Difficulty
Learning paths through this topic
Is a timing level harder on weak phones, or are weak-phone players just different?
A game that stutters on a weak phone can make a precise level unfair, but those players also tend to differ in other ways. This entry shows how to compare a level with an easy baseline so you blame the phone only when the extra gap is real.
Are players missing tap targets because the target is too small for a thumb?
Missed taps feel like the player's fault until you look at where they landed. This entry shows how to tell a near miss from a deliberate tap elsewhere and how to read a better hit area out of the data without making the game trivial.
Do players fail more when the action happens where the thumb cannot reach?
On a phone held in one hand, some parts of the screen are awkward, and a level that asks for precision there is harder than its design suggests. This entry shows how to split the screen into regions and tell a real hotspot from noise.
Is my optional hard path a fun dare or a trap that makes players quit?
Many players take an optional challenge, but that alone does not say whether it was a good idea for them. This entry shows which three numbers to read together to tell a welcome dare from a trap that sends people away.
Set the gold, silver and bronze times from what players really achieve
Time medals picked by feel are either given to everybody or nobody can reach them. Real finishing times give you fair limits, as long as you leave out the glitch runs and wait until enough players have finished.
Is my parry window too tight, or just in the wrong place?
When most parries fail, widening the window is not always the answer. This entry shows how to read early and late presses from the log to decide whether the window needs to grow or just move.
Is the Tuesday daily puzzle really harder, or are different people playing?
Comparing solve times by weekday mixes up hard puzzles with changing crowds, because the players who show up on a weekend are not the ones who show up on a Monday. This entry shows how to measure each puzzle against the person who solved it.
Does this level only seem harder because players meet it late in a long session?
Tired players solve more slowly, so a level that tends to appear late can look harder than it is. This entry shows how to measure the tiredness effect from the same levels and correct for it before you change anything.
Spot the puzzle level where people stare before they touch anything
When the first touch takes unusually long, the goal is probably unclear rather than the puzzle hard. This entry shows how to pick out such levels from a list of waiting times without one odd level distorting the yardstick.
Rate a number-grid puzzle by the hardest step it needs, not by how many cells are empty
Two grids with the same number of empty cells can be a pleasant stroll and a brick wall. This entry shows what to measure when a solver walks through the puzzle, and how to put puzzles in a fair order using it.
Count hint use on a puzzle level without counting the curious taps
A hint button gets pressed for many reasons, and most of them are not about being stuck. This entry shows how to separate real need from idle curiosity so that the list of levels people need help with is honest.
Do hints actually unblock players, or do they leave right after using one?
A hint that is used a lot is not necessarily a hint that works. This entry shows how to follow what happens in the minute after a hint and how to count progress so that a pointless move after it does not look like help.