next »

[Robert Yarger]

When it comes to designing or making puzzles, I have found that the best puzzle is the one that you feel is good.  I have made large sized puzzles, only to be told by collectors that they want them small.  I get "constructive criticism" from collectors if I make puzzles that are simple or too complex.  The same goes for ones that are elaborate and expensive vs. cheap cheap and plain.  I have also made large runs of puzzles for those who feel that there are not enough to go around, as well as smaller quntities for those who desire a smaller limit on edition.

 I have leared that everyone has thier own idea of what they want in a puzzle, and it will drive you crazy if you try to satisify every desire.  I say, collect the puzzles you like, and make the puzzles you like.  The perfect puzzle is the one that you personally choose as a favorite.

[RKB]

My system is difficulty relative to simplicity but only works for puzzles that require moves, So itis not always the most difficult puzzle that comes out top.

I need to try the 18 piece burr "Phoenix" to see what it gets.

I always liked this kind of evaluating by numbers. But I think this is a only good start as there are some flaws:

1) The system doesn't value unique solutions.
2) Puzzles that do not require any coloring schemes to render a unique solution are 'better' (at least in my opinion).
3) Defining holes can be somewhat problematic. Do puzzles like X-Ray Vision or Hollowood have holes or not? I would say they haven't... bringing your score to 78.57 for X-Ray Vision.
4) In fact your counting the internal holes twice. The completely filled shape of Cutler's Giga Burr has 81 cubes. You're only counting the cubes that are actually in the pieces.
5) Achieving 'the same' result (i.e. the same number of moves and holes) with fewer pieces will result in a lower score, whereas I would call it a 'better' puzzle.

[JWDiaz]

I apologize for this post, it doesn't belong here. And I apologize to you, RKB for writing almost over you post.

[N. MCA.]

You are right about the number of pieces, and I have made an ammendment to that, perhaps one should multiply the first two numbers? I will give it a go.

And I do know this is just a start.

[RKB]

You are right about the number of pieces, and I have made an ammendment to that, perhaps one should multiply the first two numbers? I will give it a go.

And I do know this is just a start.

Well I'm particularly interested in this. But before going to the formulas there should be a conceptual framework: what to include in the index and the (relative) importance of each of those aspects.

[N. MCA.]

RKB has got it right on the head, about having the weight of the variables at the right level.

I was thinking perhaps instead of the previous idea having it like this

Moves
(total number of pieces) + (total number of units including voids) + (voids)

all x 100

Do you think that would be better?

This one actualy acounts for the stability of the puzzle as a small puzzle with lots of holes is less stable than a large one and the smaller the puzzle is the larger the difference the voids make.