Quality & Trust in Games Development
In this article I will explore what trust is, why trust is so important for developing complex
high quality games efficiently, and how we can work towards building a high trust
environment, especially in the remote and hybrid workplace.
What is trust?
There are many definitions of trust [4], and a large body of research around it, but for the
purpose of this discussion we need to have some definition to start the exploration from.
“Therefore, we define trust as the decision to rely on another party (i.e. person, group, or
organization) under a condition of risk. Reliance is action through which one party permits its
fate to be determined by another. Reliance is based on positive expectations of, or confidence
in, the trustworthiness of another party (Rousseau et al., 1998). Risk is the potential that the
trusting party will experience negative outcomes, that is, ‘injury or loss’ (March and Shapira,
1987; Sitkin and Pablo, 1992), if the other party proves untrustworthy. Thus, risk creates the
opportunity for trust (Rousseau et al., 1998).” [3]
“Trust in another can be characterized as cognition-based trust (based on rational
calculation) and affect-based trust (based on emotional attachment). For example, trust in an
auto repair shop could come in the form of an assessment of the capabilities of the shop to
do a good job repairing one's car (cognition-based trust) or of having a longstanding
relationship with the shop's owner (affect-based trust).” [7]
To make this more digestible and colourful I created this trust diagram, inspired by a lecture
on trust I attended hosted by TrustScore [10].
Picture A: Emotional and Rational Trust Diagram [9][10]
With this we have a definition of trust we can discuss around, and explore how it affects
developing high-quality games efficiently.
Human-Machine Trust
But before we continue I want to narrow down what type of trust we are talking about. And
that is trust between human agents, at different scales.
Picture B: Trust at different scales [3]
Because you can also talk about trust for non-human agents, such as animals and machines.
“People may trust non-human agents. For instance, people may trust animals.” [7]
“When you have well-validated and/or objective measures and methods for evaluating and
designing for compliance, reliance, transparency, user-centeredness, ecological-validity, and
human performance, you do not need concepts as abstruse and messy as trust.” [5]
For animals it would seem reasonable that you can have both emotional and rational trust,
but when it comes to machines, as stated above, emotional trust would seem
counter-productive as you can measure compliance, reliance and other factors and would
not need the complexity of human emotions to factor in.
Trust and Complexity
This leads to an interesting discussion about when emotional trust is actually necessary. I
think it would be compelling to try to connect this to the Cynefin concept of complexity [12].
When you have something that is “clear” or “complicated” you could perhaps rely solely on
rational trust, but when you move into the “complex” domain it is no longer possible to
measure in a way that would allow us to only use our rational trust.
And humans, organisations and software development would all, for the most part, land in
the complex domain.
Picture C: Cynefin Framework [12]
Why is trust important for software development and quality?
So why should we care about any of this?
Picture D: Impact of Trust [1]
One important part of this, I believe, comes from intrinsic motivation [11]. If you trust people
and give them autonomy and ownership they will be more intrinsically motivated, and that
will bring many benefits [13]. Being in a trustful environment/society also makes us healthier,
which in itself brings a multitude of advantages [2].
In general there is a cost to favour control over trust in many cases. People are less
motivated to perform well if someone forces them to provide a minimum level of
performance. [15]
But there will be a lot of inefficiencies during production if there is no trust between the
different actors in that production process. Here we will deep dive specifically into QA and
quality.
Without trust, the need and cost for documentation and traceability skyrockets, and the
cost of redundancy will also increase dramatically. Especially if/when you try to, and fail at,
documenting tacit knowledge [14].
When it comes to QA and quality this can be seen in for example:
● Excessive need for test cases, instead of allowing experienced QA Specialists to
perform more efficient exploratory testing
● Test reports, and this can be pushed ad infinitum. Every step of the development
process could require some kind of report
● Double (triple/quadruple/etc.) testing - because there is no trust between agents that
the previous test activity has actually been performed or performed well
● Complex activities like risk assessment will take significantly more time as the
requirements on thorough documentation and traceability increases
● Excessive use of automated testing, due to traceability, even where it is not the
optimal solution
But there is a broader and more critical problem:
● Without trust, there is less holistic ownership of quality in the team
○ Everyone will focus on artefacts and activities they are being measured and
controlled on, instead of caring about the customers’ perceived product
quality [16]
So in conclusion the cost of distrust is high. How do we avoid this cost?
