Grey Wolf Optimizer (GWO)

GWO is a swarm metaheuristic that mimics the social hierarchy and hunting of grey wolves (Mirjalili et al., 2014). The pack ranks its solutions: alpha (α) = best, beta (β) = second, delta (δ) = third, and omega (ω) = the rest, which update toward the geometric center of the three leaders. Source: andrey-dik's MQL5 implementation (extended to allow a configurable number of leader wolves beyond the canonical three).

How it works

Three hunting stages — search (random exploration), encirclement (refine around leaders), attack (converge) — are governed by a coefficient a decreasing non-linearly from 2 toward 0:

• Ai = 2·a·r1 − a, Ci = 2·r2 (r1,r2 random in [0,1]).
• When |a| > 1 wolves explore; as a → 0 they exploit, tightening on the best-found "prey." See exploration-vs-exploitation.

Benchmark (context-relative)

In this article's early scoring, GWO scored 0.4158, ranking 4th of 6 tested. Reasonable on smooth high-dim Skin, but weakest on the sharp-peaked Forest function and poor on discrete Megacity at scale (perfect 1.0 at 2 vars → 0.022 at 1000). Dik notes "low scalability on discrete and non-differentiable functions." Earlier, small-field methodology — not directly comparable to the mature %-of-MAX rating (see population-optimization-benchmark).

The metaphor question

GWO (alpha/beta/delta/omega wolves, 2014) is a textbook case for the spoke's standing suspicion about metaphor inflation. The authoritative wikipedia-metaheuristic article records the Sörensen critique: many recent metaphor-based methods "hide their lack of novelty behind an elaborate metaphor," with "flaws includ[ing] vagueness, lack of conceptual elaboration, poor experiments, and ignorance of previous literature" — a charge frequently leveled at the wolf/bee/ant wave GWO belongs to. Stripped of the pack story, GWO's update toward the centroid of three leaders, gated by a coefficient a decaying 2→0, is a standard exploration-vs-exploitation schedule; the term "metaheuristic" itself traces to Glover, 1986 (tabu-search), long predating the metaphor wave. The benchmark verdict backs the skepticism — GWO's mid-pack-to-poor scaling is no better than the metaphor-free classics.

Why T2 (re-tiered 2026-08-25)

Held at T3 since ingest, which graded the venue — MQL5, a broker's community article site — rather than what the page rests on. This page's load-bearing claims are a mechanism description of an algorithm published by Mirjalili, Mirjalili and Lewis (2014) and a comparative score, both from andrey-dik, who did not write the algorithm and has no stake in where it lands. That is somebody measuring somebody else's work, which QUALITY.md's ladder puts at T2.

The spoke's standing finding — that in this field the scorer is usually the author — is real and does not apply here; it applies to DOAm, which stays T3 because there Dik is scoring his own modification. The confound that does remain is implementation: Dik re-implements each optimizer in MQL5, so his figure grades his code as much as the method, and his scoring scheme changed mid-series (population-optimization-benchmark). Both stay flagged in prose, where a methodological limit belongs — not in the tier.

Note the floor is unmoved by this: T1 ÷ tiered counts T1 against the same denominator either way, so the change alters the tier mix and not the spoke's floor of 0.18.

Related

metaheuristic-optimization · exploration-vs-exploitation · population-optimization-benchmark · no-free-lunch-theorem · andrey-dik · particle-swarm-optimization · artificial-bee-colony · evolution-strategies · wikipedia-metaheuristic