Carterville has 48 active listings this morning, the same number it had seven days ago, though not quite the same ones: 3 arrived and 3 left. Of the 45 that were on both readings, 15 changed their asking price.
We know that because we compared each listing with its own figure from a week ago. If we had instead asked the data which listings had changed price — which is the obvious way to do it, and the way we have done it in other posts — we would have been told 3.
What moved, and what the data said about it
| Listing | A week ago | This morning | Change | Price-change marker |
|---|---|---|---|---|
| 127 Weisbrook Ln | $170,000 | $160,000 | -$10,000 | flagged |
| 616 E Grand Ave | $149,000 | $139,000 | -$10,000 | flagged |
| 5934 Chandler Dr | $334,900 | $329,900 | -$5,000 | flagged |
| Lot 101 Elliott Ct | $34,000 | $36,000 | $2,000 | not flagged |
| Lot 112 Pendleton Ct | $34,000 | $36,000 | $2,000 | not flagged |
| Lot 102 Elliott Ct | $34,000 | $36,000 | $2,000 | not flagged |
| Lot 113 Pendleton Ct | $34,000 | $36,000 | $2,000 | not flagged |
| Lot 115 Pendleton Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 104 Elliott Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 109 Pendleton Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 116 Pendleton Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 110 Pendleton Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 111 Pendleton Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 103 Elliott Ct | $34,000 | $38,000 | $4,000 | not flagged |
| Lot 114 Pendleton Ct | $34,000 | $38,000 | $4,000 | not flagged |
Every Carterville listing whose asking price differs between our two readings.
The pattern is not subtle. Every price that came down was flagged. Every price that went up was not. And it is not that the marker is unreliable in general: each of the 3 flagged listings carries a previous price exactly equal to the figure we recorded a week earlier, so when it fires it is accurate. All 19 listings carrying the marker in Carterville today — these 3 among them, and 16 more whose price is the same on both our readings — are reductions.
The 12 it missed are all vacant lots on two streets
They are vacant lots on two adjoining streets. A week ago every one of them was listed at $34,000. This morning 8 of them read $38,000 and 4 read $36,000 — increases of $4,000 and $2,000. None of them carries a previous price. None is labelled as having changed price. As far as the data is concerned, nothing happened.
This is not a reading error on our part, and the way we know is that it is in four separate readings. These lots appear in pulls taken on November 7, November 8 and November 9 for other posts in this series, and all four agree on the price of every single lot. In none of them does any lot carry a price-change marker. Every reading from November 7 onward — four of them across six days — reads the new prices, and not one says the price changed.
There is one more trace of whatever happened. A week ago each of the 12 carried its own last-updated timestamp, 12 different values a few seconds apart. This morning all 12 carry a single identical timestamp, and it is dated earlier than the ones it replaced. A timestamp cannot get younger. Something rewrote those records, and the rewrite did not go through whatever normally sets the price-change marker.
A correction to what we published a week ago
Our Carterville post of November 6 said those lots were all asking $34,000, and built a paragraph and a figure on it. That was accurate on November 6 — both of that morning’s independent queries returned $34,000 for all 12 — and it stopped being accurate the following day.
So that post is stale rather than wrong, and the repair is a dated note on it rather than a silent edit. We would rather say that here than have somebody find the old figure and work out the rest for themselves.
And a narrower claim than we made at the weekend
Two recent posts in this series leaned on the same marker across all four counties. One reported that every one of 200 price moves was a reduction; the next reported 201 and said that no listing in the four counties had raised its price.
Every row the marker contained really was a reduction, so those sentences describe the data accurately. What this week shows is that the marker does not contain every price move. The honest version of that claim is that no listing the data flags had raised its price — which is a statement about a field, not about the market. Neither of those posts is published yet, which is lucky, and both will carry the narrower wording. We are writing it down here because the next person to use that field should know.
Three listings left, and the data says what became of one
Three listings dropped out of the active search this week. Rather than guess, each was looked for under every status this data exposes — Pending, Active Under Contract, Closed, Canceled, Expired and Withdrawn — by record identifier first and by address second.
635 Anderson St came back under the closed status, marked sold. It was asking $274,000 a week ago and the closed record reads $265,000, $9,000 lower. That is the data’s own status field, not our inference from a listing disappearing. The other two listings were not found under any status at all, so what became of them is unknown and we will not dress that up. The search was not empty, which matters: it returned 70 records across the six statuses.
One closed sale is one closed sale. No rate, share or direction is calculated from it here.
The rest of the week
3 listings arrived — at $145,000, $60,000, $40,000 — and the town’s 48 active listings now run from $36,000 to $539,900, with a median of $159,500. By kind: 22 residential, 21 land, 4 commercial sale, 1 residential income.
