For a quarter-century, the muon's magnetic wobble has stood as one of particle physics' most tantalizing puzzles. Experiments at Brookhaven and later Fermilab consistently showed the muon wobbling more than the Standard Model predicted, a discrepancy that hinted at unknown particles or forces beyond current theory. That tension appeared to dissolve in 2021 when a supercomputer-driven lattice QCD calculation from the BMW collaboration matched the experimental result to extraordinary precision, suggesting the Standard Model accounts for the muon's behavior after all.
But the resolution has birthed a fresh crisis. The older, data-driven method — which infers the strong force's contribution from electron-positron collider data — now disagrees with both the lattice prediction and the experimental measurement. Recent results from the VEPP-2000 collider in Novosibirsk have deepened the mystery: a new detector installed in 2010 produced a dramatically different pion production rate when its results were published in 2023, and that newer rate aligns with the lattice calculation. Physicists are now scrutinizing whether the discrepancy stems from experimental systematics in the collider measurements or points to something more fundamental.
What's New
The muon g-2 saga has entered a new phase defined by a clash between two theoretical approaches rather than between theory and experiment. The key developments include:
- BMW lattice QCD calculation (2021): Published in Nature on the same day as Fermilab's updated measurement, the Budapest-Marseille-Wuppertal group's lattice simulation achieved precision matching the data-driven method for the first time. Their result agreed with the experimental value, implying no new physics is needed to explain the muon's wobble.
- Independent lattice confirmations: Multiple lattice QCD groups have since reproduced the BMW result, strengthening confidence in the purely theoretical approach.
- VEPP-2000 pivot (2023): The Novosibirsk collider's new detector, installed in 2010, measured a significantly different rate of charged pion production in electron-positron collisions compared to its own earlier data. The newer measurement aligns with lattice predictions, while the older data-driven compilations — which relied on the earlier VEPP-2000 results among others — do not.
- Data-driven method under scrutiny: The traditional approach, refined by physicists like Alex Keshavarzi at University College London, uses e+e- → hadrons cross-section data to infer the strong force contribution. This method now sits in tension with both lattice QCD and the experimental g-2 value.
The discrepancy centers on the hadronic vacuum polarization (HVP) contribution — the largest source of uncertainty in the Standard Model prediction. Lattice QCD calculates this from first principles using a discretized spacetime grid, while the data-driven method integrates experimental cross-sections. For decades the data-driven approach was considered more reliable; the lattice method suffered from large systematic errors. That hierarchy has now inverted.
Why It Matters
The muon's magnetic moment serves as a quantum microscope. Because the muon couples to every particle in the Standard Model — and potentially to unknown ones — its g-factor encodes a census of nature's fundamental constituents. A persistent discrepancy would have been the clearest sign yet of physics beyond the Standard Model, possibly pointing to dark matter candidates, supersymmetric particles, or new gauge bosons.
The BMW result, if correct, closes that window. The muon's wobble is fully explained by known particles and forces. But the conflict with the data-driven method means physicists cannot yet claim consensus. The Theory Initiative — the international consortium that produces the community's consensus Standard Model prediction — has not yet incorporated the lattice results, pending further cross-checks. Until the two theoretical approaches converge, the experimental result sits in an ambiguous zone: it agrees with one calculation but not the other.
The VEPP-2000 development adds experimental weight to the lattice side. Fedor Ignatov of the University of Liverpool, a member of the VEPP-2000 collaboration, described the new pion rate as a surprise that no one expected. The measurement has undergone intense scrutiny, with Keshavarzi noting that "no measurement has been scrutinized more." So far, no flaws have been found. If the new VEPP-2000 data is correct, it suggests the older e+e- collider data — from experiments in Italy, the United States, and earlier VEPP-2000 runs — may suffer from unrecognized systematic effects.
This has practical consequences for the global particle physics program. Future g-2 experiments, including a planned measurement at J-PARC in Japan using a different technique, will push experimental precision further. But their interpretive power depends on resolving the theoretical split. The Muon g-2 Theory Initiative's next white paper, expected in the coming years, will need to adjudicate between lattice and data-driven inputs — a decision that will shape whether the community views the muon anomaly as solved or as a continuing beacon for new physics.
