How Zach Gage fine-tuned the difficulty curve of TypeShift
Rather than simply ask a player to solve a single jumbled word, or even several, TypeShift — which is available on iOS — instead asks players to create any number of words out of several tiers of letters until they’ve used every single one.
While this might sound a bit complicated, Gage told Gamasutra that the idea for TypeShift essentially came to him fully formed without any thought before he actually sat down to do it.
For Sage Solitaire, Gage says he was fiddling with single-player, poker-style games for about a year before it came together. With TypeShift, on the other hand, he wasn’t even working on word games at the time.
“Basically I just woke up one day with the game in my head, and I programmed it, and it worked,” he says. “Usually I have a problem I’m trying to solve, or a mechanic I want to explore, and oftentimes those are things that I’m working on for a long time.”
“If you play the game, it basically is what I woke up with,” he says. “You have these words stacked on top of each other, you can move columns up and down, you try and spell words in the center row, and if you use every letter one time you solve the puzzle.”
Just about everything else from the initial dream-created design made the cut, however. The core conceit remained a sort of lock picking, or safecracking, and the satisfying clicking that the letters make as they scroll through is intentional: Gage wanted players to feel as if they were really opening a lock.
Even though the core concept was virtually unchanged from the moment of conception, Gage spent almost a year fiddling with the design.
According to him, the mechanic of sliding letters back and forth on a digital combination lock is the same, but everything about how the puzzles themselves are designed has been iterated on again and again.
“I really wanted to have a puzzle where, as you worked your way through it, the puzzle got easier,” he says.
This is a common feature of crossword puzzles and the like, but TypeShift’s design makes it more complicated. In the early iterations, the puzzles never seemed to get any easier despite how many words had been solved.
“Another way to look at it is often in a crossword puzzle or Picross, there’s an orthogonal solution to a problem,” he says, “so if I don’t know the answer to one clue, I can solve some other clue and that’ll give me the answer to the one I’m having trouble with. But in TypeShift that doesn’t exist, which posed kind of a weird problem because what would happen is, you can be in a situation where you might not actually know the word that you’re looking for.”
In that situation, how would a player go about figuring it out? Finding that last word could take 20 minutes, an hour, or the player could simply never figure it out. “Even if you’ve solved 10 out of the 11 letters in the puzzle,” Gage added, “that doesn’t actually help you in any way with the 11th letter.”
Solving this problem ultimately took a three-pronged approach, the first of which was contextual hints.
“Each hint gives you one letter,” Gage says, “but if you have the word that it’s trying to give you a hint for -- or if you have part of it -- it’ll give you other letters.” Gage used the word “woods” as an example, stating that if the player already had “wo” in the correct order, requesting the hint indicates that they are correct before sliding the second “o” into place. This can be repeated until the whole word is solved.
Secondly, it was a matter of designing puzzles that could teach players to solve the more difficult ones without being too easy or too hard themselves.
“There is actually a bit of a weird skill that allows you to get unstuck when you’re missing letters,” he says, “and that skill is just understanding the structure of the English language and how words are built.”
Once you know that, according to Gage, it becomes easier to tell which ways to explore, which avenues will be more fruitful.
“The big discovery there was that if I make puzzles that are just three words,” Gage says, “even puzzles that are seven-letter words, which should be very difficult, when they only have three core words in them are very solvable by beginners. They feel very inspiring to solve, because when you first look at one of those puzzles, it looks very overwhelming. And then as soon as you put one of the words together, it often solves the other two very quickly.”
The final step to really nailing TypeShift’s design is what’s called a “clue puzzle” in the game.
When pressed, Gage didn’t have an elevator pitch on hand to describe clue puzzles — “I think I’d maybe say they’re crossword puzzles meets Rubik’s cube”.
But basically it’s the normal TypeShift combination lock puzzle with a list of crossword puzzle clues. Each clue must be solved before the puzzle is totally solved.
“When you play the game, it seems like clue puzzles are really obvious in terms of a mechanic to use,” he says, “but actually they took me like a year and a half to see. I think often when you’re working on a design, you have kind of a vision for what is good about that design, and what the design is allowing players to do, and what’s fun about it, and sometimes that can blind you to stuff that’s right under your nose, which for me it was clue puzzles.”
One other core feature that was present at the initial moment of inspiration also had to be tweaked. The game’s name was not always TypeShift. “When I woke up from the dream with the concept, the name that I had in mind was Safe Words, because you are sort of cracking a safe, but also it’s a pun,” he says.
Ultimately, Gage ditched the name for something less… loaded. “I didn’t want to name the game anything that anyone might be uncomfortable for anyone going and telling their mom to play.”