Everything else here is expert judgment recorded before the format was solved. This is the record of what happened after. 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 in play. Set median is 55.7 percent, so anything above that beat the average card. One honest caveat: this measures how you did when you HAD the card, not how easy it was to cast, which matters for the triple-hybrid cards below.
| Pair | Court | Win rate | Games | |
|---|---|---|---|---|
| BG | Elves kindred | 56.9% | 111,191 | |
| WU | Merfolk kindred | 56.5% | 94,969 | |
| GW | Kithkin kindred | 56.4% | 72,721 | |
| UR | Elementals kindred | 55.7% | 83,939 | |
| BR | Goblins kindred | 53.5% | 59,489 | |
| GR | Vivid aggro | 52.6% | 10,474 | |
| RW | Counters aggro | 52.2% | 9,871 | |
| GU | Vivid value | 51.5% | 5,482 | |
| UB | Faeries kindred | 50.6% | 9,808 | |
| WB | Counters value | 50.1% | 7,051 |
A dense set. Blight and vivid are the two that change how you build; the rest change how you play a turn.
The set's centerpiece: seven legends printed with a Lorwyn face and a Shadowmoor face. You cast the sunlit side, then pay to bring on the eclipse when the game asks for it. Treat the back half as a mana sink you are always happy to have, not as a plan.
Eirdu, Carrier of Dawn☽
Isilu, Carrier of Twilight
Oko, Lorwyn Liege☽
Oko, Shadowmoor Scion
Brigid, Clachan's Heart☽
Brigid, Doun's Mind
Ashling, Rekindled☽
Ashling, Rimebound
Sygg, Wanderwine Wisdom☽
Sygg, Wanderbrine Shield
Grub, Storied Matriarch☽
Grub, Notorious Auntie
Trystan, Callous Cultivator☽
Trystan, Penitent CullerThe cards every guide flagged. Eight of these ten landed above the 88th percentile of the set, which is a strong hit rate. Two did not, and both are noted honestly below.
Six kindred pairs and four themed pairs. The standings tell you which are real; the descriptions tell you how to build them if you get there.
Self-mill elves, then recur them and cash the graveyard in for tokens. Lethally blighting your own elf is another guaranteed way to switch the payoffs on. The board scales as the yard fills, and once stocked, recursion is your card advantage engine.
Go wide with lords and tokens, then convert a tapped board into cards and removal. Convoke is the most powerful mechanic here, so take the enablers as highly as the payoffs: a one-drop into a two-drop lets you convoke a five-mana spell on turn three.
Flood the board and cash it in with the lord effect and token-makers. Every kithkin in the set lives here, so play them all. Because it is everywhere, every other deck needs cheap interaction: their buffs happen at instant speed on their turn, so removal beats them far more reliably than blocking does.
A good curve of elementals plus payoffs for casting spells of mana value four or greater, with looting to fix the wrong-half problem. Elementals appear in every color, so these decks often drift into multicolor vivid piles.
A sacrifice deck wearing a kindred costume: the set has no sacrifice outlets, so blight is how you kill your own goblins. Blight's best home and the only deck that actively wants its creatures to die. Close fast, before the card disadvantage catches up. Only draft it when the lane is wide open.
No creature type of its own, so build to maximize vivid instead: Treasures and fixing powering out oversized attackers whose stats scale with your color count. The uncommon vivid creatures are already good at three colors and want five.
Undersized-on-entry bodies that shed their counters to grow, pump, and grant flying. Nominally the giants pair, but there are too few giants: build an assertive value deck rather than a tribe, pairing powerful blight cards with creatures that shrug the counters off.
Stretch into as many colors as the fixing allows, then cash the count in for card draw and pump. It has no gold signpost, and it should usually splash a bomb that needs only a single colored pip.
Every spell cast during an opponent's turn triggers draw, surveil, and counters. The problem is density: the payoffs are thin at low rarity, so you need roughly three of them before committing. The worst-performing deck in the format, and the clearest example that a tribe on the box is not a deck in the pod.
Creatures that enter with counters and shed them for card advantage, recursion, and removal, deliberately blighting your own board to reload. Nominally treefolk, but the treefolk payoffs are not there. A fine pile of good cards; just do not expect a synergy deck to appear.
Five hybrid uncommons, one per kindred pair, each digging four cards deep for its tribe or its two basics. They are the glue of the format and five of the six posted above-median win rates. But read the pips before you take one: one hybrid pip is pure flexibility, two narrows a card sharply, and three makes it effectively uncastable on curve outside its exact pair. Missing your three-drop on turn three is a real cost that these numbers do not show.
The guides said to fear green's tricks and ignore white's. The data mostly agrees: white's Appeal to Eirdu sits in the 6th percentile and Riverguard's Reflexes in the 22nd, while green's Blossoming Defense is comfortably above median. One correction: Thoughtweft Charge, named as the trick to fear most, actually landed slightly below median.
Plan your splashes around bodies and fetches, not lands. Aim for at least three sources, four is much better, and only splash cards that are strong late.
