All B/R update speculation.

[Lemnear]

Judging by this :

http://mtgtop8.com/format?f=LE&meta=72

Its actually 15%.

If you include online, it gets slightly worse, it goes up to 18%:

http://mtgtop8.com/format?f=LE&meta=39

BUT, look here for the non-online world in July :

http://www.tcdecks.net/metagame.php?...y&fecha=2016-7

I cannot get a pure online-online Metagame share for Miracles but, it had 36 Top4 placings out of 55 events in the last 2 months.

[Jain_Mor]

I can only look at the metagame breakdowns delivered for the bigger tournaments like GPs or whatever data websites like MtgTop8 or others deliver for live tournaments (with the later not providing full breakdowns in most cases) and use these data to get the most accurate metagame picture.I excluded Online data, as the Miracles numbers in terms of metagame share and T8 are even higher than in Paper, due to the decks ability to perform well against fringe nonsense popping up online and the tournaments only being 4-rounder, so I was only looking at the mayor live tournaments which average in 16% metagame share for miracles from the data, I was able to collect from the websites. The Top8 percentage aka average T8 placings per tournament was calculated with the T8 results listed on the websites, divided by the number of different tournaments I had metagame data of (this includes tournaments with only Day2 data as you assumed correctly, lacking full data most of the times), which (given I remember correctly now) were 45 T8 spots. In regards to your question added via edit: That is what the data is saying. I was free to set that 16% in mayor live tournaments in relation with the average T8 spots. Take it with a grain of salt, as the often lacking T16/32 as well as Day1 data paint a pretty extreme picture and I am sure it would be less drastic if I had more and better data. The lacking T16/32 data in many cases also mean that I can't set Miracles into relation to Storm or Grixis as the two decks have a much lower metagame share. Storm for example had 15 T8 and 16 T16 placings in 36 tournaments with a 6% metagame share according to MtgTop8 (didn't compare it to other websites).

Trying to decipher this is difficult.

So you used the "Major Events decks (50+ players)" option on MTG Top 8? Isn't that derived from the top placing decks though, not necessarily top 32 or top 64? Or did you comb through the events to find ones that did give top 32+ deck names?

Then this data set gave you 15-16% metagame for Miracles for top 32+ decks? (not including the top 8?)

Then you looked at how many times Miracles top 8'd those events and divided it by the number of events? To give you the average Top 8 representation of Miracles in a percentage form. And then you compared that percentage with the metagame percentage of Miracles (outside the top 8?) and found the percentage to be larger inside the top 8 rather than outside? "21.6%" compared to "16.0%" ie a 35% increase in representation.

So I guess the question is can you share your collection of data, or explain how you acquired it more clearly so we can replicate it?

[Crimhead]

...And then you compared that percentage with the metagame percentage of Miracles (outside the top 8?) and found the percentage to be larger inside the top 8 rather than outside? "21.6%" compared to "16.0%" ie a 35% increase in representation.

So I guess the question is can you share your collection of data, or explain how you acquired it more clearly so we can replicate it?

And by "outside the top8" we're including the whole event (at least day two), or just the top 16/32?

This is the most compelling evidence I've been presented with for Miracles being OP. Of course every tier one deck will see a larger percentage of top placements compared to its entry percentage (by definition), but a 35% increase seems like a lot - especially for a deck that's been the elephant in the room for so long.

[btm10]

An interesting follow up question would be to see if Miracles is even more over represented in longer events. While it's somewhat challenging to disentangle the pilot skill factor for large events (since we should expect stronger players to disproportionately be on Miracles when the stakes are higher), if we saw a larger enrichment of Miracles at the top of the field as event length grew, that would also point towards a Top ban as thr most logical remedy since you could reasonably argue that it was increased consistency rather than being overpowered that was the real culprit.

[dte]

To have an easy look, you can just look at only the top8 recorded and see the win% of miracle. Data is really easy to collect then.

Taking the "live tournament last two months" on mtgtop8:
http://mtgtop8.com/format?f=LE&meta=72

It gives you 51.67 rate win percentage for miracle. Which is good but far from busted.

The amount of data is not huge, and the method is not suitable for any deck which is not over represented, but for miracle I believe it is an ok method.
And you avoid discussion on player skill by considering only people in top8 should be on average not that bad.

[Lemnear]

Trying to decipher this is difficult.

So you used the "Major Events decks (50+ players)" option on MTG Top 8? Isn't that derived from the top placing decks though, not necessarily top 32 or top 64? Or did you comb through the events to find ones that did give top 32+ deck names?

Then this data set gave you 15-16% metagame for Miracles for top 32+ decks? (not including the top 8?)

