bootstraps

Bootstraps is the plural of bootstrap and names several concepts that share the idea of starting with minimal resources and building up to a functioning system. The related idiom pull yourself up by your bootstraps dates from the early 19th century and has become a metaphor in technical and scientific contexts for creating capability from little external help.

Computing uses “bootstrap” to describe the startup sequence that prepares a system for use. A small bootstrap program initializes hardware, loads a bootloader, and then starts the operating system kernel. Modern firmware such as BIOS or UEFI performs initial hardware tests and locates a bootloader on storage; the bootloader then loads the kernel and user environment. Bootstrap techniques also appear in software development, where a minimal compiler or environment is rebuilt from basic components.

In statistics, bootstrapping is a resampling method for estimating the distribution of a statistic by repeatedly sampling with replacement from observed data. Introduced by Bradley Efron in 1979, it allows estimation of standard errors, confidence intervals, and bias without strong parametric assumptions. Variants include nonparametric and parametric bootstrap, bootstrap-t, and bias-adjusted intervals. While computationally intensive, bootstrapping is widely used for its flexibility across diverse data types and models.

Other senses of bootstraps include physical straps or loops used to help put on boots, and the common idiom emphasizing self-reliance. In fashion and apparel, bootstraps may refer to hardware or features designed to aid wearing boots.