Mana Base Analysis

The land count comes from Frank Karsten's source-count regression for singleton (Commander) decks: a target driven by average mana value, ramp and card draw at three mana or less, and fast mana, with a credit for each commander. Modal double-faced land-backs count directly toward the actual land total rather than adjusting this target. cEDH mode lowers that target for the faster competitive format (about a 3.5-land cut, crediting fast mana fully); when there is enough cEDH tournament data for the commander, the recommendation is nudged toward that real-deck land mean and shown as a range.

Each spell's castability comes from a Monte-Carlo simulation. It draws an opening seven and applies the London mulligan, including Commander's free first mulligan — the first mulligan keeps seven at no cost; only later ones bottom a card (lowest-impact first, never lands). It then draws every turn, including turn 1. A card is judged on its mana-value turn, so by turn N it has seen its opening 7 plus one card per turn (7 + N cards). Each turn it deploys ramp as it comes online, credits fetchlands to the colors they can fetch, and applies any always-on cost reducers — then asks, on that turn, whether there is enough mana and the right colored pips together (a joint check, not two independent odds). The reported number is the share of simulated games where the card was castable on curve.

The sim plays going first for board development (one land drop per turn, no earlier plays), which is the conservative case for land sequencing — but it draws every turn, so the card count matches real games.

Colored-source findings combine a worst-driver spell (the hardest single card for a color) with a population view (how many of that color's cards are under-supported and their mean castability). The weakest color surfaces tail risk first. The commander-importance setting weights the commander's colors more or less heavily — Central holds them to a stricter, cEDH-style threshold.

The overall health reads on a four-tier scale — Excellent, Solid, Workable, Needs work — that combines Karsten source counts with the simulated avg-on-curve headline. A card that casts late only because it is expensive (a curve problem the base can't fix) never fails the verdict. Needs work means a severe or broad shortage: a color short by more than about two sources, or two-plus colors short — or a land shortfall the simulation corroborates (a color issue or broad simulated under-support riding alongside). Broad under-support on its own does not red the verdict. If the simulation says the deck functions, with no catastrophic color, no severe color deficit, and no broad color-access shortfall, a land-count shortfall plus one soft color issue is Workable, not red. Solid is a working base with minor notes; Excellent is clean.

Credit: Frank Karsten's published source-count work. The castability estimate is our own Monte-Carlo simulation, cross-checked against community mana calculators including Salubrious Snail. These figures are an estimate; treat them as a ranking aid, not a guarantee.