All these threads are a never ending stream of opinions vs opinions. Use proper statistical methods to infer solid conclusions, e.g. dimensionality reduction to define archetypes statistically, or clustering techniques to derive actual segments.
I appreciate your attempt to provide insights from descriptive analysis, but you can't take it for much more than face value.
The conclusions to be drawn from this analysis don't go much further than "Blue is a popular choice in legacy tournaments" or "Brainstorm and Force of Will are ubiquitous cards in Legacy" which unfortunately is not a very novel idea / revolutionary insight.
By stating that more skilled players tend to play blue, you assume that blue players are better at playing non-blue decks than players committed to playing non-blue decks. You then assume this is because blue offers more outs (Might just as well be that blue simply has more powerful cards, the way Lightning Bolt is more powerful than Shock). My point being, your analysis rests on 2 assumptions for which you provide no evidence. Before you know it, this thread will be as long as the B/R thread and while I wouldn't be surprised that the OP would like this, we're not going to find conclusive answers when the blind is leading the blind here.Originally Posted by tescrin
I'm not really sure where you are getting 52% from, when I calculated the average Win% across the various tournaments. I got 54.2% for Brainstorm, and 54.4% for FOW (if you consider a tie equivalent to a loss, if you remove ties entirely you get 55.3% and 55.6% for strict W to L)
That said I think your points are right on.
I agree that it's much better, but I don't know that I'm willing to consider a 4.2% / 4.4% higher win percentage "not significant."
When there are people who actually suggest that they should fetch before they draw for the turn when they need to find there one out to "improve the odds" (A fraction of a percent play) playing cards that net you a 4.2% higher win percentage seems like a no-brainier...
Significance != impact
http://stats.stackexchange.com/quest...tatistical-bac
In the post-Dig era, the Brainstorm/Force of Will decks are actually under 50%.
I'm not using "significant" in the statistically significant sense, but rather, using it to suggest that the advantage isn't banworthy.
This is not true at all!
Hypothetically, Elves could be the hands down strongest deck in a hypothetical meta. But if the other combo decks are sufficiently bad, fair decks could overall still average better than 50/50 vs combo as a whole.
This is why its ridiculous to lump decks into such broad categories for comparative purposes and any conclusions will be useless.
Here is a hypothetical meta with four decks - two blue and two green (GD1, GD2, UD1, UD2).
Here are the stats:
Gd1 vs ud1 - 55/45
Gd1 vs ud2 - 55/45
Gd2 vs ud1 - 35/65
Gd2 vs ud2 - 51/49
Ud1 vs ud2 - 40/60
Gd1 vs gd2 - 60/40
Clearly gd1 is hands down the best choice, with the highest average win rate and no unfavourable MUs. But this ill conceived method of analysis produces the irrelevant tidbit that blue decks overall are 51/49 against non-blue decks.Does this prove that you are at a disadvantage if you opt to pilot gd1?
This methodology is dismally flawed and does not work because your categories are arbitrary and the strengtn and performance of one deck has zero relevance to the other decks foolishly lumped into the same grouping. And I've now shown this with real math.
Last edited by Crimhead; 12-28-2015 at 10:05 AM.
Supremacy 2020 is the modern era game of nuclear brinksmanship! My blog:
https://fieldmarshalshandbook.wordpress.com
You can play Lands.dec in EDH too! My primer:
http://www.mtgsalvation.com/forums/t...lara-lands-dec
Yep. That's why I pushed back against the idea of calling it "insignificant" or not significant based on few data points and no statistical tests of significance...
Also turns out not everyone has a Stats background like you and I do, so when they say the word significant they might just actually mean the common tongue word significant... like...
... the OP...
I see, I read the dates wrong... and assumed all this data was post Banning. I just assumed no one would attempt to empirically prove how good (or not good) "Blue" is based on the results of 2 (count them 2 whole tournaments) each with less then 70 players...
Edit:
Now that I understand the Data better here are the W% I get.
Pre-Ban Brainstorm 57.5% FOW 59.4%.
Post Ban Brainstorm 45.7% FOW 45.3%
Assumes Draw = Loss.
Again though, I question the validity of comparing 6 mid size events vs 2 mid size events. I also question the ability to infer larger meta information from mid size events.
Last edited by GundamGuy; 12-28-2015 at 11:26 AM.
