Recommender System for Items in <i>Dota 2</i>
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Dota 2 is one of several multiplayer online battle arena games that have recently become extremely popular. A central feature of Dota 2 is that players select and purchase items that are used in the game and the selections strongly affect the outcome of the game. Item recommendations and purchase predictions in Dota 2 turn out to be interesting problems due to the many factors affecting the choice of items and due to the frequency by which items are purchased throughout the game. We present three recommender systems for recommending which items a player might buy throughout a match, based on commonly used purchasing strategies. These systems are a rule-based system, a logistic regression based system, and a logistic regression based system enhanced with clustering. Knowing only the player's hero and inventory items, the first two systems are able to predict the purchases by highly skilled players with accuracies ranging from 66.5% to 87.1%, depending on the hero where accuracy is defined as recommended items that were actually purchased in the next 5 min of the game. For the third system, the players are clustered by purchasing strategy which in turn group players by role and play style. Adding the clustering feature to the second version of the recommender system only improve the results slightly. The reason for this could be that the items in the player's inventory are already a strong indicator of a player's role and play style both of which were used to develop the first two recommender systems. An important question for future work is how the recommendations could be useful in practice for human or AI players.
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.
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 | 0.000 | 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 | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it