Between 28 January and 15 February 2026, this site published three articles about the Bitcoin price. One said it was heading to US$250,000. One said it was ranging between US$75,000 and US$95,000. One said it had fallen below US$61,000 and entered a bear market.
Eighteen days separated the first from the last. They cannot all have been right, and in fact none of them was. All three have been withdrawn.
This is the replacement: what Bitcoin actually did, and — more usefully — what went wrong in how we covered it.
What the Price Actually Did
As at 10 September 2026, Bitcoin was trading at US$109,693.
The path there over the preceding six months was not a trend in either direction. Price spent the autumn months in the low six figures, reached its highest point of that window around May, then fell substantially through June to a floor in the high tens of thousands. It moved sideways in a range through July and August before rising sharply in the first days of September to return to around US$110,000.
Those are shapes rather than precise levels, and deliberately so. We are describing a chart, and quoting levels to the dollar off a chart would be inventing precision we do not have. The one figure above stated exactly is the one we can source exactly.
What That Means for the Three Withdrawn Articles
The $250,000 article was not merely early. Seven months later the price was less than half that, having gone down before it went up.
The $75,000-$95,000 article described a range the price had already left and would leave repeatedly in both directions. It was also internally incoherent, citing figures up to $300,000 in a body sitting under a headline promising a band topping out at $95,000.
The sub-$61,000 bear market article called a direction the market did not take. Whatever the price did in February, the twelve-month picture is not one of sustained decline.
The common failure is not that three forecasts missed. It is that we published three incompatible accounts in three weeks and left all of them standing.
The Actual Problem Was Not the Forecasting
It would be comfortable to conclude that price prediction is hard. That is true and it is not the lesson.
The lesson is that a publication which prints three contradictory versions of the same story within a month is not making three predictions. It is demonstrating that nothing is checking the coverage against itself. Any reader who found two of those articles learned something more damaging about us than any individual wrong number could convey.
Worse, they sat unchanged for seven months. Nobody revisited them when the price moved. A price article with no review date is a claim about the present that quietly becomes a claim about the past without ever announcing the change.
What We Have Changed
Three things, and they are structural rather than resolutions.
The crypto section does not publish price calls. Our crypto coverage commits to regulation, custody, market structure and the practical questions a business actually faces — whether to accept crypto payments, how holdings are treated, what the reporting obligations are. That commitment was already published on the section page while these three articles contradicted it. Now the articles are gone and the commitment stands.
Figures that decay carry a review date. Where we state a rate, a threshold or a level, it is dated and pointed at a primary source. This applies across the site, not only here — the same discipline governs the Australian tax figures used in our calculators, after an earlier error in that area.
Corrections are visible. The three withdrawn articles still exist at their original URLs, carrying a notice explaining what they claimed and why it was wrong. They are removed from search but not from the record. A correction that disappears is not a correction.
If You Want to Look at Price Anyway
Plenty of readers do, and pretending otherwise would be precious. What we can offer is a chart that is honest about what it is.
Our Bitcoin rainbow chart draws the well-known power-law regression with its conventional coloured bands, and spends most of the page explaining why the model cannot tell you what happens next — that it is fitted to prices which already occurred, that the band boundaries are chosen by hand rather than derived from data, and that different published versions of the same chart disagree about which band a given price sits in.
That is the honest version of price content: show the artefact, explain its limits, and decline to convert it into a forecast.
On the figure above: US$109,693 is the price as at 10 September 2026 and is not updated automatically. It will be out of date by the time you read this, which is precisely the property that made the withdrawn articles a problem. For a live view see our Bitcoin rainbow chart, which draws from public price data in your browser.
Frequently Asked Questions
Why were the three Bitcoin articles withdrawn?
Because they made incompatible price claims within eighteen days of each other, all three were wrong, and all three remained live and unrevised for seven months.
What was Bitcoin actually worth?
US$109,693 as at 10 September 2026. The preceding six months saw a high around May, a substantial fall through June, a sideways range across July and August, and a sharp rise in early September.
Are the old articles deleted?
No. They remain at their original URLs with a notice explaining what they claimed and why it was wrong. They are set to noindex so they no longer appear in search, but the correction is visible rather than hidden.
Does BizziKit publish Bitcoin price predictions?
No. The crypto section covers regulation, custody, market structure and the practical questions a business faces. The withdrawn articles contradicted that stated position.