@bruizar
To be fair, I'm just attempting to explain the mild statistical difference. It could be variance, it could be player skill, or similar. The data is there for us to discuss.
That said, there are no assumptions on my part:
* Because firstly, I'm not making that argument. I'm just saying that the 52/48 split (or w/e) is small enough to possibly explain with some cultural explanation. That split is well within the 95% confidence interval; regardless if he applied it in his analysis.
* There is no "assumption" in "Blue having more outs." Why? Because it has more card selection. I'm fine if we're just disagreeing on what the word "outs" means, but when you can run a 2-of and it feels like a 3-4 of because of cantrips you have "more" outs in the sense that you have the ability to find the right card in the right situation and swap bad cards for good.
@crimhead
If Elves was the hands down best deck in said meta, people would play it and it would bring up the combo-game stats. You're assuming people don't already look for decks that have good MUs, but rather play only what they want to despite being measured high-level tournament statistics. People don't take Cheeri0s that often to an SCG. Why? Because it sucks. They bring the "combo" or the "Blue" decks that will perform against the expected meta.
Since his analysis mostly contains players who are serious about metagaming and serious about winning, they are unlikely to be running pure garbage and hurting the statistics. I'll admit, this is an assumption; but it should self-evident when you have like.. 20+% of people running Shardless, 20% running Miracles, and 20% or more running Delver, and similar. People are running Brainstorm Force of Will decks that they think will perform against the field.
It's not like it's made up of 47% of players running 12 copies of Cancel equivalents and running Master of Waves as a 4-of. We're analyzing people who have meta-gamed while choosing/learning previously successful decks and are mostly well-versed players.
@lordofthepit,
I suggest you also reference additional data points from Knightware tournaments. Lori puts in a good post-event wrap up that includes all archetypes played at the event.
Examples:
Knight Ware Inc, Dec 6, 2015
Knight Ware Inc, Nov 8, 2015
Knight Ware Inc, Oct 11, 2015
Knight Ware Inc, Sep 13, 2015
Knight Ware Inc, Aug 16, 2015
Knight Ware Inc, July 12, 2015
Knight Ware Inc, June 14, 2015
and so on
These provide stable points in a relatively stable metagame (Southern California) with consistent attendance >32.
West side
Find me on MTGO as Koby or rukcus -- @MTGKoby on Twitter
* Maverick is dead. Long live Maverick!
My Legacy stream
My MTG Blog - Work in progress
Can you ask Lori to send me the WER files associated with each events? My analysis is based on matches rather than standings, and all of this information is contained in the WER file. I'd be happy to provide the spreadsheets for free going forward if the TOs are interested, and I would obviously use them in my forward wide analysis.
Sorry but no! The whole point of this anaysis is to determine with data whether those decks indeed perform well or not. If we assume all the decks are in fact sound choices for the meta, then we've already reached a conclusion.
Okay, nobody said anything about total garbage. I'm assuming the blueless decks being tested are the likes of Lands, D&T, Elves, Dredge, Painter, Jund, Aggro Loam, Maverick, MUD/Post, plus maybe the odd other deck. This is an approximate list.
What if it turns out that a few (or several) of these decks are in fact at a disadvantage against the field (or the blue decks), but a couple of them are very strong? What if those weaker decks drag down the average? Are we to then jump to the conclusion about every single deck in that group? Can we say correctly that every deck in that list is at s disadvantage, and that you are handicapping yourself regardless of which deck in that list you actually play? I'm sorry, but if Maverick happens to suck against blue decks, that has zero bearing on whether or not Elves is bad vs blue decks. But in this experiment, they share the same fate.
The very premise is silly and flawed. It's not like I have to choose blue or not blue and then be handed a random deck! If that were the case, this would be a very useful experiment!
But not necessarily up above the 50% mark! I tried to explain this logically to no avail, so a couple post up I've provided an airtight mathematical counterexample. You can't argue against the facts of math my friend!
Supremacy 2020 is the modern era game of nuclear brinksmanship! My blog:
https://fieldmarshalshandbook.wordpress.com
You can play Lands.dec in EDH too! My primer:
http://www.mtgsalvation.com/forums/t...lara-lands-dec
The dominance of Brainstorm/FoW can be seen not in these numbers but by looking at the decklists of the top non-blue decks. It's not surprising that decks that play Thalia, Chalice, Abrupt Decay or mainboard Red Elemental Blasts would have good matchups against blue decks - the decks are designed to have good matchups against blue decks. This comes at a price of being weaker against random non-blue decks. And most of them - outside of those that can t1 Chalice - are weak to fast combo because they can't play FoW.
