Home Creators Posts Import Register Favorites Logout

Content

Opening Presentation

Soon after Mega Man1 was released, I went to a retail store that carried the game to see how things were going. What was displayed on the TV monitor was simply a still image of Mega Man striking a pose.

Meanwhile, the TV monitors around it were showing flashy demo plays of shooting games and action games from other companies, attracting the attention of large numbers of children.

From the beginning, Mega Man had not been a product that was strategically promoted with any particular selling point or marketing campaign as a game originally intended for the Famicom Disk System. Because of this, people said that Mega Man lacked the kind of buzz that would attract attention.

Even after its release, that understated image stayed with the game for quite some time. Putting the main characters across the entire front of the package did help convey some of that image to the target audience, but even so, there was absolutely no sense that Mega Man was gaining the kind of momentum that could lead to a hit.

I was using assembly language, a machine-level language, rather than so-called high-level languages such as those used today. Because of this, even for a skilled programmer, creating an automatic demo play required a considerable amount of time and effort compared to today.

Even if there was enough memory capacity, creating an uncertain situation was a major risk in the development of Mega Man2. I quickly gave up on implementing the demo play and switched to creating a story-based visual presentation instead. After all, there was no point unless I could get the children interested.

Vertical Double Scrolling

The opening needed movement that could have as much impact as a demo play. However, elaborate effects that required advanced programming techniques carried a high risk of delays, so I looked for something that would not be technically difficult and could be created quickly.

I had actually used this way of thinking in the ending of Mega Man1 as well. In the ending, Mega Man runs in the center of the screen while the scenery moves at the same speed as he runs. The ground beneath Mega Man and the background, however, move at different speeds, creating an effect similar to double scrolling.

When creating this kind of visual effect, you can create a sense of depth by having multiple images scroll at different speeds. For example, you can have a picture of mountains in the distance, buildings in front of them, and a character running in the foreground, with each layer moving at a different speed—the distant mountains moving slowly and the nearby buildings moving a little faster.

However, the Famicom could only display a single background image. This is where the technique known as double, or multi-layer, scrolling came in. This technique, called raster scrolling, uses the horizontal scanlines of the TV screen to divide an image that would normally be displayed as a single background. For example, by dividing the screen into upper and lower halves and scrolling the upper half slowly and the lower half quickly, the result looks as if multiple backgrounds are scrolling at different speeds.

However, we did not actually use this technique in the ending of Mega Man. It only looked as if raster scrolling was being used. We simply did not have the time or memory capacity to implement such a technique. Instead, we placed black vertical lines, made from sprite graphics, underneath Mega Man's feet and moved them at regular intervals. This made the ground beneath Mega Man appear to scroll slightly more slowly than the background, creating the illusion of double scrolling.

This method could be made to work reliably without having to create any special programming, so it was perfect for the opening. That led me to think: what if I did double scrolling vertically instead of horizontally?

In the ending of Mega Man1, I had created an effect that looked like raster scrolling. This time, I wanted to create vertical double scrolling—something that could not be divided using horizontal scanlines.

It was a clever little trick: without using raster scrolling at all, I could combine simple sprite elements to make it look as though I was using sophisticated technology.

The idea was to draw a distant cityscape and a building on the right side as a single background image. The building itself would be designed using only vertical straight lines. No stippling, horizontal lines, or diagonal lines—just vertical lines to draw the building.

When this image was scrolled vertically, the distant city would appear to move downward, while the building would appear to remain stationary. Since the building was designed entirely with vertical lines, even when it was scrolling vertically, all that was really moving was the vertical lines themselves, making the building appear to be standing still.

I then placed windows and protruding decorations on the building's walls, all drawn as sprites. These were moved vertically at a faster speed than the background was scrolling. The movement of the windows and wall decorations made it look as if the entire building were scrolling faster than the background.

In the actual timing of the sequence, once the city image disappeared below the screen, the scrolling stopped. The only things still moving were the sprite-based windows and wall decorations. But visually, it looked as if the entire building were continuing to scroll.

After the windows and wall decorations had been moving for a while, once the set amount of time had passed, the background scrolling resumed at the same speed as the windows and decorations. Then Mega Man appeared standing on the rooftop, the scrolling stopped, and the title was displayed.

Even my fellow planner, M.K., was puzzled when he saw this effect. He scratched his head in bewilderment and kept asking me, “How are you doing that?”

I was a little mischievous and simply stayed quiet and smiled at him for a while.

Although the effect had a strong visual impact, it also felt a little technical and niche. It might impress people in the game industry, but I had no idea how children would react to it. Still, I believed that the timing of the BGM, the appearance of the title, and the heroic image of Mega Man would make them feel something.

Hiragana and Kanji

There was one more new element I added: displaying the story in Japanese. In Mega Man1, we could only include alphabetic text in the ending. So for Mega Man2, I decided to display the text in Japanese.

The challenge here was displaying kanji. Looking back now, since the target users were children, hiragana alone would have been perfectly fine. But once the idea flashed into my mind, I felt that was the moment to put it into practice.

I had actually been thinking about using kanji for Japanese text for some time. Because of that, I was planning to use a somewhat larger font. I considered several possibilities, such as 16x8-dot vertically elongated characters or 16x16-dot characters. But after struggling with data capacity during the development of Mega Man1, I hesitated. Even though we had twice as much data capacity this time, was it really a good idea to use so much of it just for visual presentation?

That meant the character size would have to be 8x8 dots—one character. Even so, I just couldn't give up on my obsession with using kanji.

Then one day, I was driving a car when I noticed the lettering on a highway sign—the familiar green signs used on Japanese expressways—and thought, “Huh?”

The shapes of the kanji looked strange. I thought I might have been seeing them incorrectly, so I looked carefully and confirmed it. The kanji really did look different from the forms I was familiar with.

At the time, I had only recently gotten my driver's license, so I didn't know that the signs used a special typeface for expressways called “Kōdan Gothic.”

Kōdan Gothic was a typeface specifically designed so that drivers passing by at high speed could recognize the kanji instantly. The characters use thick strokes to make them easy to see from a distance, but with complicated kanji and characters with many strokes, those thick lines can merge together and make them difficult to read. So the designers used highly sophisticated techniques, such as reducing the number of strokes or replacing curves with straight lines, while preserving the overall impression of the kanji so that people could still recognize them correctly.

When I saw this, I thought, “This might work!”

I realized that perhaps I could draw kanji even under the extreme limitation of an 8x8-dot character.

In reality, there were kanji that I could draw and others that I simply could not. As a result, some of the Japanese expressions in the text used in the opening are a little strange.

Even so, I believe that this spirit of “If you think of something, just try it” somehow led to the creation of many different things.

Removing the Score

This was a very difficult and courageous decision to make. The role of a score in a game is to measure the results of play, allow players to compete through rankings, and provide motivation simply through the score itself.

In Mega Man1, we included a score without giving it any particular thought, almost as a matter of course. Having a score was simply considered standard for an action game. But in Mega Man, it didn't actually serve any useful purpose. It was not a good match for Mega Man's unique game system, which was something that had never existed before.

Mega Man is a game where you defeat boss enemies to obtain their special weapons, and then use those weapons to overcome stages and other boss enemies. What players compete over is not their score, but which boss enemies they defeated and which weapons they used to defeat them. And items are not obtained by increasing your score; they are obtained by defeating boss enemies.

However, the biggest advantage of removing the score was that it reduced the amount of information that had to be constantly displayed on the screen—the UI. Mega Man did not use any special processing just to display the score or life gauge. Because of this, the score, which required displaying a large number of digits, was actually one of the reasons we had to limit the number of things that could appear on screen, such as small enemies and the attack effects of boss enemies.

Implementing a Password System

This was also a groundbreaking idea for an action game. Some RPGs from other companies had already incorporated save functions using battery backup, but normally, adding something like that meant increasing the cost. There was also little feeling at the time that such a feature was necessary for action games, in particular.

Mega Man is an action game. However, as you progress through the game, you compare the attributes of enemies with Mega Man's attack attributes, much like in an RPG, and choose which stage to play and which boss enemy to fight. Having to repeat this thought process every time you got a Game Over was extremely troublesome. In Mega Man1, I felt that this made the game seem even more difficult.

The first thing that came to mind was, naturally, the Spell of Restoration from Dragon Quest. It was a system in which the game state was determined by a combination of characters. But the amount of data that needed to be saved in Mega Man was very simple. It was just which boss enemies had been defeated and which items had been obtained. So there was no need for a complicated string of characters, and a long password could also become a source of problems.

The purpose of this password was simply to restore the state you had reached at the end of your previous play session and allow you to continue playing from there. So rather than creating a complicated new password system, I thought it would be enough if entering predetermined words could reproduce the boss victories and losses and whether or not the items had been obtained.

For example, we could decide that entering “A0” would mean Boss 1 defeated, no items, while entering “B2” would mean Bosses 1 and 2 defeated, with two items obtained. There could be many combinations, but even so, it would be much easier to create than a string of random characters. The passwords would probably be shared among players immediately, but there was no reason to hide them.

