SoylentNews is people
SoylentNews
Scientists rename genes because Microsoft Excel reads them as dates:
Microsoft Excel’s automatic formatting is normally helpful for finishing spreadsheets quickly, but it’s proving to be an agent of chaos for geneticists. The Verge has learned that the HUGO Gene Nomenclature Committee has issued guidelines for naming human genes to prevent Excel’s automatic date formatting from altering data. MARCH1 (Membrane Associated Ring-CH-Type Finger 1), for example, should now be labeled MARCHF1 to stop Excel from changing it to 1-Mar.
The names of 27 genes have been changed in the past year to avoid Excel-related errors, HGNC coordinator Elspeth Bruford said. This isn’t a rare error, either, as Excel had affected about a fifth of genetics-related papers examined in a 2016 study.
Journal Reference:
Mark Ziemann, Yotam Eren, Assam El-Osta. Gene name errors are widespread in the scientific literature [open], Genome Biology (DOI: 10.1186/s13059-016-1044-7)
(Score: 2) by NateMich on Friday August 07 2020, @04:50AM (22 children)
Spreadsheets are used for everything where I used to work. I started calling it "management by spreadsheet", because no matter what they needed, I had to put it in Google sheets.
(Score: 5, Insightful) by Booga1 on Friday August 07 2020, @05:07AM (15 children)
It's because it makes nice charts and pretty colors.
Long ago I made an Excel spreadsheet that extracted all the data from our ticketing systems at work. Management then used that to decided who to fire when crunch time came. They simply picked the people who handled the least amount of cases and the lowest customer feedback scores. Unfortunately, those were all our best people. They were the ones that were able to handle the toughest cases and naturally they took longer and generally involved the angriest customers. So your reward for being the best at your job was being fired. I seriously regret ever making that spreadsheet. Had I only known....
Anyway, back to the subject at hand, this shouldn't be an issue at all. You can tell Excel to treat the cells as plain text. This is a user education issue. The fact that they're changing their naming guidelines because people can't be bothered to change a setting in a program is rather depressing to think about.
(Score: 4, Interesting) by krishnoid on Friday August 07 2020, @05:15AM (4 children)
And hence, probably the highest paid, so win-win for management. One would then hope the angriest customers got escalated straight to management, a few months [notalwaysright.com] after which, they'll figure out that maybe they shouldn't have gotten rid of them.
In a way, this all boils down to people selecting services based on cost, marketing, features, etc. rather than support and product quality. If there was a culture of evaluating those things prior to purchase, I bet things would start changing.
(Score: 2) by Booga1 on Friday August 07 2020, @05:43AM (2 children)
Yep, nailed it right off the bat there. That's pretty much how it went. It was like watching some sort of real life video game where management thought they could just trust the numbers to tell them all they needed to know about the people working for them and how things were going. The realization did come months later as the time to solve cases got worse and so did the customer feedback scores. By the time management realized what was going on, it was too late. Anyone who saw what happened to the good workers moved on, leaving us with the mediocre and true underperformers to try and save the contract. Of course, that didn't work and the remaining team of 50+ people had to be let go when we lost the contract.
BTW, that notalwaysright link you included was pretty "good" in the usual way those stories go. So many messed up stories on there...
(Score: 2) by krishnoid on Friday August 07 2020, @06:40AM (1 child)
W. Edwards Deming is frequently misquoted [deming.org] as having said that you can only manage what you can measure. He was talking about quality, but customer support/satisfaction seems like it should be another one of those things.
(Score: 0) by Anonymous Coward on Friday August 07 2020, @02:59PM
"Measure" these days is just more bullshit busywork in the form of self-eval questionaires that you dump on the losers, I mean workers. Then hand it off to "AI" or "Excel" to sort the list. As someone above already mentioned, your best people often don't score well on busywork measures.
(Score: 2) by maxwell demon on Friday August 07 2020, @08:32AM
Not if by then they moved to another company (being able to show a “positive” past performance because the negative consequences didn't yet show up). The managers that replaced them then have to deal with the situation (and probably get all the blame because, after all, it's during their time that the shit hit the wall — never mind that the previous managers set the shit in motion).
