Shipwrightery: Reaching Ultimate and Beyond

In my last article, I tried to lay out basic strategies for playing shipwrightery. I'll try to point out a couple things that might help you advance up the ranks and into an Ultimate shipwrighter, but there isn't a lot more to know. Master the strategy that I laid out last time, and you'll find yourself doing very well. Just keep in mind that, as your skill in the puzzle progresses, you'll get larger patterns to try to match. At top form, you'll be getting two 5-piece patterns, two 4-piece patterns, and two 3-piece patterns. Not until you reach that level will you be generating the best scores in shipwrightery. If you're not there yet, just keep practicing and working at it - as your standing increase, so too will the size of your patterns.

Set Up For Extra Matches

As you play shipwrightery, you'll begin to recognize many of the patterns that you see. While I don't have all of them memorized, I know quite a few. If you want to see them all, look at the YPPedia entry, here. Knowing the patterns is important because you're working on a 25 piece board and, at top level, your patterns will take up 24 pieces. That means, in order to get every pattern in your tray set up and ready to go, you're going to need perfectly matching pieces and a board with just the right number of each type of piece. In other words, it takes a lot of luck and very, very rarely happens.

One difference between a good shipwrightery player and a great one is what they do with those "extra pieces". Extra pieces are those pieces that are left over after you've set up as many pattern matches as you can. Like I said, unless you're very lucky, you're going to have some.

Worry about getting your combo of patterns set up first. Then, with whatever leftover time you have left, consider trying to create "parts" of other patterns with the extra pieces you have. Perhaps you can create a part of one of the patterns you already have in your tray, or perhaps you can create a pattern that isn't currently on your tray.

Let me steal this image from my previous article:



Notice the five "extra" pieces I have near the top-right corner. If you look at what I did with the two whites and a brown, you may recognize this as one of the 3-piece patterns, the Batten. Granted, the Batten isn't in my tray, but I have set up to clear the Halyard and the Cleat and that will give me two new 3-piece patterns. What are the odds of me actually getting the Batten? Not very good - only 2 in 16. However, a 12% chance of getting an extra match is better than the 0% chance I'd have had if I hadn't arranged these pieces in this way.

If you can't make a complete pattern, try to create part of a pattern, particularly if it's a pattern that's already in your tray and you can't fit into your combo. Have a Ballast (4 blacks) in your tray but only three extra blacks on your board? Set up the three blacks around a match that you've got set up so that, when you clear that match, if a black appears in the right place, you get a free match. What are the odds of getting the right piece? A little better than 25% (don't forget about the gold pieces). Again, not great odds, but far better than 0%.

Taking care with your extra pieces can really help you extend your combos and squeeze out some additional points.

Understand the Scoring System

Probably the biggest leap that took me from Legendary to Ultimate was gaining a better understanding of the scoring system behind Shipwrightery. Once you know the scoring system, you can actually put a value behind a puzzle attempt, and you can determine if one attempt was better or worse than another.

While the scoring systems for any of these puzzles are never revealed to the public, testing leads us to get pretty good approximations of the scoring systems. This forum thread that I started some time ago goes into a lot of great detail for the scoring system. Probably the most important bit of that thread is this table:

Step    Combo     3-Piece      4-Piece      5-Piece
----------------------------------------------------------------
1       N/A       Fair         Good         Great
2       Double    Good         Great        Excellent
3       Triple    Great        Excellent    Admirable
4       Bingo     Great        Excellent    Superior
5       Donkey    Great        Admirable    Magnificent
6       Vegas     Excellent    Superior     Artisan
7       Vegas^2   Excellent    Superior     Artisan
8       Vegas^3   Excellent    Magnificent  Paragon
9       Vegas^4   Admirable    Artisan      Paragon
10      Vegas^5   Admirable    Artisan      Master
11      Vegas^6   Admirable    Artisan      Master
12      Vegas^7   Superior     ???          Master
13      Vegas^8   Superior     ???          Master
14      Vegas^9   Superior     ???          Master

From this, you can start to see how well various combos will score. Always keep in mind that the score of a combo is the average of the scores of the individual matches of that combo.

For my own benefit, I assigned a point value to each score, like this:

Fair          1
Good          2
Great         3
Excellent     4
Admirable     5
Superior      6
Magnificent   7
Artisan       8
Paragon       9
Master        10

So, if I wanted to score a combo, I'd take the scores for each match in that combo and average them. For example, a 3-4-5-5 combo would score Fair-Great-Admirable-Superior or 1-3-5-6 for an average score of 3.75. When I last tested, you needed to be able to score about a 4 (Excellent) to reach Ultimate on Midnight. That may have changed since, but I'm guessing it's probably still pretty close.

This scoring system is probably oversimplified in the fact that it assumes all matches with the same descriptive word are worth the same amount (i.e. All 3-piece matches that score "Great" are considered equal, regardless of whether that move was a triple, a bingo, or a donkey). My guess is that those scores are actually a bit different, but my approximation seems to work pretty well.

