Rapporter • Board Game Arena

Alle rapporter

Turing Machine rapporter

#109718: "Cheating? Player not entering any question and directly testing the right code"

notabug: Dette er ikke en feil

Hva handler denne rapporten om?

Turing Machine #458628987

Hva har skjedd, eller hva gjelder det? Vennligst velg

Regler: en regel i spillet har ikke blitt respektert

Hva har skjedd, eller hva gjelder det? Vennligst velg

Regler: en regel i spillet har ikke blitt respektert

Vennligst sjekk om det allerede er en rapport om samme emne

Hvis ja, vennligst STEM på denne rapporten. Rapporter med flest stemmer er gitt PRIORITET!

#	Status	Votes	Game	Type	Title	Last update

Detaljert beskrivelse

Rapporthistorikk

Jan 6 th 2024 11:47 • Barbanera • Feilen har ikke blitt reprodusert av utviklerne enda:

Just replay the whole game, it's really short.

Jan 6 th 2024 12:02 • Barbanera • Feilen har ikke blitt reprodusert av utviklerne enda:

I checked my opponents recent games history and he/she appears to be winning all games in no time just entering the right code, same as in my game.

Jan 6 th 2024 12:04 • tan_oduvan • Feilen har ikke blitt reprodusert av utviklerne enda:

boardgamearena.com/table?table=458639202
in the my game, the person also did not select a single verifier by entering the code correctly

Jan 6 th 2024 12:25 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

So there's two things.

1. Those problems are absolutely solvable without needing any tests, just using pure deduction. They're actually relatively straightforward, I've solved them in that way myself. No cheating is required, given enough time.

2. There are very few puzzles with the settings Classic/Standard/6 verifiers, and they are all variants of the same basic puzzle. Once you have solved one, it's very easy to see at a glance what the answer will be to any of them.

It's not cheating, but it's definitely a bad idea to allow players to choose their own settings for Arena Mode, as is currently the case. There need to be fixed settings for the season so that people can not choose to use settings with such a limited range of puzzles.

Jan 6 th 2024 12:34 • Barbanera • Feilen har ikke blitt reprodusert av utviklerne enda:

I believe it is cheating and after a bit of investigation I just found how I could cheat myself. I am not sure I should post it here but if you msg me privately I will explain.

Jan 6 th 2024 12:35 • Barbanera • Feilen har ikke blitt reprodusert av utviklerne enda:

Assuming you are a developer or ufficial BGA supervisor or something, that is.

Jan 6 th 2024 14:27 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

*edit* I meant Classic/Easy/6 above, not Standard.

Jan 6 th 2024 14:29 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

No I'm just a player with too much time on my hands lol.
I started responding to "bug" reports as there were so many that weren't actually bugs, just people misunderstanding, and I can't seem to stop myself checking them 😁

Jan 6 th 2024 14:41 • Barbanera • Feilen har ikke blitt reprodusert av utviklerne enda:

Ok, thanks to poptasticboy expert help, I think that I should clarify that the "cheating" involves probably not a real cheat per se (like knowing the result by illegal means) but what I believe some people (not poptasticboy!) might be doing, i.e. having access to look-up tables of possible results.

These look-tables can be prepared beforehand by anybody, given enough time and commitment, or by other automatic means, involving no thought process whatsoever. While the former can be perhaps acceptable to some (not to me personally), the latter is probably a no no to everyone.

Anyway, having look-up tables to consult means one can solve a problem with very few questions (1-2 or, in some cases, even 0). Imho, this makes this great game one to be avoided for online games, unless you really trust your opponents to play fair (i.e. reason it all out on the spot) and play casual or select at least standard, if not even hard, mode, and hope for the best, if you really need to play arena.

Jan 6 th 2024 17:51 • Romain672 • Feilen har ikke blitt reprodusert av utviklerne enda:

I believe many people play with mainly just deduction (and potentially memory), and don't have look up tables.

In 3min, you can do most of the potential deductions.

Jan 10 th 2024 07:10 • Denlerien • Feilen har ikke blitt reprodusert av utviklerne enda:

Same - 1 round zero questions victory:

459792785
459806611
459790596

I don't know how game is working, I think that there are some limitations and not any verifiers combination is possible, but I've wrote simple prog in Python (onlinegdb.com/KmAQYPjm7) to check table in topic (458628987) and it looks like there are 34 different combinations of verifiers where are only single answer is correct. For example if verifers are 3,1,1,1,1,1 (A verifies variant 3, i.e. square>3 and others verifies first variant), then there is also single answer - 245.

