Mickaël Nadé

27 yrs · Centre-back · France
AS Saint-Étienne · Ligue 1 · Foot: left · Height: 193 cm
Contract: until 30 Jun 2028
No model evaluation for this player. MoneyballAI evaluates players with at least 450 league minutes last season in leagues where match-level player data is available from the current free source (14 European top divisions). Market value and transfer history are still shown below.
Market value
€2.0M
Updated 7 days ago
Estimated acquisition
—
Estimated · 20–80% range
Model value
—
Performance-implied · estimated
Projected (24m)
—
Resale range · estimated
Potential value gap
—
Model value − acquisition

Market value history

Market value ≠ transfer fee ≠ acquisition cost.

Highest recorded: €3.0M

Season-by-season (domestic league)

Context-adjusted values are model outputs.

SeasonClubLeaguePosAppsMinMin shareGAG+A/90Adj. G+A/90Opp. EloTeam EloLeague str.
2024/25AS Saint-ÉtienneLigue 1CB26227874%130.160.20146213070.88
2021/22AS Saint-ÉtienneLigue 1CB30256875%100.040.06143613210.86
2016/17AS Saint-ÉtienneLigue 1CB1903%000.000.05127714570.88

Historical comparables

Most similar profiles 2+ years ago (same position group) and what happened next. Chosen by similarity only — failures included.

No comparables computed for this player.

Transfer history

Fees as reported by the source; undisclosed fees shown as unknown, never estimated.

DateFromToFeeMV at time
30 Jun 2021Quevilly RouenSaint-ÉtienneFree / loan€500k
21 Jul 2020Saint-ÉtienneQuevilly RouenFree / loanData unavailable
1 Jul 2020Saint-Étienne BSaint-ÉtienneUnknownData unavailable
1 Jul 2018St-Étienne U19Saint-Étienne BUnknownData unavailable
1 Jul 2016St-Étienne U17St-Étienne U19UnknownData unavailable
1 Jul 2014Sarcelles Jgd.St-Étienne U17Free / loanData unavailable

MoneyballAI identifies statistical opportunities from available data. A player flagged as a potential bargain is not guaranteed to succeed, to transfer, to rise in value, or to generate resale profit. Market values, fees and contracts come from third-party sources; ranges are model estimates with stated confidence.