This manuscript provides a very interesting review of the process used to develop “serious” games for social learning among stakeholders in two different socio-hydrological systems in Indonesia: a mountain slope leading to lowland paddies, and a peatland “dome”. The games were developed based on hydrological studies using the Drivers, Pressure, State, Impact, and Responses (DPSIR) framework as well as analysis of the actors and stakeholders as well as the Actors, Resources, Dynamics, and Interaction (ARDI) framework. The introduction provides a good explanation of how games can influence action around water. It is particularly interesting how the study use the credibility, salience, and legitimacy framework to evaluate the two games.

The abstract concludes that “We provide clear steps in designing and adapting the game to another area…”. This is the area where more is needed to deliver on this promise.  As currently laid out, there is not enough information about what kind of hydrological or socio-economic study is needed to adapt such a game to new contexts. Was this based on quick assessment or multi-year study of the two areas described?  What number of game simulations and what number of actual players are needed? The other gap in the paper is a description of who played the game, and whether there were differences in how different types of players responded in the game, or in their interactions with each other.  For example, were all the players men? Did the players from upstream play differently than those from downstream areas, even if they were not playing the parts of their own area?

Other more detailed questions or suggestions:

Line 173 says “profit is total income minus total capital”.  But if income is on an annual or seasonal basis, shouldn’t that be the annualized cost of the capital (e.g. if there is a major outlay for pumps)?  Or should that be “minus total costs” (which is what it says in the next sentence).  In economic terms, there is a difference.  

Figure 4B X axis is labelled Amount of groundwater), but shouldn’t that be amount of surface water, or runoff? 

Figure 4C and D, what does it say that the actual choices bby the participants were so much below the simulated income, and mostly lower groundwater and runoff? 

Figure C1: how do the villages match the peat dome?

The paper needs a good copy editor throughout ,including the appendix.

In the manuscript, the authors provide a description of the H2Ours serious game developed and tested on two locations in Indonesia with an opinion on the possibility of adapting this game to other areas and conditions. While the reasons for developing such a game is clearly described and explained, the rules of the game and the flow of the game are not so clear for the reader. I found especially hard to follow so many subtitles in sections 2 and 3 (Methods and Results) that interrupt the reading flow and consequently the understanding of the game. Moreover, it is not clearly stated who should be the target group of players (students, farmers, general public etc.).

For me as the hydrologist, the content of section 2.3 about game solution space analysis is not described clearly enough, more specifically, how did you produce random choices (e.g., using some software, etc.).

Also related to the rules and flow of the game, it is not clearly described how and when the models shown in Figure 6 and Figure C3 take place in the game. Please clarify.

In line 119 it is not clear what kind of values represent discharges 5 and 3,5 m³/s (average in the mentioned year, some long-term average, something else?). Please clarify.

Table 1 in my opinion is too long/big and similarly as multiple subsections break the reading flow. I would suggest adding some summary into the text and moving the table into appendices.

In line 231, a brief explanation about both methods, i.e. Likert scale and q-method, needs to be added.

In Figure 4, Figure D1, Figure D2, and Figure D3 it is not clear what are presenting solid blue, green, and red lines. Additionally, “ml” in legends should be replaced with “masl”. In relation to these figures, why are there different thresholds used (e.g., 200, 800) than explained in lines 114–116?

In Figure 7, there are missing y-axis titles. Please add.

Please use en-dash throughout the manuscript in case of ranges and periods. E.g., 0–100 masl instead 0 – 100 masl.

Appendices should be mentioned in text in the order in which they appear, e.g., Appendix A before Appendix B.

Thank you for sharing this interesting work with me. First of all, please let me apologize if I have not properly understood those elements of the paper on which I give comments, or if my perception was wrong.

The paper documents the process of designing and applying a serious game based on a system simulation model. The application of serious games as a contribution to sustainable development receives increasing attention also raising high expectations. The presented work documents a game design process and intends to provide evidence on outcomes from playing the game. It can help to better understand the potentials and limitations of the approach and can meaningfully inform future serious game design processes. I strongly support the publication of the paper. Having said this, I want to share some thoughts which I hope can help to improve the paper.

The study has two objectives:

• Develop an adaptable serious game
• Assess if game facilitates knowledge transfer and sharing, and supports negotiation and coordination.

I am not sure whether the paper really addresses these two objectives. Either rethink your objectives or structure the paper more strongly around the ones you formulated. You documented the game development process which I consider an important contribution. It is difficult for me to judge how easy it is to adapt the developed game to a different context. The paper does not really make this clear.

I am even more concerned about the second research question. The methods section does not explain the approach in sufficient detail. An input-output process is mentioned as methodology without explaining it. You state that information was collected throughout the game development and implementation process. How was this information collected and analyzed? You mention a post-game survey. Please provide more information on: 1) who was interviewed and how have respondents been selected, 2) how many respondents were interviewed, 3) how was the survey implemented (also addressing the strong risk of social desirability bias), 4) when was it conducted?

