Deep Player Behavior Models: Evaluating a Novel Take on Dynamic Difficulty Adjustment
The DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Abstract
Finding and maintaining the right level of challenge with respect to the individual abilities of players has long been in the focus of game user research (GUR) and game development (GD). The right difficulty balance is usually considered a prerequisite for motivation and a good player experience. Dynamic difficulty adjustment (DDA) aims to tailor difficulty balance to individual players, but most deployments are limited to heuristically adjusting a small number of high-level difficulty parameters and require manual tuning over iterative development steps. Informing both GUR and GD, we compare an approach based on deep player behavior models which are trained automatically to match a given player and can encode complex behaviors to more traditional strategies for determining non-player character actions. Our findings indicate that deep learning has great potential in DDA.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Previous studies, such as [11], [20], [37], [43], address this issue by utilizing human gameplay data to define feasible behavior targets.”
From UBCL: A Reinforcement Learning Framework for Controllable and Diverse Player Behaviors (2026) · cites this paper as [11]
Citations
Cites
Mastering the game of Go with deep neural networks and tree search
Silver, D., Huang, A., Maddison, C. J., Guez, A., Sifre, L., van den Driessche, G., Schrittwieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., et al. (2016) · Nature · doi:10.1038/nature16961 · Cited by 15,612 · cited 1×
“In the meantime, machine learning approaches in video game playing that harness continual improvement through simulated play [9] have become popular.”
BackgroundMcAuley, E., Duncan, T., & Tammen, V. V. (1989) · Research Quarterly for Exercise and Sport · doi:10.1080/02701367.1989.10607413 · Cited by 2,694 · cited 1×
“…each specific enemy type, we asked for subjective assessments how strong and how balanced the particular opponent appeared, captured the player experience using the Intrinsic Motivation Inventory [5] (all 7-point Likert scales) and asked players to explain the opponent behaviors in their own words.”
MethodThe case for dynamic difficulty adjustment in games
Hunicke, R. (2005) · Proceedings of the 2005 ACM SIGCHI International Conference on Advances in computer entertainment technology · doi:10.1145/1178477.1178573 · Cited by 441
Glued to Games: How Video Games Draw Us In and Hold Us Spellbound
Rigby, S., & Ryan, R. M. (2011) · doi:10.5040/9798400658105 · Cited by 362
Glued to games: how video games draw us in and hold us spellbound
(2011) · Choice Reviews Online · doi:10.5860/choice.49-0099 · Cited by 333 · cited 1×
“Based on self-determination theory [7], we hypothesize that opponents deploying DPBMguided strategies yield high results in interest-enjoyment, due to displaying convincing, but not rigidly perfect strategies, while tension-pressure might be increased and perceived-competence might be decreased when facing near-optima…”
BackgroundDynamic Difficulty Adjustment in Computer Games Through Real-Time Anxiety-Based Affective Feedback
Liu, C., Agrawal, P., Sarkar, N., & Chen, S. (2009) · International Journal of Human-Computer Interaction · doi:10.1080/10447310902963944 · Cited by 259 · cited 1×
“In order to estimate the discrepancy between challenge and skill, various assessment techniques have been researched, such as success probability estimation [3] or biofeedback [4].”
MethodVicencio-Moreira, R., Mandryk, R. L., & Gutwin, C. (2015) · Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems · doi:10.1145/2702123.2702242 · Cited by 70
Towards Deep Player Behavior Models in MMORPGs
Pfau, J., Smeddinck, J. D., & Malaka, R. (2018) · Annual Symposium on Computer-Human Interaction in Play · doi:10.1145/3242671.3242706 · Cited by 49 · cited 2×
“Based on insights about expressive data and suitable modeling techniques from our earlier work [6], we recorded all crucial player action decisions ( attacking or switching with or to the respective element and jumping ) together with contextual data from the current situation (cf. Table 2).”
MethodSelf-citationMeasuring player skill using dynamic difficulty adjustment
Demediuk, S., Tamassia, M., Raffe, W. L., Zambetta, F., Mueller, F., & Li, X. (2018) · Australasian Computer Science Week Multiconference (ACSW) · doi:10.1145/3167918.3167939 · Cited by 18 · cited 1×
“DDA has developed from flow maximization [3] over multi-player balancing [11] up to a tool for proficiency estimation [2].”
