Time Delay Measure Against Cheaters

Yes, this is another anti-cheat measure. Yes, I have done some due diligence, and checked if a similar topic/proposal exists.

Meat of the problem: It's far easier to cheat than to be banned. The resources needed to investigate far outweigh the resources needed by cheaters to cheat.

Suggestion: Make it more expensive for cheaters to cheat.

Main Target: Troll Cheaters, those who keep on creating accounts just to troll everyone else.

Suggestions:
There are many ways to do this, but the main idea is the same. Resources (time and energy) have to be spent in order to join the main events. Players can also filter out those who haven't spent the necessary resources during match requests.

1. BRUTE FORCE: Just create a probationary server and a main server. You get promoted to the main server after 1 month and 1,000 points. For every win, bullet 2 points, blitz 4, rapid 8, classical 16. Half points per draw. One must meet BOTH criteria in order to be promoted.

2. GAMIFY: Some sort of league system, but the idea is the same, make it more time consuming to get promoted. Have all good and important events such as Shield Arenas have minimum league requirements. Also allow players to have league match search filters.

3. FILTERS: Have the option for match searches and tournaments to have minimum number of games and minimum account age. For example, have an Hourly Closed Blitz Tournament, with requirements like a minimum of 200 blitz games should have been played + account age should at least be 1 month old. A player should also have those filters when searching for matches. Those who don't have those criteria can just join the Hourly Open Blitz Tournament.

How it will work:
1. By the time the potential cheater enters the main pool, the server would have a number of game samples for AI checking to see if the player indeed cheated.

2. It would take quite a while for a troll cheater to troll the main pool.

3. If the troll cheater is dedicated enough and plays fairly in the open pool, then that's good for everyone in the open pool since he's not cheating there. And then once he gets into the main pool, he can only cheat once and have to go through the whole process all over again. We can imagine a super dedicated troll maintaining several accounts at once with different due dates. But this is resource expensive, and can be made more expensive simply by increasing the requirements to enter the main pool. I'm sure Lichess can easily figure out the optimal number of points to weed them out. Just imagine the number of man-hours required to maintain several accounts. And if he ever tries to automate it, that should be easily detectable.

4. If the troll cheater decides to cheat in the open pool, he can be eliminated before he gets into the main pool. It's also not as satisfying for the troll. Another win for the good guys.

SUMMARY:
Make it more expensive for a troll cheater to cheat.

@GnocchiPup said in #1:
>
> 1. BRUTE FORCE: Just create a probationary server and a main server. You get promoted to the main server after 1 month and 1,000 points. For every win, bullet 2 points, blitz 4, rapid 8, classical 16. Half points per draw. One must meet BOTH criteria in order to be promoted.
>

This would not promote more cheating?

@magicsacrifblunder said in #2:
> This would not promote more cheating?

Maybe, but it's fine. What wete trying to do is protect the main server from newly created accounts whose main purpose is to troll cheat.

They wouldn't be able to get into the main server if they cheat in the probationary server.

Take note that one doesn't need to get to 1k points within a month. They just need to get to 1k points. For example, player A gets 1k after 3 months, he gets promoted. Player B gets 1k after 3 weeks, he needs to wait one more week before he gets promoted.

@GnocchiPup said in #1:
> Those who don't have those criteria can just join the Hourly Open Blitz Tournament.

Tournaments have a minimum games requirement.

@GnocchiPup said in #3:
> Maybe, but it's fine. What wete trying to do is protect the main server from newly created accounts whose main purpose is to troll cheat.
>
> They wouldn't be able to get into the main server if they cheat in the probationary server.
>
> Take note that one doesn't need to get to 1k points within a month. They just need to get to 1k points. For example, player A gets 1k after 3 months, he gets promoted. Player B gets 1k after 3 weeks, he needs to wait one more week before he gets promoted.
How are you gonna get those 1k points in a pool of sharks? Or I'm missing something...

In general, I'm a proponent of additional cheat detection research (a field is quite unexplored!) instead of turning to these labyrinthine "solutions".

@Toadofsky said in #4:
> Tournaments have a minimum games requirement.

1. Not minimum enough, just 15 rated games. What I'm suggesting is probably around 50 wins (not games).
2. Also an account age requirement. This is probably the tougher of the two requirements.

The goal is for brand new accounts to not be able to join Closed tournaments willy-nilly. They can opt to join Open ones though.

Similar to game search filter. I want to filter out accounts which are less than one month old and fewer than 50 wins.

