The current Brick Alpha view
Know what deserves attention.
See what still needs proof.
A decision surface for collectors and long-horizon owners. Retirement is a catalyst—not a thesis—and every blank field is an evidence boundary, not an invitation to guess.
Research gate active
A complete 124-set economic forecast cohort and ten timing forecasts are public and frozen before outcomes. Factor score v0.1 is frozen, but zero sets are rated until verified sales and set-level decision evidence clear their gates.
Four permanent surfaces
From discovery to exit
Discover a future S-Tier candidate, wait for price or retirement to create a trigger, then monitor the thesis after purchase.
All monetary values USD · “—” means not yet supported
Rating-event discipline
What changed
Evidence before opinion
Latest research
Public research screener · canonical cohort
Screen the evidence.
See what is still missing.
Filter the frozen 25-set cohort by lifecycle, research coverage, and decision readiness. Research depth measures documentation—not investment potential—and every row opens the source-bound profile behind it.
Coverage is not a score · USD · U.S. lifecycle scope · “withheld” means a gate remains open
Canonical set comparison · one evidence cutoff
Compare evidence.
Keep the blanks honest.
Place up to four research profiles in one matrix. Identity, lifecycle, timing, price context, frozen forecasts, physical burden, and decision gates stay separate so documentation depth cannot masquerade as investment merit.
Evidence-aligned matrix
Same rows. Same meanings. Visible gaps.
Select a set to begin.The lab compares up to four canonical research profiles.
Safe interpretation
What this matrix can tell you
- Which official facts differ at the same research cutoff.
- Which sets have a frozen timing or economic forecast.
- Where current-value, package, or rating evidence remains absent.
- Which upstream evidence would unlock the next decision layer.
Do not infer
What this matrix cannot tell you
- Which set is the “best investment.”
- Whether a listed forecast is a verified current value.
- Whether a lifecycle label creates positive net return.
- Whether a missing field should be estimated from another set.
Prospective archive · July 2026 retirement cohort
124 forecasts.
Frozen before outcomes.
Compare Brick Alpha’s first complete post-retirement price cohort across 6, 12, and 24 months. Every record keeps its MSRP anchor, model reasons, wide operating interval, naive comparator, and original freeze timestamp.
Comparable research surface
Post-retirement forecast explorer
Choose a horizon, filter the frozen cohort, and select up to four sets for a side-by-side comparison.
124 forecasts12-month gross price forecast · USD unless multiple-only
Side-by-side
Comparison tray
Select up to four forecasts from the cohort.
Prospective accountability
Calibration scorecard
The schedule and provider method are fixed now. Scores appear only after each horizon matures and the preregistered BrickLink U.S.-to-U.S. new-sold unit-weighted item-price mean is captured.
No production outcomes are due yet.
Missing outcomes will remain missing; sets will not be removed because they are hard to price, delisted, or contradicted the model.
Model v0.1
Simple enough to falsify.
The rules model begins with an 8% annual gross appreciation prior, then applies small bounded adjustments for theme, wanted-to-owned ratio, community rating, MSRP band, shelf life, and price per piece. The final annual prior is clipped before compounding.
Wide operating intervals
Uncertainty grows with time.
The low and high cases are scenario bands, not statistical confidence intervals. They deliberately widen from 6 to 24 months because thin sale depth and heterogeneous condition remain unmodeled.
Naive comparator
Complexity has to earn its place.
Every set-specific forecast is scored against the same 1.04×, 1.08×, and 1.18× MSRP-multiple baseline. A later model advances only if prospective outcomes show that it adds value.
Canonical reference · official status semantics
A status is evidence.
It is not a date.
Decode LEGO.com lifecycle and availability labels without turning stock noise into retirement certainty. The result explains what the observation means, what it does not prove, and how Brick Alpha’s timing model treats it.
An explicit lifecycle signal.
LEGO indicates that the set is expected to leave its current assortment soon. The label does not reveal an exact date, future resale price, or whether anyone should buy.
No. It is an availability state.
