--- title: "Analyzer FAQ" url: "https://docs.noreja.com/en/article/analyzer-faq" language: "en" last_updated: "2026-10-23T09:16:50.105Z" --- # Analyzer FAQ Short answers to the questions that come up most when figures in the Analyzer and on dashboards don't seem to match: executions vs. cases, relationships, batching, positive and negative times, averages, filters, variants and the violation rate. The same process can show different numbers depending on where you look: on an activity, on a relationship, in the details table or on a dashboard. Usually both numbers are right; they just count different things. This page answers the questions that come up most often, with figures from the Purchase-to-Pay perspective (5,520 purchase orders). Each answer links to the page that explains it in depth; the terms are defined in the [Glossary](https://docs.noreja.com/en/article/glossary). ### Why Does an Activity Show More Than the Number of Cases? The number on an activity counts executions, not cases. One case can execute an activity several times. In Purchase-to-Pay, the 5,520 orders have 5,976 executions of "bestellposition angelegt (HW)", because one order can contain up to three hardware items. The number of cases is in the details table (**Case amount**). See [Quantities and Cardinalities](https://docs.noreja.com/en/article/quantities-and-cardinalities). ### Why Don't the Relationships Leaving an Activity Add Up to Its Number? The number on a relationship counts how many executions of the first activity continued on it. They rarely add up, for three reasons: - **Parallel steps**: one execution can continue on several relationships at once. From "Bestellung erstellt", 86% go to "rechnung erfassung", 75% to hardware items and 19% to software items, together more than 100%. - **No next step**: executions after which the case ends unexpectedly don't continue on any relationship. - **Violations**: relationships shown as Disarray, Negligence or Rework have their own numbers. The percentage next to a relationship is the share of the first activity's executions, not of all cases. See [Quantities and Cardinalities](https://docs.noreja.com/en/article/quantities-and-cardinalities). ### Why Does an Activity Done for Several Cases Count Only Once? Some activities handle several cases in one go, for example one invoice run for several invoices. Such a batched execution is one execution, so it counts once in the activity's number and once in the averages of the process model, however many cases it covers. See [Quantities and Cardinalities](https://docs.noreja.com/en/article/quantities-and-cardinalities). ### Why Don't Positive and Negative Times Add Up to the Total? With **Limit path duration to:** **Positive** or **Negative**, a relationship shows how many executions had at least one such transition duration, followed by the total in brackets. One execution that leads to several later steps, some earlier and some later, counts under both. Steps without a time stamp count in the total but under neither. On the relationship from "Software Bereitstellung" to "rechnung erfassung", 1,404 executions have a positive transition duration and 119 a negative one, out of 1,404 in total. See [Positive and Negative Times](https://docs.noreja.com/en/article/positive-and-negative-times). ### Why Don't the Transition Durations along a Path Add Up to the Cycle Time? The two are averaged over different things. The cycle time (**Cycle time (avg.)**, **Average process duration**) is averaged over cases: each case counts once. The transition duration on a relationship is averaged over executions: a step that happened three times in one order counts three times. On top of that, steps can run in parallel, and negative transition durations lower the average of a relationship. See [Average Time Calculation](https://docs.noreja.com/en/article/average-time-calculation). ### Why Does a Dashboard Show a Different Figure Than My Analyzer Tab? Every dashboard and every analyzer tab has its own filters, and all figures refer to the cases that match them. If the dashboard filters for a period or a supplier and the analyzer tab doesn't, their figures differ. Compare tabs with the same filters. See [Filter Cases in the Analyzer and Dashboards](https://docs.noreja.com/en/article/filter-cases-in-the-analyzer-and-dashboards). ### Why Do the Figures Change Slightly When I Filter? Without a filter, the Analyzer shows the process model calculated at the import. With a filter, it calculates the process model again from the matching cases. A few things look different as a result: - Averages can differ by up to one second, because they're rounded to whole seconds at different points. - The **Variants** list shows only the variants of the matching cases, sorted by their number of matching cases. The variants keep their numbers, so the list can start with #4. See [From Events to Process Model: What Is Aggregated Where](https://docs.noreja.com/en/article/from-events-to-process-model-what-is-aggregated-where). ### Why Did the Variant Numbers Change after a New Import? Variants are numbered at every import by their number of cases, the most frequent first. With new or changed data, the number of cases per variant changes, and so do the numbers and the order. A case can also move to another variant if activities were added to it. If you need to find a variant again, look at its path rather than its number. See [Process Models and Levels: Case, Variant, Process Model](https://docs.noreja.com/en/article/process-models-and-levels-case-variant-process-model). ### Why Doesn't Repeating an Activity Create a New Variant? A variant groups cases with the same structure: the same activities, connected by the same relationships. How often an activity was executed doesn't matter. An order with one hardware item and an order with three hardware items follow the same path, so they belong to the same variant. The Case view shows the difference: with **Model** the case shows each activity once with its number of executions, with **Instance** every execution separately. See [Process Models and Levels: Case, Variant, Process Model](https://docs.noreja.com/en/article/process-models-and-levels-case-variant-process-model) and [Work with the Case View](https://docs.noreja.com/en/article/work-with-the-case-view). ### Why Is the Violation Rate a Number and Not a Percentage? The **Violation rate** widget shows how many cases have at least one order violation (Disarray) or skipped activity (Negligence), on a gauge from 0 to the number of cases. In Purchase-to-Pay, that's 2,016 of 5,520 cases. Allowed deviations and Rework don't count; Rework has its own widget, **Average rework rate**, which is a percentage. See [Dashboard Widgets](https://docs.noreja.com/en/article/dashboard-widgets), [KPIs Explained](https://docs.noreja.com/en/article/kpis-explained) and [Process Patterns: Disarray, Negligence, Rework and Allowed Deviations](https://docs.noreja.com/en/article/process-patterns-disarray-negligence-rework-and-allowed-deviations). ### Related - [From Events to Process Model: What Is Aggregated Where](https://docs.noreja.com/en/article/from-events-to-process-model-what-is-aggregated-where) - [KPIs Explained](https://docs.noreja.com/en/article/kpis-explained) - [Quantities and Cardinalities](https://docs.noreja.com/en/article/quantities-and-cardinalities) - [Positive and Negative Times](https://docs.noreja.com/en/article/positive-and-negative-times) - [Average Time Calculation](https://docs.noreja.com/en/article/average-time-calculation) - [Glossary](https://docs.noreja.com/en/article/glossary)