Author round-table 016

InterviewsAuthor round-table

All about Akari

Today, we talk about the popular pencil puzzle Akari. Our guests this time are SAKAMOTO, Nobuyuki, and mimic. From Nikoli we have NyanBaz and Nob.

NyanBazToday's theme is Akari. Mimic wrote an article about heads and tails methods for solving of Akari in the blog.

mimicI did that.

SAKAMOTO, NobuyukiI was totally absorbed in reading it.

NyanBazLet me explain it simply. According to mimic, you can classify the methods of Akari into heads and tails, more or less. The head methods put light bulbs at places with numbers. The tail methods put light bulbs at cells which aren't yet lighted up. Putting light bulbs around a 4 is a head method. Putting a light bulb in other places to light up a cell next to a 0 is a tail method.

mimicI think that the basis for Akari has been the head methods recently. Because you have to search, a tail method is harder to sort out than the head methods. It becomes hard to see everything clearly if you develop tail methods. Therefore, I have come to think that it's better to make problems depending on head methods. For example like Nobuyuki's problem that was chosen for the pick of the puzzles deadline Jul 2010. That problem is applying and reapplying the diagonal 1 method. I think that problem is the Royal Road, the way to do it with a head method.

SAKAMOTO, NobuyukiWow, the Royal Road way to do it!

mimicIt's hard to use it as a method but it's really very clear.

The method with two 3s facing each other
The method with two 3s facing each other.

NyanBazThen there are the methods where it's hard to think up even a head method. For example the method of two 3s facing each other like this.

mimicThat's the method that T.Karino uses really well. It's difficult to visualize that as a head method. I like it. I call it a pincer drive myself.

Pincers of 1s
Pincers of 1s.

mimicYou get the same effect with 2. But I like 1 better because 2 will be decided regardless of what is around it, it's harder to notice than with 1.

Pincers of 2s
Pincers of 2s.

NyanBazCome to think about it, you don't use that method very much Nobuyuki.

SAKAMOTO, NobuyukiRight. But I did make a problem the other day that could almost be solved by just this method. I like repeating one method in a puzzle. So when I use it, I use it all the way.

NobThere are a lot of one method puzzles from Nobuyuki.

SAKAMOTO, NobuyukiIt's not easy to make a problem with several methods, I don't remember many methods. When I remember a method that I didn't use recently, I make a problem with just that method.

mimicWhich way do solvers like it? A problem with many methods or with just one method?

SAKAMOTO, NobuyukiPerhaps there are people of both kinds. Having both kinds of problems seems better on

NyanBazWe don't want things to be unbalanced.

mimicWith a small problem, it's difficult to pack it with many methods. I have to divide space by black cells to put many elements in a problem with Akari. It is a nice point in Akari that a light bulb influences far off cells. But that comfort disappears when I divide the space finely. Getting a good balance is really difficult.

SAKAMOTO, NobuyukiSure, nothing easy there. Maybe because with it's comfortable that the influence goes far. It may not seem that way when you solve it with paper.

mimicNot improbable. When I make a problem on paper, I confirm that every cell is lighted up to see illumination in them one by one. With one move that may take time. On a phone or PC however, the light is there in an instant. The sense of what goes on changes, I don't solve Akari with paper any more now.

NyanBazBut when you make Akari it's with paper. When you make them is there gap in how it feels?

mimicYes there is that gap. When I create Akari problems, I imagine that I'll solve it with a PC. When I create Akari, it is different from solving it, it doesn't go at once. If I have decided the placement in one place, I think about what I would do next, slowly and carefully. It takes time to make problems.

SAKAMOTO, NobuyukiI make Akari problems on the PC. I couldn't return to making Akari with paper.

mimicDo you think about putting various signs in it while you are making it? For example, when I want to make a 0 in either of two black cells, but I want to let it be decided by the development of the later half. Are there such situations? When it looks like that, I mark the black cells.

NyanBazAha, something new here. I haven't seen puzzle making software with that degree of flexibility so far.

SAKAMOTO, NobuyukiI don't make problems that require such minute planning. Because I seem to forget about it, I decide to make my decisions earlier. (laughs)

NyanBazMimic often makes problem that he has planned a huge amount of details for.

NobDo you think about solving order control?

mimicI think about it a lot. I entirely control it with 10x10 size Akari. When I fix the lower right, at times, I think about how it will become in the top left. However, I can't completely control everything with big sizes such as 24x14. My brain doesn't have that kind of capacity.