To help me see just how well I was doing, I created an Excel spreadsheet in which I had boxes for each score. As I was playing, I'd make a match and put a tally into the corresponding cell in my spreadsheet. Those cells would automatically tally up my average score and tell me how I was doing. That way, when all was said and done, I knew just how well I was doing.

While this may not make you a vastly better puzzler by itself, it will help you see how you're improving.

Gaming the System

As of the time I write this, there is no penalty for dismissing a shipwrightery puzzle attempt. At any point. That means that you can play all the way through the puzzle and, if things aren't going the way you want them to, you can dismiss the puzzle right before the end and suffer no penalty.

I'm not going to get into the ethical debate over whether or not this is a good practice, or not. I have my own opinion and I'll keep it to myself. If you want to debate it, the forums are a great place to do so. What I will say is that you absolutely *do not* have to use this "feature" to make Ultimate on Midnight. It's quite do-able without using this technique.

You see, the problem is that, in general, you're not going to consistently get more than 3-4-4-5-5 combos (often, these aren't even possible). Those combos score about 4.2, using my system. That's good enough to get you to Ultimate, but probably not good enough to get you to #1. However, the only way to get above that score is through pure luck - you need additional matches to "appear" and lengthen your combos. Sure, you can use the techniques I described above to give yourself better odds of having these things happen, but it still comes down to luck. If the right pieces/patterns appear, you score better, but you have no control over what actually appears.

So how do you increase your score? You play the puzzles in which luck is on your side and you score exceptionally well and you dismiss when the luck of the draw doesn't go your way.

Like I said, I won't go into my personal feelings about whether this is an ethical way to play, or not, but I will point out that distilling used to exhibit this same behavior and now no longer does. Now, when you play distilling, once you've started playing the puzzle, if you dismiss, you damage your score - severely.

I'll just reiterate that you do not need to use this technique to reach Ultimate, but you may very well find the #1 spot very difficult to reach.

Conclusion

As with all crafting puzzles, there's no inherent benefit to reach the Ultimate level of play. Once you've reach Renowned standing, there's no need to advance further. Those that do advance beyond Renowned do it for pride or a love of the puzzle, or both. Given that, not everyone will need the tips listed here. But, if you're looking for a goal to shoot for an Ultimate shipwrightery is that goal, I hope this article helps you along your way.

Happy puzzling!

Shipwrightery: Strategy and Scoring

When discussing shipwrightery, I tend not to talk about how to move the pieces around the board as I find that to become trivial for anyone with a bit of experience. Rather, I like to discuss two aspects of the puzzle: your strategy and the scoring system. I'll actually begin with the latter because understanding the scoring strategy drastically impacts your strategy.

Scoring

The scoring system in shipwrightery isn't always obvious to everyone and it really dictates the way one should play, so I'll talk about it briefly.

Every time you "match" a pattern, you get a score for that match. The score is determined by two things: the size of the pattern you matched and the position in the chain in which you made that match. The larger the pattern, the better you score (i.e. a 5-piece pattern will score better than a 3-piece pattern). Also, the later in a chain you make a match, the more it's worth.

Given that, there are a couple rules of thumb that you'll want to take forward:

• Build the longest chains possible


• Use your biggest matches last

While this may not be exactly accurate, I like to think of it this way (for the more mathematically inclined) - each match is worth a given score - each sequential position in a chain has a higher multiplier. By placing your best scoring matches (those with the most pieces) later in your chain, you're multiplying your biggest scores by the biggest multipliers, which results in a largest product possible.

Another thing to note about shipwrightery scoring is that your final score is the average of all your match scores. So, if you had three matches that scored Fair, Great, and Admirable, your average score would be "Great". (See the wiki for a complete list of scores.) On Midnight, it seems that you need to average approximately "Excellent" in order to score Incredible.

If you're really interested in digging into even more details about shipwrightery scoring, have a look at this forum thread.

Strategy

In terms of strategy, I find shipwrightery to be more like carpentry than any other puzzle - it really takes good "spatial organization" to do well with this puzzle. When I play this puzzle, I generally do the same few steps over and over again:

1. Plan Layout


2. Look for Shortages


3. Move Pieces


4. Execute Combo

I'll go into each step to give you a good idea of the things I look for.

Plan Layout

This is the part of the puzzle that I really find to be a lot like carpentry. When puzzling at the highest levels, you'll have two 5-piece patterns, two 4-piece patterns, and two 3-piece patterns. That's 24 pieces worth of patterns. The board is only 25 pieces large. That means it's going to take a whole lot of planning and quite a bit of luck to get them all on the board.

There's really nothing you can do about getting patterns that all fit well together or getting the correct pieces on the board to match the patterns you have. What you can impact is how you arrange these patterns on the board to maximize space. The largest patterns are the most important, so make sure you get those placed first - I'd suggest building around the outside edge to save you as much room as possible. Then move on to the next largest patterns, and so on, until you've run out of room.

