Chess notables swap ideas at "Cheating Town Meeting"

In an episode from the 1960s TV drama Mission: Impossible, master-of-disguise Rollin Hand (played by Martin Landau) impersonated a chess master competing in an international tournament.* Using a hearing aid to receive computer-generated moves from his mission colleagues, Rollin demolished his grandmaster opponents—even winning one game in a mere four moves!

While the mass TV audience cheered yet another example of gee-whiz know-how and ingenuity for which the Mission: Impossible series was justly famous, chess players greeted that particular plot device with ridicule. At the time, even the best computer programs could hardly compete with casual players, let alone professionals.

Fast-forward to 2006. In last year's best-paying public tournament, one player on the verge of winning a five-figure cash prize is ejected after getting caught with a suspicious hearing aid-like device. Another player attracts official attention and suspicion after a string of unlikely results accompanied by odd behavior. A few months later, the Mission: Impossible scenario comes full circle, when one side in the World Chess Championship match in Elista accuses the other side of surreptitiously obtaining computer help during the games.

This time, no one is laughing.

Now that computer software and hardware capable of outplaying even super-GMs is widely available at low prices, the possibility of computer-aided cheating cannot be easily dismissed—especially in events with large cash prizes or other rewards. But chess tournament culture in the United States, which relies heavily on the honor system to keep players honest, seems ill suited to cope with the emerging threat. Organizers of events that provide the greatest incentive to cheat—big-money opens and high-prestige invitational tournaments and matches like the World Championship—face a real dilemma figuring how to better secure playing environments against cheating, without alienating the mass of honest players or busting their own budgets.

To explore practical issues associated with combating cheating, New York's Marshall Chess Club hosted a "Chess Cheating Town Meeting" last December 4th. More than 40 people packed the historic club's second floor to hear an expert panel discussion followed by a question-and-answer period. Credit for the idea belongs to Marshall Chess Club President Douglas Bellizzi. I served as organizer and panel moderator.

The marquee speakers were Bill Goichberg—who is both president of the USCF and the nation's leading organizer of large open tournaments through his Continental Chess Association (CCA)—and grandmaster Alexander Stripunsky, who tied for first place in the 2005 U.S. Championship tournament.

The other four panelists were: IM Dr. Danny Kopec, Brooklyn College professor and graduate deputy chair of the Department of Computer and Information Science; Steve Immitt, a prolific National Tournament Director who the USCF honored as "TD of the Year" for 2005; Nelson Farber, New York-based attorney in private practice and an active member of the Marshall Chess Club; and myself, a non-professional tournament competitor, writer, and anti-cheating activist.

Besides the VIP panel, chess notables in the audience included past U.S. Championship brilliancy prize donor Paul Albert; science journalist and publishing consultant Paul Hoffman; Dr. Frank Brady, long-time chess writer, editor, sponsor, and organizer; The New York Times reporter Dylan Loeb McClain, who often writes about chess; USCF Executive Board member Sam Sloan; and WGM Jennifer Shahade, Chess Life Online editor. A Russian TV crew filmed part of the meeting.

CCA tightens rules for its top events

Bill Goichberg used his remarks to explain electronic-device restrictions and other new rules for the World Open and three other CCA tournaments: Foxwoods, Chicago Open, and North American Open.

Rather than the push for new USCF rules to prevent cheating, Goichberg said, "The organizers that run super-big money tournaments will be forced to be the leaders in anti-cheating."

Starting this year, publicity for the four biggest events run by Goichberg's Continental Chess Association states that:

• Players may not use headphones, earphones, cell phones or hearing aids. (This rule may be waived for players with low scores.)

• Players may not leave the tournament floor without permission. (This rule may be waived for players with low scores.)

• A player who refuses an official request to be searched for banned devices will be ejected from the tournament with no refund of entry fee.

• iPods will still be permitted—but, earphones, needed to listen to an iPod, will not.

Goichberg said he will go further and ban cell phones from tournament rooms altogether at the point when cell phones become able to run Fritz chess software. This is expected to happen about a year from now, he said. (Current CCA policy permits cell phones to be carried into playing halls, but not used inside.)

