Put another way, so far string theory is an elaborate kind of curve-fitting — a mathematical construction that matched the empirical data available at its inception, but one that to this day has never predicted a novel measurement. Unencumbered by the theory, we would not have expected any recent experiment to produce different results.