Please reach us at theedgarassociation@gmail.com if you cannot find an answer to your question.
EDGAR is cheating detection software. EDGAR analyzes bidding and card play looking for patterns that suggest unauthorized knowledge of partner’s hand. How EDGAR works is summarized here.
The EDGAR Association (TEA) is a 501(c)3 nonprofit responsible for creating and managing EDGAR. TEA is run by unpaid volunteers. Donations are welcome.
EDGAR is not open source.
We do not sell or license EDGAR. We provide EDGAR-based analysis as a service to qualified bridge organizations who request it.
TEA coordinates with bridge organizations and online platforms to acquire detailed records of bidding and play. We analyze the data using EDGAR and document the findings in a set of reports that are returned to the bridge organization.
Bridge organizations or clubs interested in obtaining EDGAR-based cheating detection services should contact us at TheEDGARAssociation@gmail.com.
No. TEA is an independent organization.
We are not affiliated with the IBA, which is a separate organization that provides arbitration services to bridge organizations such as the ACBL and USBF. Their website is https://bridgearbitration.org.
TEA has contracted with ACBL to provide Computer-based Cheating Detection in ACBL-sanctioned online games. We emphasize that TEA provides cheating detection services to the ACBL, but the ACBL is ultimately responsible for investigating and disciplining players flagged by CBCD software (see Section 502 of the ACBL’s Code of Disciplinary Regulations for more information).
Until tablets are in widespread use, EDGAR is unsuitable for F2F play. Vugraph records for F2F bridge are currently not reliable enough to use EDGAR with confidence.
We only generate EDGAR reports for qualified bridge organizations that have requested them.
Not from us. The bridge organization that requests the analysis is responsible for deciding whether and how to divulge this information.
Do not report suspicious actions to us. Report them to the sponsoring bridge organization through their normal procedures.
Do not report suspicious players to us. Report them to the sponsoring bridge organization through their normal procedures.
EDGAR does not distribute any information about its findings, except privately to the bridge organization that requested the analysis. Matters of imposing and publicizing discipline are the responsibility of the bridge organization.
Please reach us at theedgarassociation@gmail.com if you cannot find an answer to your question.
A detector is a set of rules for looking at a similar group of decisions or results for signs of cheating. For example, EDGAR has a detector which looks exclusively at opening leads vs. No Trump contracts. An EDGAR Report aggregates the evidence from many detectors.
We have published a series of articles at Bridge Winners that explain how many of the detectors work.
A list is available here.
In addition to our own internal audits, several world-class players have analyzed and endorsed our various cheating detectors, and an outside statistical consultant has evaluated our methods. Any organization which works with TEA is expected to perform their own due diligence.
Also see "Creating and Testing A Cheating Detector", and our list of Bridge Winners Articles, especially the one on Bad Landings
The minimum number of boards for EDGAR to determine a player cheated depends on how blatant the cheating – it could be anywhere from a few hundred to several thousand. Sometimes EDGAR can’t give a definitive answer, but may see sufficient evidence to flag a pair for further investigation.
EDGAR cannot prove that someone is not cheating. However, should suspicion arise about a player because of unusual plays that seem abnormally successful, EDGAR can provide tools to investigate whether the unusual, but successful actions are representative of a larger pattern or just normal good luck.
EDGAR will not detect all cheating. Its detectors monitor certain aspects of bidding and play that exhibit a large statistical separation between cheaters and honest players. More subtle forms of cheating may go unrecognized.
Also, EDGAR needs to accumulate sufficient data over many boards to confidently identify cheating; those who cheat only sporadically might not be detected. We believe that EDGAR will catch a high percentage of players who cheat regularly, making it a significant deterrent.
EDGAR is certainly faster, but it’s not easy to say which is better. EDGAR’s detectors are rules based, so EDGAR will be more consistent and objective than a human investigator. A good human investigator will have more bridge expertise, so may be a little more accurate than EDGAR on any given deal. Some cheaters will be easier for EDGAR to flag, others easier for a human investigator to flag. There are significant advantages to each of them and so they are important complements in a comprehensive system for detecting and disciplining cheaters.
Please reach us at theedgarassociation@gmail.com if you cannot find an answer to your question.
