← DevEval Head-to-head ranking

ELO for engineers. With reasoning attached.

Chess solved fair ranking decades ago: rate players by who beats whom, and track how sure you are about every number. DevEval applies the same math — Glicko-2 — to engineering work. Every rating comes with a confidence level and a written justification you can audit.

The match

A match is two developers' last 30 days, read side by side.

AI reads both developers' recent work — pull requests, reviews, discussions — and issues a verdict per dimension and overall: who did better, how confident, and why. That verdict is what feeds the Glicko-2 rating update.

Match verdict Last 30 days · Code Quality
Alice beats Bob
Code Quality · 87% confidence
Reasoning
  • + Lower variance across PRs — quality holds under deadline pressure
  • + Fewer review iterations before merge
  • + 5 substantial reviews given vs 1
→ feeds the Glicko-2 rating update

A composite dashboard says “78 vs 65” — and can't tell you why.

A verdict names the winner, the dimension, the confidence, and the evidence. Disagree with one? Open the PRs it cites and argue with the reasoning, not with a number.

The math

Glicko-2, not plain ELO.

Plain ELO gives everyone a number. Glicko-2 also tells you how much to trust that number — which is what makes the ranking usable in an organization, not just a chess club.

01 · Uncertainty is a first-class number

Every developer starts with maximum uncertainty, not a fake precise rating. The system knows what it doesn't know yet — and says so.

02 · Every comparison narrows the band

Each verdict tightens the uncertainty band around a rating. Ratings are always shown with their confidence — a fresh 1500 and a proven 1500 are not the same claim.

03 · Stable after ~10 comparisons

Ratings stabilize after roughly 10 comparisons — about two weeks of normal activity. No six-month calibration period before the numbers mean something.

04 · Only informative duels

The system schedules competitive, comparable pairs — not random ones. Once a rating is confident it stops comparing, then re-checks periodically, so ratings never go stale.

Three perspectives

Relative strength, position, and delivered value.

DevEval separates three questions that a composite score blurs together: how someone compares with peers, where they stand in the organization, and how much value they actually delivered.

RELATIVE Strength vs peers

ELO / Glicko-2 rating

Relative strength per dimension. It changes after pairwise verdicts, when one developer outperforms another.

RELATIVE Position in the organization

Percentile

Where you stand in the organization, for example “better than 87% in code quality”. Answers the review question in one line.

ABSOLUTE Value delivered over time

Absolute Points

How much value was actually delivered. Absolute Points are not zero-sum, so they can grow across the entire organization at once. They show the scale of contribution and support quarterly reviews.

Worked example
The whole org adopts AI tooling. Everyone ships roughly 2× faster.
The three numbers react differently. That's the point.
Alice's ELO
#1 #1
±0
relative position unchanged
Alice's percentile
95th 95th
±0
relative, unchanged
Alice's Absolute Points
1500 3200
+113%
absolute, the whole org moved

One system shows both answers at once: who is strongest (stable) and how much more we ship (growing). A single composite score would blur the two into a number that means neither.

Why pairwise

Why pairwise verdicts beat averages.

Composite averages are how engineering analytics ended up ignored. Verdicts are built to be argued with — and to survive the argument.

The average
Hides reasons — a 78 is just a 78, take it or leave it.
The verdict
Carries evidence. Every verdict cites the PRs and reviews behind it — auditable PR by PR.
The average
Inflates as tools improve — suddenly “everyone is above average now”.
The verdict
Relative ratings stay meaningful across the AI transition: when everyone speeds up 2×, the ranking doesn't move.
The average
Compares everyone with everyone — apples with orchards.
The verdict
Only comparable work is compared. A backend senior isn't “ranked” against an intern's onboarding PRs in an uninformative blowout.

Run the first comparisons on your own team.

Connect a repo and let the first duels run — ratings stabilize within about two weeks of normal activity.

Start free trial
30 days · no card · cancel anytime
Private beta · early access

Self-service launches soon

We're prioritising enterprise rollouts right now. Leave your email and we'll let you know the moment self-service is open.

We'll only use this to reach out about DevEval. No newsletter, no sharing with third parties.