digits of the result of Math.random()

for (let i = 0; i < 999; i++) {
    console.log("length-result:", (Math.random() + "").slice(2).length)
}

If you execute this code snippet, you will find that most of the results are less than 18.

However, theoretically speaking, the last digit of the result of random() only has a ten-percent opportunity to be 0. It means, only ten percents of the length-results can be less than 18.

Live snippet:

// Why are results of length 20 rare, rather than occurring ~90% of the time?

for (let i = 0; i < 999; i++) {
  const str = Math.random() + '';
  tbody.insertAdjacentHTML('beforeend', `<tr><td>${str.length}</td><td>${str}</td></tr>`);
}

td {
  padding-right: 20px;
}

<table>
  <tbody id="tbody">
  </tbody>
</table>

Can you tell me what caused this behavior of Math.random?