Bibliographic record
Abstract
In multiagent blind tournaments, many agents compete at an individual game, unaware of the performance of the other agents. When all agents have completed their games, the agent with the best performance--for example, the highest score, or greatest distance, or fastest time--wins the tournament. In stochastic games, an obvious and time honoured strategy is to maximize expected performance. In tournaments with many agents, however, the top scores may be far above the expected score. As a result, maximizing expected score is not the same as maximizing the chance of winning the tournament. Rather, a "riskier" strategy, which increases the chance of obtaining a top score while possibly sacrificing some expected score, may offer a better chance of winning. In this paper, we study how an agent should optimally adapt its strategy based on the size of the tournament in which it is competing. Our solution involves first approximating the agent's pool of opponents as a collection of known or estimated strategies. Second, score distributions are computed for those strategies, and the distributions are convolved to obtain a distribution for the maximum score of the agent's opponents. Finally, a strategy that maximizes the chance of exceeding the opponents' scores is computed. As a demonstration, we study optimal tournament-size adaptation in online Yahtzee tournaments involving as few as one and as many as ten thousand opponents. We find that strategies change dramatically over that range of tournament sizes, and that in large tournaments, an agent adopting the optimally risky strategy can nearly double its chance of winning.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.012 | 0.013 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".