Analysis of performance of blue decks in Legacy

[lordofthepit]

I have available data for tournaments from the Pacific Northwest, Arizona, and Sweden. These tournaments range from late August 2014 until now and range in size from 43 to 70 players. I thought it would be interesting to use them to quantify how much of an advantage blue decks have over non-blue decks.

A few notes on methodology:

• I don't have access to every single deck list, so I classify each deck based on whether or not it normally runs Brainstorm or Force of Will. For decks for which this is quite variable (mainly Aluren, Nic Fit BUG Pod, or Manaless Dredge in the case of Force of Will), this should be noted. Hypothetically, this means my tally doesn't account for the weirdo who decides to play Brainstorm in Death and Taxes.
• Intentional draws, byes, and match losses (due to late registration) are excluded in the analysis. Matches losses due to multiple game losses or disqualifications and matches influenced by one or more game losses are included in the analysis because I have no way of identifying these.
• For the purposes of calculating the performance of any specific card (e.g. Brainstorm), the matchups only consider the results of matches between two decks if one deck includes that card and the second deck does not. That is to say, the results of RUG Delver vs. ANT or Death and Taxes vs. GWb Maverick would not affect the record displayed for Brainstorm, but the matchup between Shardless BUG and Burn would.
• For the purposes of calculating the performance of any specific deck, only the results of mirror matches are excluded.

Event	Date	# Players	# Brainstorm	Brainstorm MW%	# Force of Will	Force of Will MW%
Dark Tower Legacy Open (Aug 2014)	08/24/2014	52	26 (50.00%)	39-26-1 (59.85%)	26 (50.00%)	35-27-1 (56.35%)
AZMagicPlayers.com Championship (Sep 2014)	09/06/2014	50	26 (52.00%)	38-23-1 (62.10%)	25 (50.00%)	38-23-1 (62.10%)
Card Kingdom PTQ Legacy (Nov 2014)	11/08/2014	59	43 (72.88%)	31-26-1 (54.31%)	40 (67.80%)	37-25-2 (59.38%)
Card Kingdom PTQ Legacy (Jan 2015)	01/03/2015	43	31 (72.09%)	21-13-1 (61.43%)	26 (60.47%)	23-17-1 (57.32%)
Mirkwood Legacy for Duals (Jun 2015)	06/14/2015	70	46 (65.71%)	43-34-0 (55.84%)	45 (64.29%)	42-34-0 (55.26%)
Card Kingdom GPT (Aug 2015)	08/16/2015	50	29 (58.00%)	35-26-0 (57.38%)	28 (56.00%)	37-23-0 (61.67%)
Card Kingdom GPT (Oct 2015)	10/24/2015	64	34 (53.12%)	36-44-1 (45.06%)	33 (51.56%)	33-43-1 (43.51%)
Viken Legacy Championship (Dec 2015)	12/05/2015	56	37 (66.07%)	25-38-2 (40.00%)	31 (55.36%)	37-39-2 (48.72%)
Mox Boarding House Legacy $1K (Dec 2015)	12/19/2015	59	36 (61.02%)	28-26-5 (51.69%)	35 (59.32%)	30-27-5 (52.42%)
Card Kingdom $1K (Jan 2016)	01/09/2016	67	47 (70.15%)	44-37-2 (54.22%)	37 (55.22%)	52-46-4 (52.94%)
AZMagicPlayers.com Championship (Jan 2016)	01/09/2016	64	31 (48.44%)	38-37-0 (50.67%)	26 (40.62%)	41-35-0 (53.95%)
Overall (Dig Legal)		324	201 (62.04%)	207-148-4 (58.22%)	190 (58.64%)	212-149-5 (58.61%)
Overall (Dig Banned)		310	185 (59.68%)	171-182-10 (48.48%)	162 (52.26%)	193-190-12 (50.38%)

So far, it looks like blue decks have an advantage, but not as much as I would have expected, and this is most pronounced during the Treasure Cruise/Dig Through Time era (9/26/2014 to 9/28/2015). Since the banning of those cards, non-blue decks have actually performed slightly better against the field, although this is limited.

I intend to continue to provide more information on this as I get more data. The wonderful organizers at Card Kingdom and Mox Boarding House have told me they will provide this for their new regular series, so I am optimistic I will have a greater sample size to draw from in the future!