EDGAR largely focuses on areas of the game where the impact of ability is small compared to the potential benefit of using illicit information. For example, honest experts only perform a little better than honest novices on opening lead, but cheaters using Unauthorized Information can achieve dramatically superhuman results.
When analyzing opening leads, it is important for EDGAR to digest the auction inferences available to the opening leader. For most auctions this is straightforward. Occasionally, EDGAR will be uncertain about the meaning of a call but will give the opening leader the benefit of the doubt. Rarely, an auction will be too confusing and EDGAR will not evaluate that opening lead.
When analyzing bidding, EDGAR focuses on specific outcomes rather than attempting to analyze a bidding sequence. These detectors will not flag a pair unless they outperform experts by a large margin over many deals.
Some of EDGAR’s detectors analyze individual card play. Your partner’s leads on defense, for example, won’t count in your score. Other detectors analyze specific bidding actions or final contracts, which count for both players. Whenever EDGAR flags a pair for cheating, we review the report before forwarding it to the bridge organization. If only one player is cheating, this is typically reflected in the EDGAR report, and we will document this finding along with the report.
We think it's important to strike a balance between transparency and concerns like these. That said, we see many reasons not to be overly concerned. One of them is that cheating advantageously while evading detection is extremely difficult, arguably a lot harder than playing bridge well.
Honest players have nothing to worry about: EDGAR is designed to distinguish honest play (however skillful, creative, erratic, or crazy) from cheating.
No, EDGAR does not view unusual plays as intrinsically suspicious. Bridge is a game of both skill and luck, sometimes “good” bids and plays work out poorly and “bad” bridge is rewarded. Over the long run, good play will outperform bad play. A beginner must have a run of luck that is essentially “impossible” before being flagged by EDGAR. This article goes into more detail.
No. Your opponents may draw a conclusion from just one or two suspicious deals, but there is no play, no matter how extreme or profitable, that will trigger a response from EDGAR. Instead of focusing on the occasional suspicious action, EDGAR aggregates many plays and many detectors before reaching a conclusion with confidence and will not notice until you’ve done it over and over with a highly abnormal rate of success.
No, EDGAR will not interpret your bidding success as cheating. A bidding system is one component of ability, and ability has only a limited impact on EDGAR’s findings, even at the expert level.
If this is Zia, EDGAR didn’t even raise an eyebrow – I guess you aren’t as clever as you think.
Joking aside: no, EDGAR will not flag honest players simply because of the occasional unconventional action that strikes gold. Unlike humans, EDGAR doesn’t have selective memory. EDGAR looks at every deal and considers both incriminating and absolving actions in many facets of the game. For EDGAR to flag a player for cheating, they must not only outperform a world-class expert, but outperform them by a significant margin over a large enough sample of deals.
Even if your lucky streak extends through several tournaments, and your thoughtfully creative plays work much better than usual, it is still virtually impossible you will be flagged absent actual cheating. Playing well instead of badly has little impact on EDGAR’s detectors. The detectors are designed to detect patterns of play that have extremely low probability of happening by chance. We have researched the relationship between winning and cheating and have determined that winning pairs playing honestly are not at increased risk of being flagged by EDGAR’s detectors.
This BW article summarizes some of our research.
No. EDGAR does not compare you to a club player, EDGAR compares everyone to what a world-class player, playing well, could plausibly achieve. For EDGAR to flag a player for cheating, they must not only outperform a world-class expert but outperform them by a significant margin over a large enough sample of deals. Even a skillful/creative player like Zia does not cause EDGAR to raise an eyebrow. In fact, most of EDGAR’s detectors are relatively insensitive to ability. The accuracy of opening leads, for example, differs only slightly between experts and novices. Typically, cheaters, whether expert or novice, greatly outperform noncheating players of all levels.
No, using world-class players as a baseline makes it less likely, not more likely that EDGAR will flag a club player for cheating. EDGAR does not compare your style of play to that of an expert, EDGAR compares your results to what a world-class player, playing well, could plausibly achieve. For EDGAR to flag a player for cheating, they must not only outperform a world-class expert but outperform them by a significant margin over a large enough sample of deals.
Copyright © 2023 edgarbridge.org - All Rights Reserved. TheEDGARAssociation@gmail.com