Friday, August 23, 2019

Don’t Make this Common Mistake When Prioritising Software Projects

Every company I’ve worked for has had one thing in common – they all generated good ideas faster than they were able to make them a reality. Regardless of whether they were a small start-up or a huge corporation, they all had a backlog brimming with ideas, and only so much time and resource with which to deliver them. They needed to prioritise their workload. 

Any decent product owner quickly realises the need to have a logical and defensible way to prioritise the backlog. After all, there will be many stakeholders with pet projects that they need to be delivered urgently. When it comes to something so important, arbitrary or ‘gut feel’ choices aren’t acceptable.

With this pressure comes a mistake which is simple and very easy to make. It comes so naturally and seems so logical, but is the root cause of a phenomenal amount of lost value in tech companies.

The mistake is to put the highest-value projects at the top of your priority list.

At first glance that might sound outrageous, so I’ll be more specific. The mistake is to put the highest-value projects at the top of your priority list without first considering the amount of effort required to deliver them.

The crucial thing to realise is that your goal is not simply to deliver the project with the highest value. Your goal is to maximise the cumulative value you deliver in a given amount of time. The highest value projects tend to be the largest in terms of effort, and by prioritising projects by their value alone, you may miss the chance to deliver maximum value over time.

To illustrate this, lets imagine that you are choosing between two projects, and you can’t work on both projects simultaneously:
  • Project Major is a complete rewrite of your website. It will take 12 months to complete, but once released will deliver $100,000 of revenue every month.
  •  Project Minor is a change to your payments page. It will take 1 month to complete but will only deliver $15,000 of revenue every month.
Most people, especially leaders without a technology background, will rush to deliver Project Major first, because it generates the most value. This is a reasonable instinct but misses the big picture. To see why, lets look at the cumulative value delivered if we prioritise the projects in two different ways.

First, if we prioritise Project Major first, then deliver Project Minor:



This choice delivers no value after 1 year, and $1,365,000 in value after 2 years.

What if we do things the other way around, and put the lower value project first?



This choice delivers $110,000 in value after 1 year, and $1,445,000 after 2 years.

To make the comparison even clearer, lets look at both on the same chart:


As you can see, putting the lower value project first actually generates more revenue over time - $80,000 more, to be specific. But it has other benefits too.

First, value is generated earlier. This is important for many reasons, but especially where cash flow may be an issue. Money generated can be reinvested or otherwise put to good use. In the case where the Major project is focused on first, the business would go an entire year without generating any money from the development work.

Secondly, risk is reduced. Large projects tend to be complex, and hence here is a greater chance that a mistake will be made, the requirements will change, or the project will overrun. When was the last time you heard about a large project being delivered on time? 

A simple way to prioritise, therefore, is to divide the value of each project by the amount of effort it takes to deliver it. In the case of projects Major and Minor, this is the result:

Project
Estimated Value ($ per Month)
Estimated Effort (Months)
Value / Effort
Minor
$15,000
1
15,000
Major
$100,000
12
8,333

This methodology of prioritising projects is a simple version of ‘Weighted Shortest Job First’ (WSJF), so-called because it favours short projects but is weighted towards those with the highest value.

Lets look at how we could use this same methodology to prioritise a longer backlog:

Project
Estimated Value ($ per Month)
Estimated Effort (Months)
Value / Effort
A
$       15,000
1
15,000
B
$       20,000
2
10,000
C
$       75,000
8
9,375
D
$       50,000
6
8,333
E
$     100,000
12
8,333*
F
$       25,000
4
6,250
G
$     150,000
24
6,250*
H
$       35,000
7
5,000
I
$     200,000
48
4,167
J
$         2,000
1
2,000

*NB, where projects have equal value, deliver the shortest (lowest risk) project first

As you can see, the gargantuan Project I doesn’t get a look in. In fact, if you assume that you will work on these projects sequentially, it’s nearly 3 years before you start on it. But this sequence of prioritisation delivers far more value than the alternatives.

In reality, you are probably adding new projects to the backlog all the time, and it may be more than 3 years before you start work on Project I because other projects float to the top before it does. As a Product Owner you may have to work hard to justify the choice to deprioritise the highest-value project. Nobody said this was going to be easy!