Although submitting to a search request will be mandatory for players, Goichberg said he expects to search only two or three people at the World Open. He will probably retain a security professional to perform that work.

Of course, restrictive rules of any sort involve a trade-off. As Steve Immitt, the other National TD on the panel, put it, honest players "will cooperate to a reasonable degree, but not to an unreasonable degree."

Goichberg predicted that metal detectors will never be used at mass public tournaments. He dismissed another proposed technological solution—signal-jamming equipment—because of a Federal Communications Commission rule that he said bans its use anywhere in the United States.

On the other hand, he said he will consider using signal detectors—devices that pinpoint the source of communications signals without interfering with them.

Calls for tougher penalties

Catching a cheater is one thing; applying sanctions is something else.

A chess cheater's punishment drew worldwide headlines around the turn of the year, when India's chess federation slapped a 10-year ban on Umakant Sharma, a former amateur caught receiving moves through a wireless communication device during a FIDE tournament in Delhi in December.

Many people feel strongly that the best deterrent of all would be the sight of a cheater hauled away in handcuffs. Surprisingly perhaps, that view found no support among the panel.

Both Bill Goichberg and Nelson Farber, the attorney on the panel, said police and prosecutors simply have bigger fish to fry than punishing a chess cheater. Goichberg said he once got the run-around from three separate police departments even when he supplied evidence of a run-of-the-mill crime: a World Open player had stolen Bill's checkbook and written himself several large checks.

Farber said that prosecutors do have legal authority to pursue cheaters under either criminal law ("larceny by trick" in the New York State penal code) or civil racketeering (RICO) statutes. In practice, though, "prosecuting chess cheating is not likely to be a priority of District Attorneys and U.S. Attorneys," he said.

One audience member recommended that the USCF establish a special fund to pursue legal action against cheaters. Punishment "must be attacked on a higher level than that of the individual organizer or TD," said Andrew Kalotay, a New York financial executive and former member of Canada's Chess Olympiad team. Organizers "have a short-sighted view," Kalotay said. "They just want to clean up this one tournament." The problem is that ejecting a cheater without further action produces no deterrent effect for the future.

GM Alex Stripunsky called for a three-year ban from all USCF play for a first cheating offense, and a lifetime ban after a second offense.

Even if a complaint can't be substantiated, Stripunsky recommended that the USCF should keep the allegation on file in "a special data bank of suspected cheaters." That way, patterns may become evident if reports multiply for the same suspect over time.

As computers grow ever stronger, the grandmaster called it increasingly unlikely that anyone rated below 2200 could follow a computer's move choices (excluding openings) in multiple games and tournaments. Thus, if several reports document a suspect's computer-like play, "The USCF should not be afraid to proceed with disqualification," he said.

Stripunsky drew a parallel with the performance-drugs problem in physical sports. "Using outside help in chess is comparable to drugs these days," he said. "Our federation should be very strict defending our noble game from people who want to achieve tournament victories at any cost."

He also suggested that USCF membership forms include a statement that using any kind of outside help during a rated game is illegal and is punishable by disqualification.

Covering the legal bases

Attorney Nelson Farber recommended that chess authorities—rather than seek help from law enforcement—concentrate on developing fair and appropriate procedures to adjudicate cheating complaints. For rulings that must be made "on-the-spot" during a tournament, he called for creating standard procedures that players would agree to in writing either when joining or renewing a USCF membership, and/or upon entering any tournament.

He said such policies should include:

• Consent by participants that procedures are binding.

• Immunity for decision-makers.

• Methods of protecting privacy, especially in personal searches (to be done only with probable cause/reasonable suspicion; pledge to respect the suspect's dignity).

• For large tournaments, hold on-site Disciplinary Committee hearings, employing formal procedures such as record-keeping, right of the accused to present evidence and cross-examine witnesses, and when practical, the right to an advocate, especially if the accused is a minor.

• "Cheating may be found by a fair preponderance of the evidence, which is the general standard in civil cases."

