Earth-Cult Elemental Art by Aaron Miller

Alchemy Horizons: Baldur’s Gate Limited Archetypes Overview: Meta Data, Rankings, and Card Picks

Sierkovitz looks at each of the 10 color pair archetypes in Alchemy Horizons: Baldur's Gate Limited and their plans. For each, we will give you an archetype star, hidden gem, a trap, and a sealed hero, with a bit of help from the first two weeks of data!

With two weeks of the format behind us, Alchemy Horizons: Baldur’s Gate (HGB) limited formats start to crystallize. We have enough data from 17Lands to see which cards we undervalued, which ones were overhyped and which synergies work well or not. Knowing which cards make each archetype tick gives you an edge over the competition.

Today I will look at each of the 10 color pairs. For each of them I will give you an archetype star (card with the highest win rate), hidden gem (a card that is still relatively underdrafted, but has a relatively high win rate), a trap (card picked highly that doesn’t win a lot) and a sealed hero (card doing better in sealed format than draft).

To draft successfully you need to understand the power, or lack thereof, of individual cards but also have a good understanding of the overarching plan each deck wants to have. That is why I will try to put my four selected cards in the context of the color pair and use them to showcase the plans you want to consider while playing. I will focus on the commons, as they are the bread and butter of each format and you will see them more frequently.

Impotent Izzet

I will start with the ascending win rate order for the color pairs. And Izzet has the arguable pleasure to be first here. The archetype is struggling. No clear plan, poor blue commons that don’t interplay well with the red ones, this all spelled disaster from day one and there is no happy ending this time. UR is doing very poorly.

49.2% win rate for the color pair is way below the average 17Lands user win rate of 56.5%. Best common in the pair, Valor Singer has a Game in Hand win rate (GiH WR) of 54.3%. Hardly a number to call it a star. But that points to Izzet as wanting to be aggressive even if blue doesn’t deliver the tools to be so.

For the hidden “gem” I picked Lizardfolk Librarians. GiH WR of 51.9% is not impressive but it is above average for Izzet and the card routinely wheels.

The biggest trap is playing the archetype itself, but individually Charmed Sleep is a card picked relatively highly by Arena users that has a 45% GiH WR making it a waste of a high pick in most cases.

If you play Sealed, of course you will need to have some great high rarity cards to get tempted into UR, but on common level Air-Cult Elemental delivers some solid numbers with 55% GiH WR. Mind – this is across all archetypes as UR sample size is just too small.

Disappointing Dimir

53.5% color pair win rate is much better than Izzet was, but still far from impressive. But not all is doom and gloom for Dimir decks. The win rate markedly improved in week two. Successful Dimir decks seem to lean heavily on black cards, using blue as a support color or source of bombs like Tasha, Unholy Archmage or Snowborn Simulacra. Dimir wants to survive the early onslaught and play a longer game based on attrition.

The archetype star is Sewer Plague at solid 59% GiH WR, with multiple removal spells close behind.

The hidden gem is Rimeshield Frost Giant. Picked late in drafts it still has a respectable 55.3% win rate and will help you survive till late game. The giant, Young Blue Dragon and Juvenile Mist Dragon all costing 5 mana makes the pseudo-mana dork, Clever Conjurer an interesting inclusion in the decks letting you cast those cards on turn 4.

The biggest trap in UB is Prophetic Prism. The card beloved by many is not delivering in HBG in general and with 50.1% win rate in Dimir it definitely disappoints. The format is just too fast for it and with no artifact synergies to back it up, it falls flat.

For Sealed – again based on data from all games, Armor of Shadows looks promising, being able to blank some of the removal spells for a bargain price of one mana, while also being a decent combat trick in racing situations.