By archetype, pre-Dig:

Archetype	# Players	% Field	Performance
Miracles	34	10.49%	84-65-6 (56.13%)
Death and Taxes	22	6.79%	49-52-2 (48.54%)
Reanimator	16	4.94%	48-40-1 (54.49%)
Omniscience	15	4.63%	37-41-0 (47.44%)
UR Delver	14	4.32%	32-25-0 (56.14%)
RUG Delver	13	4.01%	36-26-1 (57.94%)
ANT	12	3.70%	33-28-1 (54.03%)
Burn	11	3.40%	25-32-0 (43.86%)
Jund	11	3.40%	27-25-1 (51.89%)
Sneak and Show	9	2.78%	16-24-0 (40.00%)
UWR Delver	9	2.78%	27-21-3 (55.88%)
BUG Delver	8	2.47%	23-20-0 (53.49%)
Infect	8	2.47%	21-18-1 (53.75%)
Elves	7	2.16%	18-16-1 (52.86%)
Merfolk	7	2.16%	16-16-0 (50.00%)
Shardless BUG	7	2.16%	19-16-0 (54.29%)
Dredge	6	1.85%	15-15-0 (50.00%)
Grixis Control	6	1.85%	14-13-0 (51.85%)
Junk Depths	6	1.85%	16-18-2 (47.22%)
UWR Stoneblade	6	1.85%	20-14-0 (58.82%)
MUD	5	1.54%	10-15-0 (40.00%)
Maverick	5	1.54%	5-13-3 (30.95%)
Affinity	4	1.23%	7-15-0 (31.82%)
Deathblade	4	1.23%	12-10-0 (54.55%)
Gold Digger	4	1.23%	11-10-2 (52.17%)
Imperial Painter	4	1.23%	10-9-0 (52.63%)
TES	4	1.23%	3-10-0 (23.08%)
4c Delver	3	0.93%	10-8-0 (55.56%)
Big Red	3	0.93%	14-7-0 (66.67%)
Death's Shadow	3	0.93%	12-5-1 (69.44%)
Grixis Delver	3	0.93%	11-7-0 (61.11%)
High Tide	3	0.93%	0-10-1 (4.55%)
Manaless Dredge	3	0.93%	5-8-1 (39.29%)
Nic Fit	3	0.93%	6-9-1 (40.62%)
Zombardment	3	0.93%	3-10-0 (23.08%)
4c Loam	2	0.62%	2-6-0 (25.00%)
Esper Stoneblade	2	0.62%	4-5-0 (44.44%)
Goblin Stompy	2	0.62%	3-5-0 (37.50%)
Goblins	2	0.62%	3-6-0 (33.33%)
Junk	2	0.62%	8-2-0 (80.00%)
RG Lands	2	0.62%	1-7-0 (12.50%)
Tin Fins	2	0.62%	3-6-0 (33.33%)
UR Pyromancer	2	0.62%	8-4-0 (66.67%)
12-Post	1	0.31%	7-0-1 (93.75%)
Belcher	1	0.31%	3-3-0 (50.00%)
Bird Tribal	1	0.31%	1-4-0 (20.00%)
Deadguy Ale	1	0.31%	1-2-0 (33.33%)
Esper Mentor	1	0.31%	1-2-0 (33.33%)
Food Chain	1	0.31%	2-4-0 (33.33%)
Grixis Painter	1	0.31%	2-2-0 (50.00%)
Grixis Pyromancer	1	0.31%	2-2-0 (50.00%)
Harpies	1	0.31%	3-3-0 (50.00%)
Lands	1	0.31%	2-4-0 (33.33%)
Landstill	1	0.31%	2-2-0 (50.00%)
Living End	1	0.31%	1-2-0 (33.33%)
Mono-Blue Tempo	1	0.31%	2-2-1 (50.00%)
Mono-White Hatebears	1	0.31%	2-2-0 (50.00%)
Monoblack Aggro	1	0.31%	0-3-0 (0.00%)
Pox	1	0.31%	0-2-0 (0.00%)
RUG Depths	1	0.31%	4-3-0 (57.14%)
Rebels	1	0.31%	0-3-0 (0.00%)
Scepter/Chant	1	0.31%	0-3-0 (0.00%)
Shardless Bant	1	0.31%	2-4-0 (33.33%)
Shardless Deathblade	1	0.31%	4-2-0 (66.67%)
Stax	1	0.31%	1-3-0 (25.00%)
Sylvan Plug	1	0.31%	1-2-0 (33.33%)
Tezzeret	1	0.31%	3-3-0 (50.00%)
UW Tempo	1	0.31%	3-3-0 (50.00%)
UWR Mentor	1	0.31%	0-2-0 (0.00%)
Vial Maverick	1	0.31%	5-2-0 (71.43%)

