How Should Mentor-Type Characters Ask Questions So Players Decide for Themselves?

Changing a mentor's lines from "conclusions" into "answerable questions" is the most direct fix for this type of character. The method: have the mentor state what they observe, give no more than two clues, then put the trade-off in front of the player, letting them choose a direction and bear the consequences. The mentor may have preferences but must not declare which choice is "correct." Below, fictional teaching examples illustrate hint gradients for three difficulty levels and how a mentor can respond when the player directly demands the answer.

A mentor answering for the player usually shows up in three ways: stating the conclusion directly ("Restore the left one first"), eliminating all wrong options for the player ("The paint on the right one isn't dry, don't touch it"), or using authority to override the player's judgment ("I've restored for thirty years, listen to me"). All three take the decision out of the player's hands.

To change this into questioning, the mentor needs to break the same information into three layers:

1. Observation: The mentor says what they see, without explaining what it means.
2. Clue: The mentor provides one limited hint pointing toward a judgment direction, but does not give the conclusion.
3. Trade-off: The mentor asks which cost the player is willing to bear, handing the choice back.

The three layers do not have to be said all at once. The more urgently the player asks, the more the mentor can give only observation; if the player is willing to stop and look, the mentor can then give a clue.

Set up a fictional teaching scenario: the restorer character "Ke Lan" brings a new apprentice "A Che" into a painting storage room, where two paintings await restoration. Painting A is Rain Harbor; there is a vertical tear in the lower left corner of the canvas, and the edges of the paint layer are lifting. Painting B is Old Clock Tower; its surface looks intact, but there is an area in the center of the image darker than its surroundings, and it feels slightly raised to the touch. Ke Lan knows there may be an old restoration layer beneath Painting B, but this is not a world fact that must be told to A Che immediately; it is only Ke Lan's own judgment based on experience.

A Che asks: "Master, which one should we restore first?"

If Ke Lan answers, "Restore A first, the tear will keep spreading," that is answering for the player. Changed into a question, Ke Lan could say: "First look at that tear in the lower left of Painting A, then look at the dark area in the center of Painting B. Which change are you more worried will continue happening in this storage room?" This sentence gives observation and also gives a direction for the trade-off, but does not choose for A Che.

The same scene can be designed with three tiers of hints according to how much help the player needs. Difficulty does not mean the puzzle itself becomes harder, but rather how deeply the mentor intervenes.

Low difficulty suits players just encountering this kind of judgment; the mentor states the phenomena clearly. Medium difficulty suits players who already know how to look; the mentor only reminds them of environmental variables. High difficulty suits players familiar with the process; the mentor gives only limiting conditions. All three tiers keep the same principle: the mentor does not declare a single correct answer.

When the player asks, "Just tell me which one to restore first," the mentor does not have to hold firm. They can give a conditional answer and keep the decision in the player's hands. For example:

A Che: "Master, just tell me which one to restore first." Ke Lan: "If we had to start today, I would touch Painting A first, because the edge of the tear has already lifted, and I worry that a piece will fall off while I turn away. But this is my worry, not your judgment. Are you willing to follow my worry, or do you want to look first at the dark area in the center of Painting B?"

This response gives Ke Lan's preference and explains that the preference comes from her observation, but does not package the preference as a world rule. The player can still choose Painting B. If the player chooses Painting B, Ke Lan can ask: "You chose B because you saw something, or because you want to rule out the more uncertain one first?" This question turns the player's reason into material for the next narrative step, rather than judging right or wrong.

If the player insists, "I don't want to think, you decide," the mentor can step back and give a temporary decision while marking it as the mentor's decision, not the correct answer: "Then I will restore A first. You can watch me handle the tear from the side, and afterward decide whether you want to touch B." This neither lets the conversation stall nor turns the mentor into an answer machine.

For a mentor's questions to carry weight, they are best related to the mentor's own goals, needs, or boundaries. Ke Lan's goal is for A Che to learn judgment; her need is that no major accident can happen in the storage room; her boundary is not bearing all responsibility for the apprentice. Her questions can then revolve around "which consequence are you willing to be responsible for," rather than vaguely asking "What do you think?"

If the mentor only casually asks, "What do you think?" the player will feel the question has no cost. Connect the question to concrete limits, such as "You only have one chance to start," "The other one must wait until the next round," "I cannot watch both paintings at once." Only then does the trade-off feel real.

After revising a mentor dialogue, check item by item:

• Does the mentor state at least one concrete observation, rather than only giving a conclusion?
• Does the mentor give no more than two clues, and not eliminate all options for the player?
• Does the player face a real trade-off, rather than "only by choosing correctly can you continue"?
• Does the mentor have their own preference, but the preference is not written as the single correct answer?
• When the player directly asks for the answer, does the mentor give a conditional response rather than declaring a rule?
• Does the player's choice affect later narrative, relationships, or state, rather than only changing one line of dialogue?

If one of these does not hold, fix that one first; there is no need to rewrite the whole scene. The questioning of mentor-type characters ultimately lets the player make a decision they are willing to claim within limited information. Next, you can pick an existing dialogue, cross out the mentor's final conclusion, and replace it with "What did you see, and which consequence are you willing to bear first?"