PODC Review Process
Model the pipeline before interpreting any single review. PODC's process is proofs-first and
single-shot: a specialized program committee reads your model and your proof, and the decision is
accept or reject — there is no Major Revision, no journal-style R&R. The most consequential
mental shift for authors arriving from a software-engineering or ML venue is that there is no second
read to fix a gap in; everything that would go into a revision must be in the initial submission.
Process model
- Submission and review run on HotCRP with lightweight double-blind anonymity: identities
are hidden from reviewers (no author names/affiliations/emails in the submission), a change from
PODC's historically single-blind practice.
- Each paper is read by program-committee members (often with external subreviewers for specific
subareas) who weigh: the significance of the result, the correctness of the proof, the
precision and honesty of the model/assumptions, novelty against known bounds, and clarity.
- The committee reads the abstract and the first 10 pages as guaranteed; deeper proofs are read
at its discretion — so the merits case must land in those pages (
podc-supplementary).
- Decisions are accept / reject. Accepted regular papers publish in the ACM proceedings (≤10
pages); Brief Announcements (≤3 pages) are a separate, lighter-bar track.
- There is typically no revision round; whether a rebuttal / author-response phase runs
varies by cycle (待核实 for 2026 — confirm on the call).
What a PODC reviewer actually checks
| Dimension |
The reviewer's question |
How to pass it |
| Correctness |
Does the proof close in the stated model? |
Self-contained proofs; no assumption creep |
| Significance |
Does this advance a problem the community cares about? |
Position against the closest known bounds |
| Model honesty |
Are the timing/fault/adversary assumptions declared and consistent? |
An explicit model box the proofs never violate |
| Novelty |
Is the result / technique new relative to PODC/DISC/journals? |
A precise delta, not a topic list |
| Tightness |
Is an "optimal" claim backed by a matching lower bound? |
Prove or cite the lower bound, or don't claim optimal |
A reviewer who finds a proof gap or an undeclared assumption will reject on soundness, and
there is no revision to repair it — which is why podc-experiments (the assumption audit) and
podc-reproducibility (self-contained proofs) are pre-submission, not post-notification, work.
How PODC differs from its siblings
- vs. DISC: DISC is the other distributed-theory flagship (EATCS, LIPIcs proceedings, fall
cycle) and shares the Dijkstra Prize and Dissertation Award with PODC. The review culture is
close; the differences are calendar, proceedings format, and community center of gravity, not the
proofs-first bar.
- vs. SPAA: SPAA centers parallel computation; a message-passing fault-tolerance result is
native at PODC and a visitor at SPAA (and vice versa for shared-memory throughput/scheduling).
- vs. STOC/FOCS/SODA: these judge sequential algorithms and complexity; a PODC paper's value
must live in the distributed difficulty (faults, asynchrony, locality, concurrency) or it will
be read as a mislabeled sequential result.
- vs. software-engineering / ML venues: no artifact evaluation, no rebuttal-driven Major
Revision, no empirical-benchmark culture. The object of review is the proof.
Where author leverage actually exists
[Before submission] topic tags + a precise abstract -> reviewers who know your model (largest lever)
[Track choice] regular vs. Brief Announcement -> match the bar to the result's completeness
[In the submission] a self-contained proof and an honest model box -> pre-empt the soundness reject
[Rebuttal, if any] correct a factual misreading or supply a missing lemma/bound (待核实 it exists)
[After reject] no appeal; revise the proof and reroute to DISC/SPAA/a focused venue or a journal
Because there is no revision round, the leverage is almost entirely pre-submission. The single
highest-value action is closing every proof and declaring every assumption before 16 February.
Reading a decision
| Outcome |
What it means |
Author move |
| Accept |
Result, proof, and model hold |
Camera-ready (≤10 pages) + arXiv full version; do not weaken claims |
| Accept as Brief Announcement (if offered) |
The idea is valued but not a full regular paper this cycle |
Publish the ≤3-page BA; develop the full paper for a later venue |
| Reject — soundness |
A proof gap or undeclared assumption |
Fix the proof; do not resubmit unchanged |
| Reject — significance/novelty |
Known bound, or the delta is too small |
Strengthen (a matching lower bound?) or reroute |
Misreadings to avoid
- Expecting a revision round — there is none; the submission is the whole case.
- Treating a preprint as forbidden — arXiv posting is allowed; only concurrent submission
elsewhere is the dual-submission problem.
- Assuming a rebuttal exists — confirm per cycle (待核实 for 2026).
- Under- or over-anonymizing — follow the lightweight double-blind wording, neither leaking
identity nor stripping citations.
Output format
[Process stage] pre-submission / under review / (rebuttal if any) / decided
[Anonymity model] lightweight double-blind — submission clean of author identity?
