12
$\begingroup$

Of the following 5 strategies for getting started assembling a jigsaw puzzle:

  • Assembling the border and connecting other pieces to it
  • Working your way out from one corner
  • Trying to assemble faces or other recognizable parts of the picture
  • Picking a random piece and working your way out from it
  • Finding random pieces that fit together

which one or what combination results in the shortest solving time, on average, statistically speaking, according to the experts? Note that this is not a matter of experience. What do the studies on jigsaw-puzzle solving show?

Be sure to include the reasons why the strategy or combination of strategies is the fastest.

$\endgroup$
  • $\begingroup$ different puzzles require different strategies, there are the impossibles™ that don't have a border, puzzles with a big areas without any features and so on $\endgroup$ – ratchet freak May 18 '14 at 23:18
  • 4
    $\begingroup$ I don't solve jigsaws regularly, but don't jigsaw solvers apply a multitude of techniques (ie. you might be primarily working on the border, but you might group all the sky pieces together, or find other pieces that fit and put them together) $\endgroup$ – dwjohnston May 19 '14 at 2:30
  • 1
    $\begingroup$ I solve jigsaws (regularly) and I'm... sort of lost here, it's really kind of opinion based. Generally speaking, I solve 3D jigsaw puzzles, so... I almost always look for similar patterns. When I solve (some) 2D ones, I'll do the border or patterns, depending on what I "see" first. I don't know what would be fastest or slowest. Plus like @dwjohnston said, it's best to use more than one technique at once-- if you see a piece that belongs to the border while putting together another part of the puzzle, you'd probably separate it from the rest. It really depends on the person doing it. $\endgroup$ – Ice-9 May 19 '14 at 13:05
  • 1
    $\begingroup$ Come on! I was much more specific than the other jigsaw puzzle question! What other details are you looking for? I thought this community was mainly about solving strategy and tips about puzzle development, this question being about the first! $\endgroup$ – Brian J. Fink May 19 '14 at 16:20
  • 1
    $\begingroup$ I think some people here just hate jigsaw puzzles, closing questions regarding them indiscriminately and without one word of explanation. $\endgroup$ – SF. May 19 '14 at 17:08
11
$\begingroup$

Let's take a 25 x 25 puzzle of a color gradient(also called color ramp and color progression). No two connections the same, even it is off by a millimeter. All the pieces are in a pile, with no way to identify the pieces. The fastest would be the first solution, assembling the border first.

Number 1, which is the fastest, would require you to sift through the pile and find edges and corners. Since it is a color gradient, the top edges would take little effort and the bottom edge just more so. After finishing that, just grab a piece and put it where you think it goes. If you make a wrong guess, the piece will move around the surface you are building on into the right place, thus solving the puzzle quickest.

Number 2 is the third fastest. Not as fast as number 1, you would have to go through the pile I mentioned in my introduction to find a corner. It also depends on how you assemble from the corner. Since number 1 is building a border, this would be a block progression, starting from the one corner piece to a corner, 2 edges, and one four sided piece to a corner, 4 edges, and 4 four sided pieces, and so on so forth.

Number 3 is the second fastest. You would have to find pieces of similar color and, since there are no faces, start off with that. A color gradient would have two identical colors near each other and you could just fit them together.

Number 4 would be the second slowest, because you would have to find 2, 3, or 4 pieces that fit with it. Since every piece has a certain connection with 2-4 others, you would most likely have to sift through all the pieces to find the right one.

Number 5 would not be the fastest, but actually the slowest, because you would be required to go through all of the pieces to see ones that fit with another. Unless you got lucky, saying you chose two pieces and they just happened to fit together, you would take the most time trying to fit pieces together rather than putting all of them together.

$\endgroup$
  • 1
    $\begingroup$ Can you cite any sources? $\endgroup$ – feelinferrety Jun 20 '17 at 22:42
7
$\begingroup$

Efficiency means that you handle each tile as little as possible, and that you make as few mistakes as possible. Ideally you want to pick up a tile and fit it to at least one other rather than put it down again unmatched. Incorrect fitting wastes time when correct matches are rejected and when you have to unpick the mistake.

From experience, the most efficient strategy has 5 stages. Overall this strategy is : (i) divide the problem into manageable sub-tasks, and (ii) do the easiest tasks first. The goal is to minimise the choice of tiles and maximise the number of clues at each stage, particularly for the hardest tasks.

In the early stages colour and pattern are important, because these features are the most easily identified and distinguished. In the last stage when remaining tiles are almost identical in colour, shape becomes the dominant distinguishing feature.

  1. Turn all tiles face up and sort them into about 6 groups. One group is all edge tiles. Another group is the skyline. Other groups are the major colours or combinations of colours in the picture. Examples are : sky, sea, foliage, flowers, buildings, roads, people/animals. Then separate out tiles belonging to individual features within each group, in preparation for stage 2.

There should be no assembly until stage 1 is complete, because assembly is easiest when all the required tiles are at hand.

  1. Assemble small clusters of high-contrast patterns and easily distinguished features within each group. Typical features are faces, groups of people, windows, doors, chimneys, printed signs, the skyline, the outlines of buildings. This is your strategy #3. Regions which are mostly featureless such as the sky, or which can only be distinguished in the fine detail such as foliage, should be left to stage 5.

Assembling the border early on provides a framework for placing clusters in their approximate positions.