This begs the question – should you ever commit to huge projects in the first place?

The answer is typically no, and you should ask tough questions of every huge project that is proposed. First – can the project be broken down into smaller constituent parts, each delivering incremental value? If so, each of those projects should be valued, estimated, and prioritised separately on the backlog. You may find that some of the constituent parts then float to the top, and valued is delivered more quickly.

Secondly, if the project can’t be broken into smaller pieces, can you add more resources so as to deliver the project more quickly? Importantly, the project does not necessarily have to be reprioritised – the project can be delivered more quickly simply by working through the backlog with greater capacity available. If the expense of additional resources can’t be justified, then perhaps the project is not as valuable as first thought.

Failing that, should it be done at all? Arguably, the bigger the project, the more certain you should be as to its value before you commit to it. If the methodology used to estimate the project’s value is weak, it should be challenged, and this should be addressed before the project is accepted.

The chances are, if the project is huge and cannot be broken down into smaller parts, it is not worth doing. Even if delivered successfully, on time and on budget, market conditions are likely to have changed to the extent that the assumptions that went into estimating the project’s value may no longer be valid.

Hopefully by this point, I’ve convinced you that you should take effort into account when prioritising software projects. The same principles can be applied to any projects, not just software (indeed, many of these principles have their roots in lean manufacturing).

However, not all projects have a tangible value, and not all business value is revenue. It is not always easy to estimate the value of each project with great accuracy. Furthermore, it is not always simple to calculate the amount of effort involved in delivering a project. The next blog in this series will look at ways to estimate value for more esoteric projects, and ways to estimate effort that don’t require weeks of work themselves. In the meantime, comments and feedback on this first instalment would be greatly appreciated.

Recap

  • It is a mistake to prioritise projects by their value alone. You should also take into account the amount of effort required to deliver each project.
  • Your goal is to maximise cumulative value delivered by your organisation, not necessarily deliver a high value project quickly.
  • A simple way to prioritise is to estimate the value and effort of each project, and then divide the value by effort to get a score. Then put the highest scores at the top of your list.
  • In general, don’t commit to huge projects. Break huge projects into smaller pieces, or resource up to deliver all projects more quickly. If this cannot be justified then the project is probably not worthwhile.

Friday, November 16, 2012

Design Portfolio - Edge Loyalty Program

'Edge' Loyalty Program
Full Tilt Poker
Q4 2011 to Q2 2012




One of the first things that I wanted to accomplish after I joined the Full Tilt Poker team in February 2011 was to revamp and simplify the loyalty schemes that the company offered to its players. When I joined, the loyalty program consisted of several diverse elements that had evolved over time, and were somewhat fragmented and distinct from each other.

The existing scheme consisted of:

  • The Iron Man Challenge, which rewarded players for playing on a regular basis. Unlike many other promotions, Iron Man didn't simply focus on high-volume play, but also rewarded players for coming back to the site frequently, even if they only played a little.
  • Rakeback, which was a form of cashback promotion, with the amount of rakeback generally set at 27% of the net revenue generated by the player. Rakeback was enormously complex because of the non-transparent way in which revenue was calculated (with a complicated rake attribution mechanism and numerous deductions from the gross revenue). It was impossible for a player to reliably estimate their rakeback in advance because of these complications. In addition, rakeback was only given to players who had joined the site through certain affiliates, meaning that the majority of players did not have access to it. Rakeback was also the root cause of some common security problems.
  • Full Tilt Points (FTPs), which were a reward that players could spend on various items in a loyalty store.
  • Black Card, a relatively new promotion that rewarded very high-volume players with bonus points and other benefits, based on the number of FTPs they earned during a rolling 100-day period.

I felt that the many separate elements made the overall rewards scheme difficult to understand and advocated switching to a unified system based on rolling averages, like Black Card.

As we were planning to relaunch the Full Tilt Poker brand under the new ownership of Groupe Bernard Tapie in late 2011, we had a unique opportunity to change the rewards system, and so we began to design a unified system based around extending Black Card downwards. The system had to compete effectively against our anticipated closest competitors.