LEGO says a temporarily unavailable product is expected to return. Preserve it as dated stock evidence unless an explicit retirement label also appears.
Signal decoder · method 0.1
Interpret one dated observation
Choose the labels exactly as observed. Optional set context appears only in the export.
Examples
Availability constraint only
Temporary stockoutRestock expected—not a retirement call.
- Official meaning
- Not proven
- Current model treatment
- Decision meaning
- Next observation
Official label ladder
Availability and lifecycle are separate claims.
These interpretations paraphrase LEGO’s U.S. help page. They should be captured with a date because a product page can change later.
Private decision workspace
Your holdings.
Joined to the evidence.
Track acquisition lots, cost basis, intent, and sell discipline. Brick Alpha adds canonical identity, retirement exposure, and research readiness without turning indicative prices into portfolio marks.
Concentration
Cost basis by theme
Accountability
A holding is not a thesis.
Investment lots are strongest when confidence, intended hold, thesis, and a sell trigger are recorded. Retrospective memos remain useful, but never become pre-purchase forecasts.
Private cost-basis underwriter · method 0.1
Cost basis → required exit.
Choose an owned lot, a narrow source-backed fee path, and a return hurdle. Brick Alpha solves the gross price required to break even and to clear the hurdle without supplying a market value or forecast.
Waiting for a private position
Private scenarioSelect an owned position
The calculation runs only in this signed-in browser session.
No market-price input.Required gross prices are algebraic thresholds, not valuations, forecasts, recommendations, or evidence that a buyer exists.
Each selected unit is modeled as a separate one-set order. Income tax, labor, sale probability, time to sale, fraud, and portfolio opportunity cost are excluded. Private cost basis never becomes canonical market evidence.
Owned inventory
Position ledger
0 positionsNo current-value or unrealized-return claim without a verified mark
Private monitoring workflow · method 0.1
Know what changed.
Know what needs review.
Track a sealed set across thesis, condition, official lifecycle, evidence freshness, marketplace activity, and review cadence. Every lane keeps its own meaning; none becomes an automatic buy or sell instruction.
Attention before activity
Review queue
Operating thesis
Make LEGO investing legible.
Publish the evidence, preserve every forecast, and earn trust one resolved uncertainty at a time.
North-star metric
Verified decisions influenced
Readers who use a Brick Alpha artifact to make or avoid a transaction.
Focus now
Highest-leverage work
P0 · Market evidence
9.6
Clear the first production rating
Acquire one transaction-complete 12-sale sample, preserve the buyer-paid cutoff, and let the frozen score method decide whether the set clears.
Next 14 days
Editorial cadence
Measurement
KPI tree
| Signal | Why it matters | Now | 90-day target | Cadence |
|---|
Internal discovery benchmark · fixed corpus
Can Brick Alpha answer the questions that matter?
Twenty-five stable decisions are scored separately for responsible answer readiness and independent discovery. Partial methods receive partial credit; missing evidence stays a gap.
The free Brick Alpha field guide
Keep the return.
Question the story.
Practical research for underwriting sealed LEGO after fees, freight, time, liquidity, and storage. No hot lists. No mystery scores. Every framework ends in a decision.
Where the edge can compound
Content alpha map
Evergreen collection
Eight decisions worth getting right
The free / paid contract
The reasoning stays public.
Free readers get the methods, calculators, samples, delayed archives, and scorecards required to audit a claim.
Free forever
Frameworks · calculators · methodology · representative data · forecast accountability
Paid can add speed
Current watchlists · alerts · full datasets · saved portfolios · operational workflows
Evidence before opinion
Research pipeline
Every item moves through a defined evidence gate. No unverified number ships.
ReadyWorkingQueued
Data coverage
Source ladder
Required challenge
Every thesis gets a kill case.
Before publication, write the strongest reason the conclusion may be wrong and name the observation that would falsify it.
Publication gateThesis + counter-thesis + falsifier + timestamp
Trust before throughput
Publishing system
Nothing publishes because it looks finished. Every item must clear a content-bound, independent review and deterministic release gate.