By archetype, post-Dig:

Archetype	# Players	% Field	Performance
Miracles	35	11.29%	84-71-12 (53.89%)
Death and Taxes	21	6.77%	51-44-2 (53.61%)
ANT	19	6.13%	46-44-2 (51.09%)
Burn	16	5.16%	46-47-0 (49.46%)
Shardless BUG	16	5.16%	40-35-0 (53.33%)
Infect	15	4.84%	24-40-2 (37.88%)
Grixis Delver	13	4.19%	40-29-1 (57.86%)
4c Loam	11	3.55%	35-23-2 (60.00%)
BUG Delver	10	3.23%	20-29-0 (40.82%)
Deathblade	9	2.90%	30-23-1 (56.48%)
Jund	9	2.90%	22-19-1 (53.57%)
RUG Delver	9	2.90%	25-21-1 (54.26%)
Reanimator	9	2.90%	21-22-0 (48.84%)
Elves	7	2.26%	23-18-0 (56.10%)
RG Lands	7	2.26%	15-17-2 (47.06%)
Sneak and Show	7	2.26%	16-17-0 (48.48%)
Goblins	6	1.94%	10-17-0 (37.04%)
Grixis Pyromancer	5	1.61%	18-12-0 (60.00%)
Merfolk	5	1.61%	13-13-0 (50.00%)
TES	5	1.61%	7-16-0 (30.43%)
4c Delver	4	1.29%	13-12-1 (51.92%)
Imperial Painter	4	1.29%	14-8-0 (63.64%)
Maverick	4	1.29%	11-13-0 (45.83%)
Nic Fit	4	1.29%	11-10-0 (52.38%)
12-Post	3	0.97%	5-8-1 (39.29%)
Aluren	3	0.97%	6-10-0 (37.50%)
Bant	3	0.97%	10-8-0 (55.56%)
Dredge	3	0.97%	9-9-0 (50.00%)
Lands	3	0.97%	10-7-1 (58.33%)
Pox	3	0.97%	3-12-1 (21.88%)
Tin Fins	3	0.97%	6-9-0 (40.00%)
UWR Stoneblade	3	0.97%	6-9-0 (40.00%)
All Spells	2	0.65%	7-6-0 (53.85%)
Enchantress	2	0.65%	2-6-0 (25.00%)
MUD	2	0.65%	8-4-0 (66.67%)
NetherLands	2	0.65%	4-4-1 (50.00%)
Omniscience	2	0.65%	2-6-0 (25.00%)
Painter	2	0.65%	5-5-0 (50.00%)
BUG Thopters	1	0.32%	3-3-0 (50.00%)
Brave Sir Robin	1	0.32%	2-2-2 (50.00%)
Deadguy Ale	1	0.32%	2-3-0 (40.00%)
Doomsday	1	0.32%	2-4-0 (33.33%)
Esper Stoneblade	1	0.32%	0-3-1 (12.50%)
Food Chain	1	0.32%	0-2-0 (0.00%)
Goblin Stompy	1	0.32%	3-3-0 (50.00%)
Human Stompy	1	0.32%	4-2-0 (66.67%)
Humility	1	0.32%	3-3-0 (50.00%)
Hypergenesis	1	0.32%	4-3-0 (57.14%)
Jund Smallpox	1	0.32%	2-2-1 (50.00%)
Junk Depths	1	0.32%	4-3-0 (57.14%)
Landstill	1	0.32%	3-2-0 (60.00%)
Manaless Dredge	1	0.32%	2-4-0 (33.33%)
Mono-Black Devotion	1	0.32%	2-4-0 (33.33%)
NO BUG	1	0.32%	0-4-0 (0.00%)
Ooze Reanimator	1	0.32%	1-2-0 (33.33%)
R/W Stompy	1	0.32%	1-5-0 (16.67%)
Slivers	1	0.32%	0-3-0 (0.00%)
Spanish Inquisition	1	0.32%	2-4-0 (33.33%)
UR Delver	1	0.32%	4-1-1 (75.00%)
UW Stoneblade	1	0.32%	3-3-0 (50.00%)
Vial Maverick	1	0.32%	4-2-0 (66.67%)
Zombardment	1	0.32%	3-2-0 (60.00%)

