06-reference

commoncog tdgs accelerate maths sensemaking

2026-11-09·reference·source: Commoncog·by Cedric Chin (Commoncog)
tacit-knowledgeexpertise-acquisitiontactical-decision-gamessensemakingmathematics-educationsponsored-content

No public post URL exists for this piece — only click-tracked workshop-signup and citation links appear in the email, no "view online" permalink. This is an unplanned sixth email in the active Commoncog x Shadowbox sequence: the Nov 8 note ([[2026-11-08-commoncog-making-of-tactical-decision-games]]) was billed as the series finale, but Shadowbox extended the workshop application deadline from Nov 10 to Nov 13 to market to its own audience, and Cedric used the extra window to send "one final email about TDGs" — adapted from a Twitter thread, not a fresh Commoncog essay. Sixth in the sequence after [[2026-08-25-commoncog-why-tactical-decision-games-work]] (original, unsponsored), [[2026-11-01-commoncog-making-of-common-sense-tdg]], [[2026-11-02-commoncog-tdgs-accelerated-osmosis-expertise]], [[2026-11-03-commoncog-making-of-hindsight]], [[2026-11-07-commoncog-tdgs-in-the-age-of-ai]], and [[2026-11-08-commoncog-making-of-tactical-decision-games]]. Solely bylined and voiced by Cedric Chin — no Jared Peterson or Shadowbox-employee co-authorship in the body — so classified thought-leadership, matching the 11-07 installment rather than the guest-post installments.

Why this is in the vault

Extends the tracked sensemaking/TDG thread with a specific, falsifiable mechanism — Bessis's claim that expert mathematicians build deliberate mental representations ("objects") rather than manipulate symbols, and that this (not more problem-set repetition) is what Fields Medallists report as the actual lever — which gives a concrete design question to test against any RDCO or portfolio product that claims to teach a skill via repetition.

The core argument

Cedric grounds "sensemaking" in the Data-Frame Theory of Sensemaking (Klein et al.), which holds that people only interpret information through a "frame" — a schema, script, or mental representation already held in the head; an appraiser, firefighter, and carpenter looking at the same house see different things. Applied to maths: the question is what mental representations professional mathematicians hold that novices don't. Cedric cites David Bessis's 2024 book Mathematica — endorsed by seven Fields Medallists — which argues expert mathematicians don't see symbols on a page, they see constructed mental objects, built through deliberate daydreaming over time. Worked example: given equations for a circle and a line, a mathematician doesn't compute an intersection — they picture the line sweeping across the circle and intuit "at most two points" because they've already built that visual object. The claim compounds: each constructed representation becomes a building block for higher-level ones, and the skill of inventing new representations itself improves with practice. Bessis reports asking his own son from an early age to describe his dreams, specifically to start a vocabulary for discussing the representations forming in the boy's head before formal maths instruction begins. Cedric's conclusion: conventional maths education ("grind more problem sets") trains computation, which hits a ceiling; TDGs are offered as one domain-general mechanism for training the representation-construction skill itself, which is why some Commoncog members are adapting TDGs to teach maths, fundraising, parenting, and data skills, not just military or business judgment. A postscript extends the same frame to reading financial statements: Jay Pritzker and Warren Buffett, Cedric argues, aren't reading line items faster than everyone else — they're reading into a pre-built mental representation of a business, the same way Bessis reads a visual object instead of an equation.

Mapping against Ray Data Co

The concrete, specific connection: Scribble Works' entire product premise is paper worksheets as "an AI tutor without the screen" (project_scribble_works_product_principle_parents_site_kids_paper — parents on the site, kids on paper, no on-screen play per SW-R18). Bessis's distinction cuts directly at that premise's unstated assumption: a worksheet is only a good tutor if it trains representation-construction, not if it's a computation-drill sheet dressed up as one. This email supplies a concrete test Scribble Works doesn't currently have a documented answer to — does any given printout ask "what mental object would make this class of problem trivial" (Bessis's bar), or does it just vary the numbers on the same procedure (the "grind psets" default Cedric says most maths education silently falls into)? That's not a problem with Scribble Works specifically — it's the generic failure mode this piece names for any worksheet-shaped product — but it's the first time this tracked sensemaking thread has handed a design question that lands on a live RDCO product rather than on agent-evaluation methodology (the angle the 11-07 and 11-08 notes already took). No answer is implied here; this is an open design question worth testing against Scribble Works' actual math-adjacent content, not a verdict on it.

⚠️ Sponsorship

sponsored: true / sponsor_entity: Shadowbox. Cedric states outright: "I have no financial interest in this workshop. I did pitch ShadowBox the idea... I did have one requirement — which was that Commoncog members get first dibs. And then I helped them do the marketing for the launch." So the relationship is: idea originator, marketing collaborator, and advocate for the instructor (John Schmitt), not an equity or revenue-share stake as disclosed. The email's stated occasion is a workshop-deadline extension (Nov 10 → Nov 13, "only five seats left"), with the signup link repeated at open and close — this is a sales email wearing an essay. The Bessis/maths content is independently sourced (a published, Fields-Medallist-endorsed book, a named Commoncog members-only Q&A) and reads as credible on its own terms, but its selection and timing — released during the extended application window, closing with "here's the link one final time" — serve the sale, consistent with every prior installment in this sequence.

Related