FIDE's formula is mathematically elegant — every player's rating is updated the same way, regardless of whether they're a 2700 GM or a 1200 beginner. But the information density of a rating change is wildly different at different levels.
A 1200 player who beats a 1100 player in a school event has just demonstrated something. Was it tactical sharpness? Endgame technique? Or just that the opponent blundered a piece on move 12? FIDE's formula can't tell you. It just sees the result.
This is why NCR layers additional signals on top of the Elo math: opponent strength bucketing, colour balance, performance under pressure (final-round results weighted differently). Same underlying formula. Richer downstream report.
The numbers below 1400 move faster
One consequence of broader Elo: under-1400 players see 20–60 point swings per event. That's healthy — it means the rating is genuinely learning who they are. By 1500 it stabilises. By 1700 it's noisy only at the margin.