Good for Games Automata: From Nondeterminism to Alternation
Pith reviewed 2026-05-25 13:51 UTC · model grok-4.3
The pith
Alternating good-for-games automata recognize exactly the same languages as deterministic automata with the same acceptance conditions and indices.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We generalize all of these definitions to alternating automata and show their equivalence. We further show that alternating GFG automata are as expressive as deterministic automata with the same acceptance conditions and indices.
What carries the argument
The generalized GFG properties (history-deterministic, letter-game compliant, good for trees, good for composition) for alternating automata and the proofs that they coincide.
If this is right
- Alternating GFG automata over finite words are not more succinct than deterministic automata.
- Weak alternating GFG automata over infinite words are not more succinct than deterministic automata.
- Converting an alternating GFG Büchi or co-Büchi automaton to a deterministic one requires a 2^Θ(n) state blow-up.
- The question remains open whether alternating GFG automata with stronger acceptance conditions can be doubly-exponentially more succinct than deterministic ones.
Where Pith is reading between the lines
- The equivalence implies that, once restricted to the GFG subclass, alternation adds no extra expressive power over determinism.
- Game-solving algorithms that rely on deterministic automata could safely substitute alternating GFG automata without changing correctness, provided the acceptance condition is preserved.
- The single-exponential determinization bound for Büchi and co-Büchi cases suggests that the GFG restriction removes the usual extra exponential cost of alternation.
Load-bearing premise
Extending each GFG property from the nondeterministic setting to alternating automata preserves the original meaning and does not create unintended distinctions among the properties.
What would settle it
An alternating automaton that meets one generalized GFG definition but fails another, or an alternating GFG automaton whose language cannot be recognized by any deterministic automaton with matching acceptance condition and index.
read the original abstract
A word automaton recognizing a language $L$ is good for games (GFG) if its composition with any game with winning condition $L$ preserves the game's winner. While all deterministic automata are GFG, some nondeterministic automata are not. There are various other properties that are used in the literature for defining that a nondeterministic automaton is GFG, including "history-deterministic", "compliant with some letter game", "good for trees", and "good for composition with other automata". The equivalence of these properties has not been formally shown. We generalize all of these definitions to alternating automata and show their equivalence. We further show that alternating GFG automata are as expressive as deterministic automata with the same acceptance conditions and indices. We then show that alternating GFG automata over finite words, and weak automata over infinite words, are not more succinct than deterministic automata, and that determinizing B\"uchi and co-B\"uchi alternating GFG automata involves a $2^{\Theta(n)}$ state blow-up. We leave open the question of whether alternating GFG automata of stronger acceptance conditions allow for doubly-exponential succinctness compared to deterministic automata.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper generalizes several GFG properties (history-deterministic, letter-game compliant, good for trees, good for composition) from nondeterministic to alternating automata, proves their equivalence, and shows that alternating GFG automata recognize exactly the languages recognized by deterministic automata with the same acceptance conditions and indices. It further proves that alternating GFG automata over finite words and weak automata over infinite words are not more succinct than deterministic automata, establishes a 2^Θ(n) state blow-up for determinizing alternating Büchi and co-Büchi GFG automata, and leaves open whether stronger acceptance conditions permit doubly-exponential succinctness.
Significance. If the equivalences and expressiveness results hold, the work unifies disparate GFG characterizations in the alternating setting and clarifies that GFG alternation does not increase expressive power beyond determinism. The succinctness and determinization bounds supply concrete complexity information for finite-word, weak, Büchi, and co-Büchi cases. These contributions strengthen the automata-theoretic toolkit for games and verification.
minor comments (2)
- The abstract uses the term 'indices' in the expressiveness claim; a brief parenthetical reminder of the standard definition (e.g., parity index) in §2 would aid readers unfamiliar with the nondeterministic case.
- The open question on doubly-exponential succinctness for stronger conditions is stated clearly but would benefit from a short paragraph in the conclusion sketching why the current techniques do not extend.
Simulated Author's Rebuttal
We thank the referee for their positive assessment and recommendation to accept the paper. The summary accurately captures the main results on generalizing GFG properties to alternating automata, proving equivalences, and establishing expressiveness and succinctness bounds.
Circularity Check
No significant circularity
full rationale
The paper generalizes existing GFG definitions (history-deterministic, letter-game compliant, good for trees, good for composition) from nondeterministic to alternating automata, proves their equivalence via direct arguments, and establishes expressiveness equivalence to deterministic automata under matching acceptance conditions. These steps consist of standard automata-theoretic definitions and equivalence proofs with no reduction of results to fitted parameters, self-definitional constructions, or load-bearing self-citations. The abstract explicitly notes that prior equivalence was unproven, indicating the work supplies independent content rather than renaming or smuggling prior results by the same authors. No equations or claims in the provided text exhibit the enumerated circularity patterns.
Axiom & Free-Parameter Ledger
axioms (1)
- standard math Standard definitions of deterministic, nondeterministic, and alternating word automata together with acceptance conditions (Buchi, co-Buchi, weak).
