Where players quit
Learning paths through this topic
Count how many players reach each step of the start, even when they skip one
A chart of where new players drop out is only honest if each person is counted once per step, and a person who skipped the tutorial must not look like someone who left in it. This entry shows how to count so the numbers add up.
How much production is thrown away because storage is full, and when does it run dry?
A storage limit can quietly waste a lot of a player's work, and the same limit can leave them short a moment later. Both feel unfair, and the order in which you add and take away changes the totals. This entry shows how to simulate it so the numbers are right.
How long does a chapter really take, when many players never finish it?
Averaging the time of those who finished a chapter tells you how long the fast ones needed, and misses the slow ones who gave up halfway. The chapter then looks shorter than it feels. This entry shows how to use the time of those who left too, so the typical length is honest.
Is a long pause at a story choice a great dilemma or just confusion?
A choice that makes players stop and think is what a story wants, but a choice that makes them stop because they do not understand it looks identical in the pause time. This entry shows how to compare the pause with the time it takes to read the options and how to use the number of people who leave there to tell them apart.
Which hidden diary pages does nobody find? Count only players who entered the room
Story games hide notes, letters and photos that explain the plot. If a note is found by one player in a hundred it may simply be in a room that few people reach. This entry shows how to measure how well hidden each item is, apart from how popular its room is.
Notice when a detective story player has stopped finding clues before they quit
A stuck player does not announce it; the clues just stop coming. A fixed time limit treats a careful reader and a fast clicker the same way. This entry shows how to judge a pause in finding clues against that player's own rhythm, so a hint can arrive in time.
Spot players who go round and round the same conversation without getting anywhere
In adventure stories a hub of topics lets people ask everything, but some players loop through the same few lines and then give up, confused about what to do next. This entry shows how to detect that loop in the log without mistaking the normal return to the hub for it.
Are my story endings reached as often as I planned, or does one swallow the rest?
If nearly everyone lands in the same ending, the writing for the others is wasted and players may feel their choices do not matter. This entry shows how to check the spread of endings against what you intended, and how to avoid calling a small difference a problem.
How many players come back for the next episode, when the newest ones still have time?
After an episode ends, some players come back within days, some later, and some never. If you count everyone who has not returned yet as lost, a fresh release always looks like a failure. This entry shows how to count only those whose waiting time is really over.
Which scene to fix first: the one losing the most players or the one worse than normal?
The scene where the most players leave is often just the busy one at the start of the game, and the scene with the highest rate may have had three visitors. A better ranking asks how many players each scene loses beyond what a normal scene would. This entry shows how to compute that.
Do my hints actually help? What players do after each step of a hint ladder
Many story games give hints in steps, each more direct than the last. If most players still leave after the final hint, the ladder fails them. This entry shows how to read what happens after each step: solved, asked for more, left, or still stuck, so you know which step to rewrite.
How many hours is my story game really, for a first playthrough?
Telling people how long a story takes sounds easy, but the average is skewed by replays, by players who skipped most of the text, and by those who never finished. This entry shows which runs to count and gives a range to quote instead of a single number that is wrong for most.