Two different systems, often confused
The ICC publishes two separate sets of rankings, and they work in genuinely different ways. Team rankings rate sides in Tests, ODIs and T20Is on a points-per-match rating. Player rankings rate individual batters, bowlers and all-rounders on a 0–1000 scale. People often quote one while meaning the other, which is where most of the confusion starts.
How team ratings are built
A team's rating is, in essence, an average. The board takes the points a side has earned across a rolling window of recent series and divides by the matches played. The result is a single number, typically somewhere between 0 and 130 or so, and the side with the highest number sits top of the table.
The crucial detail is that points earned depend on the opponent's rating, not just the result. Beat a side rated far above you and you gain a lot. Beat a side rated far below you and you gain very little — and if you lose to them, you can drop sharply even though it was only one match. This is why a dominant series win against a lower-ranked team often barely moves a top side up the table.
Results are also weighted by age. Recent series count in full; older ones are weighted down before they drop out of the window entirely. A team can therefore slide down the rankings without playing badly, simply because a strong old result has aged out.
How player rankings differ
Player rankings do not average across a career. They run on a points system where each innings or bowling performance is scored against the difficulty of the context: the quality of the opposition, the match situation, the strength of the attack or batting line-up faced, and whether the team won.
A hundred against the top-ranked bowling attack in a winning cause is worth substantially more than a hundred against a weaker side in a dead rubber. This is deliberate — it stops flat-track accumulation from carrying a player to the top of the table.
Points also decay. A player who stops playing does not hold their rating indefinitely; the number erodes as their performances age. This is why returning players often re-enter the list well below where they left it.
Why the rankings sometimes look wrong
Three things account for most of the disagreement between the table and what fans feel. The rolling window means the rankings lag public perception by months. The opponent-weighting means the fixture list matters as much as form — a team with a soft schedule has fewer chances to gain. And in Tests especially, sides play wildly different numbers of matches, so a team can hold a high rating on a small sample.
None of this makes the system broken. It makes it a measure of sustained performance against quality opposition, which is a narrower thing than 'who is playing best right now'. Read it as that, and it stops looking strange.
The short version
Team rankings reward beating good sides and punish losing to weak ones, on a rolling recent window. Player rankings reward difficult performances in winning causes, and decay when you stop playing. Neither is a form table, and neither is meant to be.