We identified several weaknesses in competing loyalty programs and aimed to address them in our own. The competing programs had several traps for players which could cause them to lose out on value - a common trap being that the system was based on calendar months and years, meaning that if you started playing late in the month or year it was much harder to qualify for a VIP status than it was if you started early. Since the majority of people are paid in the last week of the month and this is when most deposits are made, this meant that many players felt disenfranchised by such programs. We decided to base our own system around the rolling periods, which had worked well for Black Card and avoided this problem.

Another common trap in competing programs is that in order to get any cash for your play, you must spend the points you have earned through playing on cash rewards in the VIP store. If you spend your points on the 'wrong' items, you get significantly less cash value than you would otherwise. We decided to avoid this issue in our own program, by not requiring players to spend points in the store to get their cash rewards.

Our new system was divided into several tiers. To keep the nice aspect of Iron Man, where you could qualify by playing little and often or by playing a lot but less often, you could qualify for each tier in the new program by maintaining an average level of play over 7, 30, or 100 days.

Once you reached the appropriate tier, you would begin earning a cash reward which would be paid on a weekly basis - much like rakeback, except available to all players and without any deductions. The size of the cash reward varied depending on how much you played and your tier within the program, and you could track it in real time in the client as it accumulated. We were the first poker site to allow the player to choose the day of the week on which their weekly reward was paid, and the first to provide detailed tools allowing the player to measure their past level of play, and make projections about what level they would reach in the future.

My working title 'Edge' (which is a synonym for 'advantage' in poker terminology) was selected as the name for the program, with 'Edge Reward' becoming the name for the weekly cashback reward.

With Edge, we were actually able to increase the rewards on offer (the highest tier offered cashback of 25% of gross revenue plus FTPs, which was significantly higher than the 27% of net revenue offered previously) while simultaneously making cash rewards available to all players, and not just a select few.

I'm very proud of the product that Edge became - we made very few compromises to bring it to market*, and in my opinion it improved significantly on the existing Iron Man and rakeback systems while also competing effectively against the other loyalty programs in the marketplace.

More information: http://www.fulltiltpoker.com/edge

*Some minor changes were made to Edge after my departure Full Tilt Poker was acquired by the Rational Group, but for reasons of confidentiality they cannot be discussed here.

Back to Index

Design Portfolio - One Click Colour Coding

One-Click Colour Coding
Full Tilt Poker
Q2 2011



Almost all online poker sites allow you to take notes on your opponents, so you can remember how they play from session to session. Along with the Ongame network, Full Tilt Poker was one of the first poker sites to offer a category selection feature (in the form of colours) that allowed you to quickly get a feel for the opponents you were facing.

The system worked from a modeless pop-up accessed via the right-click menu of your opponent, and required at least three clicks in total - one to open the dialog, another to select the colour from a drop-down menu, and a further click to save and exit. Over time it became very common for players to simply categorise players without making an accompanying text note, so we decided to improve the user experience by making it possible to easily colour code a player without needing the pop-up dialog.

My idea was to place all of the colours in a grid within the right-click menu itself, meaning that only a single click of the mouse was required to select a category for your opponent. This simple solution had virtually no risk and yet improved the user experience dramatically for those who regularly colour-code their opponents.

More Information: http://www.fulltiltpoker.com/poker/software/new-features

Back to Index 

Thursday, November 15, 2012

Design Portfolio - Robocap (Dynamic Table Limits)

Robocap (Dynamic Table Limits)
PokerStars
Q1 2010


In a live poker game, you can usually only play at one table at any given time for reasons that should be evident. However, online there is no such restriction. You can play two, four, eight or even twenty-four tables at the same time if you wish. In fact it is fairly common for players on sites like PokerStars to 'multi-table' - on average, 10% of active unique players are playing 2 or more tables at any given time. This has obvious benefits for both the operator and the player (as long as the player is a winner).

However, early on in my time at PokerStars it became obvious that multi-tabling would be a driving force behind many of the usability improvements and new features that I would design, because of one unavoidable fact - multi-tabling is hard! The 10% of players who are multi-tablers slow down the games for the 90% of players who play only one table, making the game less fun. This is not to say that every person who multi-tables significantly slows down the game

The severity of the problem was highlighted by a comparison that I performed between PokerStars (which limited players to a maximum of 24 tables by default) and it's nearest rival Full Tilt Poker (which limited players to a maximum of 8 tables by default, with this increased to 16 on request). For standard No Limit Hold'em full-ring games, Full Tilt Poker dealt 33% more hands per hour, despite the fact that more players saw the flop on average (meaning that more events occurred per hand). In fact, a 'normal' speed table on Full Tilt Poker dealt slightly more hands per hour than a 'fast' speed table on PokerStars. This was a startling statistic which needed to be explained.

