The generator below, seeded with "Example" will produce the same values as below for each of these random number distributions across all browsers.

var rng = new RNG('Example');
rng.random(40, 50);  // =>  42
rng.uniform();       // =>  0.7972798995050903
rng.normal();        // => -0.6698504543216376
rng.exponential();   // =>  1.0547367609131555
rng.poisson(4);      // =>  2
rng.gamma(4);        // =>  2.781724687386858

The underlying algorithm is RC4 and uniform number generation is about 10x slower than Math.random in V8. What you get in exchange for that is a seedable generator and additional random distributions (see example). You can still get speed and these additional distributions by using Math.random as the core uniform number generator.

var rng = new RNG(Math.random);

When no seed is provided, one is created randomly from available entropy sources. Seeds that are not strings are run through JSON.stringify() before being used.

Here's how you would replace Math.random with a seeded generator.

Math.random = RNG.prototype.uniform.bind(new RNG('my seed'));

Finally, for fun, a dice roller,

var dice = RNG.roller('4d6 + 10');
dice(); // => 17
dice(); // => 11

var RNG = require('rng-js');

This module can also be browserified thanks to browserify-shim.