Board Game Arena

SRE Turner wrote: ↑12 January 2024, 10:15

Remycad wrote: ↑11 January 2024, 18:50 Settings are now classic/difficult/6....
Isn't it a bit unreal to be able to find the code in 2 questions on almost any given puzzle, in real time mode, and with an average puzzle time of 2 or 3 minutes ?
Cause it looks that's what it's gonna take to top the arena leaderboard

I agree. I understand how you can quickly narrow down the possibilities at the easier difficulty levels. But when someone does a hard problem with only two questions, I strongly suspect they are looking up the possible answers rather than figuring it out themselves. I'm not sure why they bother, but it certainly makes it not fun for me.

Again, there's no reason you should assume that in a turn-based game. If you have time to work through the puzzle properly, plenty of Hard puzzles can still be done in 0–2 questions (though not all, of course). A good player *should* be using minimal questions, through deduction; it's not evidence of cheating.

Realtime games are different; you don't necessarily have time to think through all the deductions there. So reliably solving hard puzzles in realtime using minimal questions might be cause for suspicion.

Ultimately, the best way is just to ask them to talk you through their solution. I'm always happy to write out my deductions when asked how I reached the solution in so few questions (when I have the time to write it out of course). Someone cheating would probably struggle to do so.

poptasticboy wrote: ↑12 January 2024, 11:04

SRE Turner wrote: ↑12 January 2024, 10:15

Remycad wrote: ↑11 January 2024, 18:50 Settings are now classic/difficult/6....
Isn't it a bit unreal to be able to find the code in 2 questions on almost any given puzzle, in real time mode, and with an average puzzle time of 2 or 3 minutes ?
Cause it looks that's what it's gonna take to top the arena leaderboard

I agree. I understand how you can quickly narrow down the possibilities at the easier difficulty levels. But when someone does a hard problem with only two questions, I strongly suspect they are looking up the possible answers rather than figuring it out themselves. I'm not sure why they bother, but it certainly makes it not fun for me.

Again, there's no reason you should assume that in a turn-based game. If you have time to work through the puzzle properly, plenty of Hard puzzles can still be done in 0–2 questions (though not all, of course). A good player *should* be using minimal questions, through deduction; it's not evidence of cheating.

Realtime games are different; you don't necessarily have time to think through all the deductions there. So reliably solving hard puzzles in realtime using minimal questions might be cause for suspicion.

Ultimately, the best way is just to ask them to talk you through their solution. I'm always happy to write out my deductions when asked how I reached the solution in so few questions (when I have the time to write it out of course). Someone cheating would probably struggle to do so.

I was only talking about realtime: I assume someone could work it out in an async game. But in fairness, the one person I asked about it denied it.

Same here, only suspicion raised was about realtime games :
When I wrote my first message the leader of both arena and ELO leaderboards had won 82 of his last 84 games (could be a better ratio, it's just I stopped scrolling at this point ), for one draw and one loss !!
Those games being almost everytime realtime games classical/difficult/6, with let's say more than 75% of them solved in 2 questions and less than 3 minutes.
As you can see the timestamps of the actions in the replays, it usually takes him less than one minute to ask the 1st question, chose the 2nd one and give his solution.
And even if I was candid enough to believe it was possible in such a short time, then other players could do so too, and there should be at least 10 or 20 % of the games drawn, not 1% !!

Remycad wrote: ↑12 January 2024, 16:43 Same here, only suspicion raised was about realtime games :
When I wrote my first message the leader of both arena and ELO leaderboards had won 82 of his last 84 games (could be a better ratio, it's just I stopped scrolling at this point ), for one draw and one loss !!
Those games being almost everytime realtime games classical/difficult/6, with let's say more than 75% of them solved in 2 questions and less than 3 minutes.
As you can see the timestamps of the actions in the replays, it usually takes him less than one minute to ask the 1st question, chose the 2nd one and give his solution.
And even if I was candid enough to believe it was possible in such a short time, then other players could do so too, and there should be at least 10 or 20 % of the games drawn, not 1% !!

I would said at least 42.69% personnally.


Classical hard 6 is the arena setting, if they play arena, normal to have way more of those. And in many games top players plays a single setting, because it's the way to reach the top. If a setting is better than another to gain elo, those who only play this settings should get the top place, while those who don't, could be better but would not get the first few places.

And did you check if others players have that same winrate?
I have 800 elo in another game, and to keep it, you need to have an absurd winrate against 'weak' players (like 99% against new players, and 95% against 100/200 elos). While you still need a good winrate against strong opponent (like 60%/70%).
So unless those players only play with strong opponent, they should all have a really high winrate. And that's the case if they play arena.
And ties don't make you lose elo on that game.

*post removed by moderators

*post removed by moderators

[/quote]

Unless I'm missing something, I don't see how that can be used to cheat. It's just an interactive sheet, and you need to know the puzzle code to use it.

I've working on a computer program to play this. Isn't finished, but I was shocked at just how many possible solutions can be eliminated just by looking at what verifiers are available. About a third of the time I can identify a single solution without making any tests at all. Most of the time there are two possible solutions, one test will resolve them. Very occasionally I've been left with up to 6 possible solutions which can take 3 tests. I don't know if the top players are using human deduction to limit the potential solutions, or if they have their own computer programs, but the potentially to find the solution in zero to one test is certainly there.

Maybe the find the solution by using "Create a challenge" on https://turingmachine.info/ ?
You just need to check the solution and it is possible to finish without asking one question ?

*Moderation Edit: please do not use the Forums to call out specific tables or players by name, this is what the Moderation Reports are for. The Forums should remain a general discussion.