Then you looked at how many times Miracles top 8'd those events and divided it by the number of events? To give you the average Top 8 representation of Miracles in a percentage form. And then you compared that percentage with the metagame percentage of Miracles (outside the top 8?) and found the percentage to be larger inside the top 8 rather than outside? "21.6%" compared to "16.0%" ie a 35% increase in representation.

So I guess the question is can you share your collection of data, or explain how you acquired it more clearly so we can replicate it?

As I said, not just MtgTop8 but also TC decks, TCG Player and the like to get more decks to compare for the limited timeframe. Then I compared the metagame share these sites list with Day2/Day1 data I could find an ended with the ~16% (Some websites list Miracles at 15,3% and factoring paper & online its about 18%) for Miracles in the field in average. I did not look for T32 data as the number of events we have such complete data is marginal, so I felt forced to just look at T8 to get a decent amount of events around the globe (eliminating local factors like Miracles being even more popular in Europe than elsewhere). At this point we can compare the number of relevant events fired with the number of Miracles placing in Top8s during the same time. In the end, I compared the T8 spot percentage with the average metagame representation (because I obviously don't have the data for each specific event), which marks up to (depending on your parameters like tournament size and structure aka online/offline/both) 35% higher T8 representation than the metagame presence would imply (I took the extreme data here, just like I did for the metagame share (16% instead of the 18%) to get a peak result). The factor [T8 Percentage/metagame share] is the key component I use to define if a deck over-/underperforms in a metagame. I hope that was helpful

[Crimhead]

C we saw a larger enrichment of Miracles at the top of the field as event length grew, that would also point towards a Top ban as thr most logical remedy since you could reasonably argue that it was increased consistency rather than being overpowered that was the real culprit.

Consistency means nothing without power. 60 Plains is consistent, but it consistently does nothing.

Consitency is only valuable when a deck can consistently preform at a high power level; in which case it's not fair to blame Top rather than some of the most powerful cards to which Top allows consistent access.

Taking the "live tournament last two months" on mtgtop8:
http://mtgtop8.com/format?f=LE&meta=72

It gives you 51.67 rate win percentage for miracle. Which is good but far from busted...

...And you avoid discussion on player skill by considering only people in top8 should be on average not that bad.

This is what I've always believed (and it strongly suggests Miracles is top dog because of sign-up density more than the innate quality of the deck. Since you are looking at the just top8s (good players), the argument is more compelling.

How does this reconcile with the 35% increase from day two to the finals? It seems like Miracles is over preforming early in the events, but much less so when the rogue decks and scrubs have been weeded out. That's assuming the statistics you guys have provided are accurate and representative.

[Lemnear]

It seems like Miracles is over preforming early in the events, but much less so when the scrubs have been weeded out. That's assuming the statistics you guys have provided are accurate and representative.

This is an absolute realistic scenario of Miracles not stumbling over random crap or the infamous "Round 1 Kid" in the early rounds due to its balanced construction able to beat Combo AND aggro with the default setup.

[Jain_Mor]

To have an easy look, you can just look at only the top8 recorded and see the win% of miracle. Data is really easy to collect then.

Taking the "live tournament last two months" on mtgtop8:
http://mtgtop8.com/format?f=LE&meta=72

It gives you 51.67 rate win percentage for miracle. Which is good but far from busted.

The amount of data is not huge, and the method is not suitable for any deck which is not over represented, but for miracle I believe it is an ok method.
And you avoid discussion on player skill by considering only people in top8 should be on average not that bad.

Excuse my ignorance, how did you get the win rate number? Is it something you calculated, if so how? Or is it somewhere located on mtgtop8 that I can't seem to find?

@Lemnear, could you please include some paragraphs in the future, would help a lot. So you're saying that you included as much live tournament data you could of any event size (whether it is just a top 8 listing or top16 or 32 or etc) to find an average miracles presence of ~16%. Then you looked at the number of miracles in the top 8 of those same events? You essentially found that in the top 8, miracles is ~22% of the format and including top 8 and all the other data miracles is ~16% of the format?

Sorry if you feel like you've explained yourself clearly, but I'm not sure I understand what you've done to the point I could replicate your efforts if I liked.

[Crimhead]

To have an easy look, you can just look at only the top8 recorded and see the win% of miracle. Data is really easy to collect then.

OMFG why I never think to do this? You're a genius!

Excuse my ignorance, how did you get the win rate number? Is it something you calculated, if so how?

