The info below is simply FYI and to help someone learn and have fun with hash cracking.
This is by no means a conclusive demonstration of which wordlist or ruleset is "best".
The "best" wordlist + rules is the one that generates the plaintext!
Having a good wordlist + rules is vital to hash cracking. I use my own wordlist + rules when cracking, but for these tests, I'll only be using publicly available wordlists.
To set the test up, I compiled 1,995,899 plaintext from the stockx database, then hashed these to md5.
I used publicly available wordlist and rules, but also compared the results to my own rules.
100% of cyclone's rules have been handwritten or generated by cyclone and trained against scores of databases. These were given as a comparison. They're not always better, but as you will see below, they generally do well.
Test set:
stockx plaintext - 1.9 m
Wordlists:
rockyou + phpbb - 146 mb
hashkiller (Jan 2020) - 2.9 gb
Rules:
best64 - 77 rules (included with hashcat)
T0XlC - 4085 rules (included with hashcat)
top_1500 - top 1,500 rules (from @blandyuk's hashcatgui)
top_5000 - top 5,000 rules (from @blandyuk's hashcatgui)
d3ad0ne - 34,099 rules - d3ad0ne
OneRuleToRuleThemAll - 51,998 rules - notsosecure
cyclone_mini - 195 handwritten rules - cyclone
cyclone_250 - top 250 rules - cyclone
cyclone_1500 - top 1,500 rules - cyclone
cyclone_5000 - top 5,000 rules - cyclone
cyclone_mst - top 50,000 rules - cyclone (to replace OneRuleToRuleThemAll)
Results:
Rockyou + phpbb:
Hashkiller:
Credits to all those who were involved with creating the rules & wordlists used above.
This is by no means a conclusive demonstration of which wordlist or ruleset is "best".
The "best" wordlist + rules is the one that generates the plaintext!
Having a good wordlist + rules is vital to hash cracking. I use my own wordlist + rules when cracking, but for these tests, I'll only be using publicly available wordlists.
To set the test up, I compiled 1,995,899 plaintext from the stockx database, then hashed these to md5.
I used publicly available wordlist and rules, but also compared the results to my own rules.
100% of cyclone's rules have been handwritten or generated by cyclone and trained against scores of databases. These were given as a comparison. They're not always better, but as you will see below, they generally do well.
Test set:
stockx plaintext - 1.9 m
Wordlists:
rockyou + phpbb - 146 mb
hashkiller (Jan 2020) - 2.9 gb
Rules:
best64 - 77 rules (included with hashcat)
T0XlC - 4085 rules (included with hashcat)
top_1500 - top 1,500 rules (from @blandyuk's hashcatgui)
top_5000 - top 5,000 rules (from @blandyuk's hashcatgui)
d3ad0ne - 34,099 rules - d3ad0ne
OneRuleToRuleThemAll - 51,998 rules - notsosecure
cyclone_mini - 195 handwritten rules - cyclone
cyclone_250 - top 250 rules - cyclone
cyclone_1500 - top 1,500 rules - cyclone
cyclone_5000 - top 5,000 rules - cyclone
cyclone_mst - top 50,000 rules - cyclone (to replace OneRuleToRuleThemAll)
Results:
Rockyou + phpbb:
Code:
no rules (2.70%)
best64 (6.86%)
cyclone_mini (22.39%)
T0XlC (22.66%)
top_1500 (24.84%)
cyclone_250 (26.57%)
top_5000 (39.22%)
d3ad0ne (41.12%)
cyclone_1500 (41.58%)
OneRuleToRuleThemAll (50.77%)
cyclone_5000 (51.58%)
cyclone_mst (61.86%)Hashkiller:
Code:
no rules (34.50%)
best64 (37.98%)
cyclone_mini (52.77%)
T0XlC (54.66%)
top_1500 (58.71%)
cyclone_250 (59.01%)
top_5000 (71.54%)
cyclone_1500 (71.73%)
d3ad0ne (75.83%)
cyclone_5000 (79.25%)
OneRuleToRuleThemAll (79.64%)
cyclone_mst (86.10%)Credits to all those who were involved with creating the rules & wordlists used above.
Last edited:
