Blog

Senior Production Blog 5 (2/17/2017)

This week, I was asked by the team to come up with some alternative game modes that fit a handful of categories:

  1. They had to be easy to implement (no building massive systems from the ground up)
  2. They had to be more fast paced than the main game
  3. They had to keep the educational aspect in some way 

In early discussions with the team before I knew whether or not Rhythm of the Night, my previous project, would get cut, I discussed interest working on UI and alternative game modes for them, so I was very excited when this assignment came across my plate. Educational games and educational practices fascinate me. A lot of theories and methodology of educational software and educational games rely on repetition, pattern recognition, and using things as examples. There's a reason why flash cards work so well: repeated quizzing of flashcards, recognition of images or phrases on flash cards over and over can help cement ideas in peoples heads. 

Language learning software like Rosetta Stone and Duolingo frequently use flashcard like methodology as well as sentence construction (either typing, drag and drop, or orally) to help repeat learned words and lessons until someone gets it. 

With this in mind, I came up with 3 game modes, though only one of which will be making it into the final game.

Kanji Drop is a play on flashcard methodology without being as boring as flashcards. Similar to a flashcard quiz, in which one word or phrase is on one side, and the definition is on the other, players must match the definition to the kanji. But there's a catch: the kanji are falling down at a rapid rate, and players must slash from the definition to the matching kanji (or tap definition and slash Kanji. Exact interaction to be tested in QA), to pair the English and Japanese together. These alternative game modes are placed in transitions between the game's levels, so it can be a great mode to help cement and repeat the learned kanji from the previous section to solidify those lessons. It also is quick paced and arcade-y, giving the player a break from writing kanji over and over. The slashing like motions of the main game are carried over to this game mode meaning that players won't need to learn another control mode for this mode as well. After team discussion, Kanji Drop will be the game mode making it into the full game. However, I came up with two others.

The current battle system has players finding the opposite kanji to the one that their opponent drew. For example if your opponent drew Fire. you would draw Water or Ice. This logical thinking, especially in it's largely unexplained state currently, is difficult for people who just learned the kanji to understand what is "opposite" or what words even mean. As a proposed solution we have Opposites Attract. Players are given a kanji, and a left to right swipeable menu of all the kanji that they have learned so far. Players need to find the opposite kanji in the list. You can see a bit of an internal debate in the above design sketch: I didn't know if it would make sense to have English definitions on either of the two kanji. If it was on both, wouldn't people just be looking at the English definitions of both and finding the opposites? What if it was just on the given one? Looking through a big list of all kanji with no context might confuse players. What about putting them on the list? Would that end up being too easy? What about making it a difficulty option? Wouldn't players just then always choose the easier option? There are too many questions with this one, and to be honest it's not as engaging as Kanji Drop, so I personally feel like it's the weakest of the trio.

Lastly, we have Sentence Sentience. I'm fascinated by how language learning software instructs users to put together sentences even early on from just a few words. Even my short try of my dad's copy of Rosetta Stone Italian pretty much started off with sentences, and used those sentences to try to define words. We write a lot of words in Kanji Samurai, but without sentence construction, there's not a lot you can do with just words. You don't learn a language by just learning the words for Drive, Car, Fish, Cat, Walk, etc, but rather you learn those words in conjunction with grammar, tenses, sentence structure, special punctuation and characters ( the German Eszet (ß), umlauts, and more), and how they all work in conjunction with each other. I wanted to have Kanji Samurai have it's own sentence construction section, where players apply what they learned into actually communicating, rather than knowing the word for "old". It is, however, similar to existing gameplay, not fast paced, and may require more back end systems, but I hope it ends up in the game somewhere. Maybe the final boss could have the player write out a sentence, who knows.

For this upcoming week, I've been tasked to write a whole new tutorial from the ground up. We've changed the core first lessons in the tutorial, and I've been handed a document by Glynis detailing the lessons we have to teach and in what order, and have been left to my own devices. The team indicated that they liked the tone in my earlier tutorial rewriting, so I plan to keep it in that voice, but write from scratch instead of building upon and changing existing dialogue.

Luca Hibbard-Curto
Senior Production Blog 4 (2/10/2017)

Most of my week was consumed by my systems design class, working on a prototype system, with 16 pages worth of documentation along side it. Suffice to say, because of this I didn't have much time to work on Kanji Samurai.

However, I was asked by the team to come up with a solution for one of the UI issues we have yet to approach: the level select process.

The overall map

The overall map

We have this gorgeous piece of art for the map for our game. It starts with the tutorial and the rest of the map is covered by clouds until you finish other sections of the game, and the clouds clear. Through this we can provide a visual treatment for the players progression through the game. However, in the game's current state when a player selects a level, it pulls them away from the map to a seperate screen with a list of kanji in English on the left hand side, and a start button down below. Because the first kanji you currently draw in the game is the kanji for "one", the word "one" is in this list of words and players want to click it, thinking it's the way to start.