[jrsthethird]

[*] There is one Living End deck for the Dark Tower Legacy Open for which I do not have the full decklist, and I have no idea whether or not it plays Brainstorm or Force of Will. This is excluded in the analysis, and the denominator for the number of players used in calculating representation of each deck is 51 instead of 52.

There's no way this deck plays Brainstorm (1cmc), but it most certainly runs FOW. There's enough blue cards to support it in the cascade/cycling mechanics and how else can that deck even expect to have a chance without a counterspell that doesn't cost them mana? I'd bet he's running a couple Misdirection too.

[death]

You bet. But the thing is, he can't run Force or MisD. Same reason he can't run Show and Tell, Progenitus (blue card), Omniscience and shit.

Deck is called Living End for a reason, modern version of Living Death. And I bet it doesn't resemble anything close to Hypergenesis, which plays more like Eureka-Tell.

[lordofthepit]

Thanks, guys. I know nothing about this deck other than the namesake card.

[KobeBryan]

Thanks, guys. I know nothing about this deck other than the namesake card.

You need to consider one more thing. How many people are playing blue decks. I think thats important to gauge who is going up against who whether its a blue deck vs a blue deck or not

[jrsthethird]

You bet. But the thing is, he can't run Force or MisD. Same reason he can't run Show and Tell, Progenitus (blue card), Omniscience and shit.

Not really. You can support Force/MisD with

4x Glassdust Hulk
4x Architects of Will
4x Shardless Agent
4x Ardent Plea
4-8x Force of Will/Misdirection

And that's without digging out any non-Modern legal blue cyclers. Drifting Djinn, Primoc Escapee, and Shoreline Ranger provide large evasive bodies and aren't legal in Modern.

[death]

4x Glassdust Hulk
4x Architects of Will
4x Shardless Agent
4x Ardent Plea
4-8x Force of Will/Misdirection

And that's without digging out any non-Modern legal blue cyclers. Drifting Djinn, Primoc Escapee, and Shoreline Ranger provide large evasive bodies and aren't legal in Modern.

Demonic Dread/Violent Outburst are more focused (than Ardent Plea/Shardless Agent) towards Living End's end goal. Why should anyone waste blue duals, Force, and entry fee to register a crappier build like that over a real Hypergenesis deck, or an optimized Living End deck if under a budget?

[Crimhead]

I'm not sure what it is you are trying to establish. Is your goal to determine whether blue decks have an advantage vs non-blue decks (and to what extent)? If so, here are some problems:

1. Your data includes weak decks as an uncontrolled variable. How a tier two deck like High Tide or Tezzerat performs against a tier one deck like Elves or D&T is not very significant. In our current meta their are more tier one options that include blue than do not. If (for example) half the nonblue decks in your sample are tier one, but two thirds of the blue decks are tier one, then this is not a fair contest. All the fuss about Living End is a good example.
2. Your categories are far to broad! If it turns out blue decks as a whole average better than blueless decks, how does that help you to specifically decide if, eg, D&T is better positioned than Canadian Thresh?
3. You are ignoring important MUs. If I want to compare decks, I want to know how they fare against each and every deck in the meta - including other decks which also do (or do not) run blue cards!

Personally I can't see this data as being useful at all. Suppose I am trying to decide between playing Lands or playing Infect. Do I really care that Storm is good against Jund, or that Miracles is good against Elves? I don't know why that should affect my decision at all. But those irrelevant MUs are going to shape your conclusion. Meanwhile, all your data which actually relates to one of the decks I am condidering will be against a set of archetypes for which you have no data relating to the other deck! Totally useless!

