notes.Rmd

---
title: "MS notes etc"
author: "Jeroen Minderman"
date: "10/08/2021"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

Thus, particularly for cases involving complex socio-ecological interactions involving both natural resource dynamics as well as human behaviour, models can quickly become highly complex and technically challenging to communicate to non-specialist audiences.

It is clear that to produce models that are better able to represent complex real-world socio-ecological settings, we need a better means to both parameterise the human decision-making components in them, as well as improved means to communicate such models.

-   Models general intro

    -   Need for SE models

    -   Issues with understandability/communication/application

    -   Issues with data for parameterisation

-   Introduce GMSE [@duthie2018], basics, application, issues?
    Illustrates above issues.
    Here or later?

-   Games as novel means to link models, parameterisation, application; "games as models"

    -   General use of games in science

    -   Lack of use in terms of collecting "big data" for parameterisation.

    -   (Need to weave in idea of "big data" sourcing of data in highly variable/multidimensional settings, as opposed to "single problem", highly contained/controlled settings. "Open" games ideal for this)

    -   Parallel opportunity to improve understandability/communication.
        "Double whammy" of communication and gathering data for parameterisation.

\*General point by @grimm2020 - models often misunderstood/misused by decision makers but unclear "what is in them" - highligting need for clear description/communication.
Interactivity/game approach allows for effective way to highlight what model is, and what it does/does not do?
Also, interactivity means @grimm2020 Q3 ("is there evidence the model works") can be addressed - directly showing match of model output with empirical observations.
\*

-   Overall aim, etc.

Games (specifically, video games) have been used extensively in research settings [as a communication tool and to collect data], including for socio-ecological research and modelling, but both practical implementations and proposals for such "serious games"[^1] are often primarily or exclusively focused on educational or promotional angles.

For example XXX, XXX, XXX. Where games have been used to collect data, as far as we are aware this has been solely focused on empirical data collection for a specific purpose/research question as opposed to allowing player-users to interact with a model in a more open setting.
*[INCLUDE NOTE ON 'VIRTUAL LABORATORIES']*

Thus, casting models as games, and games as interactive models, not only has the potential to improve the communication and use of complex models in applied settings, it can further provide a platform for collecting data on human decision-making in complex socio-ecological settings, potentially allowing for a "big data" way of improving model parameterisation.

[^1]: We dislike and would discourage the use of this phrase.
    What is "serious"?
    It implies a bias against playability and "fun", which arguably is a key reason for using games in science in the first place - if it is not fun, it is arguably simply an interactive survey with participants.