A random number is a number that cannot be described. And you in the javascript code will describe it. This is contrary to the definition of a random number. But you can calculate a pseudo-random ("random type") number. How exactly - I do not know. And all the generators of pseudo-random numbers, including Math.random (), are implemented - they consider not a random number, but a difficultly predictable number.
As for the date - I myself made such a generator recently, I decided to write simple games in javascript. I had to run every time when I ran around the cells of the playing field to get a “type of random” number and, based on it, to conclude what exactly to do with the current cell of the playing field. Milliseconds instantaneous picked up, processed them (for example, took the remainder of the division by 2, that is, figured out an even number or an odd number millisecond in my hands). At first, the option seemed very cool to me, because I do not know how many milliseconds at a particular point in time it runs. But it turns out garbage in the end, and now I will explain why. The browser considers everything very quickly. If you push this "generator" code into a function and thus "generate" (calculate) "type of random" numbers in a row (that is, call this function in a row, call after call), the browser will make several passes through this function less than in 1 millisecond. You can try it yourself, and in the end you will get random numbers "in the same bundles", for example, 5 times an even number, 4 times an odd number .... The first option that I came up with was to load the browser with calculations inside the function so that it would count. I pushed the calculations of a huge factorial there, changing which, regulated the operation time of the function. But it was also necessary to regulate the probability of the events that this function generates. The probability of my event was about 1/2. To achieve 1/6, it was necessary to run the function 3 times with this factorial. In the end, yes, the function worked as I needed and gave me the probability I needed. But it takes a long time to run through all the cells of the playing field in this way - about a second. So no one does. After all, the browser is single-threaded. And I take it up for a full second. This is not an option. Everything else will slow down.
Another option that I did - you can translate the number of these milliseconds into a string and "get" each digit. And do something with it. But, all the same - the browser will issue a series of identical milliseconds. Which need to somehow weed out.
If you think well, then I think you can find an option. Now I came up with a slightly different option, I do not know whether it will work as it should. Every time I get a "0" or "1" generator, I stuff this thing into the array (i). Then again I get a pseudo random unit or zero (i + 1). And I throw random numbers into the array. Then I get the third number zero or 1 (i + 2). If all three of these numbers were in the same "party", which managed to miscalculate in less than 1 millisecond, then they will be the same. Take, for example, the second and third and delete. Only one number remains in the array. We run the function again and hammer three variables into the array again. If they are all the same, then the last two are deleted ... And so on. An array with random numbers "1" and "0" is clogged. Thus, I think you can reduce the repeatability of the same result in a series of function generator calls. In short, you just have to try.