Elves kinda deserves its own category within non-blue, because with its draw engines and tutors it sees even more of its deck than a 12 cantrip blue deck - so it doesn't suffer from the non-blue 'I don't get to play Brainstorm' problem, just the 'I don't get to play FoW' problem. You don't need to play Brainstorm when you can draw more than 3 cards and put none back. Likewise lands is also very consistent with 8 tutors and Loam as a draw engine, though it has to play a deck with a lot of limitations in order to get those effects.
This.
What if decks like Elves, Aggro Loam, and Lands are good against blue decks, but fair creature decks like Maverick, Jund, Goblins, Dead Guy, etc are suckng hind tit?
Most people who make that claim will concede it doesn't apply to lands or other blue-less DTBs of the day. The issue is all about creature based fair decks. Some folks seem to think a diverse pool of competetive midrange decks is the fundemental building block of any healthy meta.
Supremacy 2020 is the modern era game of nuclear brinksmanship! My blog:
https://fieldmarshalshandbook.wordpress.com
You can play Lands.dec in EDH too! My primer:
http://www.mtgsalvation.com/forums/t...lara-lands-dec
I think you over use the word mid-range personally, and always want to apply qualifiers to control decks that aren't Miracles to say they aren't "dedicated control" decks because... IDK.
The common thread between these Non-Blue decks is that they abuse some crazy CA engine that saw the printing years ago before Wizards decided other non-blue colors shouldn't be given CA engines... or given CA engines extremely rarely.
They print CA and cantrips in the other colors - stuff like Collected Company, Abzan Charm, various planeswalkers, Ancient Stirrings, Tassigur (can be used in BG), Kolaghan's Command,Painful Truths, and Demonic Pact are in standard.
Blue legacy decks abuse old cards as much as the other colors do. (If not more.)
Not denying that.
I meant in the context of legacy, since you know we are in the Legacy Format Discussion forum.
I guess I should have said Wizards rarely prints a non-blue CA engine that is legacy playable... since it wasn't clear.
Tassigur and Kolaghan's Command are fringe playable, but the rest are not really.
It's a combination of card selection and card advantage. Elves runs Glimpse and Symbiot + Visionary for CA, but GSZ and NO are card selection tools. Lands gets huge CA from Loam, but relies on CR and Gamble as card selection tools.
I think what's worth noting is that these decks are very synergistic and specialised, attacking the game from less common angle. This is why they can get away with narrow engines like Loam and Glimpse; as well as tutors which incur -ve CA like Gamble, NO, and CR. Even MUD and Post decks (my pet theory being that the held down by unpopularity) are specialised to run CotV (virtual CA), and also approach the game very differently.
The issue is pretty much confined to fair decks. If you want to build a fair creature based decks built around individual card quality instead of specific synergies, it seems like no amount of Bobs, GSZs, Hymns, KotRs, etc can compete with the superior card selection (and access to FoW) that you get with blue cantrips*. These are the types of decks which I think people are talking about when they say you are handicapping yourself if you don't play blue. So I agree with iatee that lumping all non-blue decks together won't provide a very meaningful conclusion.
*Aggro Loam is another deck I think suffers from lack of player support - I think CotV main simply doesn't appeal to the demographic which wants to play fair decks. Even D&T isn't considered "fair" by a lot of people, probably because of the strong prison element. But even if we agree these decks are both competitive and fair; it's still very narrow compared to your competitive options for a fair deck if you include blue.
This is a bit off topic, but that's probably my fault.
I certainly use 'midrange' as a wide category for all (non-tempo) decks with a good mix of aggression and disruption.
What I call dedicated control has been slowly disappearing from MTG - due in part to a power-creep on creatures, but also probably a reduction of player interest. Old decks like Stacks, Pox, MUC have goon to shit because they can't keep up anymore. Legacy Miracles is becoming more aggressive with Mentor. Even in Vintage MUD decks have become more aggressive with the advent of Lodestone Golem. Even Mana Drain decks are less common than in the past.
So by today's standards, Deathblade and Shardless actually look like control decks, but few will deny that they are accurately describable as midrange.
