Decks to Beat (2020-21) Discussion Thread

Oh how a month changes things! With some help from a busted new mechanic, of course.

https://i.imgur.com/LvOJ3xP.png
https://i.imgur.com/cRBq63e.png

Again, suppressed results from lack of paper event, and so on. Computationally, we should be about 41 points. However, honestly, it doesn't seem, to me, that Omni is really dominating or doing much. So, I am going to cut to the top 8.

4C Control (Snowko, Czech Pile and more!)
Grixis Tempo
RUG Delver
Eldrazi Stompy
ANT
Patriot Delver
BUG Delver
Bomberman

So, frame is intact, I will work throughout the day to update and move the threads.

Whoopsy-daisy, I forgot what day it was. The May update will come tomorrow morning, a little too late for me today to start it now.

How do the results get categorized into archetypes?

Looking back at last month, I assume that most of "Grixis Pyromancer", "Patriot Delver" and "BUG Delver" are the three variations of Lurrus Delver and "Bomberman" is mostly Zirda Combo.

4C is interesting. I guess some of it would be Yorion Control, some would be regular companion-less Snowko, and some would be Miracles (including Lurrus Miracles), but it would be interesting to see the breakdown.

Quote:

---

Originally Posted by FTW

How do the results get categorized into archetypes?

Looking back at last month, I assume that most of "Grixis Pyromancer", "Patriot Delver" and "BUG Delver" are the three variations of Lurrus Delver and "Bomberman" is mostly Zirda Combo.

4C is interesting. I guess some of it would be Yorion Control, some would be regular companion-less Snowko, and some would be Miracles (including Lurrus Miracles), but it would be interesting to see the breakdown.

---

Well, yeah, the results are still rather screwy, because Lurrus decks are still clouding results. The difference between Miracles and 4c is clear, if it has Miracle cards in there, it is Miracles and not 4c. 4c is what is left after that.

https://i.imgur.com/9PiWBSI.png
https://i.imgur.com/7bu9G92.png

The question is, what now? While Grixis and TA results were mainly fueled by the now banned Lurrus, even post post ban, Grixis has put up results. TA has not. So, I am inclined to cut TA. Our computational level should be about 37 points, but, like has been usual, our suppressed results has not given us much divergence. RUG Delver is in, along with Miracles, Infect, 4c and Lands. Now, "Fish" is actually Esper Vial and that clearly has enough points.

Also, I am inclined to exclude Gyruda for the moment, until the deck demonstrates that it can win through the Companion rule change.

RUG Delver
Grixis Delver
4c Control
Miracles
Infect
R/G Lands

We don't have a thread for a real Esper Vial deck. I could reuse Bahra's Spicy Popcorn thread, but that hardly is the same deck. Might just have to hold off on that deck while I work up a thread.

Questions? Comments? Complaints? I will move threads around later today.

Well, I took a peak in and, wow, the data for June is ragged and sparse. I am inclined to not even bother drawing anything from it. Unless anyone has some compelling reason to, I think I will just hold off and update for July on August 7th.

Quote:

---

Originally Posted by H

Well, I took a peak in and, wow, the data for June is ragged and sparse. I am inclined to not even bother drawing anything from it. Unless anyone has some compelling reason to, I think I will just hold off and update for July on August 7th.

---

Makes sense. It's a real muddle, I was looking at mtgo goldfish, and they are back to throwing in a sprinkle of f-t-f and published mtgo results, so there's not even a clear picture (well, as clear as possible) of what's happening online.

Alright, honestly, I dropped the ball. Data was ragged, data was low and frankly I forgot a couple times.

But, new year, new process I am going to try. Since we are still in a low-data era and likely to stay in one, I am going to do a bi-monthly analysis for this year.

So, with this new process (I had to manually merge the data), here is what the cutoff looks like with November and December 2020's data:

1º	4c Control	102		24º	Infect	22
2º	Death and Taxes	95		25º	Lands	21
3º	Other	83		26º	Reanimator	21
4º	Temur Delver	68		27º	Eldrazi Stompy	20
5º	Hogaak	66		28º	Temur	19
6º	Elves	59		29º	4c Loam	17
7º	Red Prison	56		30º	Turbo Depths	17
8º	Goblins	52		31º	Jeskai	16
9º	Grixis Delver	40		32º	Merfolks	15
10º	Ninjas	40		33º	Dredge	14
11º	All Spells	39		34º	UBx Shadow	14
12º	Aluren	38		35º	Painters	13
13º	OmniTell	35		36º	Pox	13
14º	Cephalid Breakfast	34		37º	Sultai Delver	8
15º	Burn	33		38º	Deadguy Ale	7
16º	Bant	32		39º	Miracles	7
17º	12 Post	31		40º	Ad Nauseam Tendrils	6
18º	Mill	31		41º	Bomberman	6
19º	Maverick	29		42º	Enchantress	6
20º	Stoneblade	28		43º	Esper Vial	6
21º	Doomsday	27		44º	Grixis Control	6
22º	Izzet Delver	25		45º	Lutri Deck	6
23º	Urza	23			Total:	1346
					Cutoff:	47.11

