Bloomburrow: The Field Guide | Second Main Phase

The Record

What actually happened 17Lands archive

Every other chapter here is expert judgment, recorded before or during the format. This one is the account of what followed. Card tiles throughout carry a win-rate chip: the games-in-hand win rate, meaning how often the player won when that card was in hand or on the battlefield. Set median is 55.6 percent.

Games analyzed

931,230

Premier draft, whole format

Average game

8.93

turns. The shortest measured.

On the play

58.0%

vs 51.2 on the draw

Play advantage

+6.8

points. The largest measured.

Colors	Family	Plan	Win rate	Games
BR	Lizards	Opponent life loss	56.0%	81,966
WG	Rabbits	Tokens and anthems	55.6%	136,832
BG	Squirrels	Food and forage	55.6%	152,863
WB	Bats	Lifegain and life loss	55.4%	107,443
UG	Frogs	Blink and bounce	54.9%	109,145
WR	Mice	Valiant aggro	54.5%	91,224
RG	Raccoons	Expend aggro	53.8%	64,428
UB	Rats	Graveyard threshold	53.6%	57,698
WU	Birds	Fliers	52.4%	31,450
UR	Otters	Spells and prowess	50.8%	38,679

Five points separate the Lizards from the Otters. Both are aggressive two-colour decks in a ten-family set where every pair got the same support. What split them is that the Lizards convert damage into cards, and the Otters need to assemble a spell engine first. In a format this fast, the deck that needs a turn off does not get one.

Nothing Here Is Gentle

The thesis Two records at once

Bloomburrow holds two records across every set in this archive, and they are the same record twice. It has the shortest average game at 8.93 turns, and it has the largest play advantage at 6.8 points, 58.0 percent on the play against 51.2 on the draw. For comparison, Aetherdrift ran 9.49 turns and Duskmourn, a horror set, gave only 4.7 points to the player on the play.

A six point tempo swing from a coin flip is enormous. It means the format punishes every turn you spend not affecting the board: every tapped land, every do-nothing enabler, every card that needs a second card to matter. The cute art is doing a lot of work to disguise what is, by the numbers, the meanest format here.

Build for it. Prioritise cheap creatures that do something the turn they land, take the removal, and be extremely suspicious of any card whose text starts working later. The set gives you ten tempting build-arounds. The archive says most of them are too slow.

**Savor**58.9%The best common in the set. Instant speed, and it does two things at once.

**Carrot Cake**58.0%Two bodies plus a scry later. Exactly the rate this format wants.

**Scales of Shale**57.6%Cheap, aggressive, and it closes games.

**Intrepid Rabbit**57.6%A one drop that turns on Valiant and never costs you a card.

**Nocturnal Hunger**56.9%Removal. Take the removal.

**Bakersbane Duo**57.3%A two drop that makes Food, which the Squirrels are delighted about.

Which Mechanic Actually Pays

Measured, not guessed Median win rate vs the set baseline

Every card with at least two thousand recorded games, bucketed by the mechanic printed on it and compared against the set median of 55.6 percent. Six mechanics, and they split neatly by whether they cost you tempo.

Forage

+1.56

The best of the six, though only eight cards carry it. Exiling three cards or eating a Food is a cost you pay from surplus, not from your curve.

Offspring

+1.33

Pay extra later for a second body. It is optional, which is the whole point: you never lose tempo you did not choose to spend. 17 cards.

Expend

+0.48

Rewards using all your mana, which you wanted to do anyway. Roughly neutral. 10 cards.

Threshold

-0.14

Seven cards in the graveyard is a real ask in a format that ends on turn nine. 7 cards.

Valiant

-1.40

Needs a second card aimed at the first one. That is two cards for one effect in the fastest format measured. 9 cards.

Gift

-1.82

The worst of the six, and the largest bucket at 22 cards. You hand the opponent a real card or a real body, and the payoff rarely covers it.

The pattern is tempo, not power. The two mechanics that pay are the two that let you choose when to spend. The two that cost you are the two that demand something up front: a second card for Valiant, a gift to your opponent for Gift. In a format with a 6.8 point play advantage, anything that trades tempo for value is fighting the format.

The Talents Are Not A Cycle

17 points end to end Ten cards, one template

Ten Talents, all enchantments that level up as you meet conditions, all built from the same template. They range from the 96th percentile to the 2nd, a spread of 17.2 points from Innkeeper's Talent at 62.9 percent down to Artist's Talent at 45.7. That is wider than the gap between the best and worst decks in the format.

Treating them as a cycle is the mistake. Innkeeper's and Hunter's are genuine high picks. Gossip's and Artist's are among the worst cards in the set, and Artist's Talent is a rare.

**Innkeeper's Talent**62.9%96th percentile. A real bomb, and the best of the ten.

**Hunter's Talent**61.3%93rd percentile, and only an uncommon. Removal attached to growth.

**Stormchaser's Talent**58.4%82nd percentile.

**Caretaker's Talent**58.1%79th percentile.

**Bandit's Talent**52.4%16th percentile. The video called this one correctly.

**Artist's Talent**45.7%2nd percentile, and it is a rare.

Six Mechanics

Learn these before you sit down

Offspring

An additional cost paid as you cast. Pay it and the creature brings a 1/1 token copy of itself, keeping its abilities.