Gruelling Gruul

Gruul is tied with Dimir in terms of win rate, but I gave it a slightly higher position as it does better with a small splash. Green has some excellent tools for splashing and you want to use those tools in the deck anyway, so frequently you can put 1-2 off color cards almost without a deckbuilding cost. In AFR RG was all about Pack Tactics (abilities that trigger when attacking with creatures total power 6 or greater) and early aggression.

In HBG it is supposed to be one of the dragons archetypes, but data strongly suggests it is neither of those. Despite two key Pack Tactics cards in Gnoll Hunter and Hobgoblin Captain still being in the set, other support for that strategy is lacklustre. Especially green creatures are not aggressive enough with defensive and mediocre two drops. The Dragons are more of an add-on and might play sometimes but as a pocket of synergy rather than the main theme. Good Gruul decks are all about ramping. And best win rates are dominated by ramp enablers, payoffs and cheap interaction that lets you live untill you can start dropping your big creatures.

The archetype star is the Undercellar Myconid at 58.7% GiH WR and is followed closely by the likes of Dread Linnorm, Owlbear or Hill Giant Herdgorger.

The hidden gem in the data is the Ettercap. A humble 2/5 for 5 with Plummet attached to it goes late in the draft and has a decent win rate of 57.5% (warning – small sample size).

The archetype’s trap is the Young Red Dragon. Picked way too early by the Arena drafters for what it does in the game – you will have to waste your earlier picks for a card that does little to promote your game plan.

Last but not least – Gruul looks slightly better in Sealed than it is in Draft and the star card in RG in sealed looks to be Heard Gorger Giant with 56.5% win rate.

Average Azorius

Blue White decks don’t do very well in HBG, especially taking into account how good white seems to be. Despite being supported by the best color in the set, Azorius is not even the best blue archetype. And that to me suggests two things: the main archetype theme does not work well and both colors try to do different things and therefore combining them does not work too well. What Azorius seems to do well is being a mid-range deck. A bit of aggression, a bit of control to be able to play both game plans.

True to that statement, best card in the archetype is the Priest of Ancient Lore with 60.5% GiH WR. Card is just a solid creature with great abilities, and doing exactly what you want it to do.

The hidden gem for me is Undersimplify, a 2 mana counterspell. As much as Oji, the Exquisite Blade is a good card, don’t lean to heavily on the main blink theme.

In fact the blink spells apart from Oji are my biggest traps in the archetype, with Blur winning at 50.2% being a highlight. The showcase for that is Pegasus Guardian – on paper a prime payoff/enabler for a blink archetype, that has only ~50% win rate. The stranger as the card has really good numbers in all the other white archetypes.

For the Sealed, note the Young Blue Dragon winning around 54% of the games it is in hand. The more impressive as sealed win rates are generally lower than the draft ones. It is relatively good in UW and better there than in other blue archetypes.

Subpar Simic

Simic, with the 54.3% win rate is the last of the lowest tier decks. Head and shoulders above Izzet, but definitely not living up to the top 5 most winning decks. On paper it is supposed to be Dragon themed, but HBG is not a paper set. I am here all week if you want more of that type of humor. Simic looks like a ramp deck, similar to RG, but in this case both colors synergise better. Red in its nature is a more aggressive color, and in RG combination it was contributing mainly very good early interaction. Blue does much more – having good early interaction (Undersimplify), great ramp payoffs (Sword Coast Serpent) and in crisis times even some ramp in Clever Conjurer.

However, it is not the blue that stars in this archetype. Yet again blue should be more of a support than a main color. The star of the archetype is yet again Undercellar Myconid. Ramping while generating 1/1 speedbumps for the opponent? Yes please. And 59.8% win rate confirms that the mushroom is the way to go.

But right on the heels of Myconid, just a mere 0.1%p lower, we have my hidden gem: Arcane Archery. A combat trick that leaves something behind, and it is what it leaves that is probably most of its value. Playing Archery to win combat and following it next turn with an 8/7 trampling Hill Giant Herdgorger is a very efficient play pattern.

