Our editorial policy says corrections take priority over publishing new guides. This page is where you can check that, because a promise to correct things is worth nothing without a list.
Every entry below is a change we made to a live page after we had already published it. Each one names the figure that was wrong, the figure the issuer actually publishes, and how the mistake was caught. We have not hidden the embarrassing ones, and one entry is a correction we made and then had to undo.
Eight times an issuer published something we said it did not
This is our most common failure by a wide margin, and it has one shape: we read the page that looked like the right page, found nothing, and wrote that the issuer publishes nothing. The figure was somewhere else on the same site.
| Date | What we had written | What the issuer publishes, and where |
|---|---|---|
| 5 September 2026 | That Port Everglades attributes a cruise line to only two of its eight terminals, and that the rest are only signposted inside the gate | Seven of the eight, in the interactive parking map linked from its own Cruise menu. The records sit in an embedded data file, so a page that is stripped of its markup looks empty |
| 6 September 2026 | That Orlando Sanford International Airport publishes no driving time to Port Canaveral | 1 hour 5 minutes, in a Nearby Attractions block on the airport’s home page listing 52 destinations |
| 6 September 2026 | That Orlando International Airport publishes no list of hotels with a shuttle | 24 of them, in a PDF linked from the references block of a page about something else |
| 7 September 2026 | That the hotel inside Miami International Airport has no page within our allowed sources | Its floor, its restaurant and its telephone number are on the airport’s own services and amenities page, and a second way to sleep inside the terminal is filed under VIP Clubs and Lounges |
| 10 September 2026 | That Tampa International Airport publishes neither a fare nor a frequency for the bus to the St Petersburg cruise terminal | $2.25 and a service window, in the airport’s own newsroom. The same paragraph gives two closing times that contradict each other, which the guide now reports rather than resolves |
| 11 September 2026 | That Melbourne Orlando International Airport publishes no readable parking rate | $1.00 per 20 minutes and a $16.00 maximum per 24-hour period. We had been reading flymlb.com, which redirects to mlbair.com, and treating the old domain’s refusal as the airport’s silence. Eight days of parking there is $128.00, which changed the ranking in our comparison of the three airports nearest Port Canaveral |
Figures that were wrong rather than missing
| Date | What we had written | What it says now |
|---|---|---|
| 10 September 2026 | That off-port parking near Port Canaveral costs $7.95 per day, multiplied out to $31.80 for a four day stay | The operator sells one package at $63.60 for the whole sailing. The $7.95 is that figure divided by an eight day cruise, so multiplying it by a shorter stay understates the cost. On a three night sailing the package wins by $16.40, the opposite of what we had published |
| 11 September 2026 | That a motorhome pays the same $20.00 per day as a car at Port Canaveral | The port publishes both answers and reconciles neither. Its parking page says $20.00 per day plus tax for both vehicles and RVs; Rule 1125 of its filed Port Tariff No. 16, effective 1 October 2025, adds a second line at $40.00 per day for oversize vehicle parking. Over eight billed days that is $320.00 rather than $160.00, so a motorhome should be budgeted at the higher figure and confirmed by telephone |
| 11 September 2026 | That five cruise lines board at Port Canaveral and exactly one of them prints a parking figure | The port publishes that it is homeport to seven, and its own directory lists eight operators. Two of them print figures: one at $17.00 per day, and Celebrity Cruises as a six step ladder from $68.00 for three nights to $153.00 for eight, every step of which divides out at that same $17.00 |
| 11 September 2026 | That the departure rate at Port Canaveral never falls back after November | It dips once, to 3.2 sailings a day in December, before January’s 3.4 and February’s 3.6. The arithmetic was ours, not the port’s |
Figures we removed because we cannot verify them
Two commuter rail operators that our guides relied on now forbid our reader in their robots.txt, and one of them is the only issuer for a set of fares we had published. We removed those fares rather than leave a number nobody can trace, on 9 and 11 September 2026. The affected guides now say what the county publishes, which is that the train gets you close and a taxi finishes the trip, and they no longer name a weekend winner that rested on a fare we cannot cite.
The practical consequence is worth stating plainly: for those corridors there will be no quarterly re-verification, because the door is shut. Where that is the case, the guide says so.
A correction we made and then had to undo
On 9 September 2026 five pages of one airport’s site appeared to change within a single day: a block of rooms disappeared, a table of figures altered seven of its eleven lines, a hotel list shrank from 34 entries to 32, and a commission changed from $0.35 to $0.27. We corrected eight guides against the new readings and rebuilt two of them.
Hours later all five pages had returned to their morning values, with the hotel list identical to the first reading byte for byte. The same address was serving two versions, and we had corrected against one that does not exist. Every one of those edits was reverted the same day, and only one change survived, because it did not depend on which version answered: three postal codes in that hotel list have a single entry, not two.
What we changed as a result: a figure that is going to carry a conclusion is now read twice, on separate requests, and the reading is kept. A number seen to move is reported in the body with its issuer and date, and it does not go into a heading.
How these were caught
Almost none of them were caught by an automated check. The pattern that finds them is duller than that: before a new guide is written, the guides next to it are read, and the writer is allowed to contradict the angle they were given by going to the issuer. Most of the entries above were found that way, by the piece being written next door.
If you find one we have missed, the contact page reaches us. A figure with the wrong issuer beside it is the kind of bug this site exists to avoid.