The Tao of math: The numbers you can count are not the real numbers.
(Score: 0) by Anonymous Coward on Friday August 07 2020, @12:15PM (6 children)
> I made an Excel spreadsheet that extracted all the data from our ticketing systems ...
What I take away from this sad-but-typical story is that your spreadsheet wasn't complete, it didn't include data about the difficulty of each problem and the skills/experience required to solve it. Perhaps this wasn't in the ticketing system, but with hindsight it should have been created somehow.
(Score: 2) by HiThere on Friday August 07 2020, @02:07PM (4 children)
You can't just create data. You can create fake information, but that's not the same thing. If the data isn't in the system, you can't create it.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 0) by Anonymous Coward on Friday August 07 2020, @05:42PM (3 children)
And yet, back to GPs story, the smart/productive members of the team got laid off. Ostensibly for reasons that were based on the presentation of the ticket data in Excel format.
So what is the way forward in this situation? Not being sarcastic. Just wondering if that fateful spreadsheet could have been done differently.
(Score: 4, Informative) by Booga1 on Friday August 07 2020, @07:11PM (2 children)
What I think should have been done differently in my opinion would have been to talk to the team leads about why certain people had poor stats. Who did they consider to be the most trustworthy and skilled team members and why. I was one of about ten team leads and they did not talk to any of us about any part of the upcoming layoffs at all.
Part of the problem was policy based. Management policy required ALL service outage cases to be assigned to one agent to do all the customer callbacks once service was restored. Team leads would pick the best sweet talking agents with nerves of steel to handle all those angry customers. That meant a single person called back dozens of angry customers and received all the negative feedback for it. People who handled outage cases were the first to be fired because "they had poor customer service skills."
The "slow" agents were in the second round of layoffs. Difficult cases were given to people you could trust to solve them. Anyone can handle a password reset, but only a few could handle complex database issues requiring deep dives into corrupted mailboxes to save a customer's important business data. Someone who knocked out 100 password resets had awesome stats and looked great on paper, but would never be able to solve a single one of the toughest technical cases. As part of that, agents that did the hardest work looked like the slowest and they were let go.
Now, all that said, cases were divided into somewhat rough but reasonable categories. To address the issues with the spreadsheet itself would have required management to assign a weight to those categories. Unfortunately they did not. It still would have only solved half the problem since categories weren't really detailed enough to tell the difference between "I can't access mail because my Outlook profile is messed up" and "I can't access mail because the database got messed up during a catastrophic server failure that required a restore from backup." Any grading for difficulty of a case would be somewhat arbitrary, but it still would have been better than treating all cases as equal.
(Score: 0) by Anonymous Coward on Friday August 07 2020, @11:34PM (1 child)
My wife worked at a small (100-ish) employee place where the owner cared about the employees (family picnics, etc.), but then he sold out to a huge company and retired. They then brought in a new CEO from halfway across the country who was cut right out of the shitty napolean complex MBA CEO types. She even brought her best friend along and put her as the HR head, but of course this person wasn't an actual employee, but she ran her own small company, so HR was suddenly run by a contractor. But my story isn't about the morally and legally corrupt goings on (you used to be able to see a lot of examples of that posted on Glassdoor, at least until they started requiring employees to post 5-star ratings), but when she came in, they clumsily did that 1-2-3 rating thing, where they would get rid of you if you got a 3, etc. It was all the rage about 10 years or more ago because "Jack Welch did it, so if I'm going to be recognized as a great CEO, I'm going to do it too". What was hilarious was that they made the managers rate everybody, but in the end there was confusion whether it was the "1" or the "3" that was bad. It was all just an excuse to get rid of people "legally" by claiming they were poor performers. It sounds like that is what was going on here. It didn't matter how well the spreadsheet reflected reality, they just needed a crutch to justify reducing headcount (which means reduced costs, which means bigger CEO bonus!).
