946 parcels on the county list. 175 worth a closer look.
We read California tax sales parcel by parcel — Riverside, San Bernardino and Santa Cruz so far — and publish every list in full, free. No sample, no paywall.
Every sale we cover, we read the whole county list, publish the reasoning parcel by parcel, and give it away free. The fastest way to catch the next one is email.
Built by one operator; every claim cites a source.
Or see the free sample of our work →
| Item | APN | City | Min bid | Flags |
|---|---|---|---|---|
| 2 | 102361007 | Corona | $32,103 | Mello-Roos possible — Mello-Roos possible: Tax-rate area suggests a special assessment. |
| 5 | 126090014 | Norco | $57,550 | Mello-Roos possible — Mello-Roos possible: Tax-rate area suggests a special assessment. |
| 3 | 103022040 | Corona | $23,424 | Mello-Roos possible — Mello-Roos possible: Tax-rate area suggests a special assessment. |
Real survivors — Riverside TC-223
One bad bid can cost more than the next ten good ones make back.
Tax-deed bidding is asymmetric: a good deal saves you some money; a paper lot, a speculator cluster, or an environmental surprise can be an unrecoverable five-figure mistake. The auction is public. The rules are codified in California Revenue & Taxation Code §§3691–3731. None of it is secret. What gets people is the traps — mixed in with the few real opportunities, row after row, with nothing marking which is which. Bid on a paper lot and you've tied up cash in something un-developable for a decade. Bid into a dead speculator's desert cluster and you've paid full price for lots the last owner already gave up on. Miss an environmental history on a commercial parcel and a $50,000 bid becomes a much larger loss. The list won't warn you.
Three layers. Only the middle one is AI.
We keep these jobs separate on purpose, because they get conflated everywhere in this space.
Rules cut the list
Six deterministic filters — string matches, threshold checks, a blocklist lookup — throw out the obvious junk. No model, no inference. Every dropped parcel records exactly one reason, so the math is auditable and you'd get the identical result running the list again tomorrow.
AI does the research
For what survives, a language model does the analyst work a human would otherwise do by hand: confirming the sale is still on and where the list actually lives this year, cross-referencing each parcel against the known risk patterns, and writing it up in plain English — with sources you can open yourself.
You decide
The model never says "bid." It produces "worth a closer look" or "probably skip, here's why." The decision is about your capital, your risk tolerance, and your local knowledge — and a model has none of those. It stays with you.
A real, closed sale — with the real numbers.
Riverside County's April 2026 sale (county sale no. TC-223) is over, and we have nothing to sell you against it — which is exactly why it's an honest case study. Here's what our pipeline did to the actual 946-parcel list.
One dumb filter did almost all the work. "Does this parcel have a street address?" removed 687 parcels by itself — about 89% of the entire reduction. Anyone selling you a neural network to find paper lots is selling you a neural network to read a field that already says SITUS ADDRESS: NONE.
946 parcels in. Six deterministic filters removed 771. One field check — does this parcel have a street address? — did 89% of the work.
| Filter | Parcels dropped |
|---|---|
| No situs address (paper lots) | 687 |
| Minimum bid out of range | 47 |
| Zip blacklist (low-liquidity) | 24 |
| Bulk-owner clusters | 13 |
| AVM-ratio (needs a valuation) | 0 |
| Surviving | 175 |
The 175 survivors had minimum bids from $10,106 to $195,409 — median $25,082, average $35,985. A tractable set a human can actually work through.
The shape of the 175
Every figure below is computed straight from the real survivor file — nothing rounded, nothing invented.
Right-skewed: a cluster near the $25,082 median, a tail to $195,409.
We publish the owner type, never the name.
| <$15k | 35 |
| $15–25k | 52 |
| $25–35k | 27 |
| $35–50k | 28 |
| $50–75k | 15 |
| $75k+ | 18 |
| Individual | 84 |
| Family group | 49 |
| Trust | 31 |
| LLC | 6 |
| Estate | 4 |
| Government | 1 |
California, one county sale at a time
We cover a few sales rigorously rather than every county thinly. Here's what we've published — and what's next.
Pins match the numbered list
- Santa Cruz County Published results June 2026 · 27 parcels as of Jun 11, 2026
- San Bernardino County Free desk analysis July 2026 · 1,614 parcels as of Jun 29, 2026
- Riverside County Free sample shortlist TC-223 · 946 → 175 as of Jun 3, 2026
- Los Angeles County Next sale · watching Oct 2026 · auction 2026B county list not yet published
We cover a few California counties rigorously — not all 58. This is the entire current list; nothing is hidden, and each entry shows when it was last checked.
Every sale we've read — the whole list, free.
No sample, no paywall. Each is a sortable, downloadable table you can work: a six-filter survivor shortlist, a pre-sale value screen, and a post-hoc results sheet. Pick one and start digging.
