Physicists once traveled nearly 3,200 miles by truck and barge from Brookhaven National Laboratory in New York to Fermilab, which is located in a building about forty miles west of Chicago. Photos of this massive ring-shaped object rolling through the Illinois night became somewhat famous in scientific circles. The move took eleven days, necessitated road closures, and produced a small but genuine public spectacle. You can learn something about the stakes from the lengths the physics community was willing to go to in order to measure something that was 2.5 parts per billion different from what a textbook said it should be. or, at the very least, how stakes are viewed by physicists.
The particle being measured is the muon, which is best described as an unstable, heavier version of the electron that only lasts for two millionths of a second before decaying. It acts like a tiny magnet, just like the electron, and this magnetism can be represented by a number known as the g-factor. With remarkable accuracy, theory predicts what g should be. The experiment determines the true value of g. Particle physics has experienced more concentrated excitement and anxiety than nearly anything else in recent memory due to the decades-long disagreement between those two numbers, which is small enough to appear to anyone outside the field as a rounding error.
| Topic | Muon G-2 Anomaly & The Standard Model Challenge |
|---|---|
| Experiment Name | Muon g−2 (pronounced “gee minus two”) |
| Primary Location | Fermi National Accelerator Laboratory (Fermilab), Batavia, Illinois, USA |
| Previous Location | Brookhaven National Laboratory (BNL), New York (data: 1997–2001) |
| Earlier Experiments | CERN, Geneva, Switzerland (began 1959, led by Leon M. Lederman) |
| What Is a Muon? | Heavier, unstable cousin of the electron; acts like a tiny bar magnet |
| What Is G-2? | The “anomalous” part of the muon’s magnetic moment — slightly larger than 2 |
| Key Anomaly | Muon measured as more magnetic than Standard Model predicts |
| Discrepancy Size | ~2.5 parts per billion — extraordinarily small, extraordinarily significant |
| 2021 Result Significance | 4.2 sigma deviation (discovery requires 5 sigma) |
| 2023 Result | 5.1 sigma from 2020 theory prediction; ~1 sigma from new lattice QCD calculations |
| Final Results Published | June 3, 2025 (after six years of data collection, ended July 9, 2023) |
| Experiment Team Size | 200+ physicists (experimental); 100+ theorists for Standard Model calculations |
| Current Status | Anomaly possibly resolved by new theoretical calculations — debate ongoing |
| Reference Website | Fermilab Muon g−2 Experiment |
The first g-2 measurement was made in 1959 at CERN in Geneva by Leon Lederman and a team of six physicists using the Synchrocyclotron. The outcomes were sufficient to support a few fundamental theories. The results of the third CERN experiment, which came to an end in 1979, were confirmed with a precision of 0.0007%, which sounds like closure. However, the theoretical predictions continued to advance, the Standard Model continued to become more accurate, and by the time Brookhaven took over the experiment in the late 1990s, there was enough accuracy on both sides to begin observing a persistent, bothersome gap.
The Brookhaven team reported in 2001 that the muon seemed somewhat more magnetic than expected. Not significantly. by a quantity that was difficult for non-specialists to understand without careful explanation. However, consistently being off by a small amount is frequently more fascinating in physics than being drastically off by a large one.
That 2001 outcome persisted. It became one of those open questions that a certain type of physicist finds genuinely difficult to stop thinking about as it sat in the literature, being reexamined and rechecked, producing theoretical papers and conflicting calculations. The difference was between 3 and 3.7 sigma, which is suggestive but falls short of the five-sigma threshold needed by the field to declare a discovery. There may be a threshold because physics has previously been burned by results that appeared genuine at three or four sigma but silently vanished when additional data became available. The muon anomaly was intriguing in the way that only a measurement that verges on confirmation can be.
Fermilab then assumed control. The massive, ridiculously meticulously transported storage ring magnet from Brookhaven was repaired, shimmed to create a magnetic field three times more consistent than before, and installed in Illinois to receive purer muon beams than Brookhaven could produce. In 2018, data collection got underway. More than 200 physicists contributed to the first results, which were published in April 2021 and showed a startling 4.2 sigma deviation from the Standard Model prediction. Not quite a revelation. However, it was close enough to elicit a response from the physics community that ranged from cautious excitement to barely contained elation. One theorist at the time stated, “Since the 1970s we’ve been looking for a crack in the standard model.” “This may be it.”
Observing the field’s reaction to that 2021 announcement gave me the impression that physicists had been waiting a long time to feel this way. The Standard Model has always had a quality of completeness that some researchers find professionally frustrating because it was meticulously put together over decades, tested against experiment after experiment, and survived everything thrown at it. It functions. Excellently. However, it does not satisfactorily incorporate gravity, explain dark matter, or explain why the universe contains so much more matter than antimatter. In a magnetic field, a muon acting a little strangely could appear as a small crack in the wall. However, new knowledge has historically tended to enter through cracks.
After that, things became more difficult. Based on the first three years of Fermilab data, the August 2023 results technically crossed the discovery threshold by pushing the experimental deviation against the 2020 Standard Model prediction to 5.1 sigma. However, a competing set of theoretical calculations using lattice QCD, a different computational technique, yielded a Standard Model prediction that was significantly closer to the experimental value.
The question shifted from whether the muon was misbehaving to which calculation of expected behavior was actually correct when two seemingly sound theoretical predictions suddenly pointed in different directions. After years of mounting tension, it’s still unclear which theoretical approach will be more trustworthy, and that uncertainty has a somewhat depressing quality.
On June 3, 2025, Fermilab released the final results from all six years of data. At the end of the experiment, the anomalous magnetic moment of the muon was measured with the highest precision ever. Which theoretical prediction you are willing to accept—the more recent lattice QCD approach or the older consensus calculation—will largely determine what those results mean for the Standard Model. After twenty years of work and one extremely carefully transported magnet, that is an uncomfortable place for the field to land.
It’s possible that the anomaly was never quite what it seemed to be and that, rather than representing true new physics, it was always partially a problem with theoretical computation. Alternatively, it’s possible that the muon is actually pointing in a direction that the Standard Model cannot account for, and that the disagreement between calculation techniques is merely postponing the recognition of that. Observing physicists negotiate that ambiguity with a combination of methodical patience and frustration is a kind of education in the real workings of science.
