Data Analysis: The Top Modern Decks & Players, End of 2021

Aliquanto

An old archetype regains some spotlight in a new shape, in no small part thanks to a player who took down multiple Magic Online events in a row. Speaking of which, the end of the year is the perfect time to check in on players' performance since January. Here's the final Modern metagame update for 2021!


shadow

Here are the Magic Online events included in today's metagame analysis.

Modern Challenge 2021-11-20
Modern Challenge 2021-11-21
Modern Premier 2021-11-21
Modern Challenge 2021-11-27
Modern Challenge 2021-11-28
Modern Challenge 2021-12-04
Modern Challenge 2021-12-05
Modern Showcase Challenge 2021-12-11
Modern Challenge 2021-12-12
Modern Preliminary 2021-11-15
Modern Preliminary 2021-11-16
Modern Preliminary 2021-11-17
Modern Preliminary 2021-11-18
Modern Preliminary 2021-11-19
Modern Preliminary 2021-11-22
Modern Preliminary 2021-11-23
Modern Preliminary 2021-11-24
Modern Preliminary 2021-11-25
Modern Preliminary 2021-11-26
Modern Preliminary 2021-11-29
Modern Preliminary 2021-11-30
Modern Preliminary 2021-12-01
Modern Preliminary 2021-12-02
Modern Preliminary 2021-12-03
Modern Preliminary 2021-12-06
Modern Preliminary 2021-12-07
Modern Preliminary 2021-12-08
Modern Preliminary 2021-12-09
Modern Preliminary 2021-12-10

Here are some quick statistics.

Number of decks in the data: 394
Number of different players in the data: 274
Number of exact archetypes in the data: 60
Number of rounds played in the data (with Top 8): 2630
Average number of Swiss rounds in the data: 6.36
Minimum number of Swiss rounds in the data: 4
Maximum number of Swiss rounds in the data: 9
Number of events in the data: 29

Archetype Presence

We should be used to the process by now. As always, we start with a look at the most represented archetypes among the top finishes: Challenges/Showcases/Premier/PTQ Top 32s plus Preliminary 3-1 and 4-0 results. They are weighted by the number of matches to make appearances in larger events more meaningful, which is relevant to determine the best decks later. This means, a deck that reached the finals of a nine-round event, adding three matches in the playoffs, has a weight of 9+3=12, whereas a Preliminary list comes at a weight of 4. Interestingly, the ranking is pretty much the same when you only take the number of appearances without weighing them.


Proportion of archetypes, out of 2,630 matches (threshold for Other: 52 matches)
Proportion of archetypes, out of 2,630 matches (threshold for Other: 52 matches)

This month, fewer archetypes made it to a presence above 2%, only twelve as opposed to November's sixteen, and no new deck entered the picture here. Jeskai Control, Living End, Mill, and BTL Scapeshift all fell below the threshold—though one could argue that Jeskai and Azorius Control should be summed together. Most decks remained quite stable from last month to this month, with the lesser represented decks falling in presence and the big two (Hammer Time and Izzet Murktide) rising further. Amulet Titan improved a little, pushing in above Azorius Control. Last, the WURG decks also diminished a little, leaving more space for the discard-based decks such as Jund Saga and, the highest rise this month, Grixis Shadow.

We already mentioned SoulStrong in the previous article when displaying a Grixis Shadow list, and he kept on achieving great results with the deck this month. He won a Modern Premier, then reached Top 4 in two Challenges in a row, then won two Challenges in a row … Following all that, he posted a guide to learn how to play the deck as well as he does. Here is his latest list, taken from yet another Challenge Top 4.


Archetype Win Rates

Now that we have the deck presence, let us check their win rates, since that is the other available metric we can use to measure their strength in the metagame.


95% confidence interval on the win rates of the most played archetypes
95% confidence interval on most represented archetypes' win rates (click pictures to view at full resolution)

As expected, the decks with the highest and lowest win rates also are among the least represented, meaning that their confidence interval is much larger. As usual, to prepare the tier list later (and because showing dozens of decks on a single graph is quite complicated) we only keep the decks with a presence above the average presence—which conveniently enough happens to be around 2% most of the time.