Reference graph
Works this paper leans on
-
[1]
Miko aj Bojanczyk and Wojciech Czerwinski. An automata toolbox. Technical report, University of Warsaw, 2018
work page 2018
-
[2]
Why these automata types? In Proceedings of LPAR , pages 143--163, 2018
Udi Boker. Why these automata types? In Proceedings of LPAR , pages 143--163, 2018
work page 2018
-
[3]
Nondeterminism in the presence of a diverse or unknown future
Udi Boker, Denis Kuperberg, Orna Kupferman, and Micha Skrzypczak. Nondeterminism in the presence of a diverse or unknown future. In Proceedings of ICALP , pages 89--100, 2013
work page 2013
-
[4]
How deterministic are good-for-games automata? In Proceedings of FSTTCS , pages 18:1--18:14, 2017
Udi Boker, Orna Kupferman, and Micha Skrzypczak . How deterministic are good-for-games automata? In Proceedings of FSTTCS , pages 18:1--18:14, 2017
work page 2017
-
[5]
Alternation removal in B\"uchi automata
Udi Boker, Orna Kupferman, and Adin Rosenberg. Alternation removal in B\"uchi automata. In Proceedings of ICALP , pages 76--87, 2010
work page 2010
-
[6]
Udi Boker and Karoliina Lehtinen. Register games. Submitted to Logical Methods in Computer Science
-
[7]
On the way to alternating weak automata
Udi Boker and Karoliina Lehtinen. On the way to alternating weak automata. In Proceedings of FSTTCS , 2018
work page 2018
-
[8]
Good for games automata: From nondeterminism to alternation
Udi Boker and Karoliina Lehtinen. Good for games automata: From nondeterminism to alternation. arXiv:..., 2019
work page 2019
-
[9]
Solving sequential conditions by finite-state strategies
J Richard B \"u chi and Lawrence H Landweber. Solving sequential conditions by finite-state strategies. Transactions of the American Mathematical Society , 138:295--311, 1969
work page 1969
-
[10]
Deciding parity games in quasipolynomial time
Cristian S Calude, Sanjay Jain, Bakhadyr Khoussainov, Wei Li, and Frank Stephan. Deciding parity games in quasipolynomial time. In Proceedings of STOC , pages 252--263, 2017
work page 2017
-
[11]
The theory of stabilisation monoids and regular cost functions
Thomas Colcombet. The theory of stabilisation monoids and regular cost functions. In Proceedings of ICALP , pages 139--150, 2009
work page 2009
-
[12]
Forms of determinism for automata
Thomas Colcombet. Forms of determinism for automata. In Proceedings of STACS , pages 1--23, 2012
work page 2012
-
[13]
Fonctions r \'e guli \`e res de co \^u t
Thomas Colcombet. Fonctions r \'e guli \`e res de co \^u t. Habilitation \`a diriger les recherches, \'E cole Doctorale de Sciences Math \'e matiques de Paris Centre , 2013
work page 2013
-
[14]
Universal graphs and good for small games automata: New tools for infinite duration games
Thomas Colcombet and Nathana \"e l Fijalkow. Universal graphs and good for small games automata: New tools for infinite duration games. In FOSSACS , 2019
work page 2019
-
[15]
Universal trees grow inside separating automata: Quasi-polynomial lower bounds for parity games
Wojciech Czerwi \'n ski, Laure Daviaud, Nathana \"e l Fijalkow, Marcin Jurdzi \'n skit, Ranko Lazi \'c , and Pawel Parys. Universal trees grow inside separating automata: Quasi-polynomial lower bounds for parity games. In Proceedings of SODA , pages 2333--2349, 2019
work page 2019
-
[16]
Alternating weak automata from universal trees
Laure Daviaud, Marcin Jurdzinski, and Karoliina Lehtinen. Alternating weak automata from universal trees. In CONCUR , 2019
work page 2019
-
[17]
Solving games without determinization
Thomas Henzinger and Nir Piterman. Solving games without determinization. In Proceedings of CSL , pages 395--410, 2006
work page 2006
-
[18]
On determinisation of good-for-games automata
Denis Kuperberg and Micha Skrzypczak. On determinisation of good-for-games automata. In Proceedings of ICALP , pages 299--310, 2015
work page 2015
-
[19]
Relating word and tree automata
Orna Kupferman, Shmuel Safra, and Moshe Y Vardi. Relating word and tree automata. Ann. Pure Appl. Logic , 138(1-3):126--146, 2006. Conference version in 1996
work page 2006
-
[20]
A modal perspective on solving parity games in quasi-polynomial time
Karoliina Lehtinen. A modal perspective on solving parity games in quasi-polynomial time. In Proceedings of LICS , pages 639--648, 2018
work page 2018
-
[21]
Efficient minimization of deterministic weak omega-automata
Cristof L \"o ding. Efficient minimization of deterministic weak omega-automata. Information Processing Letters , 79(3):105--109, 2001
work page 2001
-
[22]
Alternating finite automata on -words
Satoru Miyano and Takeshi Hayashi. Alternating finite automata on -words. Theoretical Computer Science , 32:321--330, 1984
work page 1984
-
[23]
Expressiveness results at the bottom of the -regular hierarchy
Gila Morgenstern. Expressiveness results at the bottom of the -regular hierarchy. M.Sc. Thesis, The Hebrew University, 2003
work page 2003
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.