This is easy. A deck's rank in the top8 tells you exactly how many matches it has won and how many it has lost. 5th-8th place have a 0-1 record (0% win-rate). 3rd & 4th are 1-1 (50%). 2nd is 2-1 (66.6%), and 1st is 3-0 (100%).

You can't just average the win rates from each top8 though; count every match. eg, A deck that makes 3rd place (50%) one week and 1st (100%) the next has an overall 80% win-rate - not a 75% win rate!

[Richard Cheese]

Petition to make this the official card of this thread:

[ahg113]

Petition to make this the official card of this thread:

Good suggestion, how does the reveal mechanic work irl?

[btm10]

Consistency means nothing without power. 60 Plains is consistent, but it consistently does nothing.

Consitency is only valuable when a deck can consistently preform at a high power level; in which case it's not fair to blame Top rather than some of the most powerful cards to which Top allows consistent access.

I assumed the need for a sufficient baseline power level was obvious since we're discussing a tier 1 deck. There's nothing that precludes a deck from being dominant purely because it loses to variance less than the competition, and an advantage like that would make itself felt more the longer a tournament goes. One can argue whether that consistency is itself a sort of inappropriate power, but it's definitely not the sort of power that gets cards like Necro banned.

[Sidneyious]

What y'all have to understand is this.
Thoes are big events, they likely will take the deck that has the best chance for them to top.

Local events are never like this and are never recorded and does not makeup the total share of decks at events.

My legacy night is nothing like anything I see here that y'all complain about.

Many that want bans are people that(99% sure) mtgo all day and with how mtgo is I'm not surprised at all that people play the same deck all the time.

I think legacy is just fine, it's that mtgo has caused a huge problem.

Bigger than when netdecking first started and xerox became a thing.

[Crimhead]

One can argue whether that consistency is itself a sort of inappropriate power, but it's definitely not the sort of power that gets cards like Necro banned.

I'm saying that you can't measure a deck's power independently from the deck's consisgency. Or to the extent that you can, it's irrelevant because the favourably of a match-up is depencan on both; and that a deck's positioning is a function of the favourability of each possible match-up times the probability of that pairing.

[Richard Cheese]

Good suggestion, how does the reveal mechanic work irl?

Reveal - Spend time and effort arguing about something on a forum where nobody with the power to change the situation will ever see it.
Kicker - Do math to support your point.

[Lemnear]

It gives you 51.67 rate win percentage for miracle. Which is good but far from busted.

An 51.67% winrate in the Top8 against D&T, Lands, Infect and Grixis as the other DtBs (in the last months), which we consider more or less bad matchup for Miracles, isn't something you would call busted?

@Lemnear, could you please include some paragraphs in the future, would help a lot. So you're saying that you included as much live tournament data you could of any event size (whether it is just a top 8 listing or top16 or 32 or etc) to find an average miracles presence of ~16%.

No, i took metagame numbers from the websites and from tournaments which have total breakdowns for Day1 and/or Day2. Single tournaments have no relevance for me as I don't know the metagame share for these; i just asume the numbers don't differ much from the bigger, MORE relevant tournaments

Then you looked at the number of miracles in the top 8 of those same events?

I took T8 of a much larger selection of T8s to get a decent sample size, I would be unable to get if I would just look at the few complete datasets like GPs, because otherwise such extreme examples like GP Columbus would weight far too much. I have to set such tournaments into relation with these w/o Miracles in T8

You essentially found that in the top 8, miracles is ~22% of the format and including top 8 and all the other data miracles is ~16% of the format?

It peaks at ~21% depending on the factors of your selection. You get different numbers depending on if you look at tournaments with 40/50/100/150/etc players and depending on if you factor continental differences or just take the US metagame as representative. The 16% is the statistic middle for paper tournaments as explained before

[Jain_Mor]

Right I see, that makes a lot more sense. Thank you, and good idea.
Also good point regarding ~52% win rate IN the top 8 against other tier 1 decks that are meant to be favourable against miracles that have been gunning for it for the past 2+ years

@Sidneyious, that seems like a really bad point to make. I'm sure some peoples local FNM events were fine during the cawblade era of standard, doesnt mean that standard was fine. So you're basically complaining that people who play the format more than you, and play it with much wider experience/meta/deck selection than you are complaining about the format as a whole, because your local bubble of the meta seems fine?

[nedleeds]

Survival can fit in pretty much every strategy in the combo-aggro-control-prison spectrum.

Well it's fine, then isn't it? 80% format penetration is not a problem at all if every deck ever plays the card while maintaining strategic diversity. If anything, the card would become a skill-testing pillar of the format.

Doing the lords work.

[Sidneyious]

One large meta does not equal the entire non recorded meta.