No one ruler reaches the edge. We measure the universe in rungs: parallax hands off to Cepheids, Cepheids to exploding stars, exploding stars to the expansion itself — each rung calibrated on the one below. Climb it, and watch the error compound with every step out.
Four real rungs at their true reach: parallax (~kiloparsecs), Cepheid period–luminosity (~tens of Mpc), Type Ia supernovae standard candles (~hundreds of Mpc), Hubble redshift (beyond). Each rung is calibrated on the one below, so its fractional error is added in quadrature to everything under it — the readout compounds √(Σεᵢ²) as you climb, showing why the far universe's distances carry the uncertainty of every closer rung. The distances are order-of-magnitude accurate; the compounding is the real lesson.
The ladder is drawn at representative ranges, not a live re-derivation of the distance scale; real calibration (e.g. the exact Cepheid zero-point, the SN Ia standardisation) is an active field with its own tensions (the Hubble tension is precisely a disagreement between rungs). The point stands: nothing far is measured without trusting everything near.