ITCS Workflow

ITCS runs on one annual cycle with a distinctive rhythm: submit in early September, get a yes/no in November with no rebuttal in between, and present the following January, with open-access publication in LIPIcs. There is no second round, no revise-and-resubmit, and no author-response phase, so the leverage is almost entirely before the deadline. Plan the year backward from that September date. Every date below is the ITCS 2026 snapshot (Bocconi, Milan); reopen the current call — see ../../resources/official-source-map.md.

The single-cycle calendar (ITCS 2026 anchor)

Because there is no rebuttal and no second deadline, treat the September upload as final in a way authors from rebuttal venues underestimate: the paper must answer the reviewer's objections on its own.

Working backward from the deadline

D-day (early Sep)      : full paper, complete proofs, anonymized, uploaded and re-downloaded
D-2 days               : abstract/registration + conflicts locked (earlier than the paper!)
D-1 to 2 weeks         : full version to arXiv/ECCC/ePrint (encouraged; lightweight double-blind)
                         final proof-completeness pass (no "proof omitted" on a central claim)
D-3 to 4 weeks         : freeze the model and definitions; polish the first-10-pages merits window
D-2 months             : decide ITCS vs STOC/FOCS/SODA on the novelty axis (itcs-topic-selection)

The two commonest self-inflicted failures: (1) missing the abstract/registration cutoff, which is ~2 days before the paper deadline — no registration, no submission slot; and (2) carrying a rebuttal-venue mindset that leaves objections for "later," when there is no later.

The no-rebuttal consequence, planned for

Since you cannot answer reviewers after they read, you must answer them inside the paper:

• Anticipate the "is this model empty/trivial?" question with an anchoring result up front.
• Anticipate the "is this actually new?" question with an honest related-work delta.
• Anticipate the "why should I care?" question with a first-page conceptual payoff.
• State scope and limitations yourself, so a reviewer does not raise them as your blind spot.

See itcs-author-response for how the absence of a rebuttal reshapes strategy, and itcs-writing-style for pre-empting objections in the prose.

After notification

• Accept: move to itcs-camera-ready — switch to lipics-v2021, de-anonymize, complete ACM CCS / keywords / funding, and hit the LIPIcs production deadline. Register at least one author to present. Consider submitting a Graduating Bits talk if a coauthor is near graduation.
• Reject: there is no resubmission to a later ITCS round this year (single cycle). Route the paper — often to STOC/FOCS/SODA on depth, or to a journal — via the table in itcs-topic-selection. The reviews, though not rebuttable, are usually substantive; use them to sharpen the next submission.

The Graduating Bits and community layer

ITCS is a smaller-community venue with a strong direction-setting culture. The Graduating Bits session — short talks by researchers near graduation (on either side) presenting results, plans, and themselves — is a fixture worth planning for if you or a coauthor is on the market. The conference is as much about seeding directions as reporting finished work, which is why bold, early ideas fare well here; budget time for the hallway and session conversations, not just the talk.

Reverify each cycle

• The exact abstract and paper deadlines, and the AoE/EDT timezone of the cutoff.
• Whether the current call still runs no rebuttal (the theory-venue norm, but confirm).
• The camera-ready deadline and the current LIPIcs class/version.
• The host, city, and dates — ITCS 2027 is 待核实 as of 2026-07-09; do not carry Bocconi forward.

Output format

[ITCS cycle status] on track / at risk
[Registration] title/abstract/authors/conflicts locked by the earlier cutoff? yes/no
[Paper] complete proofs present? anonymized? first-10-pages merits window done?
[Preprint] full version posted (arXiv/ECCC/ePrint)? yes/no
[Post-notification] accept -> camera-ready owner+date ; reject -> reroute target
[Calendar risks] <ordered, with dates before the early-September cutoff>