Imagine using this same method of analysis to compare fair decks vs combo decks; or creature decks vs creatureless decks. Would you put any stock into the results of such an investigation?

This looks to me like an excuse to justify hating on blue - and an attempt to back it up with (meaningless) data.

[jrsthethird]

Demonic Dread/Violent Outburst are more focused (than Ardent Plea/Shardless Agent) towards Living End's end goal. Why should anyone waste blue duals, Force, and entry fee to register a crappier build like that over a real Hypergenesis deck, or an optimized Living End deck if under a budget?

I don't understand your reasoning. Outburst is awesome because of Instant speed, but Dread is terrible. Requiring a creature to target to be able to cast it makes it dead in some matchups. It sees play in Modern is because of the colors and the smaller amount of cheap cascade cards. Plea is a stretch, but if you're going to add FOW it makes sense to include it. Shardless is the best cascade card in Legacy; in the deck it's a Living End with an extra 2/2 body attached, which Outburst and Dread are Living End with nothing else. Plea is admittedly bad, but the blue makes it considerable IF you want to jam Force in the deck. There's also the fringe benefit of not being able to pick away Shardless with Duress. If I were building the deck in Legacy, I would play Temur instead of Jund, and add 2x BBE for more redundancy.

I'm pretty sure any build of Living End is terrible in Legacy, which makes all the more sense to jam FOW if you want to try it. If you're going to cascade, play Hypergenesis. There's a reason one of them is banned in Modern. It's probably true that some scrub brought his Modern deck to the tournament for shits and giggles, but we have no proof. My point is, if someone IS wasting an entry fee to play it, and has the cards to support it, why would they NOT run FOW? It fits the deck's CMC restriction, and there are enough blue cards in the realm of cascaders + cyclers to support running it.

[death]

Shardless is the best cascade card in Legacy; in the deck it's a Living End with an extra 2/2 body attached

You would dispose of the 2/2 body after cascading into Living End. RTFC.

A Living End deck would be really terrible with those blue vanilla creatures since they have zero utility and their only function in the deck is what, pitch to Force of Will? Your deck building philosophy that 'FoW is mandatory' in this deck is flawed.

My point is, if someone IS wasting an entry fee to play it, and has the cards to support it, why would they NOT run FOW?

Because forcing Force in the deck makes you miss out on bigger and better non-blue creatures more suitable to be in a Living End deck.

[maharis]

You would dispose of the 2/2 body after cascading into Living End. RTFC.

Living end resolves before the spell that cascaded into it. The graveyard switcheroo happens before agent hits the field.

[death]

Thank you.

Here's the timing for cascade:
1) You cast a spell with cascade.
2) The cascade ability triggers and goes on the stack on top of the original spell.
3) The cascade ability resolves. If you find an applicable card that you'd like to cast, you do so.
4) The spell you cast as a result of the cascade ability resolves.
5) The original spell resolves.

[Mr Miagi]

Thank you.

Here's the timing for cascade:
1) You cast a spell with cascade.
2) The cascade ability triggers and goes on the stack on top of the original spell.
3) The cascade ability resolves. If you find an applicable card that you'd like to cast, you do so.
4) The spell you cast as a result of the cascade ability resolves.
5) The original spell resolves.

Or as one would elegantly put it

RTFC.

[lordofthepit]

<stuff>

I've always thought that the dominance of Brainstorm and Force of Will in Legacy was overstated, and I had indicated for a while that I would try to provide actual numbers when I had the data available and got around to it.

You can find the performance of different decks elsewhere, but this is the first time I've tried to actually compute the performance of Brainstorm/Force of Will decks on a whole. My job is to provide the statistics, not to editorialize or manipulate numbers one way or another by dismissing decks as Tier 2 or Tier 3. It sounds like we both agree that Brainstorm/Force of Will shouldn't be banned, but I'm not going to introduce subjective ways to fudge the data just to support this preference.

The conclusion I would make is that the blue shell was much better during the Treasure Cruise era but that on the whole, there is no significant advantage to playing Brainstorm or Force of Will.

[Crimhead]

This is fair - and I guess I misread your motivations.

There is still the problem that it is also relevant how a BS/FOW deck does against other blue decks. Obviously you can't account for that, because overall blue decks are break even against each other.

