Internet Genius Knows what you are Thinking

Active Member
✔️ HL Verified
💻 Oldtimer
Joined
Mar 13, 2005
Messages
3,877
Best answers
0
I remember this thing from long ago!


I was thinking of Spike and it gave me Paul somethinganother.
 
Last edited:
New Member
💻 Oldtimer
Joined
May 7, 2003
Messages
1,560
Best answers
0
Location
California
I never had a chance to try more, the site kept going down. Lame. I was surprised how it guessed Vegeta in under 10..rather general and/or not-very-specific questions. (I bet this thing tracks your web history the second you enter the site).
 
New Member
💻 Oldtimer
Joined
May 13, 2003
Messages
2,904
Best answers
0
Have you kept up with it over the years?
He's basically a hyperzoanoid tool for the cabal of double crossing zoalords. He's bad ass enough to beat the new Guyver 2 (Valkyria), but he's not terribly special at the moment.

To be fair, he's my second favorite villain behind
Agito Makashima|Guyver III
. And he is made of all kinds of epic badass in the first half.
Unfortunately, no. All I've seen was the 1982 OVA, the 2005 "remake" and the live action movies (David Hayter, hah). I haven't read any of the mangas. Guyver is still the first anime / manga I saw and it's still one of my favourites, even if I haven't kept up with it. :p
 
New Member
Joined
Nov 24, 2001
Messages
692
Best answers
0
I spent too much time on this yesterday and more or less designed a data schema (in my head mind you) of how I would make this.

Yes it's not that complicated, it's a series of permutations that eventually narrows it down to a small list of results where it tends to pick the more common result. The impressive part as has been mentioned is the data. It would be painstakingly time consuming to input ALL the characters he has (and make no mistake there is a ridiculous amount of characters in there). I'm sure there is also nuances in the logic that draws out the most relevant questions (the amount of times it asked me if my character was from the Warcraft series without any other indication for it to ask that was impressive). Either way, I think it's pretty cool and am determined to come up with an obscure character that doesn't exist!
Drawing out the right questions is the relevant point, why are you guys so obsessed on the data representation? You have a list of character candidates, with certain binary properties (linked to questions). You start out with the full list, then ask a question (=select a property) that no matter the answer, cuts the list of potential characters down to something as close as possible to half. Repeat until one character is left, then guess that one. If it's not correct, add the newly added character with all properties on the question-path. The impressive part is defining a list of properties that nicely discriminate your characters, and having a huge list of characters.
 
New Member
Joined
Nov 24, 2001
Messages
692
Best answers
0
You'd need some serious maths skills to come up with an effective algorithm for determining the best questions to ask to most efficiently trim down alternative possibilities and come closer to the answer. Though, it's probably already been been figured out and is fairly easy to access knowledge by now.

As has been said the most impressive part of this is the sheer amount of data, which is due to years of spreading virally and allowing users to add their own characters or details. It'd take a long time to gather that same level of information.
Not really, a simple bisection as I explained above will do and is (on average) the most efficient way to do it.
 
New Member
💻 Oldtimer
Joined
Oct 4, 2004
Messages
1,572
Best answers
0
Location
Norge
You'd need some serious maths skills to come up with an effective algorithm for determining the best questions to ask to most efficiently trim down alternative possibilities and come closer to the answer. Though, it's probably already been been figured out and is fairly easy to access knowledge by now.
I have a suspicion that you have never studied mathematics or computer science in university or college.

As has been said the most impressive part of this is the sheer amount of data, which is due to years of spreading virally and allowing users to add their own characters or details. It'd take a long time to gather that same level of information.
It's not even that impressive. The "user generated" content is also the reason this thing is ridiculously inaccurate. It's a neat project, but nothing spectacular, and nothing that hasn't been done for years.

I know of a primitive version of this that dates back around fifty years if not more (using numbers).
 
Former Forcepit Member :(
✔️ HL Verified
💻 Oldtimer
Joined
Aug 21, 2006
Messages
1,717
Best answers
0
Location
korriban
HA beat the MOFO he couldn't Guess Fasha
 
Active Member
✔️ HL Verified
🚂 Steam Linked
💻 Oldtimer
Joined
Sep 23, 2002
Messages
1,876
Best answers
0
Location
Fryslân Boppe! The Netherlands
Since when is complexity important when it comes to amusement? It's not about if it's hard to do, it's about how it's used.
 
New Member
💻 Oldtimer
Joined
Oct 4, 2004
Messages
1,572
Best answers
0
Location
Norge
That's why Prozac gets more fun out of his coffee-cup than he does out of the entire Internet.
 
Pwns Mastasurf at TF2
Retired Forum Staff
✔️ HL Verified
💻 Oldtimer
Joined
Dec 7, 2001
Messages
5,115
Best answers
0
try using characters from RObert Heinlen novels from the 1960s :p Couldn't get anyone from "The MEnace from Earth" :p
 

Users who are viewing this thread

Top Bottom