Predicting behaviour with digital simulations

Tuesday, 20. September 2016
Joost Modderman, GITP, Rotterdam

In the science fiction movie The Minority Report (Spielberg, 2002), based on the equivalent novel written by Philip Kindred Dick (1956), near-future criminal violations were being predicted via mutated humans (see Fig. 1). The police department’s PreCrime division was able to intervene with this foreknowledge just seconds before the offense (see trailer:


The novel Minority Report
Figure 1 - The novel Minority Report


While predicting future behaviour is still just not possible, it has been the leitmotif in Human Resource Management for decades. Creating forecasts for ones’ job performance using dedicated psychological instruments has become the standard procedure in the recruitment and selection practice to reduce risks and maximise performance. America’s Army Alpha test (see Fig. 2) marked the beginning of predicting job performance. This survey was developed to measure abilities of those joining the troops to fight in World War I.


Testing on a large scale with the Army Alpha
Figure 2 - Testing on a large scale with the Army Alpha (source: Google)


These so-called self-report instruments are quick and easy to administer. Nevertheless, downsides include self-report biases and faking. Recruiters and psychologists started using role-plays to complete the image of ones’ behaviour to overcome these disadvantages. Actors and observers came in play to evoke and observe ones’ behaviour as a work sample.

A new era of predicting work behaviour in Human Resource Management has arisen. A prime example in line with the aforementioned Army Alpha is an instrument nowadays used to recruit soldiers in the United States: a video game.


Screenshot America's Army - Proving Grounds
Figure 3 - Screenshot America's Army - Proving Grounds (source:


A video game like America’s Army typically let’s players move freely and decide how they want to behave. Based on the data logged and a pattern analysis, conclusions can be drawn on what to expect from player’s real-life performance. Although not applicable to any other field than war, it is a powerful illustration of the direction sampling behaviour is going.

Over the last 10 years, GITP has gained experience in developing and validating digital simulations specially for the Human Resources (HR) market. Although the term ‘serious games’ is often used, applied games as connotation is preferred. In the field of HR, the selection field in particular, applied HR games use game techniques (i.e. storytelling, freedom of choice, video characters) to enrich and simulate a working context. In this context, players are unaware of the measuring subject and are drawn into a story. Instead of reporting about one’s own behaviour, the player makes choices to complete an objective. This behavioural sample is of much value to the individual and the organisation. It is proven to be well suited to inquire complex skills and behaviours that are not easily grasped via traditional questionnaires. For example, in measuring compliant behaviour, a game incrementally predicts job behaviour and is less susceptible to social desired answering (Dubbelt et al., 2014). By using a game as an effective assessment tool, the DEVELOP platform is able to grasp the behavioural part of transversal competencies that can help predict job success, career paths and learner development. The player on the other hand is given a memorable story with lots of triggers for further development. It is the fun-element or the face validity of games that appeal to players most. 

Some challenges lie ahead. Games and simulations for assessing behaviours are typically complex to design and difficult to analyse. The development requires multidisciplinary teams and a highly iterative design process. Further, the production of high quality video material is time consuming and expensive. Their scalability and application outside their intended market is often limited. A lot of these challenges are met with a blue-print functioning as a solid framework. The key to successfully design applied simulations is ensuring that the game is based on a theoretical foundation consisting of existing psychological models with proven value.


“In order to design successful, effective serious games it is of pivotal importance that psychological paradigms are integrated in such a manner that adequate mental processes in players are provoked, while safeguarding engaging game play that is fun, i.e. possesses good flow.”  - Van Nimwegen, 2012


Currently, the creation of the framework is work in progress. The context and setting-the-scene are further conceived and tested within the consortium.

The game format is a so called first-person-narrative-game. That format is proven to be successful in capturing real behaviour without the need of real-life actors in a role play. In a first-person-narrative-game the player directly interacts with characters in the game as if the player was present in the scene. The player is part of the story. The story is realistic as possible. The actors are real filmed actors. The viewpoint of the camera is as if the player is in the movie scene.

This game type is driven by a pre-scripted scenario. Players can steer the course, the events and the outcomes of the game by choosing their own preferred answers. They can directly see the effects of their choices due the continuing story. Therefore, by selecting their preferred answer, the players can influence the plot of the game. This enhances the immersiveness and the perception of control. The main storyline is linear, the dialogs with the actors have side steps and branching. Whilst playing, the story unfolds and there are several outcomes possible.

The end-result will encompass a state-of-the art work simulation, to let the DEVELOP platform function similar to the Minority Report’s PreCrime division: predict the best next step an employee could take in his career.



Dubbelt, L., Oostrom, J. K., Hiemstra, A. M., & Modderman, J. P. (2015). Validation of a digital work simulation to assess machiavellianism and compliant behavior. Journal of Business Ethics, 130(3), 619-637.

Van Nimwegen, C., van Oostendorp, H., & Wouters, P. (2012). Game DNA: A method to structure player actions in serious games. In Proceedings of the 6th European conference on games based learning (pp. 514-XIX). Reading: Academic Conferences International Limited.