How do we build a high trust environment?
It is hard to build trust, but it is easy to destroy it, and the road back is even harder.
Picture E: Maintaining trust [3]
So what are some keys to building especially emotional trust according to research? [6][1]
● Effective Communication
○ The more people understand the process and your goals and intended
outcomes, the more they will trust in your efforts [6]
● Respect
○ The tone, content and facilitation of your engagement matters [6]
● Transparency
○ Less transparency = less engagement and trust [6]
● Sharing information
○ If people feel that information is only shared with some groups, trust will
decrease [6]
● Engaging people
○ People will show greater trust if they feel included and listened to [6]
● Recognize excellence
○ Recognize high performers publicly [1]
● Autonomy
○ Give people autonomy to do their work [1]
● Reachable goals
○ Goals that are challenging but achievable increases focus and builds social
bonds that increase trust [1]
● Leadership vulnerability
○ Being vulnerable as a leader increases trust and collaboration [1]
● Intentionally build relationships
○ Social connections build trust [1]
What is the first step on the journey towards a high trust environment?
So we want to transform the workplace into a high trust environment. How do we do that?
“The concept of organizational trust is elusive and subject to a wide range of definitions, as
well as contributing factors and circumstances. As with other organizational constructs such
as culture or climate, we cannot measure organizational trust directly. Instead, we rely on the
perceptions of individuals within the organization, who will have different views of recalled
situations and contexts based on their experience. Those views are represented as variance in
the measure, while the mean measure is considered to be a descriptive statistic for the
organization as a whole.” [9]
The first step is to measure where we are, and identify what people perceive is missing.
There is no objective measure - we need to rely on the perception of individuals within the
organization.
There are many companies providing this type of service, and it can also be done inside a
company with a proprietary framework, but TrustScore [10] is one option.
Picture F: TrustScore [10]
References
[1] The Neuroscience of Trust
https://hbr.org/2017/01/the-neuroscience-of-trust
[2] Curious about Trust: Who do you Believe?
https://ki.se/en/research/popular-science-and-dialogue/spotlight-on/spotlight-on-relationships-identity-and-sexuality/curious-a
bout-trust-who-do-you-believe
[3] The Handbook of Trust Research
https://www.researchgate.net/publication/234021311_The_Handbook_of_Trust_Research
[4] How and Why Humans Trust
https://pmc.ncbi.nlm.nih.gov/articles/PMC10083508/
[5] Trust is not a virtue: Why we should not trust
https://www.researchgate.net/publication/365271322_Trust_is_Not_a_Virtue_Why_We_Should_Not_Trust_Trust
[6] The Role and Importance of Building Trust
https://aese.psu.edu/research/centers/cecd/engagement-toolbox/role-importance-of-building-trust
[7] Trust
https://en.wikipedia.org/wiki/Trust_(social_science)
[8] The Enemies of Trust
https://hbr.org/2003/02/the-enemies-of-trust
[9] Dimensionality of Organizational Trust
https://files.eric.ed.gov/fulltext/ED504674.pdf
[10] Trust Score
https://www.trustscore.se/
[11] Quality, Testing and Agile methodologies
https://www.slideshare.net/JohanHoberg/quality-testing-agile-methodologies
[12] Cynefin Framework
https://en.wikipedia.org/wiki/Cynefin_framework
[13] Self-determination Theory
https://en.wikipedia.org/wiki/Self-determination_theory
[14] Tacit Knowledge
https://en.wikipedia.org/wiki/Tacit_knowledge
[15] Distrust - The hidden cost of control
https://docs.iza.org/dp1203.pdf
[16] What gets measured gets done or does it
https://www.forbes.com/sites/ellevate/2015/06/08/what-gets-measured-gets-done-or-does-it/
Picture A is made with support of ChatGPT.