A slight trap in my eyes is the Ambitious Dragonborn. It has some theoretical amazing scenarios with a 7/7 creature on board on turn 4, but in reality it requires playing some bad cards to sometimes get a good outcome. And in UG that will not be a good thing. I can see a merit of trying it in BG though.

As for Sealed, Simic is the worst archetype in Sealed data. It does splash well though so world is your oyster if you opened some Myconids. Why not spicing your deck with some Summon Undead if you have any tools to fill your graveyard early?

Good Golgari

Golgari is exactly like all the other ramp decks so far except for one thing. It is actually good. Not great, mind you. If it was – I would not miss such an alliteration in the header. Golgari is better than Simic or Gruul because black is just a better color. As simple as that. It also has a potential to actually do what the archetype is supposed to do.

Pairing up a couple of Dread Linnorms with Summon Undeads and some self mill is not a terrible idea and all the pieces in this sub-theme you either want in your deck anyway (large ramp payoff creatures) or go late in the draft so are easy pick-ups at low opportunity cost for your draft (Reanimate spells, self mill enablers). Golgari combines the green ramp package with black’s attrition and that is enough to make it the best slow deck in an ultrafast format.

Technically Undercellar Myconid is the archetype star, but I will recognise how close a second place Dread Linnorm was in all of those races so far especially that at 60.9% GiH WR it is just 0.1%p behind the fungus. Linnorm is one of the two most important green commons in the set and as having ramp payoffs on higher rarities is not a given, I just want to remind you not to ignore it when seen in later picks.

My hidden gem for Golgari is the Baleful Beholder. With 57.3% win rate it is performing much better than in any other archetype it is played in. The combination of ramp and playing longer games makes both abilities of the Beholder a little bit better and that might be what pushes it over the line here.

The archetype trap is Underdark Basilisk. Card that has some OK results in other color combinations performs pretty bad in BG. Is it because of the card itself? I doubt it. It is probably a sign that something went wrong. BG should not need the basilisk, having access to infinitely better black card in Guildsworn Prowler and if you don’t have that access you probably shouldn’t have been in BG to start with.

Pure Golgari is not a particularly good archetype in Sealed, but one card that stands out in it is the Grim Bounty, which not only provides a much needed removal but also helps you fixing for potentially welcome splashes.

Respectable Rakdos

We finally arrive at the archetypes that are actively good. Rakdos, with 58% win rate, is not as dominant as it was in AFR but still is pretty good. 17Lands data suggests that black commons are the key to its success, with a mid-range on the aggressive side for a plan. Incessant Provocation still works well with the sacrifice outlets so steal and sac is a valid game plan for the deck if you can get all the pieces. There is still enough of treasure generators to be able to produce some explosive starts but the quality creatures and removal at common can cope without acceleration.

The star of the archetype is a bit arbitrary. Grim Bounty is a top common at 60.1% GiH WR, but Guildsworn Prowler, Shambling Ghast and Vampire Spawn are within 0.2%p from it making any differences insignificant. All of the cards mentioned are an excellent counter play against aggressive strategies, and despite being relatively fast itself against more controlling archetypes, Rakdos get an edge against aggro by utilising them. If you ever played against BR with a Boros deck you will know how backbreaking a turn 1 Shambling Ghast can be against your hand full of 3/1 creatures.

The hidden gem of the archetype is the Earth-Cult Elemental. At a respectable 58% win rate, it does better in Rakdos than anywhere else. If you can ramp it out early it puts opponents in a precarious spot, and with a bit of a luck on the rolls can end the game on spot.

The trap I selected for the archetype is the Reckless Barbarian. This is not the early drop you want to be playing. It doesn’t contribute to your plan greatly, and you probably need to sacrifice some better picks to get it based on Arena users preference in drafting the card.

Guildsworn Prowler is not only one of the better commons in draft but has by far the best stats in Sealed with the GiH WR of 59.4%.