For answering the second research question, can you share information on the decisions of players in the game? Do you see patterns and changes over the rounds? Can you share information on what has been discussed during the game sessions?

In this context, I recommend giving more weight to Section 3.3. Interesting statements are made regarding the game dynamics. It could be combined with some statements included in the discussion (e.g. that setting up the game has built emotions). The value of the information depends, however, on the transparency of how information was generated.

On a conceptual level, I am not convinced that collecting data on credibility, salience/relevance, and legitimacy can answer the second research question. This would at least require stronger justification. Going a step further, I do not think that the survey questions are adequate indicators for Belcher’s et al. criteria. For instance, asking a player whether she sees the possibility to apply the knowledge acquired in the game does not provide information on how inclusive the process has been. I would propose a more differentiated approach regarding how you provide evidence on different criteria. Players will not be able to give meaningful answers to some of them. For instance, by playing the game, a player does not need to be aware of the underlying theory of change. I would move Table 1 to an appendix to make space for a better explanation of the methodology and the key results. 

A clear theory of change is one of the evaluation criteria and it needs to be expressed in a more explicit way. In lines 42ff, the authors argue that knowledge supports commitment (intention?) supports responsibility, supports innovation. It is not clear which type of knowledge, commitment, responsibility, and innovation the game intends to support. The learning outcomes are formulated rather vaguely. In the model framework, the authors include the assumption that income drives decisions. In the discussion, you argue that the enabling environment is the key driver and no behavioral change can be expected if regulations etc. are not adjusted. Considering all these thoughts and assumptions, what did you intend to change by letting stakeholders play the game?

Based on the provided information, it is difficult to understand the structure of the game. Key aspects are not explained, such as:

• Number of players per session; who was selected to play the game (relevant also for theory of change)
• Types of roles; in section 3.2.2 is described that there are two roles namely communities and a multi-stakeholder forum. But then one finds statements that there are upstream and downstream actors, different villages etc. Is the role definition linked to the diagnosis? Who in Table 2 creates pressure? Who is involved in the responses and dynamics? Are the relevant actors included as roles? Such game development decisions are critical for the theory of change.
• Which decisions can be taken by which player role? Table 3 and Figure 3 give the impression that only crop and forestry related choices can be made. But then it is mentioned that players can block channels. The choices of the multi-stakeholder forum are formulated very unclear and are not included in Figure 3. How could the forum make regulations or programs to prevent environmental programs. Please also better explain how you decided on the players’ choices in the game. In the discussion is mentioned that the game could not capture the complexity of the system and that few issues had to be selected to be included in the game. Please describe the process from diagnosis to selecting the issues to be included in the game. Such game development decisions are again critical for the theory of change.
• Are decisions taken collectively (a group represents one role and decides based on some agreement) or individually, creating collective outcomes. In line 433 is stated that the game demonstrated the value of collective action. How is collective action implemented in the game?
• Did the players receive real money? If yes, how was the payout determined? Did it depend on the game dynamics as is common in behavioral games? If not, make clear in Section 3.2.3 that you talk about a game endowment/play money.
• A more comprehensive explanation of outcomes resulting from choices would be important for understanding the game. For instance, in Table 3, can you provide likelihoods of flooding, water shortage, and fire events for different land use options? The choices in Table 3 indicate that unsustainable options provide highest income. At the same time, your model (Figure 3) works with the assumption that income mainly drives choices. How do you address the risk of unsustainable learning outcomes? A player’s answer in the evaluation survey that she can apply the game learnings could mean that she learnt that an upstream-all-crop strategy is best.

I do not understand the step of the game solution space analysis. What does it mean that the solution space was explored based on 3, 10, 30, 100, 300, and 1000 games with random choice? How did these 1000 games differ? Just explain this better for readers not familiar with your modelling and game development approach.

Line 74f: I challenge the statement that models and games have rarely been combined. There are numerous examples for instance on the website https://games4sustainability.org/gamepedia/. I am also aware of the following study: Lohmann, D., Falk, T., Geissler, K., Blaum, N., & Jeltsch, F. (2014). Determinants of semi-arid rangeland management in a land reform setting in Namibia. Journal of Arid Environments, 100, 23-30.

Line 80f: As a complementary approach, the authors may want to consider providing guidance on context conditions which need to be fulfilled for meaningfully using the games. Also, can you estimate the size of the areas which share socio-hydrological conditions which are featured in the presented games?

Line 185f: On which basis do you assume that decisions of players in the game represent their real-life decisions? Did you ask them to do so? Why do you need this assumption?

Can you please better link your discussion and conclusions to the objectives of the paper and the presented results.

Now that you have developed the game, can you share some thoughts on how your games could be applied to support sustainable development? Who could facilitate games with which target groups? What would be required to implement the game with a larger group of stakeholders? Ideally this way forward should be linked to the theory of change.

I hope my comments help to improve the paper. I think my concerns can be addressed in the frame of a major revision. I very much hope to see an improved version of the paper being published.