Do you think that so many win rates this much above 50% is abnormal? Don't worry, it was all planned. Or at least, it is to be expected. After all, as mentioned earlier, we only have data on the best-performing decks from Challenges/Preliminaries. So rather than the actual win rate, this is rather a metric of whether the deck tends to win tournaments or places at the lower end of the Top 32.

Compared to last month, the extreme values in win rate are a little bit lower for those decks, starting at 66% instead of 67.7%, and going up to 76.4% instead of 77.1%. On the other hand, the average win rate increased, going from 72.3% to 72.7%. To be fair, I don't know exactly how to interpret this change, but I already noticed previously that the win rate distribution evolves over time, and strictly comparing the win rate of a deck from one month to another will not be very meaningful. Instead, it should be more relevant to compare it relative to the average win rate, or its average positioning.

For instance, Jund Saga's win rate barely changed, going from 76.5% to 76.4%, but this equals a jump from third place last month to first place this month. It leads by almost a full point over the second one, WURG Blink at 75.5%, which actually happens to be one full point ahead of the third one, Grixis Shadow at 74.5%. On the other hand, last month, the three decks with the highest win rate were much closer together. We can also notice that most decks stuck much closer to the average win rate than how they did previously. Between that fact and the observation that the major decks are getting more traction compared to the minor decks, we could infer that the meta is focusing and reaching a stable state. This is to be confirmed next month.

The only exception is Belcher, whose win rate collapsed since last month, going from second place to last, with a large distance from the average both times. Simply put, as Sodeq would explain, the deck is about positioning and exploiting unprepared opposition. Now that it pushed back the four-colored decks, while tempo decks such as Grixis Shadow and Izzet Murktide thrive, it has more trouble to do well. Besides, people are now aware of the way the deck works, and are more likely to have a plan and/or specific cards to beat it. For instance, some Temur Footfalls lists adopted Leyline of Sanctity. Even though it was initially for the Burn matchup that proved tough for the cascade deck, it still happens to hit Belcher as collateral damage. Of course Belcher has ways to beat the hate, but against tempo decks with a fast clock, losing one or two turns in the race can easily mean a defeat.

Last, in general, you could say that fair decks performed much better than unfair ones this month. This might mean that people could work on tuning their decks for such matchups, such as Grixis Shadow, in the next few weeks, which could eventually lead to a resurgence of some unfair decks preying on fair decks. Green Tron is one option, or simply Belcher again if people forget it given enough time. Another possibility is Dredge, which happened to do better this month, as we see later.

Displaying Win Rate and Presence Together

Having gathered the data for both the presence and win rate, let us visualize it. At first glance, it seems that both Hammer Time and Izzet Murktide should be above the rest, followed by Grixis Shadow, and then a variety of midrange and control decks, with Amulet Titan in the mix.


Win rate by presence for the most represented archetypes
Win rate by presence for the most represented archetypes

Again, do not worry. I know you like to see what happens outside of the top tiers, so here is the complete picture (with tiers based on presence alone, the only metrics most other tier lists would use).


Win rate by presence for all archetypes
Win rate by presence for all archetypes (circle diameters based on number of different players)

I find it interesting because we can see that several decks are close to the mean, including three graveyard-based decks: Living End that fell off, Reanimator that's been stable at that position for some months now, and Dredge that rose recently. The latter most likely reappeared on the map in response to the number of grindy decks in the format that thought they could get away with two cards of graveyard hate.

Still, the diameter of the Dredge point is particularly small, because most of its results in the past few weeks come courtesy of to a single player: Violent_Outburst. He even reached the finals of a Modern Challenge with the following list.


If you remember Dredge from its past glory, you might notice that the decklist has evolved. Players run a lot of "rainbow lands" instead of fetch lands now, requiring them to give up on Life from the Loam (and reducing Conflagrate to a single copy). They also use the full four copies of Darkblast to handle Ragavan, Nimble Pilferer, plus new enablers, including the recent Otherworldly Gaze.

What about the lone point of BTL Scapeshift, the highest in terms of win rate? Its position is easy to explain: The deck only appeared in Preliminaries. Alas, we only get 3-1 and 4-0 results from Preliminaries, so any list coming from there either has a 75% or 100% win rate, hence the value close to 85% for that archetype.