This ought to illustrate the flaw with the overly broad dichotomy. Blue deck and blueless decks comprise a wide range of archetypes which each have their own strengths, weaknesses, and overall effectiveness in the meta.

I think its wholly silly to divide all Legacy decks into just two categories and try to compare those categories. You wouldn't conclude that Aggro Loam is better positioned than Elves on the grounds that combo vs non-combo matches overall favour non-combo.

Knock yourself out, of course. But I don't see how your conclusions will be practically applicable in terms of deck selection.

[Stryfo]

The point is pretty clearly not to help with deck selection. The point of the analysis is to find out whether or not the claim that one must play a blue deck in legacy to be successful has any ground in reality. Once the data set is large enough, effects from things like tier1 vs. tier2 decks etc. will not affect the averages.

The OP is separating the decks of the format into two categories because that's how the "ban brainstorm" camp paints the argument: Either you play blue (therefore brainstorm) or you are at a disadvantage. The analysis doesn't seem to point towards that conclusion in recent tournaments, though it was clearly the case in the cruise era.

There is no problem about BS/FOW decks fighting against eachother, this data is irrelevant to the point of the analysis.

[tescrin]

I think its wholly silly to divide all Legacy decks into just two categories and try to compare those categories. You wouldn't conclude that Aggro Loam is better positioned than Elves on the grounds that combo vs non-combo matches overall favour non-combo.

But that would say that playing Combo is bad in general, and thusly you may conclude that a non-combo deck is better than Elves; not necessarily Aggro Loam. Certainly you could break it down further, but the point of this is to show whether there is blue dominance or not; not by Top 8's only, but by more broad comparison.

If, for example, Blue* is 70% of the field, you expect 6-7 entries in the top 8. This means you need to normalize the data by doing what Lord of the Pit has done in some fashion. Normalization tells us that during the TC and DTT eras, Blue* decks were more dominant by a lot.

*Blue being an odd categorization as I'm sure Lemnear will eventually chime in.

Lord's services have potentially rendered the B&R discussion somewhat m00t in terms of the perpetual Brainstorm discussions. If decks running Brainstorm aren't dominating the field after normalization, it is very arguable that Blue is back in balance with the others.

Yeah.. 52/48 is a little something extra to line your pockets with, but it says that we're at a good balance point again. (or at least, seems to suggest that.) I think given that the data would fit our preconcieved notions (that it's less dominant after TC ban, and even less after the DTT ban) it looks valid on the surface.


EDIT: Also, his timing and data point to what I would believe to be an unbiased data set. It follows our intuitions and says we've entered a point of reasonable equilibrium (at least if it continues.)

One could easily argue that since 52/48 is so small, it can be that more skilled players tend to play blue just because it offers more outs; or some argument similar to that. If it were 60/40 or 65/35, it starts becoming an underwhelming argument.

[Finn]

I am one of those people who was unhappy with the stupid things Cruise and Dig did to Legacy. I am also one of the people who said that Brainstorm was the actual culprit. But I have actually been quite pleased with the shape of the format since the second ban. I think tescrin is correct.

[lordofthepit]

<truncated>

One could easily argue that since 52/48 is so small, it can be that more skilled players tend to play blue just because it offers more outs; or some argument similar to that. If it were 60/40 or 65/35, it starts becoming an underwhelming argument.

Thanks for explaining the data more clearly.

I'd also like to point out that since Dig Through Time got banned, there have been two data points in my analysis, and on a whole, the blue decks have done worse. I'm wondering if this will still hold up as I get more data.

[jrsthethird]

A Living End deck would be really terrible with those blue vanilla creatures since they have zero utility and their only function in the deck is what, pitch to Force of Will? Your deck building philosophy that 'FoW is mandatory' in this deck is flawed.

Because forcing Force in the deck makes you miss out on bigger and better non-blue creatures more suitable to be in a Living End deck.

Glassdust Hulk is pretty bad, but Architects of Will's library manipulation isn't terrible. I also quoted 3 blue flyers that fit right into the deck (3/4, 4/4, and 5/5), giving evasion that's missing from Modern builds. Who cares about a Jungle Weaver when your opponent can come back with Pyromancer and make chump tokens every turn.

Maybe it is not mandatory in the deck. It is certainly possible though.