Sweet Selesnya

WG made the top 3 in terms of win rate with 58.6%. But there is more in that, the archetype didn’t lose much in the second week of the format and in my opinion is the best thing to be drafting at the moment if cards align. As the top 2 archetypes lost some of their power due to the self-correcting nature of the draft, Selesnya stayed strong. It is also the best home for splashing the third color, giving you flexibility in draft. It also, unlike Boros, Rakdos and Orzhov, performs very well on the draw, which means you reduce the variance related to coin flipping in best-of-1.

Selesnya is supposed to be all about life gain synergies, but data shows that you should play it as a go-wide slower aggro deck with some ramp and a solid top end, as life gain payoffs are few and far between. That doesn’t mean you will not draft decks that lean heavier of life gain if you get some Prosperous Innkeepers and Trelasarra, Moon Dancer, but it is a bonus rather than the main plan you should be focusing on.

The star of the archetype is Patriar's Humiliation at 62.8% GiH WR. The cheap and efficient removal deals with many creatures even if it doesn’t kill them and is suits the deck’s game plan fantastically. You should pick this card highly anyway as it is good all around. Of note Humiliation is tied in 1st place with both the Undercellar Myconid and Dread Linnorm.

The not-so-well-hidden gem for Selesnya is the Steadfast Unicorn at 60.9% win rate. The card is a one drop of dreams. Relevant early but even better late allowing you boosting your team, racing efficiently and suiting the proactive plan of WG. It is also accidentally synergizing with the ramp of the deck, bringing you faster to the realm of double activations, making opponent’s blocking choices a nightmare.

A bit less intuitive gem is the Icewind Stalwart. If you manage to get several Owlbears, Priest of Ancient Lore, blinking your creature provides you with plenty of value and coming on a 3/3 creature, it fits perfectly with your game plan. Blink in general looks better in WG than it does in WU, where it should have been in theory supported.

The archetype trap is Dawnbringer Cleric. It does potentially gain life but don’t let it fool you. 51.5% win rate tells me to avoid it in my WG builds where possible. 1/3 body is not exactly what this aggro deck wants, especially with many 3/1 creatures being around. Your lifegain synergies should be more something that happens incidentally than something you force into your deck.

In Sealed Blessed Hippogriff shines. It has the highest win rate of any common at 60.6% GiH WR. In a format with more bombs the protection it offers is more relevant, as is 2-for-1ing your opponents. And the griff provides both and can swing games and break board stalls at the same time. What is not to like?

Brilliant Boros

Boros was a clear winner of week 1 of the format and now it is paying the price. Important cards for the archetype are picked much earlier that before so you will not likely have access to the builds as powerful as in the week 1. But still it stays one of the archetypes to beat at 59.4% win rate and, especially on the play it is a force to be reckoned with.

Boros is the definition of aggro in HBG. Both colors want to attack and were provided with great tools to do so. Good early creatures, potent and mana efficient removal package, strong uncommons and busted rares – all is there. WR has also a very flat power level. There is at least 5 great two-drops (and two one-drops) at common that can be interchangeably used.

It is maybe this breadth of early creatures that makes Blessed Hippogriff the star of the archetype as it provides two unique and very important features. The adventure allows you racing very mana efficiently and the creature provides a very needed reach (as in ability to deal last few points of damage, not ability to block flyers, although it can also do that). At 62.4% GiH WR it is quite a bit better than any other creature in Boros. It is even better in that aspect than all but two uncommons. A little horse-lion-eagle that definitely could.

The very hidden gem for Boros is an AFR reprint – Dueling Rapier. It has a really solid GiH WR of 58.3% for a card that wheels routinely. You might want to test it and see if it delivers, I would warn against playing multiples though.

There are several under-performers in Boros, but the trap I selected is the Icewind Stalwart. A gem in Selesnya does not deliver in the same way in WR. The synergy level is just not the same. Red doesn’t have Owlbears, and the plan of the deck is to end the game as fast as possible rather than play around with blinking creatures. You are much better served by Devoted Paladin as your top end creature.