  1. Extend the small clusters outwards by fitting as many as possible of the remaining tiles from the same colour group. At this stage it is efficient to pick up the most distincitive tile, locate it on the picture, and then place it in its exact place in the correct cluster, like a golfing hole-in-one. This is efficient because you are almost certain to connect it to a cluster, or at least place it beside a cluster. The reverse strategy is also efficient : use the picture and the shape of the gap to identify the missing tile among those remaining. This stage overlaps with the next :

  2. Join the small clusters together and attach them to the borders. This requires looking for the matching of patterns and colours and also the interlocking of shapes at large scales (clusters) and at small scales (individual knobs and holes).

  3. Fill in the remaining gaps. Start by completing the picture-matching strategy of stage 3, extending outwards the most easily identified features or patterns, which have the smallest choice of alternatives. Leave until last the largest areas of featureless, uniform pattern or colour, such as the sky, sea, foliage, grass, or dark shadow. These require the laborious testing of numerous options. This is your strategy #1.

The last stage requires the most organisation, method and attention to detail. Sort the pieces into 5 rows of tiles with 4 knobs down to none, with each row sorted from light to dark. Identify gaps in the puzzle where there are 2 or 3 adjacent tiles, or which have unusual shapes or the fewest remaining tiles of that shade of colour. These give the most confidence of a correct fit, and are the easiest to find.

At each stage (particularly the last) it is important to make full use of all the clues available in order to ensure that the fit is correct : check the exact match of colour and patterns and also the ease with which the tiles fit together. A fit which is too tight or too loose, or leaves an uneven gap around a tile, is almost certainly wrong.

Beside the picture, the pattern in the shapes of tiles is also a clue, to ensure that a fit is correct, or narrowing the search for a tile. In most jigsaw puzzles four tiles meet at every corner, and there is a repeating pattern of wide or narrow tiles every 6-8 rows or columns.


The above strategy - or one very close to it - is mentioned in most online advice. For example :

HobbyLark : How To Do Jigsaw Puzzles Like an Expert
Puzzle Warehouse :Jigsaw Puzzle Stragegies - A Guide to becoming a Puzzle Expert
wiki : How to Assemble Jigsaw Puzzles
Quora : How does one solve a jigsaw puzzle in the fastest way possible?
Chapter 1 : The Jigsaw Puzzle Problem (a thesis examining automation of solution strategies)


I agree with other answers here that your strategies #2, #4 and #5 are inefficient and should be avoided.

Strategy #2 : Starting from one corner can be useful when the only clue is the shape of the pieces. But generally this requires laborious searching and checking of a large number of possible options, which should be left until the options have been narrowed down (stage 4).

Strategy #4 : It is satisfying to pick up an interesting tile, locate it on the picture, and then fit it at once into its exact position in the puzzle. However, this is only likely to succeed when adjacent tiles are already in place.

Strategy #5 : Trying to fit random tiles together is highly inefficient. You are very unlikely to get a match because there are very many options, and you might try the same tiles more than once.

$\endgroup$
1
$\begingroup$

The above answer deals with a gradient puzzle. I don't have an answer analyzed by experts, but from personal experience. I have been working on a 1000 piece puzzle of the Biltmore estate. It is a night scene on Christmas. It has large area of black along the bottom and a large area of sky which is a gradient of blue. This involved using a hybrid technique for different parts for the most efficient solving time.

The building and other distinguishable features of the puzzle were quickly put together by randomly finding pieces that match and leaving them in approximate locations and building around them.

For the gradient part, like Max's answer, Method 1 was used.

For the solid black part however, It is brute force technique. Pick one piece try to match it to an open space, if it fits great, if not, put it aside and pick another one to try.

$\endgroup$
  • 1
    $\begingroup$ The question specifically asks for an expert answer, not one from experience. $\endgroup$ – Deusovi Jan 17 '16 at 21:34
0
$\begingroup$

From an engineering standpoint, it does not make any sense to build the border first. Sure, it might feel gratifying for most people, but it really is quite inefficient to do so.

  1. You set up an immediate obstruction/barrier for moving pieces around on the table and building the rest of the puzzle.

  2. When sorting the remaining pieces by color/texture/sharpness many of the pieces one needs to efficiently build that portion of the puzzle are not in front of them.

Also, why not "spread out" the gratification of assembling the border pieces during puzzle build! It surely does not hurt having some addition motivation and welcomed gratification when the puzzle building process experiences a lull. The BIGGEST challenge in not building the border first is convincing others that the instant gratification makes puzzle building less efficient. Before you disagree, give it a try next time.

$\endgroup$
0
$\begingroup$

Bottom line, the methodology may shift as you move through the puzzle. I normally work 1000 piece puzzles and start working on the border first. It helps me get a sense of the scale and colors in the puzzle and as I pull the border pieces, I get a feel for the rest of the puzzle. Once I have the border done, I normally turn over the remaining pieces, this is time consuming, but again, it helps me “know the pieces.” After sorting, I work on something large that attracts my eye, such as a specific color gradient or feature, mountain, tree line, flowers. However, I also find that if I get stuck shifting my attention to another area or me moving to a different side helps me get another perspective and continue. I also often focus on piece shape, and start grouping like pieces together, I.e., 4 heads, 3 heads, 2 heads, as the pieces decrease, I may also subgroup by color. I am always scanning the progress and at times see a random piece and that belongs on another part of the puzzle. Some of it feels like intuition, the more you work puzzles, the more fluid you become in your approach.

$\endgroup$

protected by Community Jan 19 at 6:49

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.