The Farm Service Agency - usually a backwater from a grain market perspective, seen as something of the preserve of data nerds – found itself at the centre of a Twitter storm on Wednesday after an attack of the gremlins.
"If you ran your business like the FSA, you would not have a business," tweeted Peter Meijer.
Chicago-based Top Third Ag Marketing suggested the FSA may be employing the same technicians responsible for the Hillary Clinton e-mail controversy, over use of personal accounts for official business.
The ribbing followed the release by the FSA of monthly data on US crop enrolments in US subsidy schemes which appeared to show that farmers had not after all found the unusually wet spring a setback to sowings, as markets have been trading since.
"Someone tell me how this is not a mistake," said Scott Irwin, agricultural economist at the University of Illinois.
The data, for instance, raised the "Missouri question", as reported earlier by Agrimoney.com , and which puzzled Dr Irwin, as to why just 25,185 acres of soybeans in the state - the epicentre of spring wetness concerns - appeared to have gone unseeded.
In August, the FSA data had shown a total 1.02m Missouri soybean acres claimed for prevent plant insurance.
The answer to the downgrade was that the FSA had wrongly mixed in last season's data with the latest reports – as, indeed, some investors had suspected.
Risk management consultant and farmer Garrett Toay flagged the "incredible coincidence" that the FSA's September data were "the exact same as the January 2015 final estimate, outside of Wisconsin and Wyoming".
The FSA released a corrected report later, which showed, for example that farmers in Missouri were claiming 1.05m acres in prevent plant insurance for land which had been designated for soybeans – an increase on the August figure.
Indeed, the revised report appeared somewhat bullish, in indicating a few more prevent plant acres, 2.22m, nationwide than indicated last month.
Furthermore, the total for actual soybean planted acres, at 80.69m, was below the 80.81m acres revealed a year before.
That does not appear to square at all with current USDA thinking that sowings of the oilseed are up this year, by 600,000 acres.
Indeed, "soybeans have price-supportive acreage data in these numbers", said Mike Zuzolo at Global Commodity Analytics.
Still, it was then the turn of markets to show that they can behave a little bizarrely too.
After recovering to show small gains after the first set of FSA data, they fell after the corrected, and more bullish, edition.
At broker Rice Dairy, Jerry Gidel told Agrimoney.com: "I think the corrected data could indicate that fewer soybeans were planted than had been thought.
"But the market has already come up a way in the last few sessions. I think it has gone away from buyers, who are standing off for now."
Uncle Hedge, a farmer, tweeted that "I hope the algos [algorithm-following funds] go broke trading faulty FSA numbers", adding that this would turn staff at the agency into "heroes".
It certainly looks easy enough for anyone trading fundamentals to lose a buck or two.
By Mike Verdin