The worst prediction in the history of physics is the standard model of cosmology's foundation stone. Quantum field theory, asked for the energy density of the vacuum, returns an estimate of order 10^74 m^-2 in curvature units; the cosmological constant actually measured from supernovae, the CMB, and large-scale structure is 1.1 x 10^-52 m^-2. The discrepancy spans roughly 120 orders of magnitude in natural units, and ΛCDM's response is to write the measured number into the field equations by hand: Lambda is an input, its origin unexplained, its smallness unaddressed, its near-perfect cancellation against the enormous vacuum contribution an unexamined miracle the model is built on top of.
Every attempted explanation has failed structurally. No symmetry principle cancels the vacuum energy without canceling it exactly to zero, which the nonzero measurement forbids; supersymmetry, which would have helped, is broken at scales that leave a remnant discrepancy of sixty orders of magnitude and has in any case failed to appear at the LHC; and anthropic selection across a multiverse landscape, the surviving mainstream proposal, abandons prediction altogether, explaining the value by declaring all values realized somewhere. The problem is widely treated as too hard to be on any experiment's agenda: a 10^120 misprediction folded into the model as a boundary condition and politely not discussed.
The standing is foundational and permanent within the standard framework: the model cannot derive its own most important parameter, the measured dark energy sector now shows hints of dynamics (the DESI evolving preference) that a true constant cannot accommodate, and any framework in which the cosmological term is emergent rather than fundamental dissolves the catastrophe by construction.