{
  "author": {
    "name": "Dylan Fitzgerald",
    "url": "https://dylanfitzgerald.net/"
  },
  "blyg": "0.4",
  "changelog": [
    {
      "at": "2024-09-20T03:17:00Z",
      "note": "Imported from the Ctrl+Shift archive",
      "version": 1
    }
  ],
  "content_hash": "sha256:6432a8582d3e644745a4e4cf3773c2b7b8a7fef772c41e8870f471cf515e4edc",
  "content_html": "\u003ch1\u003eConfidence as a continuum, not a boolean\u003c/h1\u003e\n\u003cp\u003eIn work as in everyday life, we tend to frame confidence as a black-and-white boolean. The key words used are “Are you confident that …?”, as in:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eAre you confident you understand the issue?\u003c/li\u003e\n\u003cli\u003eAre we confident as a team that we’ll ship X by Y?\u003c/li\u003e\n\u003cli\u003eAre you confident that the new initiative is a good idea?\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eNot only does this create a lot of pressure on whoever’s the recipient of these questions, but the data also aren’t that useful. You get a single point of data on which to build additional decisions, and it’s either right or wrong—and when it’s wrong, things feel awful for everybody.\u003c/p\u003e\n\u003cp\u003eIt doesn’t have to be like this. We already have more information; we just need to ask better questions and be open to more nuanced answers. A straightforward shift is to rebuild our questions as “How confident are you that …?”:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eHow confident are you that you understand the issue?\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eBetter yet, make scaling any given answer explicit:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eWhat percentage confident are we as a team that we’ll ship X by Y? (Ranges are even better here, but that’s \u003ca href=\"https://dylanfitzgerald.net/blog/same-words-different-meanings/\"\u003ea different discussion\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003eOn a scale of 1 to 10, how confident are you that the new initiative is a good idea?\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThis lets people give a less pressured, more precise, and likely more accurate answer—so you get better data without having to pull teeth for it, and when things don’t go perfectly, it’s a lot less personal.\u003c/p\u003e\n",
  "content_md": "# Confidence as a continuum, not a boolean\n\nIn work as in everyday life, we tend to frame confidence as a black-and-white boolean. The key words used are “Are you confident that …?”, as in:\n\n-   Are you confident you understand the issue?\n-   Are we confident as a team that we’ll ship X by Y?\n-   Are you confident that the new initiative is a good idea?\n\nNot only does this create a lot of pressure on whoever’s the recipient of these questions, but the data also aren’t that useful. You get a single point of data on which to build additional decisions, and it’s either right or wrong—and when it’s wrong, things feel awful for everybody.\n\nIt doesn’t have to be like this. We already have more information; we just need to ask better questions and be open to more nuanced answers. A straightforward shift is to rebuild our questions as “How confident are you that …?”:\n\n-   How confident are you that you understand the issue?\n\nBetter yet, make scaling any given answer explicit:\n\n-   What percentage confident are we as a team that we’ll ship X by Y? (Ranges are even better here, but that’s [a different discussion]({{\u003c relref \"same-words-different-meanings\" \u003e}}))\n-   On a scale of 1 to 10, how confident are you that the new initiative is a good idea?\n\nThis lets people give a less pressured, more precise, and likely more accurate answer—so you get better data without having to pull teeth for it, and when things don’t go perfectly, it’s a lot less personal.",
  "created": "2024-09-20T03:17:00Z",
  "id": "6m59vch71k71ns7yccat26ht0z",
  "kind": "thread",
  "media": [],
  "origin": "https://dylanfitzgerald.net/blyg/",
  "page": "https://dylanfitzgerald.net/blog/confidence-as-a-continuum/",
  "transclusions": [],
  "updated": "2024-09-20T03:17:00Z",
  "version": 1
}