There are charts about most used letters in English. What about the letter order? For example letter X has a bigger chance to appear than letter Y as the first character and something else for the 2nd character.
In terms of English language, what you are asking about are digram frequencies.
I wrote a PHP script to create a basic output.
Here are the results:
It can be better with a better source though.
This is heavily dependent on which corpus of passwords is used, but what you're looking for is Markov chains trained on common password corpora.
Markov chains are superior to simple frequency-position tables because choice of next character is often heavily influenced by more than just the previous character, but a few characters back.
For example, this model is based on the original RockYou leak.