How Trust Impacts Quality and Efficiency in Games Development

  • 1.
    Quality & Trustin Games Development In this article I will explore what trust is, why trust is so important for developing complex high quality games efficiently, and how we can work towards building a high trust environment, especially in the remote and hybrid workplace. What is trust? There are many definitions of trust [4], and a large body of research around it, but for the purpose of this discussion we need to have some definition to start the exploration from. “Therefore, we define trust as the decision to rely on another party (i.e. person, group, or organization) under a condition of risk. Reliance is action through which one party permits its fate to be determined by another. Reliance is based on positive expectations of, or confidence in, the trustworthiness of another party (Rousseau et al., 1998). Risk is the potential that the trusting party will experience negative outcomes, that is, ‘injury or loss’ (March and Shapira, 1987; Sitkin and Pablo, 1992), if the other party proves untrustworthy. Thus, risk creates the opportunity for trust (Rousseau et al., 1998).” [3] “Trust in another can be characterized as cognition-based trust (based on rational calculation) and affect-based trust (based on emotional attachment). For example, trust in an auto repair shop could come in the form of an assessment of the capabilities of the shop to do a good job repairing one's car (cognition-based trust) or of having a longstanding relationship with the shop's owner (affect-based trust).” [7]
  • 2.
    To make thismore digestible and colourful I created this trust diagram, inspired by a lecture on trust I attended hosted by TrustScore [10]. Picture A: Emotional and Rational Trust Diagram [9][10] With this we have a definition of trust we can discuss around, and explore how it affects developing high-quality games efficiently.
  • 3.
    Human-Machine Trust But beforewe continue I want to narrow down what type of trust we are talking about. And that is trust between human agents, at different scales. Picture B: Trust at different scales [3] Because you can also talk about trust for non-human agents, such as animals and machines. “People may trust non-human agents. For instance, people may trust animals.” [7] “When you have well-validated and/or objective measures and methods for evaluating and designing for compliance, reliance, transparency, user-centeredness, ecological-validity, and human performance, you do not need concepts as abstruse and messy as trust.” [5] For animals it would seem reasonable that you can have both emotional and rational trust, but when it comes to machines, as stated above, emotional trust would seem counter-productive as you can measure compliance, reliance and other factors and would not need the complexity of human emotions to factor in.
  • 4.
    Trust and Complexity Thisleads to an interesting discussion about when emotional trust is actually necessary. I think it would be compelling to try to connect this to the Cynefin concept of complexity [12]. When you have something that is “clear” or “complicated” you could perhaps rely solely on rational trust, but when you move into the “complex” domain it is no longer possible to measure in a way that would allow us to only use our rational trust. And humans, organisations and software development would all, for the most part, land in the complex domain. Picture C: Cynefin Framework [12]
  • 5.
    Why is trustimportant for software development and quality? So why should we care about any of this? Picture D: Impact of Trust [1] One important part of this, I believe, comes from intrinsic motivation [11]. If you trust people and give them autonomy and ownership they will be more intrinsically motivated, and that will bring many benefits [13]. Being in a trustful environment/society also makes us healthier, which in itself brings a multitude of advantages [2]. In general there is a cost to favour control over trust in many cases. People are less motivated to perform well if someone forces them to provide a minimum level of performance. [15] But there will be a lot of inefficiencies during production if there is no trust between the different actors in that production process. Here we will deep dive specifically into QA and quality. Without trust, the need and cost for documentation and traceability skyrockets, and the cost of redundancy will also increase dramatically. Especially if/when you try to, and fail at, documenting tacit knowledge [14]. When it comes to QA and quality this can be seen in for example: ● Excessive need for test cases, instead of allowing experienced QA Specialists to perform more efficient exploratory testing ● Test reports, and this can be pushed ad infinitum. Every step of the development process could require some kind of report ● Double (triple/quadruple/etc.) testing - because there is no trust between agents that the previous test activity has actually been performed or performed well ● Complex activities like risk assessment will take significantly more time as the requirements on thorough documentation and traceability increases ● Excessive use of automated testing, due to traceability, even where it is not the optimal solution
  • 6.