Some of the discrepancy could be explained by differences in software. For example, Full Tilt Poker allowed players less time to act, and had faster card-dealing animation. Accordingly, during my time at PokerStars I designed a number of changes to the software to speed up the games (and these were continued by my successors after I left). But clearly differences in software alone could not account for such a major variation in game speed - nor could they explain the many complaints that we would receive from recreational players who wanted to introduce games where multi-tablers weren't allowed, because of slow play.

The answer was rooted in the Peter Principle. Multi-tablers typically push themselves too far, increasing the number of tables they are playing until they are playing more than they could really handle. Once somebody feels comfortable at, say 8 tables, it is very difficult to convince them that they should play fewer. Often they won't realise that they are slowing the game down, and even if they do, human selfishness means that they are unlikely to give up the possibility of extra profit or rewards and cut back on their tables to improve the experience of others. On the other hand, there were a few multi-tablers who were able to play 24 tables without causing delays, and were capable of playing even more, but were limited to a maximum of 24.

The problem was that the existing 24-table limit was fixed in stone, and bore no relation whatsoever to how capable a multi-tabler a person actually was. People are different - one player might struggle to play 4 tables without slowing the game, while another can play 24 tables easily without causing a significant delay. The only real solution, therefore, is to have a table cap that is different for each player, and which takes into account how capable the player is of multi-tabling without causing delays.

It turns out that the average time it takes for a player to respond (to take an action when it is their turn) is about 4.5 seconds. Clearly, if somebody can maintain that speed or better while playing at 24 tables, they should be allowed to play more. However, if somebody is exceeding that time by a significant margin on a regular basis, chances are that they are playing more tables than they can really handle and their table cap should be reduced so that they don't make the game an unpleasant experience for their opponents.

Because nothing like this had been done before in online poker, I designed a configurable rules-based system. Administrators could define a rule like 'If a player's action time is greater than 11000 milliseconds, and they are playing at 10 or more tables, and their existing cap is 10 or higher, then decrease their table cap by 2'. This allowed the company to roll out the system gradually and silently so that they could see the effect that the changes would have in advance of them actually being implemented.

Dynamic Table Limits were rolled out at PokerStars after I left the company in July 2010, and at the time of writing it appears that they are being used quite conservatively (understandable, since some players are not going to like having their table limits reduced). However I expect the system to have a major positive impact on the quality and speed of ring games online in the coming years.

More Information: http://www.pokerstars.com/poker/room/features/playing-speed/

Back to Index 

Tuesday, November 13, 2012

Design Portfolio - Sit & Go Registration Improvements

Sit & Go Registration Improvements
PokerStars
Q1-Q2 2008


A common problem that occurred on PokerStars in 2008 was that you would try to enter a tournament that had a maximum number of entrants (for example, a Sit & Go with a maximum of 9 players), but by the time you had completed the registration process, the tournament had filled up and started, so your registration failed. This was very frustrating for customers who were interested in tournaments that filled very quickly, such as low-stakes or play money STTs.

To counter this problem, I designed a feature whereby the player was offered the opportunity to register in the next available identical event if the tournament had filled up by the time they completed their registration. This allowed the customer to register for a game on the first attempt every time, a huge improvement to usability.

More Information: http://www.pokerstars.com/poker/room/features/software-news/previous-years/#sfn6

Back to Index

Thursday, November 01, 2012

Design Portfolio - Text Filtering


Text Filtering
PokerStars
Q2 2008



PokerStars Tournament Filter with Text Filter at the Top

It's mid-2008. PokerStars is well-known for its wide selection of tournaments. In fact, there are now thousands of different tournaments every single day, and it is becoming difficult to easily find the ones that interest you in amongst the rabble of tournaments that don't. The simplest way to find the week's flagship event is to scroll through the list of 'Special' tournaments, looking out for an entry with the name 'Sunday Million'.