Independent passes
Six roles, one release boundary
Risk-calibrated workflow
Same truth standard. Proportionate depth.
Editorial queue
Nothing is “ready” until the gate says so
Completed-sale evidence · local workbench 0.1
Turn raw sales into an inspectable range.
Paste or import completed-sale CSV or JSON. Brick Alpha checks comparability row by row, calculates the item-plus-shipping distribution, and keeps rejected records visible instead of silently cleaning the sample.
The demo proves the method only. It never becomes market evidence.
Normalized export contains only accepted candidate fields. It does not add a license, evidence hash, approval, or canonical status.
Original artifact · prototype 0.1
Net return lab
Model the investment you actually keep—not the headline price appreciation. All inputs are hypothetical and remain in this browser.
Terminal-equivalent annualized return
—After modeled fees, shipping, packaging, damage allowance, and storage. Carry costs are end-of-hold-equivalent deductions; this is not a cash-flow-timed IRR. Income taxes and labor are excluded.
Sensitivity
What drives the result?
Decision aid only. This calculator does not estimate the future sale price or probability of sale. Validate channel fees and transaction inputs for your account and date.
Public evidence record · v0.1
Net Return Lab method + validation
This is a scenario calculator, not a price, sale-probability, or time-to-sale forecast.
Core equations
landed = purchase × (1 + tax rate) + fixed inboundmodeled proceeds = sale − variable exit − fixed exit − carryannualized return = (modeled proceeds ÷ landed)^(12 ÷ months) − 1hurdle exit = [landed × hurdle factor + fixed exit + carry] ÷ (1 − variable exit rate)IncludedEntered purchase and sale prices, acquisition tax, fixed inbound, hold-period scenario, hurdle, fee and fixed exit costs, damage reserve, stated storage, and package volume for the efficiency output.
ExcludedIncome tax, labor, unmodeled opportunity cost, portfolio effects, and forecasts or distributions for sale price, sale probability, or time to sale. Entered hold months are a scenario input, not a timing forecast.
Validation protocolThe machine gate re-executes the implementation against stored ledger outputs and binds the formula suite, UI, public evidence labels, and method record to reviewed hashes. A separate quantitative reviewer independently replicates the formulas and assumptions.
Limitations: every output is conditional on entered assumptions. Marketplace fee bases vary; storage may be cash or consistently imputed; an expected-value reserve may not become a realized cash loss.
Scenario compare · method 0.1
Compare the tradeoffs—not one magic score.
Vary purchase, exit, hold, and optional package volume across three hypothetical investments. Every column shares the current Return Lab cost stack so percentage return, absolute profit, and capital at risk stay comparable.
Hypothetical decision aid only. Entered exit prices and hold periods are not Brick Alpha forecasts. Inputs persist only in this browser. Income tax, labor, sale probability, buyer reach, and portfolio concentration are excluded.
Exit channel lab · method 0.1
Same sale. Different friction.
Compare three narrow, source-backed fee paths under one scenario. The ranking is modeled proceeds—not a claim about buyer reach, speed, or the universally “best” channel.
Original artifact · incremental decision 0.1
Hold—or release the capital?
Compare a future hold scenario with selling now and compounding the cash elsewhere. The relevant starting point is realizable net proceeds today—not what you originally paid.
Complete every required field.
Evidence interpretation
Exploratory conditional
User scenario
A hypothetical input with no independent market-evidence claim.
User scenario
A hypothetical input with no Brick Alpha forecast claim.
Asking price is not value. An active listing is neither a completed sale nor proof that the entered amount can be realized.
Fixed sensitivity
Does the signal survive?
Original artifact · liquidity capacity 0.1
Inventory is not liquidity.
Compare completed-sale flow, visible asking supply, and your position size on one dated, exact-set marketplace scope. Save up to four like-for-like analyses without collapsing them into a mystery score.
Complete a valid exact-set scope.
Evidence interpretation
Synthetic scenario
A hypothetical input used to understand the method; it makes no market claim.
BrickLink90 daysSeller concentration entered
A listing is asking supply. It is not a completed sale, realizable price, or stable inventory unit.
