The Cohen–Lenstra Heuristics

Executive summary

Groups appear with probability inversely proportional to their automorphisms

Class groups do not behave like random finite abelian groups of a given order. The Cohen–Lenstra heuristics propose instead that a group appears with weight inversely proportional to the size of its automorphism group — which strongly favours cyclic groups and makes large p-ranks rare. The predictions match computed tables closely, and they are now used in the opposite direction: as a check on whether a computation has found the whole class group.

Learning objectives

State the weighting principle behind the heuristics.
Quote the predicted frequency of odd class number divisibility.
Explain why the prime 2 is excluded from the odd-part predictions.
Use the heuristics to sanity-check computed class group tables.
Identify what has been proved and what remains conjectural.

Section 01The weighting principle

The central proposal is that a finite abelian group G occurs among class groups with weight proportional to 1/|Aut(G)|.

Automorphism counts and relative weights
Group of order 4	\|Aut\|	Relative weight
ℤ/4ℤ	2	High — cyclic groups are favoured
ℤ/2 ⊕ ℤ/2	6	Three times less likely than the cyclic group

The effect compounds: a group of p-rank r has an automorphism group of size roughly p^r², so high rank is heavily suppressed. The heuristics therefore predict that class groups are usually cyclic or close to it — which is exactly what tables show.

Why the automorphism weighting

It is the natural measure on isomorphism classes of objects with symmetry: counting each group once over-counts the symmetric ones, since they arise from fewer distinct presentations. The same weighting appears throughout enumerative mathematics, which is part of why the heuristic is persuasive.

Section 02Predictions

For imaginary quadratic fields, the heuristics apply to the odd part of the class group and predict, for an odd prime p, that the proportion of discriminants with p dividing h approaches

1 − ∏_k≥1 (1 − p^−k)

≈ 43.99%of imaginary quadratic fields have 3 | h

≈ 23.97%have 5 | h

≈ 75.45%of real quadratic fields have h = 1 (odd part)

The prime 2 is excluded, and for a reason

Genus theory determines the 2-rank of the class group exactly, from the number of prime factors of the discriminant. It is not random at all, so the heuristics deliberately apply only to the odd part. Applying them to the 2-part produces predictions that are simply wrong.

Section 03Use as a validation tool

Because the predictions are sharp and the data sets are large, deviation from them is informative. A table of computed class numbers whose distribution departs materially from the heuristics is far more likely to contain a systematic computational error than to have discovered new mathematics.

Check

Divisibility frequencies

Compare the observed proportion of h divisible by 3, 5, 7 against the predicted values over a large discriminant range.

Check

Rank distribution

High p-ranks should be rare. An excess of them suggests relations were missed, producing spurious extra generators.

Check

Cyclic proportion

The predicted proportion of cyclic class groups is high; a shortfall suggests the structure determination is incomplete.

Status of the conjectures

Most Cohen–Lenstra predictions remain unproven. Significant partial results exist — notably the average 3-torsion in quadratic fields, established by Davenport and Heilbronn well before the heuristics were formulated, and the function field analogues, which have been proved. The general statements are conjectural, and computations that agree with them are evidence for them rather than the reverse.

ReferenceFrequently asked questions

Do the heuristics apply to higher degree fields?

Extensions have been proposed for higher degree and for relative extensions, with the weighting adjusted for the Galois module structure. They are less thoroughly tested than the quadratic case, and the correct formulation in the presence of extra automorphisms is subtle.

Why do real and imaginary fields have different predictions?

Because in the real case the unit group contributes an extra constraint, effectively imposing one additional relation. The heuristics account for this by shifting the exponent, which is why real quadratic fields have a much higher predicted proportion of trivial class group.

Can the heuristics prove anything about a specific field?

No. They describe distributions over families and say nothing whatsoever about an individual discriminant. Their computational value is entirely in checking aggregate behaviour of large tables.

NavigateContinue in this stream

Curated next steps from this page. The site also surfaces algorithmically related reading below.

ProvenanceSources and further reading

This page is an original KEVOS explanatory article. It presents the underlying mathematics — definitions, algorithms, complexity results and selection criteria — in KEVOS editorial voice. No text is reproduced from any copyrighted source. Where numerical tables are relevant, KEVOS links to live authoritative databases rather than republishing static values.