At the time, PokerStars already had a basic filtering system in place so that users could hide specific games they weren't interested in. For example, to hide Omaha games the user could uncheck the box marked 'Omaha', and users could check another box so that only tournaments that were currently registering or upcoming were shown. But with thousands of upcoming tournaments at any one time, this was of limited use.

While designing the upgraded filter, I made obvious improvements such as allowing the user to filter out specific game formats that didn't interest them (like Turbo tournaments), and some less obvious ones like preventing the player from being able to enter filtering combinations which would have always returned no results. But possibly the biggest lightbulb moment came as a result of playing around with Windows Vista's search feature.

In Windows Vista, the start menu has an integrated search function. You can open the start menu and start typing, and results are shown in real time as you press each key. Microsoft has improved this further in Windows 7 and 8, to the point where you can now open a program or file with as little as three key strokes (for example, opening Word is as simple as pressing the Windows key, followed by 'W' and then Enter).

Many tournaments on PokerStars, including all of the flagship events, have names. I realised that with a similar type of search mechanism to Windows Vista, you could find a specific tournament like the Sunday Million with trivial ease. All you'd have to do is start typing the word 'Sunday', and voilĂ , there was your tournament. With the addition of a boolean NOT to remove keywords that you were not interested in, this became a very powerful feature.

More information: http://www.pokerstars.com/poker/room/features/sorting/

Back to Index

Design Portfolio - Index of Features

Introduction


This is the index for a series of posts which will come together to form a portfolio of the features I have worked on over the course of my career in software design. In my time in the industry, I have led the development of new software features at PokerStars and Full Tilt Poker, the two largest poker rooms in the world. The features I have designed reach an audience of hundreds of thousands of unique customers every day.

Although I've been designing and programming games my whole life (starting with text-based adventure games in BASIC and then side-scrolling platform games in Klik & Play and The Games Factory as a kid), I fell into designing features for PokerStars by accident. I'd shown an interest in new software features and had experience in interaction design from University, so I was given the opportunity to design a number of small features - which eventually developed into a career.

Over the course of my career I have worked on hundreds of software features for PokerStars, Full Tilt Poker and others. However, for some of these features I simply led the design team. I'm only including features in this index if I conceived the idea myself, or was substantially involved the design and/or implementation of the product.

One final note - due to the unfortunate events of 2011, many of the features that I conceived and designed at Full Tilt Poker have not yet been released. To preserve the confidentiality of these projects they will not be included here until they are publicly announced.

Features I Conceived and Designed


The following features were my own original creations:


Features I Designed


The following features were designed by me:

  • Several not-yet-released features - Full Tilt Poker - Q1 2011 to Q3 2012
  • 'Edge' Loyalty Program - Full Tilt Poker -  Q4 2011 to Q2 2012
  • Double and Triple Chance Tournaments - Full Tilt Poker - Q1 2011
  • Cap Tables - PokerStars - Q2 2010 to Q3 2011
  • Detached Chat - PokerStars - Q3 2009 to Q3 2010 
  • Synchronised Tournament Breaks - PokerStars - Q3 2009
  • Game Speed Improvements - PokerStars - Q3 2009
  • Notes Improvements (Colour Coding and Visibility) - PokerStars - Q3 2009
  • Show Hole Cards When All-In Option (Ring Games) - PokerStars - Q2 2009
  • Sit & Go Rematch - PokerStars - Q2 2009
  • Auto Add-On (Tournaments) - PokerStars - Q2 2009
  • Milestone Hands Countdown - PokerStars - Q1 2009
  • Team Pro Visibility - PokerStars - Q4 2008 to Q3 2009
  • Bounty Tournaments - PokerStars - Q4 2008 to Q3 2009
  • Multi-Currency Accounts and Games - PokerStars - Q4 2008 to Q1 2010
  • Datamining Control - PokerStars - Q4 2008
  • Auto Rebuy and Add-On (Ring Games) - PokerStars - Q3 2008
  • Fold and Show - PokerStars - Q2 2008 to Q3 2011
  • Time Zone Support - PokerStars - Q2 2008
  • Ring Game Highlighting - PokerStars - Q1 2008 to Q1 2009
  • Black Lobby Theme - PokerStars - Q4 2007 to Q1 2008

