---
title: 'Headache: New Milky Way Data Just Defies Explanation'
source: 'https://youtube.com/watch?v=FExs0euUFHI'
video_id: 'FExs0euUFHI'
date: 2026-11-11
duration_sec: 380
channel: 'Sabine Hossenfelder'
---

# Headache: New Milky Way Data Just Defies Explanation

> Source: [Headache: New Milky Way Data Just Defies Explanation](https://youtube.com/watch?v=FExs0euUFHI)

## Summary

A new study of the Milky Way's rotation curve reveals that the outer regions rotate in a 'Keplerian' manner, contradicting both the dark matter halo model and modified gravity (MOND). The findings challenge our understanding of galactic dynamics and the distribution of dark matter in our galaxy.

### Key Points

- **Dark Matter vs. Modified Gravity** [00:00] — The Milky Way is thought to contain about 10 times as much dark matter as normal matter, but new data is incompatible with both dark matter and MOND.
- **Rotation Curve Basics** [01:00] — Stars in galaxies should orbit slower with distance from the center, but observed 'flat rotation curves' indicate stronger gravity, leading to the dark matter hypothesis.
- **Keplerian Drop Beyond Solar System** [02:01] — New data shows the rotation curve drops in a 'Keplerian' way beyond the solar system, meaning the dark matter halo predicted by models is not present.
- **MOND Also Fails** [02:33] — The data are also incompatible with MOND, which predicts stronger gravity in outer galaxies. The paper states this challenges both ΛCDM and MOND.
- **Milky Way's Dark Matter Fraction** [03:25] — If correct, the Milky Way would be only about three quarters dark matter, less than the cosmological average, which is hard to explain with dark matter.
- **Error Bars and Measurement Challenges** [03:54] — Error bars are large because measuring the Milky Way's rotation curve is harder than for distant galaxies, but future observations will reduce uncertainty.
- **Bullshit Meter Rating** [04:38] — The paper gets a 3/10 on the bullshit meter; it's nuanced and doesn't oversell results, but the flexibility of dark matter is a problem.

### Conclusion

The new Milky Way data may indicate that both dark matter and MOND are wrong, potentially leading to a paradigm shift in astrophysics.