What this does not tell you
It does not tell you that 12 sellers raised their prices. The asking prices in this data moved up; whether somebody raised a price or corrected one that had been entered wrongly, the records do not say, and the point of this post holds either way — the change happened and the marker missed it. It does not say why any price moved, up or down. It does not tell you what any of these properties is worth, and an asking price never did. It does not tell you whose listings these are: the records carry no office field. And it describes two mornings seven days apart — a price can change again this afternoon, and apparently without saying so.
Common questions
Did Carterville asking prices go up or down this week?
Both, and the split is the interesting part. Of the 45 listings on both readings, 3 came down — by $5,000, $10,000 and $10,000 — and 12 went up, all of them vacant lots on two adjoining streets, by $2,000 or $4,000. The town’s median asking price across all 48 listings is $159,500.
What is wrong with the price-change field?
Nothing is wrong with it when it fires: each of the 3 listings it marked carries a previous price exactly equal to the figure we recorded a week earlier. The problem is what it does not do. All 19 listings carrying the marker in Carterville today are reductions, those 3 included, and the 12 increases we measured carry no marker at all — no previous price, no price-change label, nothing. On this week’s evidence the field is a record of price cuts rather than of price changes.
You said last weekend that no listing had raised its price. Was that wrong?
It was true of the field and we should have said so. Two posts, on 11/08 and 11/09, reported that every one of 200 and then 201 price moves in the four-county data was a reduction. Every row the field contained was indeed a reduction. What this week shows is that the field does not contain every price move: here are 12 that it does not hold. The honest version of that earlier sentence is that no listing the field flags had raised its price, which is a narrower thing.
Your post of November 6 said those lots were all $34,000. Is that wrong now?
It was right on November 6 and it is out of date now, which is why this post says so rather than quietly editing it. Both of that day’s independent queries read all 12 at $34,000. Every reading since — four of them now — reads 8 at $38,000 and 4 at $36,000. The figure in that post describes the day it was written and no longer describes today.
What happened to the listings that disappeared from the search?
Three listings left the active list. Each was looked for under every status this data exposes — Pending, Active Under Contract, Closed, Canceled, Expired and Withdrawn — by record first and by address second. 635 Anderson St was found under the closed status, marked sold, at $265,000 against the $274,000 it was asking a week ago. The other two listings were not found under any status, so what became of them is simply unknown, and a listing leaving a search is not a sale. The search was not empty: it returned 70 records in all.
💬 Greg Holthaus on checking the number yourself
“The lesson here is not about Carterville, it is about where a number comes from. A field that says ‘price changed’ is somebody else’s summary of the data, and this week it summarised 15 moves as 3. Comparing two readings against each other is slower and it is the only way we would have seen the other 12. If you are being shown a list of price reductions by anybody, including us, the fair question is whether that list is what moved or what got flagged. And if a figure we published is out of date, you should hear it from us first.”
Want the current asking price on something specific in Carterville? Call or text 618.925.8654 and we will read it off this morning’s data rather than last week’s.
Where these numbers come from
Two readings of local MLS listing data for Carterville, seven days apart — November 6 and November 13, 48 active listings each time. The search was not retyped between the two: it is loaded from the earlier reading’s own record of what it asked and sent back unchanged, and the new reading is checked to have echoed it exactly. That check exists because a retyped search once came back with zero results and no error at all. A second, cosmetically different form of the same search was run again this morning and returned the identical 48 records.
Listings were matched one to one on a stable record identifier, never on the address text, and each listing’s asking price was compared with its own figure from a week ago. That is the whole method behind this post: it does not ask the data what changed, it compares the data with itself. Asking the data what changed is precisely what would have missed 12 of the 15 moves.
The 12 increases were checked against three further readings taken on November 7, November 8 and November 9 for other posts in this series. All four agree on every price, and in none of them does any of the 12 carry a price-change marker. A one-off reading error would not survive that.
About these figures: everything above is local MLS listing data read on the mornings of November 6 and November 13, 2026. A change in an asking price is not a sale and not a valuation, and nothing here is described as sold except the one record the data itself marks sold. This page makes no claim about why any price moved, in either direction, and none about who owns or lists any of these properties — the records we read carry no office field, so that cannot be read from them at all. No days-on-market figure appears because this data does not carry one and obtaining one would mean opening a detail record per listing, each full of agent contact fields we do not read. No individual listing is named except the 15 whose prices moved and the one closed record this post is about, and no agent, no brokerage and no photograph from the data appears anywhere. Nothing on this page is a loan offer, a rate quote or any statement about financing: no interest rate, no monthly payment and no lender appears in it. It names no broker other than Greg, and no other brokerage. Figures describe the records as they stood on the mornings they were read; a listing can change its price or leave the market the same afternoon.