So without knowing how real game is working there is no way logically find out answer only using logic.

Jan 10 th 2024 07:15 • Denlerien • Feilen har ikke blitt reprodusert av utviklerne enda:

p.s. "So without knowing how real game is working there is no way logically find out answer only using logic." I mean find solution without any questions, only by verifiers combinations.

Jan 10 th 2024 09:33 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

@Denlarian "So without knowing how real game is working there is no way logically find out answer only using logic."

It is clear you do *not* know how the game works!
You are likely missing the 2 most important rules:

1) No verifier can be redundant
2) There must be a unique solution.

Let's give a simple example: if the Verifiers for "Yellow compared to Purple" and "Which colour is smallest" are both in a puzzle, you know right away, without any tests, that Blue is smallest. This is because any other result would make the other verifier redundant (i.e. it would provide no new information), which is not allowed in a valid puzzle.

Using the 2 listed rules, it is usually possible to solve a Classic puzzle in under 3 tests, and frequently they can be solved in 0 tests.

Jan 10 th 2024 11:23 • Denlerien • Feilen har ikke blitt reprodusert av utviklerne enda:

@poptasticboy, I've checked so that it would be unique solution, of course. But you are right - I have forgotten to check are there any redundant verifiers. I've fixed it - and yes, for table 458628987 only one answer is correct. So using such program someone can easy find solution without any questions. And of course it can be done logicaly.

Jan 11 th 2024 19:44 • mdross1 • Feilen har ikke blitt reprodusert av utviklerne enda:

Same thing is happening with player DJBisdaGOAT at table 460594398, for instance.

It may be a matter of BGA only using a limited number of game sets, but this is definitely against the spirit and written rules of the game.

Jan 11 th 2024 21:02 • mdross1 • Feilen har ikke blitt reprodusert av utviklerne enda:

460117764 As well, a single verifier used to win.

Jan 11 th 2024 21:32 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

@mdross1 you seem to be missing the point. Those are both turn-based easy difficulty games. I haven't taken the time to study those specific puzzles myself, but in most cases anyone good at the game *should* typically be solving those puzzles in 0-2 questions. That's not evidence of cheating, as a skilled played *should* be getting those results.

Now if they can reliable do the same in hard/realtime games you might have a point.

Jan 11 th 2024 21:34 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

@mdross1 you seem to be missing the point. Those are both turn-based easy difficulty games. I haven't taken the time to study those specific puzzles myself, but in most cases anyone good at the game *should* typically be solving those puzzles in 0-2 questions. That's not evidence of cheating—a skilled played *should* be getting those results.

Now if they can reliable do the same in hard/realtime games you might have a point.

Jan 11 th 2024 21:35 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

(sorry for double post, I lost signal)

Jan 12 th 2024 07:01 • DogEgg258 • Feilen har ikke blitt reprodusert av utviklerne enda:

@Barbanera Please provide sufficient evidence when you point out that someone else cheated. For example some game that must not be solved with 0 question but they did. Instead of imagining a possibility out of thin air.

Jan 12 th 2024 18:21 • mdross1 • Feilen har ikke blitt reprodusert av utviklerne enda:

@poptasticboy

Thanks! To be honest I am new to the game and think I was missing a level of nuance here.

The rules say "You will need the criteria of all verifiers to find the code. No criterion is superfluous", which to me suggested that at an absolute minimum any number of questions that's fewer than the number of verifiers is suspect. That being said, the fine folk at boardgamegeek have certainly identified a number of game states that can be solved with 0 questions ( boardgamegeek.com/thread/3166002/no-question-challenges-5-cards-classic-easy-mode-o ) so I'm happy to concede that a 0 question solution isn't a guarantee of cheating.

I still have uncomfortable feelings about 0 question solutions on BGA as it's certainly possible that there's a limited set of easy puzzles that people are able to look up for quick wins and the losers wouldn't be able to tell, but if BGA is using a large catalogue then I guess it's just time to sharpen my pencil!