San Bernardino — July 2026 tax sale
- Worth a look
- 148 of 1,250
- Bid ÷ assessed
- 0.14–0.49
- Platform
- County MyTaxSale (Grant St.)
Santa Cruz — June 2026 tax sale
- Sold
- 14 of 18 at auction
- Winning bids
- $2,400–$115,556
- Platform
- Bid4Assets
Riverside County — TC-223
- Survivors
- 175 of 946
- Min-bid range
- $10,106–$195,409
- Platform
- Bid4Assets
What a survivor row looks like
Every parcel that clears the filters comes with its structural risk flags rendered as small badges, plus the reasoning behind them. These are illustrative example rows in the shortlist format — open the Riverside table to see the real survivors and try the filters yourself.
Format illustration — not live listings
| Min bid | Address pattern | Flags | Reasoning (short) |
|---|---|---|---|
| $24,500 | Residential street, mid-tract | none | Clears every structural filter. Next check: county GIS for lot size and a comp scan. |
| $31,200 | Numeric + "Pl" in a 55+ community | Condo / HOA risk — Condo / HOA risk: Likely condo — check HOA dues. | Title likely conveys free of pre-sale HOA back-dues, but you'd owe assessments forward. Pull the dues schedule and reserves before pricing. |
| $48,900 | Commercial corridor, owner LLC "…auto…" | Commercial risk — Commercial risk: Business-owned — possible commercial/environmental history. | Owner and address suggest commercial history. Run a Phase I — the standard environmental records check, about $3K — before bidding. Cheap insurance against a $50K mistake. |
| $19,750 | Standard tract, non-baseline tax code | Mello-Roos possible — Mello-Roos possible: Tax-rate area suggests a special assessment. | The tax-rate code suggests a possible extra special tax (Mello-Roos). Confirm on the secured tax bill; the code alone never gives you the dollar amount. |
Illustrative formatting, not live listings. Six filters cut the list; the badges mark the five trap patterns (plus likely occupancy) on what survives — all explained in the research library.
Fair questions
Is this just another tax-lien-course grifter?
Reasonable suspicion — the niche is full of them. Here's how we're different: the entire reduction is six rules you can read, not a black box; we publish a full worked example on a closed sale with the real per-filter numbers, including the parts that didn't work; and we tell you flatly what we don't do (no buy/sell calls, no title search, no data guarantee). There's no upsell ladder, no $2,000 "masterclass," and nothing to buy — the research is free and the reasoning is all shown. Read the methodology and judge for yourself.
Is the research really free?
Yes. Every research post and the full sample brief are free to read, and the shortlist CSV is a free download — nothing is paywalled and there's no account to create. We're not running an upsell funnel or a paid course. If we ever publish a paid brief for a specific sale it will be clearly marked and entirely optional; the free research stands on its own.
Do you tell me what to bid?
No. We produce "worth a closer look" or "probably skip, here's why," never "bid" and never a number. Your maximum bid depends on your capital, your risk tolerance, and diligence only you can do on a specific parcel. A model has none of that, so it doesn't get a vote on the decision.
When is the next sale?
Whenever a California county posts its next list — usually about 90 days before the auction. Subscribe to the newsletter and we'll email you the day a list drops, with the sale dates and a first read on what's in it.
Which counties do you cover?
California tax-deed counties, one sale at a time, when a county posts its list. Our published worked examples so far are Riverside, Santa Cruz, and San Bernardino; subscribers hear which sale we're reading next. We're not trying to cover every county in the state; we'd rather do a few sales rigorously.
What if parcels disappear between the brief and the auction?
They will — counties pull parcels as owners pay their taxes, sometimes right up to the sale. That churn is normal, and it's why the diligence checklist ends with a sale-morning re-check of the live auction list. The brief tells you what cleared the bar when the list was published; the final word on what's still biddable is always the auction platform.
Do I have to be in California?
No for most sales — many California counties run their auctions online; each brief names the platform for that sale. Out-of-area diligence is harder, though, and the physical-inspection step doesn't go away.
Is the data guaranteed?
No. The county PDF is the source of truth for the list, and we don't independently verify every owner name, tax-rate-area code, or bid amount against live assessor records. Automated valuations are a third-party model's estimates and are frequently wrong on atypical properties. Treat every number as a starting hypothesis to verify, not a fact.
Do I still need to do my own diligence?
Yes, always. The filtering gets you from "946 parcels, any of which could waste your money" to "175 that clear the structural bar, here's the reasoning on each." Pulling title, checking environmental records, and inspecting the property are still on you. The deed is yours; the eviction, if there is one, is still your problem.
Spend your diligence where it counts.
Get the next California sale the day its list drops — the shortlist, the reasoning, and the traps, free in your inbox.