From Data to Tier List

Now that we have the data on both metrics for the most represented decks (above the average presence), we can use the same normalization formula as before on both of them. Summing up gives us a value we can use to rank the decks.


Sum of the normalized win rate and presence for the most represented archetypes
Sum of the normalized win rate and presence for the most represented archetypes

Following that we generate the tier list by setting the average of those values as the threshold for tier 1.5, and then adding or subtracting standard deviations to change tiers.


tier list

Players of the Year

Looking at deck results is nice, but how about players? We can do that too. Win rate, presence in results, et cetera. All that is doable. Still, I chose not to use the win rate for the analysis this time, as I don't really want to highlight people that only spiked a 4-0 once in a Preliminary this year without doing anything else. So let us use a few metrics to find out who had the best results on Magic Online this year, depending on the type of event. Let us start with the number of Challenge (and Showcase/Premier/PTQ, the major events) Top 8s. Five players lead the dance.


PlayerTop 8 Finishes
Do0mSwitch10
Laplasjan9
Sodeq9
WaToO9
kanister9

What decks did they play the most? Do0mSwitch used Izzet Murktide four times, Laplasjan almost only ran Hammer Time, Sodek was mostly on Living End, WaToO ran control decks based in white and blue most of the time, as expected, and kanister used Four-Color Elementals/Blink in more than half of his matches.

If we check the entirety of the Top 32 appearances, we find the following five players on top.


PlayerTop 32 Finishes
CrusherBotBG28
WaToO27
Laplasjan26
kanister26
Darthkid25

CrusherBotBG exclusively plays Hammer Time. Darthkid ran a large panel of decks, even though he also played Hammer Time the most (six times).

In Preliminaries, we find some different players. This time we showcase six of them since two of them tied for fifth place.


Player3-1/4-0 Finishes
kanister28
HouseOfManaMTG22
McWinSauce20
Xwhale19
triosk17
xfile17

Here, kanister's decks are even more diverse, but the one he played the most was Amulet Titan (five times). This is also the deck HouseOfManaMTG almost exclusively plays. On the other hand, McWinSauce is more of a control player, his favorite variant being WURG Control. Xwhale also played a wide spread of decks, but the one he ran the most was by far Hammer Time. triosk seems to bring the balance back to control with BTL Scapeshift as deck of choice but also running a few others, whereas xfile also picked a lot of different archetypes yet mostly stuck to Burn.

Combining both Preliminaries and major events, here are the five players with the most appearances in the results.


PlayerPublished Finishes
kanister54
CrusherBotBG44
HouseOfManaMTG35
SoulStrong32
musasabi29

SoulStrong is the last new name to appear in this section, even though it is not the first time we mention him in the article. As you'd expect if you read the first part of this article, most of his results were with Grixis Shadow.

Of course, there are many more players with lots of accomplishments in Modern, some of them hardly correlated to tournament results! If you want to find their positioning based on their results, you can check the full details on this Google sheet.

Conclusion

That's it for 2021! It seems like the Modern metagame is getting closer to a stable state after the cataclysm Modern Horizons 2 was. The next three Standard sets only brought small additions. Unfair decks appear to be on the decline, while fair decks are thriving. Still, this dynamic might change if fair decks enter an arms race and don't respect unfair strategies.

See you next year for the last update before Kamigawa: Neon Dynasty!


Opinions expressed in this article are those of the author and not necessarily Cardmarket.



3 Comments

To leave your comment please log into your Cardmarket account or create a new account.

Hopper-MTG(27.12.2021 21:04)

Very good work, thanks for sharing! I still would like to point out that graphs are very rich in details and not easily readable as rasterized,non zoomable pictures. Furthermore, like any article on cardmarket, links to cards might annoy readers as not showing the card image immediately (like on any other mtg site).

Tuardian(20.12.2021 02:12)(Edited: 20.12.2021 02:20)

Great analysis with a lot of work involved, thks for sharing it! All very interesting stuff to digest over Christmas!

Cheers and happy holidays Anaël

Aliquanto
Aliquanto(20.12.2021 11:09)

Tuardian thanks, I wish you the same!

cardPreview