REVIEW 1 major objections 3 minor 25 references
Meeting and coalescence times for random walks in the largest component of the Erd\H{o}s-R\'enyi random graph
T0 review · 1 major / 3 minor · reviewed 2026-08-02 · deepseek-v4-flash
Pith's one-line read Two independent random walks on the largest component of an Erdős–Rényi random graph meet in expected time of order n, the ambient vertex count, across strictly supercritical, slightly supercritical, and critical regimes; the same linear sc
desk verdict New Θ(n) meeting/coalescence results on the Erdős–Rényi giant across all three regimes; referee must check the unproved uniform expansion lemma. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central object is the random target time t_⊙(G)=|E(G)| Σ_{u,v} π(u)π(v) R_eff(u,v), which equals Kemeny's constant and is the degree-weighted average effective resistance. The proof bounds this quantity uniformly in each regime—using the core-and-attached-trees structure of the contiguous configuration-model surrogate for the supercritical giant, and using volume, diameter, mixing, and resistance estimates in the critical window—then converts it to a worst-case meeting-time bound via a general comparison inequality. The lower bound rests on the variational extremal characterization of hitting times: a low-energy test function that vanishes except when the two walks are close on the core
What would settle it
Find a sequence (n, ε_n) with ε_n^3 n → ∞ such that the degree-weighted average effective resistance of the largest component of G(n,(1+ε_n)/n) is not of order ε_n^{-1} while its edge count is of order ε_n n; then the random target time would not be O(n) and the meeting-time upper bound would fail. Numerically, computing E_{π⊗π} τ_meet for such sequences and checking that the ratio to n stays within fixed positive bounds would provide evidence; a clear divergence from linear scaling would refute the claim.
Extended reading notes
Core claim
The central claim is that with high probability, for the largest component C_1 of G(n,p) in any of the three regimes, c n ≤ E_{π⊗π} τ_meet ≤ t_meet(C_1) ≤ C n, where c,C are absolute constants in the supercritical cases and may depend on the window parameter in the critical case; a corollary extends the same order to expected coalescence and voter consensus time. The upper bound follows by showing that the random target time (Kemeny's constant) is O(n), using degree-weighted average effective-resistance estimates on the auxiliary configuration-model decomposition of the giant component, and then applying a general comparison between worst-case meeting time and the sum of mixing time and rand
Load-bearing premise
The supercritical proofs rely on transferring estimates from the auxiliary configuration-model decomposition to the actual Erdős–Rényi giant via contiguity; if a high-probability property of the auxiliary model fails to hold for the actual giant in the required direction, the linear meeting-time bounds would not apply to G(n,p).
Editorial extensions
If this is right
- Two-walk meeting time on the giant component is Θ(n) even at criticality, where the component has only n^{2/3} vertices, so the ambient scale n is the correct normalization.
- Coalescence of one walk per vertex and voter-model consensus inherit the same Θ(n) scale, so opinion dynamics on these networks reach consensus in linear time.
- The constants in the supercritical regimes are independent of the edge probability parameter over a wide range, indicating a robust universal scale.
- The proof strategy—resistance-based target-time bounds plus variational lower bounds—may transfer to other random graph models with a similar core-and-trees decomposition.
- The critical-window result closes the gap between known subcritical and supercritical behaviors, giving a complete picture across the phase transition.
Reading between the lines
- The linear meeting-time scale likely extends to other two-particle quantities, such as the expected number of meetings by time t, which would behave as t/n.
- A testable extension: in the supercritical regimes the meeting-time distribution may be approximately exponential with rate Θ(1/n) because the process mixes on a much shorter time scale.
- The resistance-based techniques could give sharp bounds on cover times or other additive functionals on the critical component, where the ambient n scale emerges from the edge count times resistance.
- The contiguity transfer suggests that any high-probability, isomorphism-invariant resistance or mixing estimate that holds in the auxiliary configuration model will also hold in the actual Erdős–Rényi giant, which may simplify future analyses of random-graph walks.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves that the stationary and worst-case expected meeting times of two independent continuous-time simple random walks on the largest component of the Erdős–Rényi random graph G(n,p) are of order n in three regimes: fixed supercritical (p=λ/n, λ>1), slightly supercritical (p=(1+ε)/n with ε→0 and ε^3 n→∞), and the critical window (|n^{1/3}(np_n−1)|≤A_0). The proof combines the DKLP/DLP contiguous configuration-model decomposition with electrical-resistance identities for Kemeny's constant, Aldous's meeting-time comparison, variational lower-bound test functions, and, in the critical window, the Nachmias–Peres estimates. The meeting-time bounds are then extended to coalescence times and voter-model consensus times via comparison results of Oliveira and Kanade–Mallmann-Trenn–Sauerwald.