Friday, July 06, 2012

Proactive Responsible Gaming



I recently read an article in another magazine by Simon Burridge, the CEO of Virgin Games. In the article, Burridge laments the fact that gambling businesses are often perceived by the public and portrayed by the media to be dirty and evil, despite the fact that the gambling industry has much tighter controls than other ‘vice industries’, such as the alcohol and the tobacco businesses. For example, you can self-exclude yourself from a chain of bookmakers if you are a problem gambler, but can you self-exclude yourself from purchasing alcohol at Tesco if you’re a problem drinker?

Burridge also argues that the National Lottery is unfairly perceived as more reputable than other types of gambling (especially casino games), and doesn’t have the same stigma attached, despite the fact that it is available to children as young as 16, from a diverse range of outlets, with no control to prevent problem gamblers from taking part.

There is a difficult truth to gambling and other so-called ‘vice industries’ however. Any business, be they a clothes retailer, an eBay seller or a car dealership, wants loyalty and repeat business. The easiest way to assure repeat business is to have a product that your customers can become addicted to – which is something that the alcohol, tobacco and gambling industries have – but of course the social consequences of addiction are extremely severe. What is unique to the ‘vice industries’ is that we must find a way to retain customers and ensure repeat business without causing addiction.

While I strongly agree with Burridge that the gambling industry’s controls are significantly stronger than those of alcohol and tobacco, and that the public’s perception is extremely unfair, I also believe that there is more that we can do, and that frankly the responsible gaming systems of even the best operators today are still only in their infancy.

Today’s responsible gaming methods are mostly retroactive, which is probably because online gaming operators have a difficult quandary to resolve – on the one hand, we must not accidentally exclude good customers with responsible gambling habits; but on the other hand, we must be seen to give problem gamblers help and the tools to control themselves.  The problem with retroactive methods is that the player is already experiencing a problem by the time they are exposed to the controls. For example, by the time a player reaches their deposit limit or table limit, they have probably already lost a sum of money which would be considered highly significant to most people; and using the self-exclusion tools requires a player to first admit that they have a problem and should take a break from the games.

What I’m going to suggest is not an expansion of retroactive methods (although, as I have argued for security purposes, I believe that competing companies should work together to share a blacklist of self-excluded players), but the introduction of proactive methods that would allow us to detect the signs of problem gambling before the player caused themselves significant harm. In the online sector we are uniquely positioned to do this as we have access to a wealth of information about a player’s financial and playing habits, and can easily process this data en masse.

The benefits to detecting a problem gambler early are obvious. First and foremost, protecting players who are at risk is simply the right thing to do. By stopping somebody early, you might prevent them from harming themselves or others, and you might be able to offer them help and advice that will help them to control their addiction. If nothing else, this should help you sleep at night.

Stopping problem gambling early also helps safeguard the reputation of your brand, and the entire industry, in particular by guarding against the type of media stories that we see all the time in the Daily Mail, of people who destroyed their families by running up gambling debts that they couldn’t afford.

Another less obvious benefit to detecting a problem gambler early is that you can guard against fraud and chargebacks. Problem gamblers are much more likely to chargeback on transactions as their losses inevitably mount, and are also more likely to attempt to use credit cards belonging to other people to finance their addiction. Detecting a problem gambler early ideally allows you to prevent such issues in the first place, but at the very least allows you to implement greater security controls and perform KYC checks in advance of any problems rather than cleaning up afterwards.

But how do you detect problem gambling proactively, without negatively impacting players who are able to gamble responsibly? Simple – you look for the tell-tale signs.

