# Engineering Performance Is Mostly Decision-Making, Not Technical Skill ## Claim Most of a team's performance comes from the quality of its decisions rather than the depth of its technical skill, so improving output means improving how choices get made under pressure. <!-- atomicity: 1 idea ✓ --> ## Evidence **Source**: [[3. Resources/Readwise/Articles/Reading the Game Mental Skills for High-Performing Engineering Teams|Reading the Game: Mental Skills for High-Performing Engineering Teams]] The claim arrives by analogy from elite sport, where the split is well established: > [!quote] > if you equate that to engineering teams, you're saying that 20% of our performance as a team is actually coming from technical skills. And the other 80% is the decision making The corresponding shift in where attention goes is the actionable half: > [!quote] > the biggest shift with for me was stop watching the ball and start watching the people That reframes what a lead is even looking at. Watching the ball means tracking the artifact — the PR, the deploy, the incident. Watching the people means tracking who is deciding what, with what information, under what pressure. The second is where the 80% lives, and it is invisible in every dashboard. The talk extends this to agents explicitly: > [!quote] > they will continue adjusting especially when you have AI teammates in there too and you also need to adjust to them ## Scope & conditions - The 80/20 split is transferred by analogy from sport, not measured in engineering. Treat the *direction* as the claim and the ratio as illustration. - Holds most strongly for teams whose technical floor is already adequate. Where genuine skill gaps exist, technical investment dominates and this advice misdirects. - Hardest to act on because decision quality is not instrumented. Most teams measure artifacts precisely and decisions not at all. - A related discipline from the same source: reflection immediately after a high-stress event captures adrenaline, not stable patterns — so retros held in the aftermath are measuring the wrong thing. ## Links - [[Product & Startup Instincts MOC]] — relocates team performance from technical depth to how choices get made under pressure - [[3. Resources/Permanent Notes/Individual Agent Adoption Does Not Become Team Capability by Aggregation]] — the same gap between individual skill and team output - [[3. Resources/Permanent Notes/Tool Access Does Not Aggregate Into Team Capability]] — extends: capability is not the sum of what people can technically do - [[3. Resources/Permanent Notes/Institutional Response Speed Distinguishes Tolerated From Addressed Harm]] — contrasts: decisions under pressure judged by their latency rather than their content