JEE Rank Predictor

Estimated rank ≈ N × (1 − S / S_max)^k,  reported as a band of roughly ±15%

• Syour score out of 300
• S_maxmaximum marks, 300 for JEE Main
• Nassumed number of candidates appearing
• kcurve exponent fitted to past score-to-rank data; steeper values crowd more candidates into the middle of the range

• Treat the band as the answer and the point estimate as an artefact. Real score-to-rank tables shift by thousands of ranks between years at the same score, because the pool grows or shrinks and the paper gets harder or easier. Any predictor claiming a precise rank is overstating what the data supports.
• JEE Main is conducted in multiple sessions and ranks come from normalised percentile scores, not raw marks, with the best of a candidate two attempts counted. That normalisation is computed after all sessions close, so no prediction made from a raw score can reproduce it.
• Error grows fastest at the top. Between 280 and 290 marks a handful of candidates separate three-digit ranks from four-digit ones, so the relative uncertainty at the sharp end is far worse than the same percentage band suggests lower down.
• JEE Advanced is a different paper with a different maximum, its own difficulty profile, and a candidate pool restricted to roughly the top 2.5 lakh from Main. A curve fitted to Main scores does not transfer to it.
• Rank is not admission. Category reservation, home-state quota, and branch-wise closing ranks decide seats, so the useful next step is checking JoSAA opening and closing ranks for your category across the last three rounds rather than refining the rank estimate.