Later levels have the core 3 game modes: practice, spar, and battle, which could theoretically have different sets of kanji, so they are presented in their own columns with a brief description of gameplay. My solution for this clutter and confusion is this:

The proposed solution scraps the second screen where players confirmed that they wanted to play that mode and instead uses a scroll that drops down from the top of the screen when players make a selection. Each kanji is listed out, but this time with a graphical representation, English translation, and anglicized translation. This no longer has the confusion of players wanting to click "one" as the context around "one" makes it clear that it isn't a button. Not in the above sketch, but discussed later with programmers is arrow buttons on each side of the "In This Level" rectangle, allowing selection of game modes, and moving around elements in the UI can allow for a brief description of game modes as well. 

If players don't want to play the level they selected, they no longer have to hit the back button and wait for the scene transition to bring them back to the map screen, they can instead just tap the map again, and the scroll will disappear, allowing them to make the choice that they wanted. This echos mobile app design, and is used in apps like Uber, Googles suite of apps, Tweetbot, FitBit, Facebook, and others. 

It also cements the idea of UI elements dropping down from the top of the screen, our agreed upon UI scheme for gameplay, so UI is consistent across all aspects of the game and players know where to go when they want to bring the UI down.

This next week I've been asked to come up with VDD's and Design sketches for alternative game modes to help sell the learning aspect even more, and we have a meeting tomorrow at 1 PM.

Luca Hibbard-Curto
Senior Production Blog 3 (2/3/2017)

This week, the team challenged through Greenlight, meaning a lot of documentation had to be done, addressing various requested changes. While I had done the main element I had been assigned last week (reforming the tutorial), Glynis was working on finishing up the Game Design Document, fully committing to changes making the game more "game like", adding scores, ranking, and bonuses. The Design Document, presented in powerpoint format, was pictureless, so my job this week was to take the Design Document and give it a graphical treatment. 

The VDD fully encapsulates what is in the powerpoint Design Document: the 3 different modes, the scoring system, stroke order, ranking players kanji, and the bonuses they get in various modes.

The scoring system seems interesting: it assigns each node a value of 5 points, and each node you get wrong subtracts 5 points. This then factors into your rank, and if you make mistakes, you lose out on bonuses. However, you can still get everything right, and not get all bonuses. Stroke order, while important in actually writing the Kanji is kind of boring and overly complicated for a video game adaptation of a language. Thus, stroke order is now relegated to a bonus feature. It also means that a player can literally be perfect at our game. By getting the kanji right, using no undos, in proper stroke order, the player can perfectly draw each kanji. In order to keep scores competetive, a time bonus is added, so unless players are drawing all the kanji in less than a second, each players score will be slightly different, meaning that the player has room for improvement, and encouragement to keep playing. After creating the VDD, we had a brief meeting, blocking out who is going to talk where during the greenlight presentation, and that's what we've got done so far. We have a meeting coming up tomorrow blocking out the next two weeks of work for everybody, so I'm excited to see what comes next.

Luca Hibbard-Curto
Senior Production Blog 2 (1/27/2017)

And we're off!

Sunday came, and a massive 2 hour meeting with it, helping us understand what steps to make going forward. Before the meeting, I had worked with Julia, one of the artists on adapting some of her sketches into a UI Prototype.

The original sketch from Julia

The original sketch from Julia

Making interactive PDF's is something I set myself down to learn last year, I adapted this sketch and make some photoshop mockups using ingame assets I found poking around the repo. And then from there I took screenshots of every existing UI menu in engine, and stitched them all together.

GIF.gif

It's rough, but it's much cleaner than the current UI:

Buttons are scattered all over the place and aren't really necessary all the time. The checkmark button is going away, as programmers this week have been working on an auto detect system that automatically checks and accepts correctly drawn kanji. We really only need 3 permanent buttons: play audio (pronunciation of kanji), undo, and hint. These 3 buttons are now permanently affixed to the bottom right of the screen, and any additional buttons can be moved down there with small glyphs as well. Something else about the above screenshot is how weighted it is to the left. The game takes place in the center, while the tutorial takes up 80% of the right 1/5th of the screen. This means that players are looking to the left every few moments for a game that takes place in the center of the screen. We plan on moving this tutorial UI above or on top of the game board, with freshly updated text.

The tutorial as it stands is wordy, overly formal, and too tutorial-y. It's in a game about samurai, with mention to them once at the very end. So this week I was giving the task of transcribing the entire player experience for the tutorial:

What words show up when, what animations for hints play when, when is the player left alone, and simplifying and decorating that old crusty text.

Even this is subject to change, as we have made the decision to go with smaller, starter kanji and teach combinations of them that make different kanji (think letters not words), but it's much more flavorful, gives encouragement,  and is less redundant. Tomorrow we meet again to discuss what's left in terms of getting things ready for Greenlight, but we've already made a lot of progress, and our initial steps in cleaning up the rough stuff in the original capstone prototype.

Luca Hibbard-Curto