So, this means we have 7 Decks to Beat:
4C Control
Death and Taxes
RUG Delver
Hogaak (Placeholder thread, seems we don't actually have a dedicated Hogaak thread?)
Elves
Dragon/Red/Moon Stompy
Goblins

I'll move the thread a little later today. Any comments welcome. Also if anyone knows a more appropriate thread for Hogaak, let me know.

Whoops, forgot what month it was. I should have an update out this week, with the combined data for January and February.

OK, here we go, the combined data for January and February of 2021. Obviously with Valki getting hit with the Cascade nerf-bat, I did not consider it, so we will take the other top 8 decks:

1º	Temur Delver	86		27º	Turbo Depths	25
2º	Ninjas	77		28º	Sultai	22
3º	Death and Taxes	70		29º	Merfolks	21
4º	All Spells	61		30º	Doomsday	19
5º	Valki Cascade	61		31º	Mono Black	15
6º	Elves	57		32º	Urza	15
7º	Aluren	56		33º	Bomberman	14
8º	OmniTell	54		34º	Mill	14
9º	Maverick	50		35º	Hogaak	13
10º	4c Control	49		36º	Mono Blue	13
11º	Jeskai	47		37º	Slivers	13
12º	Dredge	45		38º	Buried Phoenix	9
13º	Stoneblade	45		39º	Deadguy Ale	9
14º	Other	42		40º	Tin Fins	9
15º	12 Post	40		41º	Sultai Delver	8
16º	Burn	40		42º	UBx Shadow	8
17º	Bant	39		43º	Esper Vial	7
18º	Turbo Depths	38		44º	Food Chain	7
19º	Infect	36		45º	Sneak and Show	7
20º	Eldrazi Stompy	34		46º	The Epic Storm	7
21º	Red Prison	34		47º	Jund	6
22º	Cephalid Breakfast	33		48º	Landstill	6
23º	Izzet Delver	33		49º	Mono Black Aggro	6
24º	Goblins	30		50º	Pox	6
25º	Temur	30		51º	Tibalt Cascade	6
26º	Reanimator	27

Our cut-off here is 51 points, meaning that Maverick just makes it and 4C Control just misses.
RUG Delver
Ninjas!
Death & Taxes
Oops, All Spells!
Elves!
Aluren
Omnitell
Maverick

As always, I welcome any comments, questions or concerns. Interesting results that, having not really looked into the data, would not quite have expected.

Quote:

---

Originally Posted by H

1º	Temur Delver	86		27º	Turbo Depths	25
2º	Ninjas	77		28º	Sultai	22
3º	Death and Taxes	70		29º	Merfolks	21
4º	All Spells	61		30º	Doomsday	19
5º	Valki Cascade	61		31º	Mono Black	15
6º	Elves	57		32º	Urza	15
7º	Aluren	56		33º	Bomberman	14
8º	OmniTell	54		34º	Mill	14
9º	Maverick	50		35º	Hogaak	13
10º	4c Control	49		36º	Mono Blue	13
11º	Jeskai	47		37º	Slivers	13
12º	Dredge	45		38º	Buried Phoenix	9
13º	Stoneblade	45		39º	Deadguy Ale	9
14º	Other	42		40º	Tin Fins	9
15º	12 Post	40		41º	Sultai Delver	8
16º	Burn	40		42º	UBx Shadow	8
17º	Bant	39		43º	Esper Vial	7
18º	Turbo Depths	38		44º	Food Chain	7
19º	Infect	36		45º	Sneak and Show	7
20º	Eldrazi Stompy	34		46º	The Epic Storm	7
21º	Red Prison	34		47º	Jund	6
22º	Cephalid Breakfast	33		48º	Landstill	6
23º	Izzet Delver	33		49º	Mono Black Aggro	6
24º	Goblins	30		50º	Pox	6
25º	Temur	30		51º	Tibalt Cascade	6
26º	Reanimator	27

---

This is the greatest goddamned thing I've ever seen.

