What does 2 mana buy? Magic creature power creep, measured
The power-creep argument has raged for thirty years on anecdotes. Power and toughness are printed on the card, so we measured every new-design creature Magic ever released, in its debut year. The verdict: for two decades the typical 2-drop was worse than Grizzly Bears — and then the mid-2010s repriced the whole curve. No market data, just the stats.
Pushed 2-drops by year
Bar = share of that year’s new 2-mana creatures with 5+ total stats (a 3/2 with upside, a Watchwolf, or better). Right columns: mean total stats at 2 and 3 mana.
The curve got repriced, once
The data kills two popular versions of the story. Power creep is real — but it is not a smooth 30-year drift, and it did not start with Alpha’s famous overpowered spells. Creature stats were remarkably FLAT (and modest) from 1993 through the mid-2000s: the average new 2-drop carried fewer total stats than a vanilla 2/2. The reprice happens in one visible step around mid-2010s, and the new level has held since — bigger bodies, and abilities stacked on top rather than traded for stats. It is the same era Magic’s output exploded (see the Reprint Ratio): more cards, and stronger ones.
| Era | Mean stats @2 mana | Pushed 2-drops | Mean stats @3 mana |
|---|---|---|---|
| 1993-2004 (the long baseline) | 2.88 | 6.7% | 3.78 |
| 2005-2014 (creeping, slowly) | 3.15 | 6% | 4.13 |
| 2015-2019 (the reprice) | 3.58 | 8.7% | 4.55 |
| 2020-2025 (the pushed era) | 3.64 | 11.4% | 4.87 |
Method & sources
- Every NEW-design creature via the open Scryfall API, counted once in its debut year (a card’s first printing is not a reprint). English paper; tokens and memorabilia excluded; creatures with non-numeric stats (*/X) skipped.
- “Total stats” = printed power + toughness. “Pushed” = 5+ at 2 mana / 6+ at 3 mana — the statline that made Watchwolf a tournament card in 2005. Stats measure BODIES only; abilities have also grown, so this measurement is conservative about the true power level.
- No market data is used. Known facts reproduce: 1994 and 2005 medians both sit at 3 total stats at 2 mana; the shift lands mid-2010s. Valued 2026-07-07; reproducible from the open API.
FAQ
Is Magic power creep real?
For creatures, measurably yes — and it arrived late. The median new 2-mana creature carried 3 total stats (power + toughness) from 1993 through the mid-2000s: literally worse than Grizzly Bears. It reached 4 in the mid-2010s and stayed there, while the share of "pushed" 2-drops (5+ total stats, Watchwolf-class) went from 6.7% of new designs to 11.4% in the 2020s.
What was the baseline for a 2-mana creature?
Grizzly Bears — a vanilla 2/2 for two mana, printed in Alpha. The surprise in the data is that for Magic's first decade the AVERAGE new 2-drop was below that bar (stats came with drawbacks or went to abilities). Watchwolf (a 3/3 for 2) was considered format-defining in 2005; in the 2020s roughly one new 2-drop in 9 matches or beats those stats, usually with abilities attached.
Did 3-mana creatures creep the same way?
Yes, on the same schedule: median total stats went 4 → 5, and 6+ total-stat 3-drops went from 10.1% to 29.6% of new designs. The creep is a body-per-mana repricing across the curve, not a single famous card.
How was this measured?
Printed power and toughness of every NEW-design Magic creature (each card counted once, in its debut year) via the open Scryfall API (printed creature stats via the open Scryfall API (new designs, debut year), valued 2026-07-07); English paper, tokens and memorabilia excluded, non-numeric stats (*/X) skipped. No market data is used. Known facts reproduce: the 1994 and 2005 medians sit at 3 total stats at 2 mana, and the modern shift lands in the mid-2010s — matching the design era players point to.
Binder full of pre-reprice creatures? See what they're worth.
Foilio values any card across Pokémon, Magic, Yu-Gi-Oh, One Piece, Lorcana and more, live prices in EUR & USD, your whole collection tracked. Free, no signup.
Related research
Pokémon HP Inflation
The sibling study: Pokémon's ceiling exploded while its floor crept.
16,322The Reprint Ratio (Magic)
The same mid-2010s inflection in output: more new designs in 11 years than the first 22.
2.69×The Modern Horizons Tax
Where pushed designs get printed: MH rares out-price normal-set rares.
Cite or embed this study
Free to use with attribution (CC BY 4.0). Paste the embed on your site for the chart card + a link back, or grab the plain citation.
<a href="https://foilio.net/studies/mtg-creature-creep?utm_source=share&utm_medium=study_embed&utm_campaign=viral" target="_blank" rel="noopener">
<img src="https://foilio.net/studies/og?stat=6.7%25%E2%86%9211.4%25&label=share%20of%20new%202-mana%20creatures%20with%205%2B%20total%20stats%20(Watchwolf-class)%2C%201990s%20vs%202020s&title=What%20does%202%20mana%20buy%3F%20Magic%20creature%20power%20creep%2C%20measured" alt="What does 2 mana buy? Magic creature power creep, measured, Foilio" width="600" style="max-width:100%;height:auto;border-radius:12px" />
</a>
<p style="font:13px/1.5 system-ui,sans-serif;color:#555;margin:6px 0 0">
Source: <a href="https://foilio.net/studies/mtg-creature-creep?utm_source=share&utm_medium=study_embed&utm_campaign=viral">What does 2 mana buy? Magic creature power creep, measured, Foilio</a> · data CC BY 4.0
</p>Foilio (2026-07-07). What does 2 mana buy? Magic creature power creep, measured. Retrieved from https://foilio.net/studies/mtg-creature-creepThe big picture, across every game we measure: The Chase Multiple · The Price Census · The Promo Premium · The Foil Premium