Refuting one bogus "legal" nostrum popular on Internet chat boards, Farber stated: "The finding of the (on-site disciplinary) committee should be final for purposes of the tournament, and a sanction imposed. In other words, if it is later determined that the accused did not cheat and that he or she was deprived of $20,000 by the ruling, quite simply, tough."

He explained that a tournament organizer would run little risk of losing a lawsuit in those circumstances, because courts have held that the need for "finality" may outweigh "correctness" when determining the outcome of a sporting event. He added that courts are generally reluctant to overrule decision-makers unless the decision was "arbitrary, capricious, or fraudulent."

Still, Farber said rulings issued at a tournament site should be reviewed by the USCF Ethics Committee and Executive Board.

National TD Steve Immitt spotlighted three relatively recent developments that make cheating a threat to organized chess competition as never before:

1. Chess software has advanced to the point that "chess may be close to being solved, if it hasn't already been."

2. Smaller, cheaper and more powerful computer and communications hardware.

3. Larger prize funds. "When the payoff is in the tens of thousands of dollars, it undoubtedly will inspire more to take a shot" at cheating, Immitt said.

This places organizers in a bind. "People who do not cheat," Immitt explained, "will not enjoy being subjected to the multitude of restrictions and limitations on their freedoms necessary to eliminate totally any possibility of cheating—invasive searches, continuous monitoring, even in the restroom, total prohibition on talking, being confined to the playing area at all times, etc. People will cooperate to a reasonable degree, but not to an unreasonable degree."

The best answer, he said, is to work toward "a reasonable compromise," that would avoid over-regulating the mass of honest competitors, while making it harder for cheaters to communicate with an accomplice.

Downfall of the "royal game"?

IM Dr. Danny Kopec, Brooklyn College computer professor and the author of many chess books and DVDs, portrayed cheating as a natural outgrowth of a creeping erosion in respect for the culture and conditions of tournament chess—even among participants themselves.

Dr. Kopec cited faster time controls, inconvenient and stressful playing schedules, increased reliance on databases and opening preparation, and the elimination of adjournments, among changes he said have diminished chess'stature as "The Royal Game."

"There was a time when one entered the hall of a chess tournament and it was like entering the hallowed halls of a prestigious academic institution," he said. "One felt comfortable that the ethics, morals, and justice, that the essence of the logic of chess which is innately tied to human interest in the game, would dictate all human behaviors at a chess tournament."

That contrasts sharply with current chess culture, in Dr. Kopec's view. Today, he said, "It seems that like in most other endeavors in life, the clock and monetary reward seem to dictate chess, while the science and pure love of the game/sport suffers."

He divided the blame among organizers, chess professionals, and amateur players young and old.

In my own remarks, I warned that the chess world may eventually face a "professional" sort of cheater: a team of dedicated thieves using hidden wireless devices to transmit moves between players and computer-equipped accomplices. Such a cheating squad might be able to swipe the first prize in several sections of a single big tournament, and no one would be the wiser.

Among about a dozen past cheating incidents I found written up in various news reports and online discussions, not one suspect seemed to put much effort into avoiding detection. Some, like "John von Neumann" at the 1993 World Open (see sidebar), were so conspicuous that it almost looks as if they were actively trying to get caught. The biggest danger I see is not that prospective thieves will upgrade their technology for cheating, but that they'll upgrade their attitude. As Michael Corleone (Al Pacino) said in The Godfather, Part III: "Our true enemy has not yet shown his face." What's more, as wireless devices become ever cheaper and less detectable, a day may come when visual or even physical inspection will be useless for preventing cheating.

"Within five years or so, it will be a simple matter to embed a computer inside the human body," says Clint Ballard, a Seattle-area software executive and tournament sponsor. To get around the various economic, technological and legal hurdles associated with watching or inspecting players during competition, Ballard suggests examining completed games.

"One idea for a deterrent is to require big money prize winners give a lecture on their critical game(s)," he wrote in an email to this writer. "Any real chessplayer wouldn't mind demonstrating how they crushed their opponent and a cheater who knows they have to go in front of a crowd and explain their moves, might think twice about cheating."