## Transcript

Our galaxy contains about 10 times as much&nbsp; dark matter as normal matter. That’s what we&nbsp;&nbsp; thought. But a team of astrophysicists&nbsp; just found that new data about the&nbsp;&nbsp; Milky Way isn’t compatible either with&nbsp; dark matter or its biggest competitor,&nbsp;&nbsp;
modified gravity. Ok now then what? Let’s have&nbsp; a look. Stay around until the end for a special&nbsp;&nbsp; offer that will bring your coding skills to an&nbsp; entirely new level with boot.dev. Dark matter&nbsp;&nbsp;
is a supposedly invisible substance that fills&nbsp; up the universe and makes itself noticeable only&nbsp;&nbsp; through its gravitational pull. Astrophysicists&nbsp; believe that it makes up about 85% of matter,&nbsp;&nbsp; on average. But it is not evenly distributed&nbsp; and clumps in galaxies where it forms big,&nbsp;&nbsp;
mostly spherical, halos that usually extend far&nbsp; beyond the reaches of the galaxies’ visible stars. think that dark matter exists is&nbsp; that galaxies rotate too fast.
Going by standard gravity, one expects that&nbsp; as you move away from the centre of a galaxy,&nbsp;&nbsp; the gravitational pull gets weaker. And if&nbsp; you do the math, this means that the stars&nbsp;&nbsp; out there should orbit slower around the centre.&nbsp; This relation between the star’s distance from&nbsp;&nbsp;
the galactic centre and its orbital speed&nbsp; is called a ‘rotation curve’. In reality,&nbsp;&nbsp; the speeds of stars don’t drop as fast with the&nbsp; distance from the centre as they should. Instead,&nbsp;&nbsp; they become almost constant, leading to&nbsp; what is called a ‘flat rotation curve’.
This means the gravitational pull of the&nbsp; galaxy is stronger than we think. Hence,&nbsp;&nbsp; dark matter must do it. It’s not just that, from&nbsp; the rotation curves one can extract an approximate&nbsp;&nbsp;
distribution of the dark matter. And the huge&nbsp; spherical dark matter halos fit the data best. This is also where the estimate for the amount&nbsp; of dark matter in the Milky Way comes from,&nbsp;&nbsp; from the rotation curve of the stars around&nbsp; us. But a new paper that now appears says&nbsp;&nbsp;
that actually the rotation curve beyond the&nbsp; approximate location of our own solar system&nbsp;&nbsp; drops in a “Keplerian” way, that is, with the&nbsp; slope you get from standard gravity. You see&nbsp;&nbsp; the issue in this figure. This blue line&nbsp; is what dark matter predicted. This data&nbsp;&nbsp;
here doesn’t fit. The dark matter halo&nbsp; that we’ve been told surrounds us doesn’t&nbsp;&nbsp; seem to be there. If nothing else, it’s the&nbsp; largest missing-object report ever filed. It’s not just that the dark matter hypothesis&nbsp; fits badly with the data. The data are also&nbsp;&nbsp;
incompatible with the alternative explanation,&nbsp; modified Newtonian dynamics, MOND for short. In MOND, the gravitational pull in the outer&nbsp; parts of galaxies is stronger than what Einstein’s&nbsp;&nbsp;
general relativity predicts. Dark matter and MOND&nbsp; disagree about the reason, but both explain why&nbsp;&nbsp; the rotation curves are almost flat. The new Milky&nbsp; Way measurements say that the curve just isn’t&nbsp;&nbsp;
flat. As they write in the paper, if correct,&nbsp; this would “challenge both ΛCDM and MOND”. Though really it is a bigger problem for MOND.&nbsp; Because the data do not say that dark matter is&nbsp;&nbsp;
absent in the Milky Way. It still seems to be&nbsp; there near the centre. It’s just that if you&nbsp;&nbsp; go further out, away from the centre, then&nbsp; the supposed halo doesn’t seem to be there. If the data and its interpretation are&nbsp; correct, then the Milky Way would only&nbsp;&nbsp;
be about three quarters dark matter, which&nbsp; is less than the cosmological average. This&nbsp;&nbsp; is hard to explain with dark matter because it&nbsp; has it that galaxies form because dark matter&nbsp;&nbsp; clumped there. However, dark matter is such a&nbsp; flexible and underdetermined theory I am sure&nbsp;&nbsp;
one can explain it somehow. It’s more difficult&nbsp; for MOND exactly because MOND is more predictive. That said, as you may have noticed, the&nbsp; error bars on these data points are quite&nbsp;&nbsp;
large. You’d think that the data from the&nbsp; Milky Way would be the best rotation curve&nbsp;&nbsp; data that we have but in reality that’s&nbsp; not the case. This is because if we look&nbsp;&nbsp; at other galaxies we preferably look away&nbsp; from the Milky Way and that gives a quite&nbsp;&nbsp;
clean measurement. If we look at the Milky&nbsp; Way itself we have to figure out where all&nbsp;&nbsp; these stars in the foreground are and where&nbsp; they’re going. It’s actually much harder. On the flip side though, once the observations&nbsp; become better and the error bars shrink,&nbsp;&nbsp;
there is less to argue about. If you are&nbsp; looking at a distant galaxy you can always say,&nbsp;&nbsp; oh, maybe we’re looking at it from a weird&nbsp; angle, or maybe it’s actually two galaxies,&nbsp;&nbsp; or maybe we have the distance entirely wrong.&nbsp; With the Milky Way there are no such excuses.
I give this paper a 3 out of 10 on the bullshit&nbsp; meter. It is actually very nuanced and doesn’t&nbsp;&nbsp; oversell the results as much as I did.&nbsp; It’s a 3 because I think astrophysicists&nbsp;&nbsp;
should acknowledge that the flexibility of&nbsp; the dark matter hypothesis is a problem,&nbsp;&nbsp; not an asset. The good news for dark matter is&nbsp; that it can probably survive this. The bad news&nbsp;&nbsp; for dark matter is exactly the same. In the&nbsp; end though, it may be good for astrophysics.&nbsp;&nbsp;
Because finally dark matter and MOND may&nbsp; have something in common: they’re both wrong. Young people in science frequently ask me&nbsp; what they should learn to improve their&nbsp;&nbsp; career options. And I always tell them, no&nbsp; matter what area of science you work in,&nbsp;&nbsp;
learn to code. And today's sponsor, Boot.dev, is&nbsp; a great place to get started. If your research is&nbsp;&nbsp; any good, spreadsheets just won't be enough.&nbsp; You'll have to figure out how to manage your&nbsp;&nbsp;
data and how to write a user interface so&nbsp; that other people can use it. And that's&nbsp;&nbsp; exactly what you'll learn on boot.dev,&nbsp; backend development, database management,&nbsp;&nbsp; user interfaces. They have hands-on lessons in&nbsp; which you write code directly in the browser&nbsp;&nbsp;
and you solve challenges to level up. They also&nbsp; have what they call the training grounds where&nbsp;&nbsp; you can figure out things without pressure.&nbsp; This is knowledge which will come in handy in&nbsp;&nbsp; pretty much any career. If you decide to go&nbsp; for the full experience, use my code SABINE&nbsp;&nbsp;
to get 25% off your first year. So, go and check&nbsp; this out. Thanks for watching. See you tomorrow.
