INSIGHT ARTIFICIAL INTELLIGENCE
Andy, how is machine learning being utilised to recognise and address problematic gambling and how effective is it? How does ML work in the context of understanding player behaviour?
Machine learning can be a very strong tool to understand complex relationships between types of behaviour. Te benefit of machine learning is not only in developing models, but using models to identify behaviours that can be incorporated as features into our current analytical or heuristic models.
Understanding is a human trait (or not sometimes), while machine learning is a tool to be used by an evolved, experienced mind. It can track, show changes and relationships between data points, but to understand and explain is beyond machine learning.
How can machine learning help achieve the sustainable development of the gambling industry long term? How does machine learning help gambling companies understand the customer and allow for growth and responsible gambling at the same time?
In one word: width. As the industry grows and regulations increase, machine learning solutions can help add the scalability needed to ensure current structures can deal with the increase in demand and responsibility.
Tere are literally billons of datapoint, and ML can help with the selection of relevant variables and changes in behaviours, but deploying them in the real world is still restricted based on the time/processing power available to run these in a commercial business setting and the sheer amount other processes, which need to take place in a modern iGaming company. Academia has both the resource and, moreover, the time to help the industry help itself, and that is where most of the current thinking comes from.
Do you think that we will come to a point where machine learning can be used on its own as a way to prevent gambling related harm?
No, the human element will always be a part of the process when dealing with customer harm. As we have seen in industry and research, the use of machine learning as a tool to help identify problematic behaviour patterns or customers can make the process more efficient. However, the nature of the decision boundaries and error rate means the human element will always be required in the process.
To explain, we need to understand and interpret. In order to do that we must be able to have the actions and reactions described to us simply. Our ‘Markers of Harm’ are 180+ triggers, by brand, vertical, market and velocity linked to player behaviours, over multiple timelines. Gaming peers give people who are not data scientists the ability to make solid, balanced and accurate judgement calls on changing a person’s gaming behaviours to stop harm in the first place.
P104 WIRE / PULSE / INSIGHT / REPORTS
Andy Masters Chief Operating Officer, Crucial Compliance
“The human element will
always be a part of the process when dealing with customer harm. As we have seen in industry and research, the use of machine learning as a tool to help identify problematic behaviour patterns or
customers can make the process more efficient, however, the nature of the decision boundaries and error rate means the human element will always be
required in the process.” Andy Masters
What are the disadvantages of machine learning in the context of minimising gambling related harm? For instance, can poor or incomplete data deliver an unsatisfying experience for the customer and operator?
Poor and incomplete data is a problem for both analytical modelling solutions as well as machine learning counterparts. Generally, gaming has access to a large amount of detailed data compared to other industries. Te real issue, in terms of data, is identifying early-life problem gamblers without incorrectly classifying other cohorts as problems gamblers. Tis can be minimised in ways such as a model providing a prediction and a confidence score of the prediction’s validity. Tis allows more informed decisions to be made on the outputs, as we do at Crucial.
An issue with machine learning solutions is the potential for lack of interpretability that can come with using black box types of solutions. Tis needs to be taken into consideration when designing a model to ensure the descriptiveness of the outputs meets the operational needs. When dealing with gaming related harm, being able to interpret the outputs without a full technical understanding is crucial.
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96 |
Page 97 |
Page 98 |
Page 99 |
Page 100 |
Page 101 |
Page 102 |
Page 103 |
Page 104 |
Page 105 |
Page 106 |
Page 107 |
Page 108 |
Page 109 |
Page 110 |
Page 111 |
Page 112 |
Page 113 |
Page 114 |
Page 115 |
Page 116 |
Page 117 |
Page 118 |
Page 119 |
Page 120 |
Page 121 |
Page 122 |
Page 123 |
Page 124 |
Page 125 |
Page 126 |
Page 127 |
Page 128 |
Page 129 |
Page 130 |
Page 131 |
Page 132 |
Page 133 |
Page 134 |
Page 135 |
Page 136 |
Page 137 |
Page 138 |
Page 139 |
Page 140 |
Page 141 |
Page 142 |
Page 143 |
Page 144 |
Page 145 |
Page 146 |
Page 147 |
Page 148 |
Page 149 |
Page 150 |
Page 151 |
Page 152 |
Page 153 |
Page 154
