainted by our biases. Some experience take disproportionate significances in our memories.

Top8s, taken out of the context of proportion to representation, is almost useless in evaluating a deck's strength vs the meta. A decks chance of top8ing is proportional to the number of pilots pushing it (scaled relative to its win-rate).

Let me take an example. Omnitell was roughly 20% of the field on day two of GP Kyoto (I don't have data for day one). The deck saw three spots in the top sixteen - which is also just under 20% of that bracket. Looking at just the three top16s, you'd think this deck was well positioned. But if you see that those three spots are in proportion to its overall percent of the field, you can guess it's going to be roughly a 50/50 deck. Turns out, Omni has a win-rate of 47%

You can go by win-rates, or you can go by top8s:entrants - you'll generally get the same conclusion. But for the love of god, don't just look at top8s alone! This is good to help you predict what you might be likely to run into (people blindly flock to these decks), but very misleading for anything else. Look at all the data!