How bad are precon manabases? We diagnosed all 169

August 31, 2026

The first advice every new Commander player hears is some version of "fix the manabase." It gets repeated so often it has become background noise. Nobody checks whether it is actually true, or how true, or which decks deserve it most.

So we checked. All 169 official precons, 2011 through 2026, run through the same Fix my mana engine that went live on the site this week, at default settings, on August 31. The numbers below cover the 165 hundred-card decks. The four Throne of Eldraine decks are 60-card Brawl products, and there is a note about them at the bottom.

Not one precon has a clean manabase

The engine calls a manabase healthy when three things hold. The land count sits in the recommended range for the commander. No color is short on sources for its pips. And no land always enters tapped.

Zero of 165 pass. And every single precon ever printed, the four Brawl decks included, runs at least one land that enters the battlefield tapped no matter what. The average precon runs eight of them. The median is seven.

That is the strict version of the finding, and the bar is honestly a high one. One tapland is not a broken deck. Twenty-one is a different story.

The slowest manabase Wizards ever shipped

Subjective Reality, the Aminatou deck from Commander 2018, runs 40 lands. Twenty-one of them always enter tapped. Guildgates, bounce lands, refuges, gain lands, a tri-land, three cycling lands. Fewer than half of the deck's lands enter untapped, the only precon ever printed under 50%.

Notice the years. Three of the six worst are from Commander 2017, and that is not a coincidence. 2017 is the worst mana year in precon history, averaging 15.2 always-tapped lands per deck. The Edgar Markov deck, the one that launched a thousand upgrade guides, came out of the box with 18.

Draconic Domination gets the extra indignity of the worst commander castability score in the whole line. By the engine's math you can cast The Ur-Dragon on time about 62% of the time with the manabase as printed. The deck's entire pitch is a nine-drop dragon, and more than a third of games the mana is not there for it on curve.

Precons fixed their mana while nobody was watching

Split the line at 2020 and the trend is as clean as the bracket trend we found last time.

The 2026 sets average 6.4 taplands per deck, less than half of 2017. And the best manabases ever shipped are almost all recent. The Fantastic Four deck from the Marvel line runs exactly one land that always enters tapped. Buckle Up from Kamigawa runs one. Quick Draw from Outlaws and The Hosts of Mordor from Lord of the Rings run two each.

The color math got better too. In the 2010s nearly half of precons were short on sources for at least one of their colors. Today it is one in seven. The worst offender ever is Land's Wrath from Zendikar Rising, which the engine scores nine green sources short of what its green pips ask for. When a precon is short a color, by the way, it is usually green. Green leads the deficit table with 15 decks, ahead of white at 9.

More lands, slower lands

Here is the pattern underneath all of it. 76 of the 165 precons run more lands than the engine recommends for their commander. The number running fewer is zero.

My read: those two facts are the same fact. When your lands enter tapped, you compensate by playing more of them. The padding is rational, but it means the average precon spends extra slots on mana and still hits its turns late. Better lands fix both problems at once, and most of the engine's suggestions are exactly that trade, a tapland out and an untapped land in.

What the engine actually cuts

Across all 165 runs, the most-cut land in precon history is Myriad Landscape, cut in 35 decks. The reasons the engine prints are blunt. From the Sultai Arisen run: "Enters tapped; the deck has 45 black pips and only 22 black sources."

The second most-cut card will annoy people: Path of Ancestry, cut 24 times. It is a beloved card, and it is in 82 of the 169 precons for a reason. The engine has a reason too. From the Avengers Assemble run: "Enters tapped; only 74% of this deck's lands enter untapped." In a deck already drowning in taplands, the scry is not free. The engine does weigh a land's utility, and it still cuts Path when the tapland count is bad enough. The swaps are suggestions with reasons attached, not commands, so if Path earns its slot in your deck, keep it.

The adds are less controversial. Cavern of Souls tops the list at 22 decks, then Boseiju, Who Endures and Yavimaya, Cradle of Growth at 19 each. Untapped, on-color, and hard to punish.

What the fix costs

The engine proposes up to six swaps by default. Over the 89 decks that got swaps, the median sticker price for the adds is about $76. One fix is literally free. Eldrazi Unbound is a colorless deck running three always-tapped lands, and the engine's replacement for all three is the humble basic Wastes. Total price of the fix, zero dollars.

At the other end, leave the budget field empty and the engine assumes money is no object. Its dream fix for Urza's Iron Alliance includes an Underground Sea and a Tundra, $1,700 of Reserved List duals bolted onto a precon. It is not wrong, exactly. There is a per-card budget field in the tool, and this is why.

What this changes at your table

The advice was right. "Upgrade the manabase first" survives contact with the data, and now it has numbers attached. If your precon is from 2017 or 2018, it deserves the advice twice.

But the advice has aged better than the decks it was written about. A 2024 precon with seven taplands and clean color math does not need $76 of lands before you sleeve it. It needs a few targeted swaps, and the specific swaps matter more than the category.

Which three is the part you do not have to guess anymore. Paste your deck into the checker, open Fix my mana, and it will name the cuts, the adds, and the reason for each, against your commander and your budget.

How this was measured, and where it is soft

Every number is from a single run on August 31, 2026, against the live production engine at default settings, no budget cap, no cherry-picking, no reruns. The decklists are the official ones from the Archidekt_Precons account, the same lists in our precon library.

Things you should know before quoting it:

• "Always enters tapped" means unconditionally. Checklands, battle lands, and anything with an untapped mode are not counted in the tapland numbers, though the untapped-share stat counts conditional lands as half.
• The four Throne of Eldraine Brawl decks were run but sit out of every stat. They each carry seven always-tapped lands, which is why the "every precon ever printed" line above holds. But the engine's land ranges are calibrated for 100-card Commander decks, so its land-count and color math for a 60-card deck is not meaningful and none of it is counted here.
• The six-swap default is a floor for a lot of decks. 65 of the 89 decks that got swaps hit the cap, so the full fix for those is bigger than what one click shows you.
• Prices are the cheapest paper printing on the day. Reserved List prices especially move around.
• The health bar is our engine's opinion. The land ranges and color math are derived per commander, and the accuracy page covers how the engine gets audited. Every cut and add ships with its reason printed next to it, so you can disagree with any of them on the merits.

Every precon in this post is in the library with its full rating, decklist, and a Fix my mana button.

Browse all 169 precons

Card art is the property of Wizards of the Coast, images courtesy of Scryfall. CommanderBracket is unofficial Fan Content and is not endorsed by Wizards.