Powerful because it is optional. Cast it on curve early, and pay the Offspring cost later when you have the spare mana.

Forage

A cost on green and black cards: either exile three cards from your graveyard, or sacrifice a Food.

Any green or black deck can eventually forage, but most forage cards are weak without enablers, so take the Food makers alongside them.

Gift

A kicker variant. You promise the opponent something good as you cast, and if the spell resolves they get it.

The gift is usually a card or a tapped Fish, both real costs. Scores worst of the six, so demand a genuinely large payoff.

Valiant

Triggers when a creature becomes the target of a spell or ability you control, but only once each turn.

Free and repeatable if you have abilities that target, expensive if you need a whole card to turn it on.

Expend

Rewards spending a set amount of mana in one turn. Almost every instance here is Expend 4.

Rewards what you already wanted to do. Red green has the mana creatures to reach four a turn early.

Threshold

Returning mechanic. Upgrades a permanent once seven or more cards are in your graveyard.

It actively fights forage, since every forage exiles three cards out of the graveyard. Do not try to do both.

The Ten Families

Every pair, in finishing order Plans from the guides, results from the archive

LizardsBlack red 56.0%

Aggressive, and it turns opponent life loss into card advantage and hand disruption. The best deck in the format, built from the worst colour.

RabbitsGreen white 55.6%

Flood the board with tokens and convert width into damage with anthems. Fast, simple, and it does not need a second card to function.

SquirrelsBlack green 55.6%

The grindy Food and graveyard deck. Make Food, forage for value, out-attrition them. The most drafted deck in the set.

BatsWhite black 55.4%

Trigger life gain or life loss every turn to fuel the payoffs, while fliers do the actual work of pressuring them.

FrogsBlue green 54.9%

Pair strong enters-the-battlefield creatures with blink and bounce to reuse them. The best of the blue decks, which is a low bar here.

MiceRed white 54.5%

Be aggressive and target your own creatures to trigger Valiant. Held back by Valiant scoring below the median, and by red being the weakest colour.

RaccoonsRed green 53.8%

Ramp on two, spend four mana a turn to trigger Expend, attack with buffed bodies. The plan is fine, the cards are not quite there.

RatsBlue black 53.6%

Fill the graveyard to seven for threshold upgrades. Seven cards is a lot to find before turn nine, which is the whole problem.

BirdsWhite blue 52.4%

Win in the air while buffed ground creatures also attack. Evasion is real, but the deck is slower than everything above it.

OttersBlue red 50.8%

Cheap spells to trigger prowess and pingers. Last of ten by 1.6 points, made of the two weakest colours, and it needs setup the format will not grant.

The Scorecard

Four videos vs the archive

Median 95th

The prerelease bombs

Thirteen called, and they land high. Dreamdew Entrancer and Sword of Fire and Ice both hit the 99th percentile. The softest was Toski at the 55th, still above median.

9 of 10 correct

The always-draft list

Nine finished above the set median, median 75th percentile. The one miss is Playful Shove at the 28th.

8 of 10 correct

The never-draft list

Eight land in the bottom quarter, median 12th. Stargaze at the 42nd and Long River's Pull at the 31st are below average rather than unplayable.

Inverted

The colour ranking

Called red the best colour in the set. Red finished last of five. Black, ranked third, finished first. Only green, at second, was placed correctly.

The colour miss deserves its nuance. Red really does have the best common removal, which is what the ranking was based on, and the best deck in the format is black red. But across every mono-coloured common and uncommon, red averages 53.0 percent against black at 56.1, a 3.1 point gap that is twice the colour spread in Duskmourn. Red is a fine second colour and a poor first one. The lesson repeats: rank the decks, not the colours.

Always Draft These

Nine of ten confirmed Commons and uncommons

**Bushy Bodyguard**56.0%57th percentile. Needs a filled graveyard or a Food.

**Head of the Homestead**55.9%55th percentile. Castable in green or white.

**Playful Shove**53.7%28th percentile. The one clear miss on the list.

Never Draft These

Eight of ten confirmed In the order the video gave them

**Hoarder's Overflow**45.0%1st percentile. The worst card on the list.

**Stocking the Pantry**48.9%4th percentile.

**Nightwhorl Hermit**50.0%7th percentile.

**Barkform Harvester**51.9%12th percentile.

**Otterball Antics**52.5%19th percentile.

**Long River's Pull**53.9%31st percentile. Weak, not unplayable.

Where the list overreached. Stargaze finished at the 42nd percentile, which is below average but perfectly playable, and the video itself conceded it is fine when you are already stabilised. Long River's Pull at the 31st was judged on a two colour mana base, so a heavier blue deck can support it. The other eight are bottom quarter and the advice holds.

The Short Version

If you read nothing else

This is the fastest format in the archive and it gives the biggest advantage to the player on the play. Cheap, proactive, board-affecting cards over everything.
Offspring and forage pay. Gift and Valiant do not. The difference is whether the mechanic lets you choose when to spend.
Black is the best colour and red is the worst, yet black red Lizards is the best deck. Red is a fine support colour and a poor lead.
The Talents are not a cycle. Two are high picks, two are among the worst cards in the set, and one of those is a rare.
Blue red Otters finished last by 1.6 points. Both its colours are weak and its plan needs a turn off that the format will not give you.
Do not try threshold and forage together. Forage exiles the graveyard that threshold is trying to fill.