Jan 14 th 2024 14:07 • ollyfish2002 • Feilen har ikke blitt reprodusert av utviklerne enda:

everyone can cheat using this even without asking questions
github.com/ThomasdenH/turing-machine-ai

Jan 15 th 2024 00:37 • kirin0205 • Feilen har ikke blitt reprodusert av utviklerne enda:

Turing Machine n°462165339

J ai testé le bon code du 1er coup par hasard.
Mais il y avait 6 cartes de tests, donc impossible de conclure.

Le joueur en face a fait 2 tests et validé LA BONNE RÉPONSE.

Je pense que dans les lignes de code, il apparaît le numéro qu il y a dans le jeu.
La personne en face me semble avoir triché et utilisé le livret de réponses inclus dans le jeu.

Jan 15 th 2024 03:27 • poptasticboy • Feilen har ikke blitt reprodusert av utviklerne enda:

@kirin0205 your puzzle can be solved in 2 questions with no cheating involved; I just did so myself. Verifiers B & C can be deduced easily without tests. With more thought, you can also deduce A. After testing D you can deduce F as well, leaving just E for the second test, to solve the puzzle.

However unless I'm mistaken, the specific test your opponent gave to E didn't actually give the required information. I was left with 354 or 435 at that point, and their specific test would not help to narrow that down (but a different test code would).

So in conclusion, getting the correct answer in 2 tests is completely possible, purely by deduction. However unless I'm mistaken, your opponent's specific tests would actually have left 2 possible answers, so they only had a 50/50 chance of getting the correct answer.

Jan 24 th 2024 15:46 • SraTacoMal • Feilen har ikke blitt reprodusert av utviklerne enda:

I solved table #458320917 with no verifiers. imgur.com/a/IUfFjum

My solution wasn't a guarantee (actually I tried solving it again just now and got it wrong), but at the time it felt like the most likely solution, and it ended up being correct. Here was my logic:

-Verifier C gives the only piece of information for purple, so purple should be the smallest
-Verifier D is questioning how many even numbers there are. Usually, but not always, in this case there are 2 even numbers.
-There's not a lot of info about what blue and yellow are. I doubt that they're the same number, though, because unless they're both 4, there's not enough info to tell what they are. And if they are both 4, then there's not enough info to tell what purple is (unless Verifier D says that all 3 are even numbers, which I doubt). If yellow is 5, then blue must be lower and could be 2-4. If yellow is 2 or 3, there's the same issue with multiple possibilities for blue. If yellow is 4, blue could be 2 or 3, or 5. 5 is the only one that is sequestered off, so I believe Verifier B will say that blue is greater than yellow, and Verifier A will say yellow is 4.
-Since purple is the smallest, it must be the only even number remaining: 2.

The only thing I was positive about was that purple had to be the smallest. The rest could have easily been wrong.

Aug 30 th 2024 17:07 • vfbfzfgz • Feilen har ikke blitt reprodusert av utviklerne enda:

easy puzzles often have one possible solution, and it can be deduced without asking a thing.

Dec 19 th 2024 15:11 • Tisaac • Utviklerne ønsker mer informasjon for å kunne reprodusere denne programfeilen:

Please reach out to me in PM if you have any sensitive information on how to cheat. From my code review, I don't think any information is leaked.

Dec 24 th 2024 10:36 • Tisaac • Dette er ikke en feil:

Legg til noe i denne rapporten

Legg til her noe som synes relevant for å reprodusere denne feilen eller forstå forslaget ditt:

En annen bord ID / flytt ID
Løste F5 problemet?
Oppstod problemet gjentatte ganger? Hver gang? Tilfeldig?
If you have a screenshot of this bug (good practice), you can use a picture hosting service of your choice (snipboard.io for example) to upload it and copy/paste the link here.

Avbryt Oppdater rapport

#109718: "Cheating? Player not entering any question and directly testing the right code"

Hva handler denne rapporten om?

Hva har skjedd, eller hva gjelder det? Vennligst velg

Hva har skjedd, eller hva gjelder det? Vennligst velg

Vennligst sjekk om det allerede er en rapport om samme emne

Detaljert beskrivelse

Rapporthistorikk

Legg til noe i denne rapporten

Endre rapportstatus til:

Hjelp

Bidra

Navigere

Følg oss