Recently we have been seeing a steady stream of organisations asking about figures. How many people download BetBlocker? How many users click through from our domain to BetBlocker? Can you help us establish the size of the black market?
Operators, regulators, other support agencies, they all want to put a number on harm so that they can tally it in their spreadsheets and provide soundbites about the good they’re doing.
This is a completely understandable drive. Coming from a background in mathematics, numbers can help make a complex world simple. They provide easily understood data points. They help explain to non-experts in the field how we’re making a difference. If the graph goes up, we’ve done more good.
Unfortunately, however, when discussing gambling harm, numbers can often be misleading. That intuitive data point that the CEO, press and MP wants to see, misrepresents the reality of the situation.
They’re a lazy attempt to simplify something that can’t be simplified, resulting in decisions being made based on flawed information. As the famous phrase, popularised by Mark Twain, goes "there’s lies, damned lies and statistics."
Examples:
Why quantifying fails
Quantifying in this manner is based on a fundamentally flawed base assumption: that all gambling addicts are experiencing the same levels of harm.
The reality is that levels of gambling harm will fit a normal curve. This is the bell-shaped graph shown below. On the left-hand side are the people who gamble but experience no problematic behaviours whatsoever.
In the middle is the bulk of players. These are the people who show some markers of harm, but aren’t destroying their lives by engaging in gambling.
The further right on the graph we go the more substantial the harm experienced until we reach the far right of the graph where we have the people whose lives are severely negatively impacted by gambling. Those who face bankruptcy, imprisonment and potentially even suicide.
The extreme end of this sample represent a small number of people compared to the bulk in the middle, but these are the people whose lives are being ruined and who are the most likely to cause significant harm to others, the most expensive to treat and the most likely to end in tragedy.
What does this have to do with why quantifying fails? Let’s discuss the examples we gave a little earlier:
I) Large traffic = small traffic
Time and time again we’ve seen operators conclude that because they have huge traffic and a comparatively small amount of that traffic clicks through to BetBlocker (or indeed other responsible gambling support services) that these resources aren’t making a lot of difference. This is then used to justify giving less or stopping donations altogether.
We all naturally see figures as a solid facts. Reliable. Dependable. But perhaps we need to start asking a few more questions and thinking about the questions we ask
It is notoriously difficult to get players to use blocking software. It’s easy to understand why, in the position of the operator, it can be concluded that not many users are actually engaging with blocking software. It must be an ineffective solution.
But the truth about blocking software is that it is almost universally used as a crisis response option. So many users only look to engage with blocking software once they’ve already experienced significant harm and are scrambling to stop the rot.
What does this mean? This means that while the proportion of an operator’s traffic that’s referred to a blocking software may be unimpressive compared to their overall traffic, it represents the far right side of the normal curve.
The people that are most addicted and at highest risk of harm. The people experiencing the highest levels of harm and therefore those that are most likely to benefit from using blocking software.
The small numbers don’t mean we aren’t making a lot of difference. Our support is being taken by those that need it the most.
II) Jolly Roger on the starboard show
We’ve recently seen a number of regulators approaching us looking for information on the size of the black market. The clear intention has been to establish what proportion of the players in their country are choosing to play with unlicensed operators.
It seems like a reasonable data point for any regulator to want. Unfortunately, it’s a data point that is very misleading for anyone who doesn’t know better.
Looking to the UK market as an example – this jurisdiction has Gamstop, the national self-exclusion register. This allows a player to self-exclude from all UK-licensed operators in one go. But it’s also allowed for more nefarious activity.
Unscrupulous operators and affiliates engineer their websites to rank in Google for search terms like ‘casinos not on Gamstop.' Traffic searching this search term is effectively 100% confirmed gambling addicts.
These are users that have signed up to Gamstop because they have experienced difficulty controlling their gambling and are now trying to get round the restriction they put in place. As markers of harm go, this behaviour has to come very close to the top of the list.
What point am I making? Like the previous example, I’m highlighting that while play with the black market may represent a small proportion of the overall market size, the black market targets the most vulnerable subset of the market.
We’re not talking about an evenly distributed cross-section of the market. We’re talking about a focused attack on the most vulnerable section of the market.
The propensity for the black market to cause harm, both in the human terms of despair and tragedy, and in the cold financial terms of the cost to the taxpayer, is huge.
While the black market may be proportionally small, it’s capacity far outstrips that which we would expect when we consider the problem based on the proportion of the market alone.
Numbers Lie
That’s not true. But they can be misunderstood with a lack of context or knowledge. And they can be misrepresented, in this case, to make the problems of gambling harms appear smaller than they actually are.
We all naturally see figures as a solid facts. Reliable. Dependable. But perhaps we need to start asking a few more questions and thinking about the questions we ask. Because if we ask the wrong questions, the numbers will end up taking support away from those that most need it.