Quote:

---

Originally Posted by H

TABLE

---

Turbo Depths appears twice --- 18th place with 38 points and 27th place with 25 points.
Presumably, either these are meant to be two different decks or Turbo Depths actually has 63 points and should be higher?

Edit: There's also Valki Cascade at 5th place and Tibalt Cascade at the bottom of the table. Same deck?

This article is an excellent discussion of control in legacy right now.

Post MH2 win rates are varying wildly challenge to challenge. I think delver can be contained by Bant miracles, which got significantly powered up by endurance and prismatic ending.

I think we are in a cyclical A. blue tempo (rug/ur delver) -> B. blue control (Bant) -> C. blue a+b (s&s/doomsday) -> A. tempo meta. Which is pretty healthy imo.

Fair non-blue is doing quite well too in the B position (maverick/lands/elves) or c position (titanpost)

Do people think I am correct? Do people like this or prefer brewing to beat an obvious best deck? Which deck can take advantage of this the best? How should one construct your sideboard?

For example, curses is very favoured against C, very unfavoured vs A and around even vs B. Should I make it less bad vs A? Or concentrate flex’s on B. (Urza’s saga/choke/other control hate) and hope to delver is getting endurance’d to death?

Whoops, I am real late on this. Kind of slammed here at work though, so I think I will just do a 4 month recap in July, catching all of March, April, May and June. Also, I'll try not to miss anything on the merged data.

March, April, May, June 2021 merged:

1º	Izzet Delver	235		29º	Tin Fins	37
2º	Death and Taxes	200		30º	UBx Shadow	31
3º	Elves	189		31º	Eldrazi Stompy	30
4º	Ninjas	150		32º	4c Loam	29
5º	Turbo Depths	121		33º	Doomsday	29
6º	Reanimator	90		34º	Landstill	28
7º	Miracles	85		35º	Esper Vial	26
8º	OmniTell	85		36º	Pox	23
9º	Maverick	68		37º	Sultai	23
10º	Other	66		38º	Grixis Delver	22
11º	Stoneblade	66		39º	The Epic Storm	22
12º	Lands	65		40º	Nic Fit	17
13º	Sneak and Show	65		41º	12 Post	16
14º	Temur Delver	64		42º	Humans	16
15º	Food Chain	57		43º	Mill	16
16º	Infect	54		44º	MUD	15
17º	Burn	51		45º	Esper	14
18º	Dredge	50		46º	Slivers	14
19º	Painters	49		47º	Bomberman	13
20º	All Spells	46		48º	Deadguy Ale	13
21º	4c Control	45		49º	Jeskai	11
22º	Ad Nauseam Tendrils	41		50º	Affinity	9
23º	Goblins	41		51º	Jund	8
24º	Hogaak	40		52º	Buried Phoenix	7
25º	Cephalid Breakfast	39		53º	Enchantress	7
26º	Aluren	38		54º	Merfolks	6
27º	Red Prison	38		55º	Mono Black	6
28º	Grixis Control	37

Oh boy, that was a good bit a data. Seemed fairly clear, aside the nebulous category of "Bant" which really were all Miracles decks, that just happen to run Green. So, with that merged, Miracles makes it (which makes perfect sense, really). Clean cut off too, 90.5 points, giving us a clear top 8. Sorry Maverick, :frown: (still love you).

UR Delver
Death & Taxes
Elves!
Ninjas
Turbo Depths
BR Reanimator
Miracles
Omnitell

Is GW depths counted within Turbo Depths?

Quote:

---

Originally Posted by Purple Blood

Is GW depths counted within Turbo Depths?

---

Yeah, Turbo Depths is basically just a catch-all for all Depths decks, unfortunately. But we don't even have a thread for GW Depths here anway.

This is a pretty good summary of the format.

I believe in paper decks like food chain and aluren are very good and should be grouped with 4c control

Quote:

---

Originally Posted by Reeplcheep

This is a pretty good summary of the format.

I believe in paper decks like food chain and aluren are very good and should be grouped with 4c control

---

It's been pretty hard to get really good data lately, TCDecks data is not very reliable, events are missing there and sometimes the classification is off.

I am more than happy to reconsider how we do the Decks To Beat here, but what should we consider as a process? It's not clear to me how to "extract" it from the Data Collection program exactly. Any ideas? Or should we just move to a purely subjective system?

Yah that was more of a subjective take, but since it was from one of the better grinders I thought it was wor th mentioning.

If you are fine with the data collection project data you would just rank all the decks by the winrate CI min, which naturally takes into account sample size. Or if you believe tiers are popularity not power just rank it by play percentage. The time period could be anything you want.