Eldrazi Mirror Matchup Card Analysis

No card clearly stands out. The sample is small, so treat this as a loose signal.

I looked at Ramp Eldrazi mirror matches in Modern over the last 12 months. That gave 128 decisive match results, with draws excluded. For each card I compared the winrate of lists playing it against lists not playing it.

Biggest winrate gains in the mirror (cards in at least 40 matches):

Misty Rainforest (main): 56.3% with it vs 37.5% without (+18.8 pts). I think this is mostly a deck-building artifact, because it's a land and tends to travel with certain list builds.
Verdant Catacombs (main): 52.4% vs 43.5% (+8.9 pts). The same caveat applies.
Nature's Claim (side): 53.1% vs 46.8% (+6.3 pts). This is the top non-land card.
Shifting Woodland (main): 51.2% vs 45.7% (+5.5 pts)
Soulless Jailer (side): 52.4% vs 47.7% (+4.7 pts)
Dismember (side): 52.3% vs 47.6% (+4.7 pts)

Cards that did worse in the mirror:

World Breaker (main): 45.8% vs 55.6% without (−9.8 pts)
Kozilek's Return (side): 44.2% vs 51.8% without
Disruptor Flute (side): 45.8% vs 52.2% without

Takeaway: Nature's Claim is the best candidate among cards that would actually affect the mirror. It's a +6 pt edge on about 49 matches, which is within noise. The fetchlands almost certainly reflect which list the player ran, not the cards themselves. I can't say which card improves the mirror the most.

I used the "Ramp Eldrazi" archetype because it has by far the most mirror data. The other Eldrazi variants have too few mirror matches to analyze.

Sources: Data from 461 Ramp Eldrazi tournaments over the last 12 months (66% MTGO, 34% Melee). Examples: