What if you could just narrow it down a bit. Third character is in this range, fourth in this range, and so on.
Characters of the private key? Well then your number of overall possibilities for the correct private key would simply be "number of possibilities for second character" * "number of possibilities for third character" * "number of possibilities for fourth character" * "number of possibilities for fifth character" * ... * "number of possibilities for the last character".
Once you have that number then you need to work out how many times you can perform elliptic curve multiplication, SHA-256, RIPEMD-160, and then two more SHA-256s each second, to turn each private key in to an address. DaveF made some benchmarks for doing this a couple of years ago:
https://bitcointalk.org/index.php?topic=5112311.msg50823897#msg50823897. With the top end GPUs on the market - GeForce RTX 2080 Super/Ti - you are looking at about 2 - 2.5 billion keys per second.
Are you asking what are the possibilities of successfully finding the private key of an address by brute forcing? If so, as @hosseinimr93 said, 1 in 2256.
It's actually 1 in 2
160, on average, since there are 2
96 private keys per address, on average.
Specifically, if you had that list and a machine that performs 100 trillion ECDSA and SHA256 hashes per second
It is elliptic curve multiplication, not ECDSA, that is needed here, and there are more hashes required than just a single SHA-256 as I've outlined above.