Fixed crowding stresses
What if absorption weakens?
Public method · v0.1
Mechanical measures, visible denominators.
Every result is reproducible from the entered stock, flow, window, and position. Missing flow suppresses ratio outputs instead of producing infinity.
monthly flow = sales ÷ window days × 30.4375months active supply = listings ÷ monthly flowposition market days = units ÷ (sales ÷ window days)buyer-paid IQR = (Q75 − Q25) ÷ medianBrowser-local comparison
Compare capacity on equal terms.
Save two to four analyses. Incomparable channels, cutoffs, windows, or evidence classes are displayed but never ranked.
0 saved
Original artifact · analyst discipline 0.1
Write the thesis.
Write what kills it.
Turn a LEGO idea into a falsifiable, timestamped decision record. The structure can be complete even when the conclusion is wrong—and that is what makes later learning possible.
Brick Alpha doctrine · v0.3
The underwriting standard
A repeatable decision process designed to survive hindsight, transaction friction, and uncomfortable outcomes.
Primary outcome
Net annualized return
((net proceeds ÷ landed cost) ^ (365 ÷ days held)) − 1
“Net” means after platform fees, payment fees, shipping subsidy, packaging, damage allowance, and directly attributable storage.
Confidence policy
Precision must be earned.
Low
Incomplete comps or unstable inputs. Directional only.
Medium
Multiple sources and defensible comparables. Wide range.
High
Validated sales, stable methodology, demonstrated calibration.
Proportional evidence · no paralysis
Good baseline now. Exact measurement when the decision needs it.
Core ranking
Identity, realized-price evidence, liquidity, exit costs, and frozen forecasts drive the core view. Missing box dimensions do not block it.
Estimated storage
A source-bound sealed-box value can support a clearly labeled baseline with a ±15% volume sensitivity range.
Exact storage + shipping
Exact profit per cubic foot, capacity planning, and dimensional-shipping conclusions require a verified U.S. retail package.
The range is an operating assumption, not a confidence interval. If no credible package baseline exists, only the storage-sensitive output stays blank.
Price evidence · useful before perfect
Market context can appear before a rating.
Tier 1Transaction-complete
Stable sale IDs, explicit shipping, exact condition, and U.S. scope can support a verified current-value range.
Tier 2Triangulated secondary
Two independent aggregates can support a labeled indicative range with wide sensitivity—not a target or action.
Tier 3Offer reference
A retailer quote supports entry-price context only while it is fresh and direct. Affiliate or stale snapshots remain references.
A rating still requires a prospective exit-value model and transaction costs. Better baseline data narrows uncertainty; it does not erase it.
Comparable decisions · factor score v0.1
A score exists only when economics—not enthusiasm—support it.
Published scores
0 / 24
Method frozen · live ratings withheld
35%Expected net return
Probability-weighted, after-friction annualized return.
20%Margin of safety
Frozen acquisition price versus verified buyer-paid value.
20%Liquidity
Transaction-complete sale depth and interquartile spread.
15%Downside resilience
Net return under the preregistered downside scenario.
10%Retirement catalyst
Frozen 90-day retirement probability, never a value substitute.
Every score carries a wide operational uncertainty band and a visible factor breakdown. Missing dimensions do not block the core score; missing verified sales or economic forecasts do. A score is still not a buy or sell action.
Portfolio accountability · split boundary
The method is public. Individual holdings stay private.
Public research
Definitions, decision gates, score methods, benchmark rules, correction history, and reviewed aggregate calibration.
Internal portfolio record
Receipts, cost basis, position size, condition, unpublished theses, raw sale statements, and position-specific sell triggers.
Existing acquisitions without a timestamped pre-purchase memo may receive a retrospective accountability record, but they never count as forecasts made before the buy.
Corrections · append-only
Errors are corrected visibly, never silently.
Material corrections preserve the original statement, the corrected statement, the reason, date, and effect on the conclusion. A correction changes the content hash and requires fresh specialist review. Frozen forecasts are superseded, never rewritten.