Deck construction math
Why your deck wants 8-12 ramp pieces, not 6
The opening-hand probability that decides whether you land turn-3 plays.
TL;DR: Most EDH decks run 6-8 ramp pieces and wonder why they can't cast their 5-drops on turn 4. Math says 8 ramp pieces is the floor for reliability, 10-12 is where the curve feels smooth, and going lower trades consistency for false deckbuilding flexibility. The difference is opening-hand probability — not deck quality.
The question nobody asks
You shuffle up. You draw seven. You count lands. You count spells. You look for something that lets you play ahead of curve.
The question is: how often does that "something" actually show up?
Most EDH decks run 6-8 ramp pieces. Six is the old consensus default. Eight is the new one, sort of. Neither number comes from math. Both come from vibes — "feels like enough" without taking up too many slots.
Here's what the hypergeometric distribution says: at 6 ramp pieces in a 99-card deck, your chance of drawing at least one in your opening seven is 38%. At 8 pieces it's 48%. At 10 it's 57%. At 12 it's 65%.
Flip those percentages. At 6 ramp pieces you miss in 62% of opening hands. At 8 you miss in 52%. That's the difference between "I sometimes have ramp" and "I usually have ramp." The inflection point sits at 8 — the number where you cross 50/50 odds and ramp becomes the expected case instead of the lucky case.
The math (shown, not asserted)
Hypergeometric distribution answers this question: "In a deck of N cards with K success cards, what's the probability I draw at least X successes in a sample of n cards?"
For an opening hand in EDH:
- N = 99 (deck size)
- K = number of ramp pieces
- n = 7 (opening hand size)
- X = 1 (we want at least one)
The formula for drawing exactly x successes is:
P(X = x) = [C(K, x) × C(N - K, n - x)] / C(N, n)
Where C(a, b) means "combinations: a choose b."
To get "at least one," calculate 1 - P(X = 0). That's easier than summing all the success cases.
Here's the table for K = 6, 8, 10, 12, 14:
| Ramp pieces | P(0 ramp in hand) | P(at least 1) |
|---|---|---|
| 6 | 62.1% | 37.9% |
| 8 | 52.4% | 47.6% |
| 10 | 43.8% | 56.2% |
| 12 | 36.2% | 63.8% |
| 14 | 29.6% | 70.4% |
The jump from 6 to 8 gains you 10 percentage points. The jump from 8 to 10 gains you 9. Diminishing returns don't kick in until you hit 12 — and even there, going from 12 to 14 still gains 6 points.
Most decks stop at 8 because they feel the pressure of "only 99 slots." But if you're honest about what turn-3 acceleration does for your game plan, 10-12 is the target. Below 8 you're hoping. At 10 you're expecting.
Ramp tiers (what actually counts)
Not all ramp is created equal. The tiers matter because fast ramp compounds and slow ramp catches up.
Tier 1: 0-mana and 1-mana (the dream)
- Sol Ring (in 83% of decks on EDHREC as of writing)
- Mana Crypt (if you own one)
- Mana Vault (speeds combo, less good in midrange)
- Jeweled Lotus (situational, banned in some metas)
These let you cast a 3-mana spell on turn 2 or a 4-mana spell on turn 3. They break parity. Every deck wants Sol Ring. Most decks can't afford Mana Crypt. That's fine. One fast rock is enough to justify building the rest of the curve around it.
Tier 2: 2-mana (the workhorses)
- Arcane Signet (universal, in 73% of decks)
- Signets (Azorius Signet, Izzet Signet, etc.)
- Talismans (Talisman of Progress, etc.)
- Mind Stone (cycling matters late-game)
- Fellwar Stone (politics-dependent)
These let you cast a 4-mana spell on turn 3 if you hit your land drops. Two-mana ramp is the backbone. A deck with 8 total ramp pieces should dedicate 5-6 slots to this tier. They're reliable, they're cheap, they don't ask you to splash green.
Tier 3: 3-mana (green's land ramp)
- Cultivate (puts one land in hand, one on the battlefield)
- Kodama's Reach (same)
- Three Visits (one land to battlefield, faster but no card advantage)
- Nature's Lore (same)
- Rampant Growth (2-mana, technically tier 2, but weaker than rocks)
Cultivate
Green decks want 3-4 of these on top of 2-mana rocks. The advantage is resilience: land destruction is rare, artifact removal is common. The disadvantage is speed: three mana for one mana of acceleration is a bad rate turn-over-turn. But it fixes your next three draws and it survives board wipes.
