Gaming Clairvoyance in Puzzle Quest

Hooya

Well-Known Member
OP
Member
Joined
Aug 25, 2006
Messages
1,878
Trophies
0
Age
42
Location
Central Illinois
Website
Visit site
XP
317
Country
United States
I got into Puzzle Quest after everyone else on the board did, so this might be old news, but I searched and didn't find anything about this specifically. People complained that the computer was all-knowing (clairvoyant) but I'm pretty sure that it is, and I have made a video proving that it can be.

The puzzles are not randomly generated. At all. I have the same puzzle the first (and second puzzles are always the same, and the third, but I only memorized a bit of the first). I'm uploading a youtube video right now of this and will edit this post with the url once it's been posted.

I still think this is a fun game, but the fact that there is nothing random in the puzzles annoys me, it could have been much better if it were random like Bejeweled was.

Here's the video:

http://www.youtube.com/watch?v=pQoDtPOJwlA
 

Hooya

Well-Known Member
OP
Member
Joined
Aug 25, 2006
Messages
1,878
Trophies
0
Age
42
Location
Central Illinois
Website
Visit site
XP
317
Country
United States
If they weren't random what's the point of downloading puzzles via wi-fi or the ability to "create your own"?
Actually,random puzzles never crossed my mind....

those are for single clear out puzzles, like when you are attempting to capture a monster, it's the same every time and there's a specific solution. This has to do with the "random" endless puzzle cascade that should happen when you take part in normal battles. There's no real solution because there shouldn't be an ending, but theoretically I could memorize the best course of action for that initial puzzle and be able to completely rape any opponent because I'll always be able to get 4 of a kind or better. After that fight is over, leave the game, and the next time I start it's the same puzzle again...

The point is that the creators of the game claim that the computer isn't all knowing and can't predict future cascades to do better than a human player, but I prove that this cannot be the case because the puzzles are pre-conceived.
 

Jeda

Well-Known Member
Member
Joined
Nov 7, 2002
Messages
329
Trophies
0
Age
41
Location
Germany
Website
Visit site
XP
361
Country
Gambia, The
There is no true random in a computer. Pseudo randomizers can be initialized from several things, like the system clock for example. The DS likely uses something in the savegame.
So of course if your savegame doesn't change, the RNG will be initialized from the same data resulting in the same board and the same new stones getting created until a single change happens.

You're saying "somehow the first move stops that chain", but if you look close to what the first stone that gets created over the red one is, you'll see that it is different one. Now if the whole board was indeed preset it would be the same.

So you've proven that boards are random instead....

Note that the computer could still be knowing what block will get created by precalculating the RNG's next outputs
 

Hooya

Well-Known Member
OP
Member
Joined
Aug 25, 2006
Messages
1,878
Trophies
0
Age
42
Location
Central Illinois
Website
Visit site
XP
317
Country
United States
There is no true random in a computer. Pseudo randomizers can be initialized from several things, like the system clock for example. The DS likely uses something in the savegame.
So of course if your savegame doesn't change, the RNG will be initialized from the same data resulting in the same board and the same new stones getting created until a single change happens.

You're saying "somehow the first move stops that chain", but if you look close to what the first stone that gets created over the red one is, you'll see that it is different one. Now if the whole board was indeed preset it would be the same.

So you've proven that boards are random instead....

Note that the computer could still be knowing what block will get created by precalculating the RNG's next outputs

but the result is the same every time. I could make the video again and make sure that I control the board and show that I can see ahead. It changes based on what has happened, but it changes the same way every time, so it's not random in any sense at all, not even computer sense (I know there is no true random when it comes to computers). The fact that I knew I would get that +5 skull after a chain on my first move shows that. So even though it might be very hard if not impossible for a human to predict for every variable, the computer truly can see in the future for this game, which goes against what the developers have claimed (read a few discussions about this on gamefaqs).
 

Fieryshadowz

Well-Known Member
Member
Joined
Feb 27, 2007
Messages
398
Trophies
0
XP
85
Country
United States
Start your game. play one game.. save. turn off
Reboot start another game.. the 1st game you play will always be the same each time you start up. If you make the same move as the first time then the same blocks will be generated.. and the comp will make the same move.. you keep doing the same things and the same things will keep happening.. HOWEVER.. if you make a different starting move.. the comp. will do something different as well.. as diff. gems will be made.

(also.. after the first initial puzzle the rest is supposed to more "random" again)
 

Psyfira

Credit: 0ml. Insert tea to continue
Member
Joined
Dec 31, 2003
Messages
3,886
Trophies
0
Location
England
XP
270
Country
The point is that the creators of the game claim that the computer isn't all knowing and can't predict future cascades to do better than a human player, but I prove that this cannot be the case because the puzzles are pre-conceived.
Not quite: you've only proven that the board layouts are pregenerated. This does not prove that the computer AI uses that fact to it's advantage.