Boros is not only a top 2 archetype in Draft. It excels in Sealed as well with 57.5% win rate second to Orzhov only. The card that beats all the commons in Sealed is the Patriar's Humiliation. 64.8% win rate is no joke, it actually beats all the uncommons in that respect, so if you open a pool with several of them and a good creature suite – consider playing a fast, aggressive deck and you may be positively surprised.

Outstanding Orzhov

What happens when you mix the two best colors in the set? Well it seems you get a winning archetype. Despite a complete lack of a theme (sorry Minthara of the Absolute, that is the truth) Orzhov gets there on the sheer quality of the cards. With best win rate in Draft (59.4%) and Sealed (58.2%) you can’t ignore it. Good Orzhov deck uses its card quality to out-grind the opponent.

Capable of aggression but with powerful card draw and a selection of 2-for-1s it can play an adaptive gameplay: able to capitalize on the opponent stumbles but also able to weather an early offensive and take over the late game. Orzhov has a stunning range of good uncommons, with 14 of them having a GiH WR higher than 60%. Not that it is bad in the common department either: 9 of them exceed 60%.

In this quality field, Priest of Ancient Lore is the one that stands out. 63.8% GiH WR is amazing for a common. Just imagine if it had flying. Priest is a perfect card for the archetype. Fixes your game with the draw, provides a body, gives you a natural 2-for-1 for the attrition games. Later it can also become a sac fodder if need be. Given the high quality in the archetype, any draw effect becomes that little bit better as it draws you gas, not mediocrity.

That is probably why my hidden gem is Valiant Farewell. 59.5% for that combat trick is head and shoulders above its numbers in any other color pair. Combination of being able to deal with larger creatures with the extra value the boon priovides with the draw a card effect is again something that Orzhov wants to be doing. All from a card that frequently wheels – that is a bargain.

The trap for that pair is the Flaming Fist Officer. Unlike their colleague, the Flaming Fist Duskguard, which is good in any archetype and a great card altogether, Officer is a disappointment. But especially so in Orzhov where technically it was supposed to be at its best and synergize with the archetype theme. With a paltry 51.4% win rate it didn’t deliver, and yet it is picked quite highly by Arena users. According to my observations, cards like the officer tend to do so. Unruly Mob was also a crowd favourite despite bad numbers. Probably it offers a dream of turning a small cheap creature into a behemoth – a dream that maybe sometimes comes true, but is not worth the many failures needed to achieve it.

In Sealed there can be only one card which can win. Vampire Spawn. With a whooping 64.2% GiH WR spawn looks to be a Sealed all-star common. Even it the best archetype it stands out. The 2/3 body, the life swing, the capacity to stabilise the board, all is in favour of the Spawn. The card that was initially overlooked in AFR, but later in the format became one of the top commons, delivers also in the HBG and this time it doesn’t surprise.

Closing Conclusions

Look, I know it is a tautology, but at least I stay in the alliteration theme, OK? I really hope that this article gives you a slightly better grasp of the format and also gives you an idea on how to use the 17Lands data to interpret the card power and help you see some synergies in the archetypes that will guide through your next drafts. HBG is not a perfect format but so far it was fun to play and knowing which traps to avoid and which gems to pay attention to can bring more fun to your experience too. Even though blue is not great, it is by no means unplayable. Even though white is strong it is by no means unbeatable. And if you follow the advice above you may realise it yourselves. Let me know if you do!

Card Kingdom - Double Masters 2022
Default image
Sierkovitz

I am a limited player, who mainly skips playing in order to analyse the limited data using 17Lands.com. I run a podcast: Magic Numbers, where I try to use data to let you improve your limited game play, find out which heuristics work out and which common ideas are not well supported by data.

Articles: 4