Friction Score Calculator
Enter the engineering signals your team hit in the last 7 days, grouped by severity. Uses the same formula as the Deviera dashboard — lower is better.
Signals in the last 7 days
CI failures on main/master · failed deployments
Stale PRs · CI failures off main · changes requested · review time exceeded
PRs opened · issues opened · oversized PRs
Pushes · approvals · TODO commits · first contributions
Total signals: 100
Enter your signal counts and calculate to see where your team lands.
How the Friction Score is calculated
Every engineering signal Deviera detects carries a severity. Over a rolling 7-day window, those signals are weighted — critical ×4, high ×2, medium ×1, low ×0 — summed, then divided by the total number of signals and scaled to 0–100.
The division is the part most people miss, and it is what makes the metric useful. A raw weighted count would simply track team size: a busy team would always look worse than a quiet one. Dividing by total signals turns the score into a measure of severity mix. The practical consequence is that healthy activity pulls your score down — 200 routine pushes alongside 3 critical failures scores far better than 3 critical failures alone.
The reference points are worth knowing: a window that is 100% critical scores 100, 100% high scores 50, 100% medium scores 25, and 100% low scores 0. No signals at all also scores 0.
- Critical (×4): CI failure on main/master, failed deployment
- High (×2): stale PRs, CI failure on other branches, changes requested, review time exceeded
- Medium (×1): PRs opened, issues opened, oversized PRs
- Low (×0): pushes, approvals, TODO commits, first contributions
This calculator uses the same function as the product, so the number here is the number you would see on your dashboard for the same signal mix. See how Friction Score works in Deviera for the live version.
Frequently asked questions
- What is a Friction Score?
- A Friction Score is a single number from 0 to 100 — lower is better — representing how much engineering friction a team is carrying. It is computed from the delivery signals of the last 7 days, weighted by severity, so one figure replaces reading five separate dashboards. A score of 0 means no friction; 100 means the team is under critical load.
- How is the Friction Score calculated?
- Each signal from the last 7 days is weighted by severity: critical counts 4, high counts 2, medium counts 1, and low counts 0. The weighted sum is divided by the total number of signals in the window and scaled to a 0–100 range. Because it is a ratio rather than a raw count, the score measures your severity mix rather than how busy the team is.
- Why do low-severity signals lower my Friction Score?
- Low-severity events — pushes, approvals, TODO commits — add nothing to the weighted sum but are still counted in the total the sum is divided by. That is deliberate: a team shipping 200 routine events alongside 3 critical ones is in better shape than a team whose only 3 events are all critical. Healthy activity dilutes friction, so a calculator that ignored low-severity signals would badly overstate your score.
- What counts as a good Friction Score?
- Roughly: 0–25 is Healthy (shipping cleanly), 26–50 is Moderate (some friction worth addressing soon), 51–75 is Stressed (friction is actively slowing delivery), and 76–100 is Critical (immediate action needed). The goal is a low, stable score that trends down — a single week is less informative than the direction over a month.
- Which engineering signals count as critical?
- Only two categories: a CI failure on your main or master branch, and a failed deployment. Both block everyone rather than one person, which is why they carry the heaviest weight. CI failures on other branches, stale pull requests, and requested changes count as high; opened pull requests and issues count as medium.