NobDo you run out of brain capacity? I thought that never happened to you.

mimicThat does happen, naturally. (laughs) I may even become haphazard.

NyanBazHow about you Nobuyuki?

SAKAMOTO, NobuyukiMy methods are haphazard from the beginning. (laughs) I'm not conscious of any order to solve anything very much. I pack a lot into a method that I want to use. I think that it's a success if you use the method a lot. I think different people use different ways to go along.

mimicFor Akari, we still get new difficult methods. I'm anxious about where Akari will get to.

SAKAMOTO, NobuyukiI feel that's right. Akari is basically an easy puzzle. So it's hard to know where Akari will evolve into.

NobWell. Easy problems are important, but when they are only easy, you get tired of them. So we need hard problems, too. There are many people who like thinking about difficult methods among puzzle authors. It's a necessity to develop hard methods.

SAKAMOTO, NobuyukiI can handle that about Akari. It's different with Ripple Effect. If someone uses hard methods for Ripple Effect, I can't solve them. (laughs)

NobAre there hard methods that haven't been discovered yet for Ripple Effect?

SAKAMOTO, NobuyukiI think that there still is because it hasn't developed very much.

NyanBazLet's get back to Akari. Hard methods are developed. But do you think that Akari has enough medium difficulty methods?

mimicI think that became a topic a while ago, the diagonal 1 method is medium.

NyanBazAre there any medium difficulty methods that appeared recently?

SAKAMOTO, NobuyukiProbably no. Most of the new methods are hard. There is a gap in difficulty between what my image is and what gets published. I feel the diagonal 1 is a hard method.

mimicReally? I'd call it medium.

NobThe diagonal 1 is the same as corner 2 of Nurikabe, in the reasoning. In Nurikabe, that method is ranked as easy.

SAKAMOTO, NobuyukiSomething there, it surely isn't hard. I tend to judge methods to be difficult. I think about difficulty as whether a person who learned the rules of a puzzle for the first time would notice it.

NobThat's the way we look at it too. Well put, great.

NyanBazLet's switch to something lighter. What is a number that you like in Akari?

NobDo you like 0s, Nobuyuki? You use 0s a lot.

SAKAMOTO, NobuyukiHmmm... It's difficult to say. I think that it's either 4s or 0s.

NyanBazWhy 4s?

SAKAMOTO, NobuyukiThat was a topic a while back, I'm weak in controlling the order. But I can control the 4s. I know that if I put a 4 somewhere, the solver will put those 4 light bulbs first.

NobVery true, you can control that completely.

SAKAMOTO, NobuyukiBecause 0 is a safe number to fall back on, I like 0. It becomes an entrance to the solution. It's useful both in the middle stage and the end of the game. That's why I was at a loss with 4 or 0.

NyanBazHow about you mimic?

mimicI love 1s. It's usable in all sorts of methods. 1s are usable in diagonal methods, the pincer drive method, and there are more. If I have to make an Akari problem with one kind of number, I'd want to make it with 1s. I wouldn't want to make with 2s and 3s.

NyanBazYou like there to be no end to the variety.


SAKAMOTO, NobuyukiMy preferences are clearly different from yours, versatile 1s or settled 4s and 0s.

NobAs I recall with 0s, I rarely see problems where you decide on cells where you can't put a light bulb with a 0, and have a 3 in the neighbor cell. It's the same as the 03 rules in Slitherlink. I think that it's a way into the puzzle that's easy to spot.

SAKAMOTO, NobuyukiThere is something in that. I forget all about that because, in Akari there is 4, that's an entrance that's easier.

mimicThere is something there. It's possible to have like the Slitherlink's 031313 in Akari, it's a regular sequence for Slitherlink.

NyanBazIt's possible to have even a 1 where three direction are enclosed in 0s.

SAKAMOTO, NobuyukiThere really is depth to Akari. It hasn't reached its end of its development, completely, yet.

NyanBazNow comes the last question. Please share your thoughts about Akari with us.

mimicI've solved and made them ever since it arrived on the website. I am proud to have contributed something to the development of conventional Akari. For me Akari is the puzzle which is most important next to Slitherlink.

SAKAMOTO, NobuyukiFor me, Akari is a stage for expression. Akari has a lot of flexibility. So it's easy to express that I want to do. I can make problems where light shines through from top to bottom. I can make careful messy problems. I can reveal myself comfortably. I'll reveal myself here from now on.

NyanBazThank you for today.

Discussion on Mar 2011 Published on Mar 19, 2014

Related link