However, having to enter characters every time you started the game was quite a hassle. In an action game in particular, forcing the player to do this kind of operation doesn't leave a very good impression. If we used character input, we would have to display all the characters on the screen and let the player move a cursor around, select a character, and confirm it. That meant we needed data for every character. Even characters that weren't actually used would have to be included as fakes.

This would waste both memory capacity and the player's time. It wasn't a very effective solution for an action game. I wondered if there was a way to make the operation simpler while also reducing the amount of data required. That was where I began thinking about a completely different form of password.

I wrote down the types of data that needed to be remembered and the number of possibilities for each in a notebook. There were actually a huge number of possible combinations. But as I mentioned above, the results themselves were extremely simple, like “A0” or “B2.” So one possibility was to create symbols in advance to represent “A0” and the other combinations, and simply have the player select them instead of entering characters.

But then it would no longer really be a password. If you could simply select the symbol you wanted from the beginning and start playing, it would have essentially the same meaning as selecting a difficulty level. What we needed was a way to arrive at a simple result through simple operations.

So instead of displaying all the characters and selecting them, I wondered what would happen if I placed the same characters in different positions. It was the opposite way of thinking. I could take a predetermined set of characters—A, B, C, D, E, F, and G—and place them inside a table divided into square cells.

This would limit the number of characters to only those actually needed. It would also eliminate the hassle of moving a cursor back and forth across a large number of characters. Since the amount of data we needed to remember was small, this kind of approach would be sufficient.

That's when I realized something: did we even need characters at all?

If the contents of the stored data changed depending on where a character was placed, then there was no reason for it to be a character in the first place. We could simply use circles (○) for everything.

If the circles were placed in predetermined cells within the table, the corresponding data could be retrieved. From a programming standpoint, this would be easy to handle because we could store the possible patterns in a table in advance and simply compare them.

With this method, all the player had to do was place a few circles inside a table of square cells. It was a very simple operation.

When I actually drew it in my notebook, it became a remarkably futuristic-looking password screen unlike anything I had ever seen before. The interesting thing was that the arrangement of the circles looked almost like a picture. Even for an action game, it felt more appropriate than a string of characters like the Spell of Restoration.

As I was drawing many different patterns in my notebook, I noticed something.

“Maybe I could memorize this.”

Normally, passwords are written down on paper and kept for later. But what if you could remember them without writing them down? That could become a major element.

When you started the game and the password screen appeared, you could quickly place the circle-shaped characters on the screen. If this operation could be done smoothly, you could continue playing from where you left off without any stress.

I tried this repeatedly, changing the number of cells in the table. 4x4. 5x5. 6x6. 7x7. As I increased the number of cells, it became more difficult to remember the patterns. I also realized that I needed to consider how many circles there should be.

I prepared several tables with different numbers of cells and placed a few circles in them in advance. These would be the passwords.

In the morning, I would look at the password and copy the circles onto an empty table. I would draw the circles in exactly the same positions.

Then, when it was time to finish work, I would take a new empty table and draw the circles in the positions I had memorized, without looking at the original.

After repeating this for many days, I began to understand what kind of table could be memorized naturally.

The number of cells that could be remembered naturally, without causing stress, was 5x5 or smaller. If I really tried to memorize them, I probably could have remembered larger patterns. But what I wanted to find was the kind of pattern that could be memorized naturally through repeated practice.

And that is how the 5x5 password screen that everyone knows was eventually created.

At the time, I remember thinking that since it wasn't made up of words, perhaps we could call it a “Pass Chart.” But it was never publicly announced or given an official name.

(Automatic translation by AI)

Files

Comments

Sheacu

In the Lion King, the monkey mentor Rafiki, says change is good but not easy. This article shows that. I wouldn't have guessed that the building was part of the background. This is why I like animation. It makes the impossible into a reality. I think that is the best part of being a content creator. I am surprised to learn that even Japan have different forms of Kanji. When I was learn basic Japanese Kanji, I learned that is was not just writing but an art form. To be honest, the score system in the first game doesn't make sense. I know it is used to keep score but there is no in-game reward for them at all.

AK’s Blueprints

The subject of kanji is extremely deep and quite complex. Kanji was originally imported from ancient China and adapted to fit the sounds of the Japanese language. Some of the kanji used in Kōdan Gothic are not technically correct. They are characters that should not normally be written that way in everyday writing. However, they are displayed more like symbols, with priority given to making them instantly recognizable. This is possible because kanji are not phonetic characters like the alphabet, but logographic characters in which the characters themselves carry meaning. The ancient hieroglyphs of Egypt and the cuneiform script of Mesopotamia were also logographic writing systems. Kanji is one of the few surviving writing systems of the same kind.