Can a cheater's moves give him away?

Another after-the-fact approach, mentioned in my Town Meeting remarks, is engine move matching–applying the old truism, "It takes a thief to catch a thief."

Many online chess servers try to root out cheaters by comparing suspects' moves with an engine's first or second choices. Engine move-match percentages were widely discussed in connection with the biggest upset in the 2006 World Open (Chess Life, October 2006, pages 24-25), and were used by Veselin Topalov's manager to buttress his public accusations that Vladimir Kramnik had computer help in the World Championship unification match last autumn.

However, naïve use of move match numbers can lead to wrong conclusions, says IM Dr. Kenneth Regan, associate professor of computer science and engineering at the State University of New York at Buffalo.

For high-level play, Dr. Regan says, the percent of moves that match an engine's first choices is too simple a statistic to reliably indicate whether a player had computer help. "It's not the fact of whether you match that matters, it's a question of how many different options you have that match," he says. In other words, the observed percentage of matching moves needs to be adjusted based on how "forcing" was the opponent's play.

Dr. Regan has published statistical analysis of engine move match data for the Kramnik-Topalov championship match. He describes his results as a "statistical exoneration" of Kramnik's apparent high rates of matching Fritz move choices. For example, when presenting his figures for Game 3, he wrote, "Based on comparisons, the data do not support cheating at all." (See www.cse. buffalo.edu/~regan/chess/fidelity/ )

Dr. Regan continues to refine his approach, which rests in part on statistical principles comparable to those employed by the authors of Freakonomics—the 2005 bestseller that applied unconventional economic thinking to a broad range of societal activities.

For now, his method is too labor-intensive for large-scale uses, such as open tournaments. For the analysis to be performed automatically, he says chess engine makers would have to make their programs more "scriptable."

Clint Ballard also is designing software to detect statistical evidence of cheating. His effort relies on modeling players' individual styles and error rates, then pinpointing deviations from each.

However, National TD Steve Immitt cautioned that engine matching can only be "a secondary or tertiary form of evidence" to back a cheating complaint. There are too many legal problems to rely on it as the primary evidence, he said. .

* "A Game of Chess," Mission: Impossible episode aired January 14, 1968. Writer: Richard M. Sakal. Director: Alf Kjellin.

Chess Cheating Web Reports

The following Web links give details of several incidents going back to 1993, in which one or more players were accused of cheating. The incidents run the gamut from formal accusations backed by physical evidence (a player caught with Pocket Fritz or a concealed wireless voice receiver), to anonymous rumors and hearsay.

1.) On http://www.hindu.com/2006/12/27/stories/2006122704042200.htm:
An Indian newspaper's report of the 10-year chess banishment imposed on Umakant Sharma, a local player whose FIDE rating zoomed from the mid-1900s to 2480 in a matter of months—until he was caught receiving moves through a wireless device concealed in his woolen cap during a December 2006 FIDE tournament in Delhi. Meanwhile, a close associate of his, an international master who played on India's Olympiad squad in Torino 2006, also came under suspicion of getting computer help, and as of January was the subject of a formal complaint. (www.chessbase.com/ newsdetail.asp?newsid=3595 )

2.) http://www.seniorchess.zoomshare.com/
The "Blockade Chess Cheaters" Web site, set up in November 2005 mainly to showcase a petition sent to the USCF at that time. Signed by a couple of prominent chess figures including IM Dr. Danny Kopec, that petition asked chess authorities and big organizers to start studying how to provide better enforcement that could improve public confidence that games and prizes couldn't be stolen by cheaters. The site includes the full text of the petition, plus the full text of the security and ratings (anti-sandbag) policies published for the 2005 HB Global Chess Challenge.

3.) http://beta.uschess.org/frontend/magazine_124_122.php
Chess Life's account of two suspected cheating incidents at the 2006 World Open.

4.) From the Daily Dirt blog, here and also here
Two lengthy discussion threads that got started with the 2006 World Open cheating incidents, and developed into a raging debate over tournament prize structures and the relative morality and integrity of amateurs compared with professional chessplayers. Especially interesting is the back-and-forth between IM Ben Finegold and several posters who disputed his views about class prizes and cheating.