Significance. If the result holds, it is a substantial contribution: it establishes a uniform Θ(n) scale for two-walk meeting times across the entire supercritical-to-critical transition, with constants that are absolute in the supercritical regimes, and it transfers this to coalescence and voter consensus. The paper is well organized, gives self-contained proofs of the electrical and variational estimates, and makes clever use of existing random-graph structure results. The main proof strategy is convincing and the parameter-uniform formulations are carefully stated. However, the upper bounds rely on an unproved 'uniform form' of an external lemma, which is load-bearing and must be resolved.
major comments (1)
- [Section 5, Lemma 10 (after Eq. (25))] The proof of the upper bound rests on the assertion that 'the uniform form of Benjamini–Kozma–Wormald [5, Lemma 5.3] applies: it covers every even degree sequence satisfying 3≤min_i d_i ≤ max_i d_i ≤ N^{0.02}.' No statement or proof of this uniform version is provided, and the cited reference is not shown to contain it. This is not a cosmetic issue: Lemma 10 is used to obtain the spectral-gap bounds (23)–(24), which drive Lemma 11, Lemma 13, Lemma 14, and Lemma 17, and hence the upper bounds in Theorem 1(i)–(ii). Please supply a proof of the uniform extension or an exact reference with a theorem statement. This is a missing-support concern and should be resolved before acceptance.
minor comments (3)
- [Section 5, proof of Lemma 10] The parity argument for the small-Λ range (0<Λ<1) is terse: the constants c, C in 'cΛ≤P(Y_Λ is odd)≤CΛ' are not tracked, and the relation between the conditioned-Poisson variable D_Λ and Y_Λ could be made more explicit. Since the lemma claims uniformity, please spell out the constants.
- [Section 8, proof of Lemma 28] The inequality 't^{ct}_{meet} ≤ δ t^Q_{meet} ≤ C(t^{ct}_{meet}+δ)' is stated without referencing the two displayed inequalities (108)–(109) from the proof of Lemma 26. Adding the explicit cross-references would aid readability.
- [General] The paper is long and the exploration argument in Lemma 15 is intricate. A short intuitive summary of the 'arrival-ticket' coupling, or a figure analogous to Figure 1, would help readers follow the proof.
Circularity Check
Meeting-time bounds are derived from independent structural and Markov-chain estimates; no prediction reduces to its fitted input or to a self-citation chain.
full rationale
The paper's central meeting-time result is not circular. The upper bound for the slightly supercritical giant is obtained by combining Kemeny's identity (Lemma 5), Aldous's meeting-time comparison (Lemma 6), and random-graph estimates: the DKLP auxiliary model, kernel expansion (Lemma 10), weighted core resistance (Lemma 13), and average target time (Lemma 14). These lemmas are proved from stated graph-theoretic and Markov-chain inputs, not from the meeting time itself. The lower bound uses the variational Lemma 7 with an explicitly constructed core-projection test function, whose stationary mean and Dirichlet energy are bounded via Lemmas 16 and 9; it does not presume the desired bound. The critical-window and fixed-supercritical regimes are handled by independent estimates from Nachmias–Peres [20] and by the DLP/DKLP decompositions, with no parameter fitted to meeting-time data. The coalescence/voter corollary is derived from meeting-time bounds through Oliveira's and KMS comparison inequalities, which are external results applied to the already-established meeting-time scale, not definitions of the meeting time in terms of coalescence. The paper's use of self-citations (DKLP, DLP, Nachmias–Peres) is to import prior structural theorems about component sizes, mixing times, diameters, and contiguity; these are independent of the meeting-time conclusion and do not constitute a uniqueness argument or an ansatz that presupposes the result. The only notable support gap is Lemma 10's invocation of a 'uniform form' of Benjamini–Kozma–Wormald [5, Lemma 5.3] for every degree sequence satisfying max degree ≤ N^{0.02}; the paper does not prove this uniform extension, and it is load-bearing for the kernel spectral gap. This is a question of unverified external support, not circularity, since even if the uniform form fails the argument does not reduce to assuming its own conclusion. No step was found in which a 'prediction' is equivalent by construction to a fitted input or to the defining assumption of an auxiliary model.