Do you think the category is too broad, and yo0u would like to distinguish between more aggressive midrange decks vs more controlling midrange decks. That's fair. But by the same logic I want to distinguish between control decks which are more aggressive vs control decks which are more dedicated to the answers and disruption with very little in the way of potential aggression. I simply don't understand your reluctance to acknowledge what is a very fair and widely recognised distinction. distinction.
Supremacy 2020 is the modern era game of nuclear brinksmanship! My blog:
https://fieldmarshalshandbook.wordpress.com
You can play Lands.dec in EDH too! My primer:
http://www.mtgsalvation.com/forums/t...lara-lands-dec
Have you even played with Painful Truths?
At Tales of Adventure this weekend I saw at least 5 different people cast that card, in both blue and non-blue decks. It's still too early to tell what impact it will have on the format, but that's over 10 percent of the field and 5 more people than I saw casting Kolaghan's or Tasigur. Gurmag sees more play in Legacy, because the ability to break a Goyf standoff and the inability to be countered by Counterbalance are worth the extra mana.
I Top 4'd the second Source tournament with an Abzan deck that oozed card advantage. Bob, SFM, Souls, Bitterblossom, Library, Truths, etc. There were long games that I had less cards in my library than my blue cantripping opponents.
I think your pretty much spot on on this assessment. Can't find much to disagree with, the only thing I'll say is that I think Post and to some extent MUD decks are held back here by the existence of Wasteland and Bloodmoon (& to a lesser degree Magus of the Moon). This is just my pet theory though.
Really this is just a philosophical difference about what control and aggro means. IMO control and aggression aren't the opposite ends of a spectrum. Who's the Beat down and all that Jazz.
I personally find pretty meaningless to split hairs in this way, the difference between a more aggressive and more controlling mid-range deck is often minor and pretty fluid, and can rapidly change with a few sideboard cards. The real truth to magic is that you need to know your role in the match... even if you are playing burn you aren't always the aggressive deck.
I know this isn't all that helpful, but IMO magic players split hairs to much about what decks are... and get too hung up on that... and forget that the role you play changes from match to match. I've stolen a handful of matches from Burn when they didn't realize they were actually the control deck, and I've stolen a handful of matches from Miracles when they didn't realize they were actually the aggro deck.
I think that's one of my two favourite articles on MTG strategy (along side Tempo and Card Advantage). But the point of that article is not that control/aggro labels are meaningless or useless. Rather that it's a mistake to take those labels as ironclad guides for playing those decks.
We call a deck aggro or control for two reasons:
An example!e is Burn. Burn is almost always on the beat down plan in almost every MU. When Burn is forced to play control (using creatures to block and/or spells to kill threats), this usually is bad news for the Burn player.
- In the meta, the deck expects to take on a particular role (control or beatdown) the majority of it's MUs.
- The deck is designed to play that side of a MU. When forced to take the opposite role, the deck is going "against the grain".
Similarly, in a meta with little or no dedicetd control your most controlling aggro/control decks will almost always take the control side of a MU, and hence people start to think of them as simply being control decks.i
This is exactly my point! Aggro-control hybrid decks are designed to be much more fluid than linear aggro or dedicated control. The can easily "change gears" between control or beatdown as the situation demands.
It's good that we have decks like this, and I can understand their appeal. But this doesn't mean that pure aggro and hard control styles are invalid as concepts! If you can't see that Legacy has moved a way from more polarised aggro or control strategies towards more moderate and fluid midrange and tempo strategies, I don't think you are looking far enough back.
Edit:
If we define control vs aggression as the likelihood (or frequency) that your deck will end up going control vs beat down in practise, then they are very much opposite ends of a spectrum!
Of course this is meta-dependant, but that's my entire point! D&T, Shardless, etc, are considered control decks because they are among the least aggressive decks in the meta. If we had more "dedicated control" (Pox, draw/go, Stacks, etc), those decks would be "going beatdown" a lot more often.
Last edited by Crimhead; 01-12-2016 at 03:40 AM.
Supremacy 2020 is the modern era game of nuclear brinksmanship! My blog:
https://fieldmarshalshandbook.wordpress.com
You can play Lands.dec in EDH too! My primer:
http://www.mtgsalvation.com/forums/t...lara-lands-dec
There are currently 1 users browsing this thread. (0 members and 1 guests)
Reply With Quote