(Score: 2) by Booga1 on Saturday August 08 2020, @12:22AM
I could understand that being the case in a lot of situations. Ours was different. The problems were not due to a management change or corporate buyout. Our collapse was in no small part related to the economic downturn at the time. Getting rid of the highest paid people was probably part of it, as krishnoid mentioned above. It was definitely not about manager/CEO bonuses since eventually the whole contract was canceled and everyone on that contract was let go, managers included. So, everybody got the chopping block eventually, but the spreadsheet was just used to make the cuts in ways that made very little sense to any of us(as team leads).
Only a couple of very small side contracts that had no overlap with ours survived.
(Score: 1, Insightful) by Anonymous Coward on Friday August 07 2020, @03:04PM
> it didn't include data about the difficulty of each problem and the skills/experience required to solve it
Oh cool. Just add a few more pages to the self-eval forms.
(Score: 4, Interesting) by Jay on Friday August 07 2020, @04:20PM (1 child)
You can sometimes tell excel to treat the cells as plain text. And sometimes it's somehow magically already changed not just the formatting, but THE ACTUAL DATA IN THE CELLS so that when you tell it to change to plain text, it gives you the plain-text of the date format it "helpfully" converted your data to.
I don't know if this is a recent bug or a "feature", but I've run into it twice lately. It was absolutely mind-blowing to see data turned into dates, force the dates to plain text, and be left with the dates in the same format. Cutting and pasting into notepad got me the dates, showing it was the underlying data at that point!
I've taken to reading and writing all .csvs in R, because I just don't trust excel to not mangle the data, even when explicitly told how to handle it.
(Score: 2) by Booga1 on Friday August 07 2020, @06:35PM
It seems that you're not alone running into this. A recent question on Super User [superuser.com] was having trouble with CSV imports being treated differently than they used to be.
Looks like there is an option to force it back to the old way according to this answer: https://superuser.com/a/1574923/897970 [superuser.com]
(Score: 0) by Anonymous Coward on Friday August 07 2020, @10:10PM
In all fairness, you probably should have seen it coming, what else were they going to do with that data?
(Score: 2) by drussell on Friday August 07 2020, @05:26AM (5 children)
Indeed... Spreadsheets are intended to be electrified ledgers.
Sadly, I've seen them used for far to many insane purposes.... (facepalm)
(Score: 3, Insightful) by krishnoid on Friday August 07 2020, @06:41AM (2 children)
That's the problem with a powerful, flexible, reliable tool -- once the hackers get their hands on it, it will be used for purposes you never anticipated.
(Score: 3, Funny) by DannyB on Friday August 07 2020, @03:26PM (1 child)
Hackers getting their hands on a spreadsheet is not a problem.
The problem is MBAs getting their hands on a spreadsheet.
* OMG! Excel is the best database evar!
* OMG! I can write VBA macros!
* I'm now the greatest programmer evar!
* I can write separate "query" functions to search the rows of each table for the row I want!
* It's fast and efficient! Even on HUGE databases that have HUNDREDS of rows!!!
* I don't see what's the big deal about "complexities" that my programmers complain about. It's all so easy!
To transfer files: right-click on file, pick Copy. Unplug mouse, plug mouse into other computer. Right-click, paste.
(Score: 3, Touché) by krishnoid on Friday August 07 2020, @05:46PM
* Ok, maybe this is starting to get a more complex
* Now that I've shown results and head to greener pastures, I'm can dump this pile of technical debt off onto some nerd
* So long, everyone
* Wow, this MBA stuff is easy!
(Score: 2) by DannyB on Friday August 07 2020, @03:21PM (1 child)
Electrocuting bean counters doesn't sound so insane.
To transfer files: right-click on file, pick Copy. Unplug mouse, plug mouse into other computer. Right-click, paste.
(Score: 2) by krishnoid on Friday August 07 2020, @05:47PM
Brings new meaning to the term 'protected' cells. Click on the wrong one and your mouse zaps you. I like it!