Non-green decks don't have access to this tier in volume. That's why artifact-heavy strategies (Esper, Izzet) lean harder on tier 2 and why green-based decks can get away with slightly fewer rocks.
The "second land drop" problem
Here's the thing nobody talks about: ramp doesn't just speed you up. It also smooths your curve when you miss land drops.
You keep a two-land hand with a Signet. Turn 2 you cast the Signet. Turn 3 you miss your land drop. You now have two lands and a Signet — three mana. You can cast your 3-drop. Without the Signet you'd be stuck on two mana, probably discarding to hand size by turn 5.
That scenario happens in 15-20% of games (depending on how many lands you're running — most decks sit at 36-38 lands, which means missing turn-3 land drops in about one game in six). Ramp pieces act as insurance against flood and screw both. They're not just acceleration. They're variance smoothers.
Decks that skip ramp die to mulligan variance more than they die to interaction. You mulligan because you drew three lands and four 6-drops, or because you drew six lands and one spell. Ramp pieces convert excess lands into usable mana and convert awkward hands into keepable ones. The more you run, the more hands you can justify keeping.
At 6 ramp pieces you mulligan more. At 10 you mulligan less. The difference is about one mulligan every four or five games. That's huge over a 20-game sample.
What this means for your 99
If your deck wants to cast 4-mana plays on turn 3 and 6-mana plays on turn 5, you need 10-12 ramp pieces. If you're fine casting on-curve (4-mana on turn 4), you can get away with 8. Below 8 you're playing a different game — you're the player who drops Avacyn on turn 8 while the Simic player is already closing with Craterhoof.
Count your current list. If you're at 6, find two more slots. Cut a mediocre draw spell or a win-more combo piece. If you're at 8 and your curve tops out at 6 mana, you're fine. If your curve tops out at 8+ mana, go to 10.
The math doesn't lie. The opening-hand probability is the difference between decks that feel smooth and decks that feel clunky. You've felt it at the table — now you know why.
What to cut to make room
The hardest part of adding ramp is finding what to cut. Here's the priority order for cuts:
- Win-more cards. Anything that's only good when you're already ahead. If a card makes you win harder but doesn't help you get there, it's a cut.
- Cute synergy pieces. The card that triggers twice if you control your commander and it's a full moon. Synergy is good. Requiring three pieces to align is bad.
- Redundant removal. Most decks run 8-10 removal spells. You can go to 7-8 if it means your deck casts its threats a turn earlier.
- The sixth board wipe. Four to five is plenty. Six is insurance you don't need.
Do not cut card draw to make room for ramp. Draw and ramp are the two pillars. You need both. Cut the air between them.
Three Visits
The exceptions (when 6 is fine)
Some archetypes don't want 10 ramp pieces:
- Aggro / low-curve decks. If your curve tops out at 4 mana and you're trying to win by turn 6, you don't need Cultivate. Six ramp pieces (mostly 2-mana rocks) is enough.
- Land-matters strategies. Decks running Azusa, Exploration, Oracle of Mul Daya, and Burgeoning already have 8+ functional ramp pieces. Don't double up.
- Storm / ritual-based combo. Decks that win off Dark Ritual into Peer into the Abyss don't need permanent ramp. They need fast mana and protection.
For everyone else, 8 is the floor and 10 is the target.
Why pods feel faster now
EDHREC data shows that average ramp counts have been climbing since 2020. In 2019 the median was 6.2 pieces per deck. In 2023 it hit 8.1. In 2025 it's at 8.7 and rising.
Part of that is power creep — Dockside Extortionist, Jeweled Lotus (before bans), Arcane Signet's printing in 2019. Part of it is format education. Players have figured out that the turn-3 Rhystic Study or turn-4 Smothering Tithe is what wins casual games. You can't cast those spells on curve without ramp. So the average deck adapted.
If your list is still at 6 pieces, you're playing 2019 EDH in a 2025 meta. That's fine if your pod is also at 6. It's miserable if your pod is at 10 and you're the one stuck on four mana on turn 5.
This is a Commander Pulse evergreen. For weekly EDH data digests delivered to your inbox, sign up here.