Price Evaluator

Data source

Estimates use a market snapshot from Doginal Dogs listings and recent sales (refreshed on a timer, typically every ~12 hours). Live prints can move between refreshes, so numbers are a model output on that snapshot, not a live order book.

Step 1 — Per-trait blend (used inside the anchor)

For each trait value we combine listing floor and sale comps into one blended anchor price: mostly floor-weighted when both exist, with sale depth nudging the weight; top sales can only lift the blend so far unless there is enough trade history. It is not simply “max of floor and last sale.”

Step 2 — Compound base (max of the rarest layers)

• We take non-synthetic traits with a known count of ≤ 2,500 out of 10,000, sort from rarest upward, and keep the top three.
• We compute the blended price above for each of those layers.
• Base price = the maximum of those blends, so a strong market on one layer (e.g. Shiba) is not ignored just because another layer is numerically rarer (e.g. Poncho).
• Thin-market nudge: If the layer that wins that max has < 100 supply and the snapshot shows no listings and no sales for that value, base is multiplied by 1.08. If there is still no numeric base, we may seed from the collection floor × 1.08 when the overall rarest layer qualifies the same way.

Step 3 — This dog’s own sale

If this dog appears in recent sales in the snapshot, that sale price is merged into base with max(blend base, own sale) before bonuses, so realized demand for this inscription lifts the anchor.

Step 4 — Trait bonus (stacking beyond the anchor)

Traits that are not in the three-layer compound anchor, have count ≤ 500, and show a top sale or floor above the current base add 20% of the difference, each. (So the anchor layers are not double-counted here.)

Step 5 — Core estimate

(Base + trait bonus) × rarity rank multiplier × trending multiplier. Rarity uses overall rank bands: top 1% → 1.5×, top 5% → 1.25×, top 15% → 1.1×, top 25% → 1.05×. Trending uses the strongest admin-configured trait demand multiplier that applies to this dog.

If there is no usable base at all, we fall back to the average of available per-trait listing floors, then apply rarity and trending the same way.

Step 6 — Floors and scarcity bump

• Collection floor: the estimate is not left below the collection listing floor.
• Many thin trait books: if three or more trait values have zero listings and zero sales in the snapshot, the estimate is multiplied by 1.15 (illiquidity / thin-tape adjustment).

Step 7 — Optional multipliers (when they apply)

• True 1/1 traits (or configured display overrides): 3× on the running estimate.
• Color harmony: when the same color theme shows across two or more layers, an admin-set multiplier applies (Black Classic dogs skip double-counting with a black color match).
• Minimal / aesthetic tiers: e.g. ultra-clean, minimal, visor-minimal, black classic (configurable in admin).
• Custom combos: admin-defined trait combinations can multiply the result (multiple matches multiply together).
• Angel / vanity dog numbers: quad or triple repeaters and listed trend numbers (configurable caps).

Step 8 — Community floors and lore

• Refused community offer floor: when a minimum refused offer is recorded (USD converted live to DOGE), the final estimate is not left below that floor.
• Community lore: when a lore entry carries a signed price multiplier, it adjusts the estimate after the steps above.

Listing vs estimate

If the dog is listed, we may show ask vs estimate (undervalued / fair / overpriced). That verdict is for display; it does not change the model estimate.

Sale price attribution (per trait, for comps)

When building each trait’s comp stats, attributed top sales are capped so one grail sale does not paint an entire common trait:

• Ultra-rare (count ≤ 100, plus Diamond & Shiny): full attributed sale.
• Moderately rare (101–500): capped at about 2× the trait’s median sale or the listing floor, whichever helps more.
• Common (> 500): capped toward the trait’s median sale.

What this still can’t measure

Sentiment, hype, personal taste, and “grail” narratives move prices in ways no snapshot model fully captures. Use this as one structured view of the tape, not a price guarantee.

⚠ This tool provides a data-driven estimate, not financial advice. Actual sales depend on buyers, timing, and conditions the model does not see. Do your own research before buying or selling.