--keyspace calculation showing larger number for 4-word passphrases than 5-word passphrases

[huertanix]

Not sure if this is a bug or just a misunderstanding from my end on how prince derives permutations, but I've been using the --keyspace operator to determine permutation counts for various passphrase sizes (e.g. four-word passphrases versus five-word passphrases) and have run into an issue where the calculated keyspace for five-word passphrases is significantly lower than for four-word passphrases, and this feels like it should not be correct. Here's the wordlist I'm using: https://github.com/freedomofpress/securedrop/blob/develop/securedrop/wordlists/en.txt and here's the example commands and output:

Four-word passphrases

./pp64.bin --dupe-check-disable --elem-cnt-min=4 --elem-cnt-max=4 --keyspace < ~/wordlists/en/securedrop/en.txt
408718275456

Five-word passphrases

./pp64.bin --dupe-check-disable --elem-cnt-min=5 --elem-cnt-max=5 --keyspace < ~/wordlists/en/securedrop/en.txt 
109051768512

I'm using Ubuntu 22.04.1 on a 64-bit Intel CPU, glibc version 2.35 in case that helps.

[roycewilliams]

https://hashcat.net/wiki/doku.php?id=frequently_asked_questions#what_is_a_keyspace