The cosmic star formation history is the universe's production curve: the mass of new stars formed per unit time per unit volume, traced across redshift through ultraviolet and infrared luminosity functions. The curve rises from early times to a pronounced peak at cosmic noon, redshift 2 to 3, then declines by roughly an order of magnitude to the present (Madau and Dickinson 2014).
The shape is the problem. In ΛCDM, star formation should track hierarchical halo assembly, vigorous when structure is assembling fastest, yet the observed history misbehaves at both ends. At high redshift, JWST keeps finding star formation already proceeding efficiently in massive systems when the model's halos are barely assembled, the same strain that produces the early-galaxy mass crisis, with the most extreme systems requiring baryon-conversion efficiencies near 50 percent where roughly 20 percent is the model's observed ceiling at any later epoch (Bouwens et al. 2016; Xiao et al. 2024). At low redshift, the steep post-noon decline outpaces what halo-assembly rates alone explain, requiring quenching prescriptions tuned after the fact. The curve's gross features can be fit, but only by letting feedback do whatever each epoch requires.
The standing is sharpened by every JWST cycle: the high-z end keeps filling in with earlier, more efficient star formation, stretching the gap between when the model can build the fuel tanks and when the engines are observed running.