[Reviewer-axis map] each review point -> correctness | significance | model-honesty | novelty | tightness
[Leverage plan] the pre-submission or (rebuttal) action that can actually change the outcome
[Forbidden moves] identity leak / claiming a revision round that does not exist / unsupported optimality
1---2name: podc-review-process3description: Use when reasoning about how an ACM PODC submission is evaluated — covering lightweight double-blind review, a program committee that reads proofs, the accept/reject decision with no journal-style revision round, the Brief Announcements track's lighter bar, and how PODC's process differs from DISC, SPAA, and the sequential-theory venues STOC/FOCS/SODA.4metadata:5 source: https://skillmd.com/skills/brycewang-stanford/podc-review-process6---78# PODC Review Process910Model the pipeline before interpreting any single review. PODC's process is **proofs-first and11single-shot**: a specialized program committee reads your model and your proof, and the decision is12**accept or reject** — there is **no Major Revision, no journal-style R&R**. The most consequential13mental shift for authors arriving from a software-engineering or ML venue is that there is no second14read to fix a gap in; everything that would go into a revision must be in the initial submission.1516## Process model1718- Submission and review run on **HotCRP** with **lightweight double-blind** anonymity: identities19 are hidden from reviewers (no author names/affiliations/emails in the submission), a change from20 PODC's historically single-blind practice.21- Each paper is read by program-committee members (often with external subreviewers for specific22 subareas) who weigh: the **significance** of the result, the **correctness of the proof**, the23 **precision and honesty of the model/assumptions**, **novelty** against known bounds, and clarity.24- The committee reads the abstract and the **first 10 pages** as guaranteed; deeper proofs are read25 at its discretion — so the merits case must land in those pages (`podc-supplementary`).26- Decisions are **accept / reject**. Accepted regular papers publish in the ACM proceedings (≤1027 pages); Brief Announcements (≤3 pages) are a separate, lighter-bar track.28- There is typically **no revision round**; whether a **rebuttal / author-response** phase runs29 varies by cycle (**待核实** for 2026 — confirm on the call).3031## What a PODC reviewer actually checks3233| Dimension | The reviewer's question | How to pass it |34|---|---|---|35| Correctness | Does the proof close in the stated model? | Self-contained proofs; no assumption creep |36| Significance | Does this advance a problem the community cares about? | Position against the closest known bounds |37| Model honesty | Are the timing/fault/adversary assumptions declared and consistent? | An explicit model box the proofs never violate |38| Novelty | Is the result / technique new relative to PODC/DISC/journals? | A precise delta, not a topic list |39| Tightness | Is an "optimal" claim backed by a matching lower bound? | Prove or cite the lower bound, or don't claim optimal |4041A reviewer who finds a **proof gap** or an **undeclared assumption** will reject on soundness, and42there is no revision to repair it — which is why `podc-experiments` (the assumption audit) and43`podc-reproducibility` (self-contained proofs) are pre-submission, not post-notification, work.4445## How PODC differs from its siblings4647- **vs. DISC:** DISC is the other distributed-theory flagship (EATCS, LIPIcs proceedings, fall48 cycle) and shares the Dijkstra Prize and Dissertation Award with PODC. The *review culture* is49 close; the differences are calendar, proceedings format, and community center of gravity, not the50 proofs-first bar.51- **vs. SPAA:** SPAA centers **parallel** computation; a message-passing fault-tolerance result is52 native at PODC and a visitor at SPAA (and vice versa for shared-memory throughput/scheduling).53- **vs. STOC/FOCS/SODA:** these judge **sequential** algorithms and complexity; a PODC paper's value54 must live in the *distributed* difficulty (faults, asynchrony, locality, concurrency) or it will55 be read as a mislabeled sequential result.56- **vs. software-engineering / ML venues:** no artifact evaluation, no rebuttal-driven Major57 Revision, no empirical-benchmark culture. The object of review is the **proof**.5859## Where author leverage actually exists6061```text62[Before submission] topic tags + a precise abstract -> reviewers who know your model (largest lever)63[Track choice] regular vs. Brief Announcement -> match the bar to the result's completeness64[In the submission] a self-contained proof and an honest model box -> pre-empt the soundness reject65[Rebuttal, if any] correct a factual misreading or supply a missing lemma/bound (待核实 it exists)66[After reject] no appeal; revise the proof and reroute to DISC/SPAA/a focused venue or a journal67```6869Because there is no revision round, the leverage is almost entirely **pre-submission**. The single70highest-value action is closing every proof and declaring every assumption *before* 16 February.7172## Reading a decision7374| Outcome | What it means | Author move |75|---|---|---|76| Accept | Result, proof, and model hold | Camera-ready (≤10 pages) + arXiv full version; do not weaken claims |77| Accept as Brief Announcement (if offered) | The idea is valued but not a full regular paper this cycle | Publish the ≤3-page BA; develop the full paper for a later venue |78| Reject — soundness | A proof gap or undeclared assumption | Fix the proof; do not resubmit unchanged |79| Reject — significance/novelty | Known bound, or the delta is too small | Strengthen (a matching lower bound?) or reroute |8081## Misreadings to avoid8283- **Expecting a revision round** — there is none; the submission is the whole case.84- **Treating a preprint as forbidden** — arXiv posting is allowed; only concurrent *submission*85 elsewhere is the dual-submission problem.86- **Assuming a rebuttal exists** — confirm per cycle (待核实 for 2026).87- **Under- or over-anonymizing** — follow the lightweight double-blind wording, neither leaking88 identity nor stripping citations.8990## Output format9192```text93[Process stage] pre-submission / under review / (rebuttal if any) / decided94[Anonymity model] lightweight double-blind — submission clean of author identity?95[Reviewer-axis map] each review point -> correctness | significance | model-honesty | novelty | tightness96[Leverage plan] the pre-submission or (rebuttal) action that can actually change the outcome97[Forbidden moves] identity leak / claiming a revision round that does not exist / unsupported optimality98```