Particle Physics Discoveries that Disappeared

Particle physicists seem to constantly announce discoveries which then disappear. What is going on? How seriously are you to take the recently announced Fermilab muon g-2 anomaly or the B-meson anomaly at the Large Hadron Collider? In this video I explain what's going on.

I have a longer comment about the recent muon g-2 measurement here:

Addendum to what I say at 5 mins 30 seconds: That's only for a normal distribution. The relation between confidence level and standard deviation can differ for other distributions.

Correction to to what I say at 7 mins 18 seconds ("had no known interpretation"). This only refers to the superjets, not to the pentaquarks (which, well, were interpreted as pentaquarks). Sorry about that.

0:00 Sponsor Message
1:00 Intro
1:42 How do the predictions work?
3:12 Where do anomalies come from?
5:46 Anomalies that disappeared
8:48 Other problems
9:55 What's with that B-meson anomaly?

This video was sponsored by DataCamp.

#science #physics #data

[robertansley]

Those _researchers that perform very creative analysis_ @9:00 reminds me of my accounting professor who said _Numbers will confess to anything if you torture them_ !

[MedlifeCrisis]

As soon as I saw these stories break I was like "I need to know what Sabine says about this before I know how to think"😂 It's not that I'm blindly trusting of her, it's just that it's way outside my domain and I know it's Sabine's field. Also I'm blindly trusting of her.

[fluffy_tail]

physicist: we use 5sigma as the best confidence level, but people start getting interested around 3 sigma
me, a biologist: 👀

[Aufenthalt]

It is very important for the public to get an overview of the failures and not only the successes. Only so it is possible to get a non emotional view of the events. And Sabine make an excellent job in giving us an overview of the failure histories. Great channel.

[marcochimio]

Excellent analysis, Dr. H. As you might imagine, this type of thing comes up in totally unrelated sciences, as well. The medical literature is littered (no pun intended) with reports of new and exciting biomarkers (molecules whose high or low levels indicate some particular disease) that only vanish when they are re-investigated by other labs. Part of the problem is we look at tens of thousands of proteins and use only 2-3 sigma to verify. About 10 years ago (I'm writing this from memory, so I may have some details wrong), some researchers re-tested at about 75 "exciting" new markers of heart disease from the journals. They found that 74 of these could not be reproduced, and the remaining one biomarker was "iffy" at best. It's just as bad or even worse with new cancer markers.

[creightonfreeman]

Oh Sabine, you are so tactful. "Some independent researchers use, very creative data analysis methods." lol

[flockofwingeddoors]

The content you make is SUPER important! I don't hear this kind of analysis in the news that filters down to me usually, but having reality checks like this is important to keep us grounded and impartial

[mailtorajrao]

Thank you, @datacamp for sponsoring these oh so excellent Sabine videos!

[phdtobe]

Sabine, I’ve been waiting for you to cover this topic, as I have been sceptical about the strength of the claims made for some of the YouTube coverage of these observations. Thanks.

[faarsight]

This is such a good video. Rarely do we hear about when a speculative new possible exciting discovery turns out not to have been true.

[picksalot]

Thanks Sabine for shedding light on this important subject. A 5th category may need to be added to the list to cover "Dishonesty." It would take into account things like sensationalism, seeking funding, fame, notoriety, and other incentives.

[gmaf]

When I was a kid it was Saturday morning cartoons. Today, it's Saturday morning Sabine :)

[x_abyss]

The moral of the story: if you have a new particle discovery, don't let the New York Times publish about it. Jokes aside, fluctuation is the bane of any form of research. Great video as always Sabine.

[clarkh]

*breathes, slight camera zoom* "very creative methods of analysis"

that gave me a LOL

[nickthurn]

Love it. 🙃❤️🙃

Serious question: would it be fair to say that there is pressure to publish preliminary results because that can lead to more funding? So articles in the popular media, although annoying and often misleading, are just part of the process of keeping research going.

[RogerM]

Nice to see you back. Hope some video about gravitational waves as the future trend study in Physics.

[toddboothbee]

I'll never look at my paranormal coincidences with the same confidence again.

[ApeinSpace]

I so enjoy spending time with Sabine, smart, funny and to the point.

[jagatiello]

"Noise happens to look like a signal."
A great euphemism for 'I don't like this kind of music'.
Great video Sabine, as usual...

[eljcd]

Dr. Hossenfelder, thank you kindly for giving us an original view on the Physics world instead of regurgitating what others already said. Amazing how always can give us something new to learn in each video!