The fact that the programmers were capable of making the computer predict casades doesn't prove that it actually does. The computer makes the same move every time because, for any board it's given, it's using the same algorithm to work out what the best possible move is. Give a computer (with a basic AI) the same board 50 times over and it will make the same decision every time.

I'm not saying you're wrong, you just need more proof than that.
happy.gif


Note that the computer's first move was to match 3 skulls, and in any board it makes sense to always go for the skulls. The interesting experiment would be to see what the computer does instead if you go first and take the skulls. Is it ever in a situation where it avoids a good move, makes a poor one instead but gets a chain from pieces it shouldn't be able to see? Try it.
 

Jeda

Well-Known Member
Member
Joined
Nov 7, 2002
Messages
329
Trophies
0
Age
41
Location
Germany
Website
Visit site
XP
361
Country
Gambia, The
Not quite: you've only proven that the board layouts are pregenerated. This does not prove that the computer AI uses that fact to it's advantage.

You can even see in the video that the layout isn't pregenerated. I asume that by pregenerated you are talking about more than the initialy visible field.
 

Psyfira

Credit: 0ml. Insert tea to continue
Member
Joined
Dec 31, 2003
Messages
3,886
Trophies
0
Location
England
XP
270
Country
Jeda: I was talking about the whole board, hidden and visible at the very start of the game (sorry that's what you get when a programmer thinks about it, lol). Of course the moves you make will slide those pieces around in different ways afterwards, what Hooya is talking about is that the whole game grid is generated at the start and you just move the pieces, rather than the pieces being generated as they fall into the visible area. That's what I meant by pregenerated, the fact that the start grid it makes is identical every time. (not sure if it really is identical, I didn't look that closely at it to know if there were any differences. It's getting late here
happy.gif
)
 

Jeda

Well-Known Member
Member
Joined
Nov 7, 2002
Messages
329
Trophies
0
Age
41
Location
Germany
Website
Visit site
XP
361
Country
Gambia, The
Jeda: I was talking about the whole board, hidden and visible at the very start of the game (sorry that's what you get when a programmer thinks about it, lol)

Well I was guessing that's what you were talking about. However:
In the Video, the skull move on the left has no influence on the first hidden stone in the column of the red move.
Doing the red move first (first part of the video) produces a skull gem there, while doing the skull move first (second part) produces a purple exp gem there. I'm talking about the first stone appearing, not the end result. If the hidden stone was pregenerated it would be the same in both cases.
 

Psyfira

Credit: 0ml. Insert tea to continue
Member
Joined
Dec 31, 2003
Messages
3,886
Trophies
0
Location
England
XP
270
Country
Doing the red move first (first part of the video) produces a skull gem there, while doing the skull move first (second part) produces a purple exp gem there. I'm talking about the first stone appearing, not the end result. If the hidden stone was pregenerated it would be the same in both cases.
Oh I see what you mean now. So you're absolutely right, it's not pre-generated. My bad
smile.gif
 

Hooya

Well-Known Member
OP
Member
Joined
Aug 25, 2006
Messages
1,878
Trophies
0
Age
42
Location
Central Illinois
Website
Visit site
XP
317
Country
United States
maybe the entire board is not pre-generated, but the results of a given move are the same every time. So if I move that skull and then the red gem, the results would be the same as what you saw. So even though the gems aren't totally pre-generated, it's an algorhythm that determines what will come next, and the result is predictable (maybe not by a human, but by the computer). I guess that was my point. Just because the computer is clairvoyant doesn't mean that it always makes the right choice though. There are difficulty choices on the game, and that might have a lot to do with how the computer thinks ahead, not just the spell level of you enemies and such.
 

Hadrian

Literally as TIGHT as a gnats chuff.
Former Staff
Joined
Oct 12, 2004
Messages
16,991
Trophies
3
Age
42
XP
5,032
Country
United States
I always thought that the board generated according to the stats of both players, like if one players cunning was higher than another and so on.
 

Jeda

Well-Known Member
Member
Joined
Nov 7, 2002
Messages
329
Trophies
0
Age
41
Location
Germany
Website
Visit site
XP
361
Country
Gambia, The
maybe the entire board is not pre-generated, but the results of a given move are the same every time. So if I move that skull and then the red gem, the results would be the same as what you saw. So even though the gems aren't totally pre-generated, it's an algorhythm that determines what will come next

Yes of course there is an algorhythm. Nothing works without one. My point is that this algorithm is a simple call of the (pseudo) random number generator. Basicly my point was that your video showed exactly what one is to expect from a RNG based game.
 

Psyfira

Credit: 0ml. Insert tea to continue
Member
Joined
Dec 31, 2003
Messages
3,886
Trophies
0
Location
England
XP
270
Country
Just because the computer is clairvoyant doesn't mean that it always makes the right choice though. There are difficulty choices on the game, and that might have a lot to do with how the computer thinks ahead, not just the spell level of you enemies and such.
You keep using the word "clairvoyant", but you still haven't proven that the computer player AI is clairvoyant.