    But there isa broader and more critical problem: ● Without trust, there is less holistic ownership of quality in the team ○ Everyone will focus on artefacts and activities they are being measured and controlled on, instead of caring about the customers’ perceived product quality [16] So in conclusion the cost of distrust is high. How do we avoid this cost?
  • 7.
    How do webuild a high trust environment? It is hard to build trust, but it is easy to destroy it, and the road back is even harder. Picture E: Maintaining trust [3] So what are some keys to building especially emotional trust according to research? [6][1] ● Effective Communication ○ The more people understand the process and your goals and intended outcomes, the more they will trust in your efforts [6] ● Respect ○ The tone, content and facilitation of your engagement matters [6] ● Transparency ○ Less transparency = less engagement and trust [6] ● Sharing information ○ If people feel that information is only shared with some groups, trust will decrease [6] ● Engaging people ○ People will show greater trust if they feel included and listened to [6] ● Recognize excellence ○ Recognize high performers publicly [1] ● Autonomy ○ Give people autonomy to do their work [1] ● Reachable goals ○ Goals that are challenging but achievable increases focus and builds social bonds that increase trust [1] ● Leadership vulnerability ○ Being vulnerable as a leader increases trust and collaboration [1] ● Intentionally build relationships ○ Social connections build trust [1]
  • 8.
    What is thefirst step on the journey towards a high trust environment? So we want to transform the workplace into a high trust environment. How do we do that? “The concept of organizational trust is elusive and subject to a wide range of definitions, as well as contributing factors and circumstances. As with other organizational constructs such as culture or climate, we cannot measure organizational trust directly. Instead, we rely on the perceptions of individuals within the organization, who will have different views of recalled situations and contexts based on their experience. Those views are represented as variance in the measure, while the mean measure is considered to be a descriptive statistic for the organization as a whole.” [9] The first step is to measure where we are, and identify what people perceive is missing. There is no objective measure - we need to rely on the perception of individuals within the organization. There are many companies providing this type of service, and it can also be done inside a company with a proprietary framework, but TrustScore [10] is one option. Picture F: TrustScore [10]
  • 9.
    References [1] The Neuroscienceof Trust https://hbr.org/2017/01/the-neuroscience-of-trust [2] Curious about Trust: Who do you Believe? https://ki.se/en/research/popular-science-and-dialogue/spotlight-on/spotlight-on-relationships-identity-and-sexuality/curious-a bout-trust-who-do-you-believe [3] The Handbook of Trust Research https://www.researchgate.net/publication/234021311_The_Handbook_of_Trust_Research [4] How and Why Humans Trust https://pmc.ncbi.nlm.nih.gov/articles/PMC10083508/ [5] Trust is not a virtue: Why we should not trust https://www.researchgate.net/publication/365271322_Trust_is_Not_a_Virtue_Why_We_Should_Not_Trust_Trust [6] The Role and Importance of Building Trust https://aese.psu.edu/research/centers/cecd/engagement-toolbox/role-importance-of-building-trust [7] Trust https://en.wikipedia.org/wiki/Trust_(social_science) [8] The Enemies of Trust https://hbr.org/2003/02/the-enemies-of-trust [9] Dimensionality of Organizational Trust https://files.eric.ed.gov/fulltext/ED504674.pdf [10] Trust Score https://www.trustscore.se/ [11] Quality, Testing and Agile methodologies https://www.slideshare.net/JohanHoberg/quality-testing-agile-methodologies [12] Cynefin Framework https://en.wikipedia.org/wiki/Cynefin_framework [13] Self-determination Theory https://en.wikipedia.org/wiki/Self-determination_theory [14] Tacit Knowledge https://en.wikipedia.org/wiki/Tacit_knowledge [15] Distrust - The hidden cost of control https://docs.iza.org/dp1203.pdf [16] What gets measured gets done or does it https://www.forbes.com/sites/ellevate/2015/06/08/what-gets-measured-gets-done-or-does-it/ Picture A is made with support of ChatGPT.