In the recent article “Anti-Confession: Andrey Gromov - Anna Gulevich”, Anna Gulevich makes the argument that CAT mishandled their case and is falsely accusing them of ethical violations, i.e. of collusive cheating.

In this article, CAT seeks to establish two items:

1. We can confirm that Anna Gulevich and Andrey Gromov are indeed on the CAT disinvitation list.
2. We maintain that their placement on this list is justified.

CAT’s list is private for a reason; contrary to what some believe, we are absolutely not in the business of publicly accusing players of cheating.

Our primary goal has always been to uphold the integrity of online competitions by using various forms of evidence to make judgments about the likelihood of whether a player is honest. If a player arouses some suspicion, we monitor that player; if that suspicion is deeper, we present our investigation internally and hold a vote on whether we ought to advise online organizers to not extend an invitation to that player.

Gromov-Gulevich implicitly waived their right to privacy by posting details of their case publicly. This article is our response.

CAT suspects Gromov-Gulevich of sharing information about their hands with one another.

There are several hallmarks of collusive cheating that differ from self-kibitzing:

1. Declarer play appears honest. In the case of a pair where the players are of unequal ability, the stronger player may coach the weaker one. In that case, we expect to see the weaker player play at a level that is similar to the stronger one, but still at a level that appears honest.
2. Collusive cases feature cheating that appears far less consistent than we see in self-kibitzing cases. Cheating on every hand starts to feel like cheating; cheating sporadically feels more like help.
3. Opening leads are not perfect, but far better than typical expert level. Defense after the opening lead is typically very strong.
4. Dubious bidding sequences that appear to work well on the actual pair of hands (though possibly not on the actual layout). Bidding misunderstandings that seem to get sorted out without issue. Overall bidding judgment that is much stronger (and luckier) than expected.

CAT made the judgment that Gromov-Gulevich was colluding using two forms of evidence: hand analysis and statistical evidence. The following is a partial presentation of that evidence.

In the case of Gromov-Gulevich, our belief is that they:

1. Usually bid honestly during the auction (not always).
2. Occasionally exchanged information on some key deals or in situations where a bidding misunderstanding was likely.
3. Often exchanged information when the correct opening lead or defense was unclear, especially against high level contracts.

We analyzed hands the pair played between April and July of 2020. While it was clear that they often played honestly, we felt there was sufficient evidence to believe they exchanged information on at least some hands.

We present a sample of hands played by Katt (Gulevich) and Agromov017 (Gromov) that led to our conclusion.

Example 1:

4♠ is an understandable underbid, but the 6♣ cuebid is unusual and perfect facing the North cards. In a poll of 15 experts, 14 bid 5♠ with the South cards and one doubled.

Example 2:

The pass of 3NT is interesting; the opening lead and switch are noteworthy. How can North determine their side is more likely to cash a heart than a club?https://tinyurl.com/y6qdkbza contains an example of the pair leading 9 from 109x.

Example 3:

A notable auction in that West judged his hand strong enough to bid RKC, but even more that EW are on the same page that 4♠ is actually keycard for hearts. West had only bid notrump and made a spade preference.

Example 4:

East makes a fortuitous pass of 2♠, as partner is 2=2 in the minors -- most would compete. West finds the ♦Q lead, the only lead to beat 3♠. In a poll of 15 experts, 13 led a spade, one led a club, and one thought it was close between a spade and a club. No expert considered leading the ♦Q.

Example 5:

The choice to bid Stayman on a 3=4=2=4 7-count was curious, but here it allowed them to reach the superior 2♥ partial.

Example 6:

A highly aggressive and successful 5♣ bid, followed by a perfect opening lead to beat 5♠-X a trick.

Example 7:

After 3♠ “nat FG”, and a lazy 4♠ signoff, West finds an inspired cue-bid to get their side to slam. In a poll of 10 experts, all ten passed 4♠.

Example 8:

Example 9:

East’s decision at trick one on both of these hands was unusual and lazy, but did not cost.

Example 10:

West makes no attempt to get to hearts; this time partner has a maximum with a singleton heart, and 3NT is better from the West side. In a poll of 15 experts, 14 responded 3♥ and one passed.

Example 11:

West upgrades to an 8-12 2♠ opening with an unremarkable 7-count (the alternative would have been 2♦ Multi), then judges to save against 6♣ -- right, as partner has a singleton club and little defense.

Example 12:

A remarkable auction to recover their 4-4 spade fit at the 5-level.

Example 13:

Excellent partnership understanding to be on the same page with respect to the double of 2♦. 5♣ is simply an overbid, but a decent shot opposite partner’s actual hand.

Example 14:

West finds a highly aggressive off-shape double to reach their cold game. In a poll of 15 experts, 13 passed, 1 doubled, and 1 abstained.

Example 15:

East finds an opportune reopening double. In a poll of 15 experts, pass was the unanimous choice.

CAT does not believe that a few suspicious hands are, by themselves, sufficient grounds for disinvitation. However, CAT also receives invaluable help from statisticians to determine the likelihood of foul play. Below we present an excerpt of Nicolas Hammond’s findings.

Hammond analyzed 1310 hands that Gromov-Gulevich played online in a report dated October 22, 2020. We received permission to publish excerpts of that report.

A summary of their online data:

A summary of Gromov and Gulevich’s face-to-face data:

To give context to these numbers, if Gromov-Gulevich (while playing online) had these numbers in F2F bridge, Gromov would be the third best defender of all time, and Gulevich would be the fourth. They would be behind Lotan Fisher (#1) but ahead of Cezary Balicki (#5).

Online, Gromov and Gulevich are very similar on declarer play: in the bottom third of top experts. While Gromov’s declarer play is roughly of the same strength online as in F2F, Gulevich is stronger online.

Both players are exceptionally strong opening leaders online. If their live leads were as good as their online leads, they would be the third-best pair in the world for opening leads. Their actual leads from F2F bridge put them in the bottom 2% of the world among top expert pairs. Both players lead remarkably better online, but especially Gulevich.

Quoting Hammond’s report: “Their sudden and dramatic improvement in defensive ability in on-line play, with no perceived improvement in declarer play over the same time, is strongly indicative of collusive cheating on defense.”

Although there has been a lot of speculation about the use of statistics and connection data in CAT cases, we want to make it clear that we are, first and foremost, bridge players. When we see unusual occurrences or reports of suspicious hands, we follow up on them and investigate. If we deem the evidence is sufficient, we hold a vote, and make a recommendation if necessary.

Our investigative procedure begins with expert analysis and statistical analysis of hundreds of hands. All of us are experienced players, and we know what high level bridge looks like -- and, perhaps more importantly, what it does not look like.

CAT does not want this article to be construed as a cheating accusation. In the interests of transparency, we are simply providing a glimpse into the way we operate.

CAT welcomes scrutiny, but to reiterate, we are not in the business of publicly accusing people of cheating. This article is an exception to our policy of not publicly discussing any investigations.

While this article is only a glimpse of our report for Gromov-Gulevich, we feel it is enough to form an impression on whether or not CAT performed its due diligence and acted in good faith. It is certainly our opinion that we took this matter seriously, just as we take all our cases seriously.

Accordingly, CAT unanimously voted to disinvite Andrey Gromov and Anna Gulevich in August 2020. We stand by that decision today, and we welcome an independent investigation of their hands.