The game board generator and the CPU player AI are two completely separate parts of a game. You've proven that the board generating part works in a predictable way, but you cannot just assume that means that the CPU player AI part knows how.

You're right about the difficulty settings; on higher difficulties the computer AI probably will plan ahead a few moves. But you have not yet proven that the computer AI is using more pieces than those you can see when it makes those plans. Maybe it does, maybe it doesn't, but the evidence you've given here isn't enough to make a call either way.

(omg Critical Thinking AS is actually useful? *faints*)
 

Hooya

Well-Known Member
OP
Member
Joined
Aug 25, 2006
Messages
1,878
Trophies
0
Age
42
Location
Central Illinois
Website
Visit site
XP
317
Country
United States
Just because the computer is clairvoyant doesn't mean that it always makes the right choice though. There are difficulty choices on the game, and that might have a lot to do with how the computer thinks ahead, not just the spell level of you enemies and such.You keep using the word "clairvoyant", but you still haven't proven that the computer player AI is clairvoyant.

The game board generator and the CPU player AI are two completely separate parts of a game. You've proven that the board generating part works in a predictable way, but you cannot just assume that means that the CPU player AI part knows how.

You're right about the difficulty settings; on higher difficulties the computer AI probably will plan ahead a few moves. But you have not yet proven that the computer AI is using more pieces than those you can see when it makes those plans. Maybe it does, maybe it doesn't, but the evidence you've given here isn't enough to make a call either way.

(omg Critical Thinking AS is actually useful? *faints*)

I see your point there, ultimately without looking at the source code that would be improvable (unprovable?).

One thing I have found that's interesting though, I started a new game with a new character and the first matches are again the same board layout, so it's not even based on your character in any way.... I just got Bejeweled again for my Windows Mobile device and I just don't see how the randomness of that couldn't be included in Puzzle Quest... I just think it would make for a more satisfying experience.
 

Site & Scene News

Popular threads in this forum

General chit-chat
Help Users
  • AncientBoi @ AncientBoi:
    eeewww
  • K3Nv2 @ K3Nv2:
    I thought it was the toilet
  • AncientBoi @ AncientBoi:
    okies. Time to go watch YT paranormal ghost things. L8er my luvs :D
    +1
  • K3Nv2 @ K3Nv2:
    I got a massive clue
  • BakerMan @ BakerMan:
    this mf def ain't watching ghost shit, he boutta beat his meat fr
    +1
  • K3Nv2 @ K3Nv2:
    Nah he's about to be the ghost in your bedroom
    +1
  • Xdqwerty @ Xdqwerty:
    @K3Nv2, and leave ectoplasm all over the place
  • BakerMan @ BakerMan:

    this is him being described
    +2
  • Xdqwerty @ Xdqwerty:
    Sigh
  • Xdqwerty @ Xdqwerty:
    Yawn
  • Xdqwerty @ Xdqwerty:
    @SylverReZ, I dislike the kind of drm where you have to play single player games online all the time bc of some verification bs
    +1
  • SylverReZ @ SylverReZ:
    @Xdqwerty, Don't use games that have Easy Anti-Cheat as its been exploited many times.
  • Xdqwerty @ Xdqwerty:
    @SylverReZ, my PC can't run most AAA games so i wont
    +1
  • Xdqwerty @ Xdqwerty:
    Most of the modern AAA games
    +1
  • Xdqwerty @ Xdqwerty:
    @SylverReZ, I also heard one of the Prince of Persia games was so unfinished that it required the "24/7 online" drm so a puzzle could be done and the game could be finished. And that when the Ubisoft servers were closed the (cracked) game was impossible to finish or something like that
  • SylverReZ @ SylverReZ:
    @Xdqwerty, That's extra scummy. Ubisoft nowadays ship out incomplete games like Skull and Bones which was being worked on for nearly a decade now.
    +1
  • Xdqwerty @ Xdqwerty:
    @SylverReZ, i think they have been doing that since late 2000s
    +1
  • Xdqwerty @ Xdqwerty:
    Either that or their old games were unfinished aswell but we can't notice it
  • Psionic Roshambo @ Psionic Roshambo:
    I like that games can be fixed after the fact, hate that it's being abused via beta tests... And DLC... I was a 7800 owner back in the day and loved Impossible Mission, turns out I couldn't beat it because it was actually impossible lol
  • Psionic Roshambo @ Psionic Roshambo:
    I never knew about it at the time but a fixed version was available but you had to mail in your broken copy lol
  • Psionic Roshambo @ Psionic Roshambo:
    So that version is semi rare
  • Xdqwerty @ Xdqwerty:
    @Psionic Roshambo, I have a rom of the ds version of impossible mission
    Xdqwerty @ Xdqwerty: @Psionic Roshambo, I have a rom of the ds version of impossible mission