Here's a picture of a typical starting board:



The first step is to look at my patterns and try to determine where they'd best fit on my board. Given the pieces I have, this is how I think I'd like to arrange them:



So the next logical question is, how the heck did I come up with that?

My approach is pretty simple. Start with the biggest patterns you have, whether those be 5-piece patterns or 4-piece patterns. These are the most important pieces to place (because they score the best) and they're also going to be the most constraining. Once you have those placed, your options for the smaller pieces will be more limited and it'll take you less time to decide where to put them.

When determining where to put your largest patterns, try to find ways to put them together that won't "waste" any spaces. Notice how I put together the berth and the gangway in the example above. The two pieces fit together in such a way that no spaces are trapped where they can't be used.

Some patterns fit together better than others. The berth and gangway work quite well together. In other cases, like the helm and capstan shown below, there aren't always good ways to put them together. Notice the green block that's been trapped in the upper-left corner - that block can't be used in any other patterns and thereby limits your options for additional pattern matches.



Other pairs, such as the mooring with the nest can be downright awful. Which two patterns you get isn't under your control. How you try to arrange them to save the most space is up to you. With ten 5-piece patterns available, I won't go into the best way to match them all, but here are a few examples of useful ways I've found to match patterns. Note that the same pattern may get moved around the board based on the pattern it's paired with. For example, the shot works best in the lower-right when paired with the nest, but it works best in the lower-left when paired with the mooring. Here are a few arrangements to help you get the idea - remember, the goal here is to fit your largest patterns on the board and leave as much space as possible for other patterns.



Once you have your biggest patterns in place, move on to the next largest patterns (the 4's, generally), and so on until you've placed everything you can. By placing the largest patterns, which have the least options available to you, you make it easier on yourself when placing the smaller patterns.

Look for Shortages

I kept this as a separate step, but I really do this while performing steps 1 and 3. Once I have my plan set from step 1, I usually start moving pieces around. While I'm doing this, I'm looking at the pieces I have on the board - will I have enough to complete my patterns? Am I going to be short a specific color?

If I go back to my original example, I wanted to do this:



If you look close enough, you can see that I can't possibly do this. The berth requires 2 wood, the gangway requires 2 (it would normally require 3, but I'm using the gold for that), and the halyard would normally require a wood, but I'm using a gold there, as well. So that means I need 4 wood for these three pieces. Take a look at that picture. There are only 4 pieces of wood on the entire board. That means I have no way to create the pattern for a yard, which would require two additional pieces of wood that I simply do not have.

So what happens? I revise my plan.

While moving pieces about to fit my plan as best I can, I have to determine what I can do with that upper-right corner where I had been planning on putting the yard. Fortunately, because my plan already accounts for all the other patterns, I only have one option to evaluation - can I get a cleat in there? Every other pattern is done and the yard won't work due to lack of wood so the cleat is my only option. Does it fit? Certainly. This is what I ended up with:



Not far from the original plan, but it does show that it's important to be flexible while implementing your plan.

Move Pieces

Once your plan is in place, begin moving your pieces to implement that plan. I won't get much into the actual moving of pieces around. The basics are simple - getting good and efficient at it simply takes practice. Practice makes perfect and, once you get used to moving the pieces, you won't even really think about it any longer - then you can dedicate yourself to thinking about your plan and how to get those most out of the patterns and the board in front of you.

When you're moving pieces, you're constantly looking at your plan. Will it work? Do you have all the pieces you need? Do you need to modify it? While I have these first three steps separated, they really all happen together. Often, I plan for a couple seconds and then immediately begin moving pieces. While moving, I'm revising that plan and looking for shortages and adjusting accordingly.

Execute Combo

Finally, after all that work, it's time to sit back, relax, and build your combo. While you might feel stressed and pressed for time during the previous segment, this is a part of the puzzle you simply should not rush. Every time you clear a pattern, your timer starts over. Your board is already set up for the next pattern, so take a moment and slow down.

Rather than simply firing off your entire combo in one fell swoop - you should be looking for new, "intermediate", clears that crop up after clearing existing patterns. These intermediate clears help you push those ever-important big patterns further and further back in the combo. The longer you can hold those big patterns, the better off you'll be.

Here's an example:



In this case, when I clear the nest shown in the left image, it leaves a piece of cloth in just the right place so that I can now clear a boom. When these happen, you don't want to miss them. Try to make your chains as long as possible and keep the biggest patterns as far back in the chain as possible.

So, when it comes time to execute your chain, take time after each clear to see if something new has cropped up. If it has, lucky for you! Take advantage of it and get yourself a better score!

When you simply can't match any more patterns, your chain is over and you have to start over at step 1. Rinse and repeat until complete!

Shipwrightery: Required Reading

Shipwrightery is a puzzle that I feel is often overlooked. It doesn't seem to have a large following and the pay you get for excelling in the puzzle is usually overshadowed by the pay you might get from distilling. Nonetheless, I've had some requests asking me how to perform well at this puzzle so I'll do my best. Here are some basics to get you started:

Shipwrightery Basic Info
Shipwrighting Scoring Discussion