> For the complete documentation index, see [llms.txt](https://docs.multitudes.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.multitudes.com/knowledge-base/why-we-use-percentiles.md). # Why we use percentiles We talk a lot about percentiles in Multitudes, so this page provides an intro to percentiles – what they are, why they're useful, and which ones you should use for what. To set your default percentiles in Multitudes, jump to this section: [How to set preferred percentiles](#how-can-i-set-my-preferred-percentile). ## What are percentiles? A percentile tells you what percentage of the data sits under a given number. If the 75th percentile (P75) for age in a room of 100 people was 39, that means 75 people in the room are 39 or younger. Some percentiles have special names: * P25 = first (or lower) quartile * P50 = median, or second quartile * P75 = third (or upper) quartile Percentiles give you a clearer picture of your data's shape, because you always know how many observations sit above or below a given point.
## Where do you use percentiles in Multitudes? Charts for certain metrics in the app show the data aggregated to percentiles instead of the average: * `Change Lead Time` * `Coding Time` * `Review Wait Time` * `Editing Time` * `Deploy Time` * `Lines of Code Changed (PR Size)` * `Files Changed (PR Size)` You can set your preferred percentile for these charts to be P50, P75, P90, or P95 (we recommend P75, or P50 if you're just getting started). See more here: [How to set your preferred percentile](#how-can-i-set-my-preferred-percentile). ## Why not just use the average (mean)? Both the mean and the median describe what's "typical" in a dataset (they're both [measures of central tendency](https://www.abs.gov.au/statistics/understanding-statistics/statistical-terms-and-concepts/measures-central-tendency)) – but they respond very differently to outliers. The **mean** (or "average") is calculated by averaging all the numbers together: sum all the numbers and divide by the count of numbers. Because the value of every number is included in the sum, the mean is susceptible to outliers. For example, if your largest number is 10x the next-highest value, that will drag the mean up. The **median** doesn’t have that issue. It’s rank-ordering the numbers, so whether your largest number is 10x the next-highest or just 1.1x the next-highest, the rank doesn't change – and so the median likely won't be impacted. As long as you have 3+ numbers in your dataset\*, one outlier won't change the median. Another benefit of using the median is that you always know that 50% of the values sit above it and 50% of the values sit below it. This makes it much better for making real-world decisions. For example, say you want your team's code changes to get a human review in under 4 hours. If you measured that goal with the average, one review that takes weeks would drag the average up, even if most reviews take under 4 hours. That single outlier would make the goal look further out of reach than it really is. Your P50 or P75, on the other hand, won't move if your dataset is big enough (in this case, at least 6 numbers.) *\*Note:* If you only have one or two numbers in your sample and one is very large, that will impact the median. When you have one value, the median is that value – and for a dataset of 2, you'll average the two values to get the median. That said, it's hard to draw any conclusions from datasets that are this small anyway. ## How do you calculate a percentile? First, sort the values in your dataset from smallest to largest. There are several accepted formulas for calculating percentiles, but the core idea is the same: cut the sorted data at the relevant point. For quartiles, you split the data into 4 equal groups; for deciles, into 10 equal groups. A percentile is just a more general version of this – it tells you the cut point for any percentage you choose. For more about how to calculate percentiles, read our blog post here: [What are percentiles – and why we recommend P75](https://www.multitudes.com/blog/what-are-percentiles-and-why-we-recommend-p75) ## What percentile should you use? This depends on your goals. For Multitudes metrics, we generally recommend P75. Here's why: * P75 gives you confidence that most datapoints are under that level – because 75% of the numbers will be less than that. If P75 looks good, most of your work looks good. * That said, the median (P50) can be a good starting point for teams early in their goal-setting journeys who want to start with something more achievable – since P50 just looks at what half the data did. P75 walks a nice line between showing you what most (75%) of the work looked like, while still leaving room for outliers. ## How can I set my preferred percentile? Your options depend on your permission level in Multitudes: * [Owners and Managers](#owners-and-managers) can set the organization default * [Anyone](#everyone) can update the percentile they see when they log in ### Owners and Managers Owners and Managers can set the organization default. * Setting a default percentile for your organization will determine with percentile is used in Multitudes's notifications, and will set the default when a team member first logs into Multitudes. * To set this, go to [Settings > Organization Settings > Default settings](https://app.multitudes.co/team-settings/organization-settings?section=default-settings) and choose your preferred percentile. (Screenshot below.)
### Everyone Anyone – from Members to Managers and Owners – can update the view they see when they log into the app. We use saved filters in Multitudes, which means that we'll save your most recent filters and load those the next time you visit the app. So to change your percentile: 1. Go to one of the pages that has a filter for percentiles: The [Flow of Work](https://app.multitudes.co/flow-of-work) or [AI Impact](https://app.multitudes.co/ai-impact/impact) pages 2. Find the "Percentile" filter at the top of the page and choose your preferred percentile (screenshot below). We'll then show you this percentile the next time you log in.