**The rules clearly state that each TEAM only has X submissions in a given round.** (In Round 1 it is 5 submissions, see the [Leaderboard Rules](syn6156761/wiki/400649)).   The members of a team must not make submissions as individuals, otherwise a team with N individuals would get N-times more submissions than other teams! (If only one team member is making submissions and submitted as individual by accident it's no problem, the point is that each team can only have 5 submissions total).   The same is true for **members of the same research group/lab** even if they don't form a synapse team, because it would be hard to believe that they are not working together and sharing results.   **If multiple members of your team/group/lab were making submissions as individuals**, or if you have any doubts about this rule, please contact @daniel.marbach immediately. As we are early on in the challenge we can find a solution.   However, if you are not contacting us and we discover such an issue later on in the challenge, it would be taken extremely seriously.   We do have mechanisms to detect potential cheating using multiple synapse accounts. Since DREAM Challenges are a scientific experiment that will be published, **cheating amounts to scientific fraud**. As this challenge has a low number of submissions per team, this issue is particularly important and we will thus be very vigilant.   See also: * [DREAM Principles](syn6182468/wiki/401779) * [DREAM Challenge rules](syn4294018/wiki/232126) * [Module Identification Challenge Rules](syn6156761/wiki/401428)

Created by Daniel Marbach daniel.marbach
The [DREAM Challenge Rules](syn4294018/wiki/232126) are very clear: * Pargraph 1b: Cheating by, for example, misleading other participants with decoy submissions, **or using fake identities to make more submissions than allowed**, is tantamount to scientific fraud. Sage Bionetworks and DREAM take incidences of cheating extremely seriously. In instances where Sage Bionetworks and DREAM confirm clear cases of cheating, the individual/team in question will be disqualified, and depending upon the gravity of the particular cheating incident, may be subject to any or all of the following: (1) individual/team banned from participation in future DREAM Challenges; (2) individual/team banned from future access to Synapse, and (3) individual/team?s cheating reported to all related and/or relevant parties, including, for example, publishers, funding agencies, University authorities, administrators, and colleagues. * Paragraph 4d: However, merging and disbanding Teams solely to circumvent limits on the maximum number of entries per team (or purchased data per team) is grounds for disqualification. > The spirit of the rules is perfectly clear, it's unethical to try to gain an advantage over other teams using some trick to increase the number of submissions that you or your team can make. **Thus, individuals working together and sharing results must submit as a team** (doesn't matter whether they're in the same research group or not, if you collaborate you're a team). We cannot technically prevent this (e.g., if we disabled the option to submit as individual, you could still form a team including just one member), but we do have checks and follow up suspicious behavior.   In the challenges that I was involved in so far, I always had a very positive experience and great community. I'm very optimistic that this will also be the case in this challenge!
==================================================== Any mechanism to prevent this in our module disease challenge? ==================================================== to be frank, other than self control and consciousness, practically no. they tend to tolerate this behavior, to encourage more partipation; and as i said such behavior might get one rewards. if i were the organizers, it will be different. and that's why i cannot be an organizer--because this is not something rare.   it truly stuns, disappoints, and hurts me constantly how one is willing to do anything to get a name on a paper. usually, even if i am the best-performer, i choose to decline the authorship if i have to work with such people. but i also now realize that it is a luxury to have such choices at the beginning and not everyone has this choice.   anyway, i found i am often too emotionally involved in this topic and make everyone hate me. (i suggested real-PI-name system to the organizers before, which made them really hate me and put a line against it in the dream principles.....) so i should stop here.
========================== if you look at past competitions, yes i have seen big teams who assemble and disassemble and then team up in the last round, so that they figure out the test cohort, structure while everyone else cannot. and yes, they get their prize and their talk and eventually their nature biotech paper. ========================== Any mechanism to prevent this in our module disease challenge?
hi, dong,   i am just speaking out of the voice of experience rather than from the organizers: this is my impression of dream compared to several other such competition platforms, e.g. kaggle, which i also coach kids on (but not active as my own, since my school doesn't like me spending too much time in competition). among all platforms, dream is the most generous one. when such behavior is observed in kaggle, one's account will immediately and permanently be terminated together with all your teammates, and your cell phone you used for registration. and your name will be blacklisted.   But, dream will find a way to tolerate and fix this behavior, which i believe it has an educational component in it. to me, and to be honest, i am not sure whether it is truly a good thing.   if you look at past competitions, yes i have seen big teams who assemble and disassemble and then team up in the last round, so that they figure out the test cohort, structure while everyone else cannot. and yes, they get their prize and their talk and eventually their nature biotech paper. and in rare cases in a competition with data leak that the answers are accidentally released, there will be teams willing to cheat the leaderboard and fit into the answers and top the leaderboard.   but to answer your question, i think one can think whether s/he is willing to win that way? what kind of people would like to win that way? does winning that way really bring you happiness? (i think it is out of question that it will bring you success) bearing these three questions in mind, i am sure we are all smart people, and everyone can easily find answers to any questions they have for the limits of leaderboard submission.   yuanfang
Hi Daniel, A new hand may be confused by option about individual or team when submitting. To be clearer, there should be just one option: as a team. And the limit (5) is the number of submissions **per team per subchallenge**? Our team prepared for both subchallenges. Another suggestion is that all teams must be non-overlapping:) like the modules we submitted. Lucky to solve it earlier instead of at late stage to prevent some people submit as individuals at first then form a team at last round.
Yes, we do not intend to disqualify teams that contact us now, but find a fair solution in the sense as you suggested. --daniel
hi, daniel,   i would suggest that no need to disqualify these teams, even that this was an item specified in rules multiple times. since usually it is the first time for someone to make such mistake.   BUT it is only fair to other teams if all submissions are deducted from the total. As it is clear to everyone, this challenge is all about how to figure out the appropriate size with in limited submissions. And I hope most participants will stand by me (and the right) on this point.   i would like to mention that speaking from the perspective a long-term participant (and kind of annoying one) of DREAM, I trust synapse has a full function of detecting such issues, as IP addresses from IDs are easily compared and matched and grouped by labs, units and then institutes (as we are all coders, and we all know that implementing this takes probably less than 1 hour).   And in the rare case that a submission indeed end up as the winning solution, how s/he achieves such position will be scrutinized by not just the organizers by also the community, including every leaderboard submission from this lab, and reading through every line of your code.   good luck for everyone's submission! y. g.

========= IMPORTANT: Members of the same team/group/lab are not allowed to submit as individuals!!! ========= page is loading…