By Sarah Bell
This column draws on my experience with the English Bridge Union (EBU). I chaired the EBU team investigating allegations of unfair play online. We tried to identify what, if anything had happened and reported back to the EBU’s Laws and Ethics Committee on how we’d come to our conclusions. Lots of people have asked about the process we used to do this work, and so here’s a brief overview.
Cheating is a tricky topic to discuss because it’s one that people feel strongly about. In my time working with the EBU, half of the people I spoke to thought we were policing the online game too aggressively and the other half felt we weren’t doing enough. People also often changed their mind about the process if someone they knew became involved; when they believed someone to be cheating who wasn’t quickly thrown out of the game, or when a friend they trusted was being investigated. Most people who play bridge online were impacted by cheating to some extent and investigations felt intensely personal for those being investigated. It’s no surprise such strong emotional responses resulted.
Setting up a process was difficult and there were a lot of questions to consider as we did it. To what extent were existing offline protocols for investigation fit for application to online cases? Who would investigate allegations of unfair play online? How would they go about this? What would the standard of proof be? What about reasonable and proportionate sanctions? Every NBO had to grapple with these questions, and many came up with different approaches. There’ve been a lot of discussions about this on online forums.
I’m convinced many people cheated online because they felt it “didn’t really count” or that online bridge “isn’t really bridge”. I have some sympathy with this insofar as prior to Covid, there weren’t many highly prestigious events taking place online and online bridge clubs tended to be casual. A top English player was quoted in a newspaper describing banning players for cheating online as like imprisoning someone for stealing a tin of baked beans. We’d regard that crime as being more serious, however, if that tin of baked beans was all anyone had to eat. When online bridge became the only form of bridge that any of us could play, cheating online became just cheating.
Central to our process was to consider all the different strands of evidence. No one piece of evidence was likely to prove anything, but many separate pieces of evidence could add up to a strong case or, indeed, show that that one suspicious-looking incident didn’t fit with the general picture.
You may be aware of Nicolas Hammond’s work in using statistics to detect cheating; statistical measures played their part as pieces of our evidential puzzle. One example that shows this nicely is the opening lead. How often does a player make the double-dummy best lead? If you have access to a large number of hands that someone has played it is easy to calculate their double-dummy lead success rate over those hands, using a computer program. An expert calculated this statistic, and others, over tens of thousands of hands from honest players who volunteered and then looked at how that related to their rank on the EBU’s grading system, finding that it correlated pretty closely. This helped us to contextualise data from players who were being investigated – if someone was leading significantly better than even the strongest players over a large number of hands that was suggestive that something might be amiss.
Obviously, not everyone who leads unusually successfully over lots of hands is cheating, and there are other influencing factors: forms of scoring, partnership aggression in the auction, whether they were on lead to suit contracts or no trumps etc. But it formed part of our evaluation. Something. We used other statistical measures too, and these were especially useful when two large comparative datasets were available, for example comparing data from events where kibitzers were and weren’t allowed.
We also looked at hands. It’s all very well to say that someone was leading or playing statistically outside the norm but what if they’d just run into a set of hands that suited their style? Or where the normal action was the best one? We examined hundreds of hands per case to see whether the player had taken unusual actions, consistent with illicit knowledge of the deal. If they did a lot of unusual things that sometimes worked and sometimes didn’t, it suggested they weren’t cheating. If they routinely overcalled on four-card suits at the two level, vulnerable, when partner had a big fit but never when they didn’t, that suggested something else. Clearly, this is subjective, which is why multiple analysts looked at the hands separately to reach their own conclusions before coming together to discuss what they thought was going on. This is the sort of analysis that could be developed through crowdsourcing. I can imagine a future in which hands could be made available online to a network of analysts, who’d be able to work through them as and when they have time. This would make it easy to throw “dummy” hands in, where no player was under suspicion, and to reduce bias by anonymising all hands and randomising their order.
BBO were also hugely helpful in providing information about kibitzing activity for our investigations. I’m not going to go into detail about this, both because I’ve already written more than I promised the editor, but also to avoid giving something away that could help people avoid detection. That said, it’s no exaggeration to say their input has changed the game of bridge, bringing cases to the fore that would never have been heard were it not for their input.
This might all sound a bit bleak, and it’s easy to lose sight of the fact that dishonest players are a small minority. The truth, however, is that a lot of NBOs and individuals have done fantastic work in promoting honest play. On a personal note, one positive that has come out of my work is that I have met and worked with genuinely awesome people from around the world, who care hugely about protecting integrity and honesty in the game we all love. I hope that this is an area where people from different countries can continue to co-operate with and learn from each other, and that that might promote deeper links between bridge nations.
Sarah Bell is a part time professional bridge player and full time teacher. She writes a regular column in English Bridge. She will represent England in the European Championships in Madeira this June, playing in partnership with Michael Byrne in the mixed team. She has served on the EBU's Laws and Ethics Committee and was founding chair of their Online Ethics Investigation Group.
We are sorry that this post was not useful for you!
Let us improve this post!
Tell us how we can improve this post?