Handbook application: from concept to controlled practice

Purpose. This expanded section turns the original page into a practical handbook. It preserves the supplied material and adds a repeatable way to apply, check and review The Cohen–Lenstra Heuristics. It does not replace a contract, legislation, a controlled standard, competent engineering judgement or specialist advice.

The operating aim is to turn a compact mathematical statement into a usable chain of definitions, claims, examples and checks. Read the original explanation first, then use the workflow and checks below to convert knowledge into evidence.

Treat The Cohen–Lenstra Heuristics as a network of definitions and implications, not as a list of formulas. The working vocabulary on this page—heuristics, section, cohen-lenstra, class, groups—should be made explicit before any proof or computation begins. Record the ambient set or structure, the permitted operations and the equality or equivalence relation in use. A compact theorem often changes meaning when the base field, finiteness condition, commutativity assumption or direction of an action changes.

For a proof, write the hypotheses as a checklist and mark the line at which each one is used. For a computation, state the representation of the input, the arithmetic model, the termination condition and the output invariant. For a classification problem, distinguish existence from uniqueness and distinguish an object from its representation. These separations prevent a correct local calculation from being mistaken for the general result.

A useful worked example should be small enough to inspect completely but rich enough to exercise the main mechanism. Compute the result in two ways where practical: symbolically and by substitution, structurally and numerically, or directly and through a normal form. Then include one near-miss example in which a hypothesis fails. The contrast explains why the theorem is shaped as it is and gives the reader a diagnostic pattern for later problems.

Verification is part of the mathematics. Check domains and codomains, substitute proposed solutions, test identity and zero cases, compare dimensions or cardinalities, and confirm that maps respect the required operations. In numerical work, report precision, conditioning and a residual rather than digits alone. In algorithmic work, separate mathematical correctness from implementation complexity and resource limits.

Step-by-step operating method

Fix the setting. State the objects, ambient structure, notation and assumptions before manipulating symbols.
Separate claims. Distinguish definitions, hypotheses, conclusions, equivalent conditions and consequences.
Choose a method. Select proof, construction, calculation or algorithm according to the question actually asked.
Work a small case. Use the smallest non-trivial example to expose the mechanism and test edge behaviour.
Verify independently. Substitute back, check invariants, test boundary cases or use an alternative derivation.

Worked-example protocol

Illustrative method—not a source theorem. Start with a small admissible input and list the definitions it must satisfy. Carry out each transformation on a separate line, citing the property that permits it. Preserve exact values until approximation is necessary. At the end, verify the output against the original definition and one invariant such as dimension, degree, determinant, order, norm or residual. Then alter one hypothesis and observe which step ceases to be valid. This protocol creates a reusable example without inventing a theorem-specific numerical answer.

Stage	Record	Quality check
Input	Objects, domain, notation, assumptions	Every symbol is defined
Method	Permitted operation or cited result at each step	All hypotheses hold
Output	Exact result and representation	Correct type, domain and form
Verification	Substitution, invariant or alternative derivation	Independent agreement
Boundary test	Zero, identity, degenerate or failed hypothesis	Scope is understood

Common failure modes and recovery actions

1. Watch for

Using a theorem without checking every hypothesis.

Recovery: Return to the governing definition or requirement and restate the decision in one sentence.

2. Watch for

Treating a suggestive example as a proof of the general case.

Recovery: Separate evidence from assumption, assign an owner and set a date for validation.

3. Watch for

Changing notation or conventions part-way through an argument.

Recovery: Run a small counterexample, boundary test, pilot or independent check before proceeding.

4. Watch for

Hiding a division-by-zero, convergence, finiteness or commutativity assumption.

Recovery: Record the consequence, decision and rationale, then update the controlled baseline.

5. Watch for

Reporting a computed result without a residual, substitution or structural check.

Recovery: Escalate when the issue affects safety, compliance, acceptance, material value or an agreed tolerance.

Review checklist

Can every symbol be traced to a definition or prior result?
Which hypothesis does each major step use?
Does the method cover zero, identity, degenerate and boundary cases?
Can the conclusion be checked by a second representation or calculation?
Are mandatory requirements distinguished from recommendations and illustrative values?
Are sources, assumptions, units, dates and versions recorded closely enough to reproduce the decision?
Have safety, legal, ethical, stakeholder and operational consequences been considered at the appropriate level?
Is there a named owner and a trigger for review, escalation, change or retirement?

Questions for deeper application

What is the most important distinction a practitioner must preserve when applying The Cohen–Lenstra Heuristics?