> For the complete documentation index, see [llms.txt](https://docs.siminsights.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.siminsights.com/scoring-and-mastery/how-scores-work.md).

# How scores work

How a score shown in a simulation differs from a competency model, and what skill mastery means.

A score and a competency model do different jobs. A score is shown to the learner inside the simulation and is not logged to user data. A competency model is logged to the learner's account and is used to track performance over time.

|                     | Score                             | Competency model                                    |
| ------------------- | --------------------------------- | --------------------------------------------------- |
| Where it appears    | In the simulation only            | In the learner's account and in HyperSkill Platform |
| Logged to user data | No                                | Yes, to the learner's account                       |
| Purpose             | Immediate feedback during the run | Tracking performance over time                      |

The rest of this page explains how quiz counts and skill mastery are calculated.

## Quizzes in lessons

A quiz, or a single multiple-choice question, is scored as a count: right answers out of questions asked. The lesson's results screen shows that count, for example 7 / 10. Open-response questions and flashcards are not scored, so a lesson that mixes them with multiple choice says so on its results screen.

## Simulations with a competency model

A simulation with a competency model scores skills, not questions. If a learner's skill level has not been observed, it starts at 50%. Each time the learner does something the simulation was built to observe, that skill's value moves, and the change is carried up to the latent skills above it in the model. At the end of the run each skill has a value from 0 to 100%, and the overall mastery is the average of the model's user attributes, all weighted equally.

These values are what appears as **Skill mastery** and **Overall mastery** on a run, and what is averaged into a simulation's, a session's or an organization's mastery figures.

A simulation without a competency model records no skill mastery. Its runs still show the timeline, event log, replay, and other data.

## Spark conversations

A Spark simulation scores the conversation with agents the author set up. Each agent judges one thing, such as use of filler words, and can set a skill value the same way a simulation does.

## Related pages

* Competency models: the skill trees behind simulation scoring.
* Run detail: where skill mastery appears for a single run.