Our Take
The muon g-2 story illustrates how progress in fundamental physics often advances not through clean victories but through the migration of puzzles from one domain to another. The original theory-experiment tension has been replaced by a theory-theory standoff, with experimental data from VEPP-2000 now acting as a tiebreaker that favors the lattice approach. This is a healthy development: it forces the community to confront the assumptions and systematics of the data-driven method, which has underpinned precision Standard Model tests for decades.
However, caution is warranted. The data-driven method rests on a vast body of cross-section measurements from multiple independent experiments over many years. A single new result from one detector at one collider, however precise, does not automatically invalidate that edifice. The VEPP-2000 result needs independent confirmation — ideally from other e+e- machines such as BESIII in Beijing or Belle II in Japan, or from the upcoming CMD-3 and SND analyses at VEPP-2000 itself. Until then, the lattice result, while impressive and independently replicated, carries its own systematic uncertainties: finite-volume effects, discretization artifacts, and the challenge of extrapolating to physical quark masses.
Editorially, we view the BMW lattice calculation as a landmark achievement in computational particle physics. A decade ago, lattice QCD was an order of magnitude less precise than the data-driven method; today it rivals or exceeds it. That trajectory mirrors the evolution of lattice QCD in other domains, such as the calculation of the strong coupling constant and light quark masses, where it has become the gold standard. The muon g-2 HVP contribution appears to be following the same path.
But the stakes are uniquely high here. The muon anomaly has driven a generation of model-building and experimental proposals. If the lattice result holds, a major motivation for certain classes of beyond-Standard-Model physics evaporates. If the data-driven method is ultimately vindicated, the anomaly returns with even greater force, because the experimental precision will only improve. The community's responsibility is to resist premature closure and let the cross-checks play out.
FAQ
What is the muon g-2 anomaly?
The muon g-2 anomaly refers to a long-standing discrepancy between the measured value of the muon's anomalous magnetic moment (g-2) and the Standard Model prediction. Experiments at Brookhaven (2001) and Fermilab (2021) found the muon wobbles more in a magnetic field than theory predicted, a difference that reached 4.2 sigma — close to the 5-sigma threshold for a discovery claim.
How did the BMW lattice calculation change the picture?
The BMW collaboration used lattice QCD — a supercomputer simulation of the strong force on a discretized spacetime grid — to calculate the hadronic vacuum polarization contribution from first principles. Their 2021 result, published in Nature, agreed with the Fermilab measurement, suggesting the Standard Model fully explains the muon's wobble without new particles. Independent lattice groups have since confirmed this result.
Why does the data-driven method disagree?
The data-driven method infers the strong force contribution from experimental measurements of electron-positron collisions producing hadrons (mainly pions). This approach has been refined for decades and was long considered more reliable. However, recent results from the VEPP-2000 collider in Novosibirsk, published in 2023, show a significantly different pion production rate than earlier measurements. The new VEPP-2000 rate aligns with the lattice prediction, suggesting the older collider data used in the data-driven compilations may have unrecognized systematic errors.
Has the muon g-2 mystery been solved?
Note definitively. While the lattice calculation matches the experimental result, the Theory Initiative — the international body that produces the consensus Standard Model prediction — has not yet adopted the lattice result, awaiting further verification. The conflict between the two theoretical approaches means the community lacks a single agreed-upon prediction. Until that convergence occurs, the anomaly's status remains ambiguous.
What happens next?
Several paths forward are underway. Other e+e- colliders (BESIII, Belle II, CMD-3, SND) will test the VEPP-2000 pion rate. Lattice groups continue to refine their calculations, reducing systematic uncertainties. The Muon g-2 Theory Initiative will issue its next white paper incorporating the latest lattice results. Meanwhile, Fermilab's final Run-2/3 analysis and the future J-PARC g-2 experiment will push experimental precision further. The resolution will likely emerge from this multi-pronged cross-checking over the next few years.