5.) www.chessninja.com/dailydirt/2005/ 07/u2000_intrigue_at_hb.htm
A leading chess news site's report, followed by another lengthy discussion thread, about a player who forfeited from a class section of the 2005 HB Global Chess Challenge in Minneapolis. That player fled the tournament after being caught making cell phone calls during his games several times. Officials suspected he was receiving moves by phone from an accomplice sitting at a computer. Six weeks later, he returned to win a $5,800 section prize at the World Open.

6.) www.chessbase.com/newsprint. asp?newsid=693
A tongue-in-cheek account of a 2003 cheating incident at a tournament in Germany. Although the incident itself apparently was real, Chessbase News opted to make light of it in this report. (Chessbase is the German company that produces and markets Fritz.) For a great laugh, read the items beneath the news story, under the headings, "Editorial Comment," "Dentophonics" and "Nice One, Centurion."

7.) The "Nice One, Centurion" comment under the preceding item refers to an earlier cheating incident from 1999, in which a German amateur apparently used computer help to defeat one or more grandmasters. People became suspicious when, playing against a GM in a complicated position, the amateur cavalierly but correctly announced "mate in 8." Unfortunately, complete details are hard to come by, in English, at least. A Google search of the suspect's name turned up mostly unidentified, years-old bulletin boards posts.

8.) www.chessninja.com/dailydirt/2005/11 /topalov_accused_of_cheating.htm
Late in 2005, East European Internet sites broadcast rumors that an unnamed participant at the FIDE World Championship tournament in San Luis, Argentina, had accused winner Veselin Topalov of receiving computer assistance. It later emerged that no San Luis participant lent his name to the scurrilous rumor: several were asked, and each denied making the accusation. However, a Russian grandmaster named Dolmatov did go on the record to accuse Topalov of cheating. Since Dolmatov at one point was a trainer with Russia's Olympic chess team, and by some accounts is close to World Champion Kramnik, this fueled speculation that the barrage of public charges that the Topalov camp later unleashed against Kramnik during the 2006 championship match, were a form of "payback."

9.) www.worldchessnetwork.com/English/chessNews/evans/040112.php
An old Larry Evans column briefly described the 1993 "von Neumann" incident, as part of a more general comment about computers in chess. John Von Neumann is the name of a major figure in early computer science, who worked on the Manhattan Project developing the A-bomb. He died in 1957.

10.) www.chessbanter.com/showpost.php?p=58809&postcount=13
A Usenet post mentions both the "von Neumann" and "mate in 8" incidents (see items 8 and 6, above).

11.) www.thechessdrum.net/newsbriefs/2002/NB_PeterMoss.html
Still another chess news site's report of a 2002 World Open cheating incident, which led to a player being forfeited from a class section during a late-round game with prize money on the line. This one didn't involve computers, but good old face-to-face consultation (in Russian) about a game in progress.

12.) www.chessreporter.com/your_ cheating_heart.htm
A 2004 post gives a secondhand account of a player (apparently outside the U.S.) who caught his opponent cheating with chess software running on a mobile phone.

13.) www.chessreporter.com/cheating_in_chess92004.htm
Another secondhand account, detailing how a cheater at his board and an off-site accomplice with computer could communicate both sides' moves to one another, using an earpiece and cell phones. Although an anonymous posting about a chat with an anonymous online acquaintance is hardly authoritative, one party's assertion that "he knew players who cheated and won several thousand dollars" in this manner, does not sound far-fetched.

14.) www.chessbase.com/newsdetail.asp?newsid=3633
Perhaps nothing will damage chess more in the public eye than the discovery of cheating at the highest levels. This report about an accusation against Topalov at Wijk aan Zee appeared in one of Germany's largest newspapers, Süddeutsche Zeitung.

15.) www.chessbase.com/newsdetail.asp?newsid=3654
Andrew Martin, in his playchess.com Internet radio show, takes a light-hearted look at various cheating methods.