Assumptions & free parameters
assumptions (6)
- domain assumption DKLP contiguity theorem for the slightly supercritical giant [9, Theorem 2]
- domain assumption Strictly supercritical contiguous decomposition of Ding–Lubetzky–Peres [13, Theorem 1]
- standard math Mixing-time bounds for near-critical and strictly supercritical giants [12, Theorem 1; 5, Theorem 1.1]
- domain assumption Nachmias–Peres critical-window estimates [20, Theorems 1.1–1.2, Lemma 5.4, Propositions 5.5–5.7]
- standard math Comparison inequalities of Aldous, Oliveira, and Kanade–Mallmann-Trenn–Sauerwald [1; 22; 16]
- standard math Standard electrical-network and Markov-chain identities: commute-time identity, Kemeny constant resistance formula, Nash–Williams inequality, variational hitting-time characterization
Cite this review
Pith. "Pith review of Meeting and coalescence times for random walks in the largest component of the Erd\H{o}s-R\'enyi random graph." pith.science (2026). https://pith.science/paper/CV7MBUZT
@misc{pith2026260713183,
author = {Pith},
title = {Pith review of: Meeting and coalescence times for random walks in the largest component of the Erd\Hos-R\'enyi random graph},
year = {2026},
howpublished = {\url{https://pith.science/paper/CV7MBUZT}},
note = {Machine review of arXiv:2607.13183}
}
abstract
We prove that the stationary and worst-case expected meeting times of two independent continuous-time random walks on the largest component of the Erd\H{o}s-R\'enyi random graph $G(n,p)$ have order $n$ throughout the strictly supercritical, the slightly supercritical and the critical regimes. Using these bounds along with a fine-tuned combination of comparison inequalities due to Oliveira (2012) and Kanade-Mallmann-Trenn-Sauerwald (KMS, 2023), we deduce that expected coalescence time and full voter-model consensus also have order $n$ throughout these three regimes.
Figures
Reference graph
Works this paper leans on
-
[1]
D. Aldous,Meeting times for independent Markov chains, Stochastic Processes and their Applications38(1991), no. 2, 185–193. doi:10.1016/0304-4149(91)90090-Y
-
[2]
Aldous and J
D. Aldous and J. A. Fill,Reversible Markov Chains and Random Walks on Graphs, unfinished monograph, 2002; recompiled version, 2014. Available at https://www.stat.berkeley.edu/~aldous/RWG/book.html
2002
- [3]
-
[4]
M. T. Barlow, J. Ding, A. Nachmias and Y. Peres,The evolution of the cover time, Combinatorics, Probability and Computing20(2011), no. 3, 331–345. doi:10.1017/S0963548310000489. arXiv:1001.0609
arXiv 2011
-
[5]
I. Benjamini, G. Kozma and N. C. Wormald,The mixing time of the giant component of a random graph, Random Structures & Algorithms45(2014), no. 3, 383–407. doi:10.1002/rsa.20539. arXiv:math/0610459. 44
arXiv 2014
-
[6]
C. Cooper and A. M. Frieze,The cover time of the giant component of a random graph, Random Structures & Algorithms32(2008), no. 4, 401–439. doi:10.1002/rsa.20201. arXiv:0803.0929
arXiv 2008
-
[7]
C. Cooper, A. M. Frieze and T. Radzik,Multiple random walks in random regu- lar graphs, SIAM Journal on Discrete Mathematics23(2009), no. 4, 1738–1761. doi:10.1137/080729542
-
[8]
J. T. Cox,Coalescing random walks and voter model consensus times on the torus in Zd, Annals of Probability17(1989), no. 4, 1333–1366. doi:10.1214/aop/1176991158
arXiv 1989
Show all 25 references
-
[9]
J. Ding, J. H. Kim, E. Lubetzky and Y. Peres,Anatomy of a young giant component in the random graph, Random Structures & Algorithms39(2011), no. 2, 139–178. doi:10.1002/rsa.20342. arXiv:0906.1839
2011 arXiv