Background
Research works printed but not linked: 1
- Alexander Streicher and Jan D . Smeddinck . 2016 . Personalized and Adaptive Serious Games. In Entertainment Computing and Serious Games, Ralf Dörner, Stefan Göbel, Michael Kickmeier-Rust, Maic Masuch, and Katharina Zweig (Eds.). Lecture Notes in Computer Science, Vol. 9970 . Springer International Publishing , Cham, 332--377. Alexander Streicher and Jan D. Smeddinck. 2016. Personalized and Adaptive Serious Games. In Entertainment Computing and Serious Games, Ralf Dörner, Stefan Göbel, Michael Kickmeier-Rust, Maic Masuch, and Katharina Zweig (Eds.). Lecture Notes in Computer Science, Vol. 9970. Springer International Publishing, Cham, 332--377.
Cited by
Dungeons & Replicants: Automated Game Balancing via Deep Player Behavior Modeling
Pfau, J., Liapis, A., Volkmar, G., Yannakakis, G. N., & Malaka, R. (2020) · 2020 IEEE Conference on Games (CoG) · doi:10.1109/cog47356.2020.9231958 · Cited by 48
“It was trained over 1000 epochs, based on insights from previous work [16]–[18], [42]; benchmarks prior to the study also indicated diminishing returns when further increasing the range of parameters.”
MethodSelf-citationPfau, J., Smeddinck, J. D., & Malaka, R. (2020) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376423 · Cited by 55
“In previous work, we were successful in showing that player model agents can yield significantly higher motivation compared with heuristic opponents in a short-term online study using the 2D platform fighter Korona:Nemesis [9, 39].”
BackgroundMethodSelf-citationThe Arousal Video Game AnnotatIoN (AGAIN) Dataset
Melhárt, D., Liapis, A., & Yannakakis, G. N. (2022) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2022.3188851 · Cited by 67
“predictive models of player behaviour [25], [26], playtime [27], churn [28], [29], or player experience [5], [9], [30].”
BackgroundPfau, J., Liapis, A., Yannakakis, G. N., & Malaka, R. (2022) · IEEE Transactions on Games · doi:10.1109/tg.2022.3167728 · Cited by 26
“The network was initialized randomly, contained four hidden ,basedoninsightsfrompreviouswork[16]–[18],[50], [51];benchmarkspriortothestudyalsoindicateddiminishing returnswithmoreparameters(moreandwiderlayers)and epochs.”
MethodSelf-citationBot or not? User Perceptions of Player Substitution with Deep Player Behavior Models
Pfau, J., Smeddinck, J. D., Bikas, I., & Malaka, R. (2020) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376223 · Cited by 20
“In previous work, we were successful in showing that player modeling agents yield significantly higher motivation potential than heuristic opponents [36].”
BackgroundMethodInfluentialSelf-citation
Metadata
- Authors
- Pfau, J., Smeddinck, J. D., & Malaka, R.
- Affiliations
- University of Bremen (DE)
- Newcastle University (GB)
- Venue
- Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems
- Publisher
- ACM
- Published
- 2019-05-02
- Pages
- 1-6
- OpenAlex
- W2942569678
- Cited by
- 17
- References
- 9
- Citations by year
- 2026: 2 · 2025: 2 · 2024: 1 · 2022: 2 · 2021: 4 · 2020: 6
- FWCI
- 1.15
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) (publishedVersion)
- Funders
- DFG/German Research Foundation (1320)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence in Games
- Social Sciences › Sociology and Political Science › Digital Games and Media
- Social Sciences › Psychology › Developmental and Educational Psychology › Educational Games and Gamification
- Keywords
- dynamic difficulty adjustment, player behavior modeling, deep learning, player motivation, player experience, non-player character behavior, game development
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Deep Player Behavior Models: Evaluating a Novel Take on Dynamic Difficulty Adjustment. https://app.cochat.ai/papers/10.1145/3290607.3312899
- The paper
- Pfau, J., Smeddinck, J. D., & Malaka, R. (2019). Deep Player Behavior Models: Evaluating a Novel Take on Dynamic Difficulty Adjustment. Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems. https://doi.org/10.1145/3290607.3312899
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.