Gamblers Anonymous famously has a list of ‘twenty questions’, to which problem gamblers are expected to answer positively to at least seven. This isn’t a 100% scientific method of determining whether somebody has a problem, but simply something predictive, built on years of experience. The questions are:

1                     Do you lose time from work due to gambling?
2                     Is gambling making your home life unhappy?
3                     Is gambling affecting your reputation?                          
4                     Have you ever felt remorse after gambling?
5                     Do you ever gamble to get money with which to pay debts or otherwise solve financial difficulties?
6                     Does gambling cause a decrease in your ambition or efficiency?                       
7                     After losing, do you feel you must return as soon as possible and win back your losses?
8                     After a win do you have a strong urge to return and win more?        
9                     Do you often gamble until your last pound is gone?
10                 Do you ever borrow to finance your gambling?
11                 Have you ever sold anything to finance gambling?
12                 Are you reluctant to use gambling money for normal expenditure?                
13                 Does gambling make you careless of the welfare of your family?                      
14                 Do you gamble longer than you planned?
15                 Do you ever gamble to escape worry or trouble?
16                 Have you ever committed, or considered committing an illegal act to finance gambling?
17                 Does gambling cause you difficulty in sleeping ?
18                 Do arguments, disappointments or frustrations create an urge within you to gamble?
19                 Do you have an urge to celebrate any good fortune with a few hours gambling?       
20                 Have you ever considered self - destruction as a result of your gambling?

These questions can give us clues as to the type of behaviours to look for. The fact that the list contains 20 questions, of which problem gamblers are expected to answer with 7 or more with ‘yes’, also suggests that detecting only one type of behaviour probably isn’t enough. Instead, as with the detection of collusion, bot use, and other types of unethical play, detecting problem gambling is best done by automatically searching for multiple types of behaviour, scoring each one, and then flagging the highest-scoring accounts for manual investigation and follow up by a trained professional.

The types of behaviour to look out for might include:

·        Unhealthy Session Times: The Gamblers Anonymous questions suggest that playing during working hours (question 1) or playing when you should be sleeping (question 17) are signs of problem gambling. A simple way to detect this would be to look for players who are playing during the typical working or sleeping hours in the country that they are based in. For example, in the United Kingdom, a player who plays at 3pm on a Monday afternoon will typically be playing in working hours, and a player who plays at 2am on a Thursday morning will be playing in normal sleeping hours. These types of session, when combined with other factors, could be indicative of problem gambling.

·         Unhealthy Session Length: A player who plays excessively long sessions is clearly not playing responsibly, and Gamblers Anonymous suggest that gambling longer than planned (question 14) is a sign of problem gambling. So scoring a player progressively based on the length of their session (i.e. a 14-hour session scores higher than a 12-hour session) could be a good way to identify unhealthy playing habits.

·         Destructive Playing Habits: The Gamblers Anonymous questions identify several types of destructive playing habits. For example, a player who plays until their account is empty is effectively answering ‘yes’ to question 9. A player who loses all their money and then immediately deposits to try and win it back, or who chases their losses by moving up in stakes, is answering ‘yes’ to question 7. Even moving up in stakes after a big win could be indicative of a problem.

·         History of Self-Exclusion: A player who has a history of excluding themselves or imposing betting limits may be experiencing remorse over their gambling (question 4 of the Gamblers Anonymous questions). Such a person may suspect that they have a problem but be having difficulty controlling it – something which they could use your help with.

·         Indicative Chat: As any experienced operator will know, players make some very extreme admissions in the chat – everything from suicide threats, to sexual fantasies, to gambling problems. Monitoring the chat for certain key words and phrases associated with an unhappy home life (question 2), remorse (question 4), or problem gambling generally could help to identify players who are having difficulties.

·         Financial Habits: Monitoring the financial habits of your players can help you to detect tell-tale signs that something might be wrong. A player with a history of chargebacks is more likely than usual to be a problem gambler. A player with a large number of failed deposits, or any attempts to deposit with a card belonging to another person, could be indicative that a player is turning to fraud to fund their habit (question 16). Similarly, receiving a large number of inbound transfers from other players could indicate that the player is borrowing money from friends to finance their gambling (question 10).

As should be obvious, few of these factors are strong indicators of problem gambling when seen in isolation. For example, a player receiving lots of inbound transfers could simply be being staked, and a player who plays at 3pm in the afternoon could simply be unemployed, retired, or a professional gambler. However, in combination with each other, these indicators can add up to make a convincing case. It can’t hurt to call a player on the phone if they are showing some of the signs of an impending problem, and check that everything is alright. If doing so prevents even one player from experiencing the nightmare of a fully-fledged gambling addiction, it would be worth it.