-
[10]
J. Ding, J. H. Kim, E. Lubetzky and Y. Peres,Diameters in supercritical random graphs via first passage percolation, Combinatorics, Probability and Computing19 (2010), no. 5–6, 729–751.doi:10.1017/S0963548310000301. arXiv:0906.1840
2010 arXiv
-
[11]
J. Ding, J. R. Lee and Y. Peres,Cover times, blanket times, and ma- jorizing measures, Annals of Mathematics (2)175(2012), no. 3, 1409–1471. doi:10.4007/annals.2012.175.3.8. arXiv:1004.4371
2012 arXiv
-
[12]
J. Ding, E. Lubetzky and Y. Peres,Mixing time of near-critical random graphs, Annals of Probability40(2012), no. 3, 979–1008.doi:10.1214/11-AOP647. arXiv:0908.3870
2012 arXiv
-
[13]
J. Ding, E. Lubetzky and Y. Peres,Anatomy of the giant component: The strictly supercritical regime, European Journal of Combinatorics35(2014), 155–168. doi:10.1016/j.ejc.2013.06.004. arXiv:1202.6112
2014 arXiv
-
[14]
van der Hofstad,Random Graphs and Complex Networks
R. van der Hofstad,Random Graphs and Complex Networks. Volume 1, Cambridge Series in Statistical and Probabilistic Mathematics, vol. 43, Cambridge University Press, Cambridge, 2017.doi:10.1017/9781316779422
2017 doi
-
[15]
T. E. Harris,Additive set-valued Markov processes and graphical methods, Annals of Probability6(1978), no. 3, 355–378.doi:10.1214/aop/1176995523
1978
-
[16]
Kanade, F
V. Kanade, F. Mallmann-Trenn and T. Sauerwald,On coalescence time in graphs: When is coalescing as fast as meeting?, ACM Transactions on Algorithms19(2023), no. 2, Article 18, 46 pp.doi:10.1145/3576900. arXiv:1611.02460
2023 arXiv
-
[17]
D. A. Levin and Y. Peres,Markov Chains and Mixing Times, 2nd ed., with contributions by E. L. Wilmer, American Mathematical Society, Providence, RI, 2017. doi:10.1090/mbk/107
2017 doi
-
[18]
T. M. Liggett,Stochastic Interacting Systems: Contact, Voter and Exclusion Pro- cesses, Grundlehren der Mathematischen Wissenschaften, vol. 324, Springer, Berlin, 1999.doi:10.1007/978-3-662-03990-8
1999 doi
-
[19]
Lyons and Y
R. Lyons and Y. Peres,Probability on Trees and Networks, Cambridge Series in Statistical and Probabilistic Mathematics, vol. 42, Cambridge University Press, Cambridge, 2016.doi:10.1017/9781316672815. 45
2016 doi
-
[20]
Nachmias and Y
A. Nachmias and Y. Peres,Critical random graphs: Diameter and mixing time, Annals of Probability36(2008), no. 4, 1267–1286. doi:10.1214/07-AOP358. arXiv:math/0701316
2008 arXiv
-
[21]
C. St. J. A. Nash-Williams,Random walk and electric currents in net- works, Proceedings of the Cambridge Philosophical Society55(1959), 181–194. doi:10.1017/S0305004100033879
1959 doi
-
[22]
R. I. Oliveira,On the coalescence time of reversible random walks, Trans- actions of the American Mathematical Society364(2012), no. 4, 2109–2128. doi:10.1090/S0002-9947-2011-05523-6. arXiv:1009.0664
2012 arXiv
-
[23]
R. I. Oliveira,Mean field conditions for coalescing random walks, Annals of Probability 41(2013), no. 5, 3420–3461.doi:10.1214/12-AOP813. arXiv:1109.5684
2013 arXiv
-
[24]
Penrose,A generalized inverse for matrices, Proceedings of the Cambridge Philosophical Society51(1955), 406–413.doi:10.1017/S0305004100030401
R. Penrose,A generalized inverse for matrices, Proceedings of the Cambridge Philosophical Society51(1955), 406–413.doi:10.1017/S0305004100030401
1955 doi
-
[25]
X. Wang, J. L. A. Dubbeldam and P. Van Mieghem,Kemeny’s constant and the effective graph resistance, Linear Algebra and its Applications535(2017), 231–244. doi:10.1016/j.laa.2017.09.003. 46
2017 doi
Reviewed August 2, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.