@GnocchiPup said in #1:
> Yes, this is another anti-cheat measure. Yes, I have done some due diligence, and checked if a similar topic/proposal exists.
>
> Meat of the problem: It's far easier to cheat than to be banned. The resources needed to investigate far outweigh the resources needed by cheaters to cheat.
>
> Suggestion: Make it more expensive for cheaters to cheat.
>
> Main Target: Troll Cheaters, those who keep on creating accounts just to troll everyone else.
>
> Suggestions:
> There are many ways to do this, but the main idea is the same. Resources (time and energy) have to be spent in order to join the main events. Players can also filter out those who haven't spent the necessary resources during match requests.
>
> 1. BRUTE FORCE: Just create a probationary server and a main server. You get promoted to the main server after 1 month and 1,000 points. For every win, bullet 2 points, blitz 4, rapid 8, classical 16. Half points per draw. One must meet BOTH criteria in order to be promoted.
>
> 2. GAMIFY: Some sort of league system, but the idea is the same, make it more time consuming to get promoted. Have all good and important events such as Shield Arenas have minimum league requirements. Also allow players to have league match search filters.
>
> 3. FILTERS: Have the option for match searches and tournaments to have minimum number of games and minimum account age. For example, have an Hourly Closed Blitz Tournament, with requirements like a minimum of 200 blitz games should have been played + account age should at least be 1 month old. A player should also have those filters when searching for matches. Those who don't have those criteria can just join the Hourly Open Blitz Tournament.
>
> How it will work:
> 1. By the time the potential cheater enters the main pool, the server would have a number of game samples for AI checking to see if the player indeed cheated.
>
> 2. It would take quite a while for a troll cheater to troll the main pool.
>
> 3. If the troll cheater is dedicated enough and plays fairly in the open pool, then that's good for everyone in the open pool since he's not cheating there. And then once he gets into the main pool, he can only cheat once and have to go through the whole process all over again. We can imagine a super dedicated troll maintaining several accounts at once with different due dates. But this is resource expensive, and can be made more expensive simply by increasing the requirements to enter the main pool. I'm sure Lichess can easily figure out the optimal number of points to weed them out. Just imagine the number of man-hours required to maintain several accounts. And if he ever tries to automate it, that should be easily detectable.
>
> 4. If the troll cheater decides to cheat in the open pool, he can be eliminated before he gets into the main pool. It's also not as satisfying for the troll. Another win for the good guys.
>
> SUMMARY:
> Make it more expensive for a troll cheater to cheat.

There are lots of faults in this.
1. There are a few people who might start cheating just after they join the main server.
2. Not everybody likes to play chess but they want to just play a particular variant only, so the points system makes no sense.
3. About the filters, there are some variants like Horde for example, where its impossible to get a person to play against you, especially if you don't like 1/2+0. This is because 99% of people who send horde seeks are either 1/2+0 or too slow like 10+0 or slower. Those who play proper horde (3+0 and etc) are all 2200+ rated, so new players will not like the variant, as they would always lose. So the only way to get a game with someone of your level there would be to ask a friend to play, but those friends might not be online all the time you are. Also, new players may not know whom to speak with to get some games. Hence, 200 games seems impossible.
4. There are also people like Magnus Carlsen who decide to make an account just one day before the titled arena and want to play it without anybody knowing its them. So again, 200 games seems impossible in less than 24 hours (it would also give time for others to figure out its magnus).
5. I also have noticed that there were a few cheats (not gonna name them for obvious reasons) who would play without engine in all fast TCs (1+0 and lower) and cheat in slower tc's. So, the cheat can play 200 games of 1+0, then start playing 2+1 in the main server and start cheating.

So basically, this system which you proposed doesn't work.

@cFlour said in #8:
> There are lots of faults in this.
> 1. There are a few people who might start cheating just after they join the main server.
> 2. Not everybody likes to play chess but they want to just play a particular variant only, so the points system makes no sense.
> 3. About the filters, there are some variants like Horde for example, where its impossible to get a person to play against you, especially if you don't like 1/2+0. This is because 99% of people who send horde seeks are either 1/2+0 or too slow like 10+0 or slower. Those who play proper horde (3+0 and etc) are all 2200+ rated, so new players will not like the variant, as they would always lose. So the only way to get a game with someone of your level there would be to ask a friend to play, but those friends might not be online all the time you are. Also, new players may not know whom to speak with to get some games. Hence, 200 games seems impossible.
> 4. There are also people like Magnus Carlsen who decide to make an account just one day before the titled arena and want to play it without anybody knowing its them. So again, 200 games seems impossible in less than 24 hours (it would also give time for others to figure out its magnus).
> 5. I also have noticed that there were a few cheats (not gonna name them for obvious reasons) who would play without engine in all fast TCs (1+0 and lower) and cheat in slower tc's. So, the cheat can play 200 games of 1+0, then start playing 2+1 in the main server and start cheating.
>
> So basically, this system which you proposed doesn't work.

> 1. There are a few people who might start cheating just after they join the main server.
Yes, I even gave an example that it can happen and most probably will happen. I've also mentioned that it specifically targets troll cheat accounts. Those who just create new accounts just to cheat.

2 and 3. Those are just examples, the main point is a month of non-cheating behavior worth x number of wins.

4. Titled players automatically get into the main server.

5. These are different types of cheaters altogether. I call them the determined cheater, as opposed to the troll cheater.

In any case, try looking at it from the cost perspective.

Current
There's practically no cost to cheating. One gets banned, creates a new account. Done in less than an hour.
But it costs a lot of time and resources to investigate, and then ban a cheater.

Proposed
The cost for cheaters = one month of good behavior. Yes, they can do that and cheat again, but this time the cost balance is reversed. It's now more costly to repeatedly cheat than to repeatedly ban.

So it doesn't work if your definition of work = absolute cheating prevention. That's just impossible.

But my definition is simple, just make the cheaters use up more time and energy vs. the investigators. And one month of good behavior should be more time and energy-consuming enough.

@Arckai said in #5:
> How are you gonna get those 1k points in a pool of sharks? Or I'm missing something...

Just ignore my points system.
The general idea is to have at least one month of good behavior quantified by let's say 100 wins without cheating.

For example, I would like to play in a tournament where everyone's account is at least 1 month old, and everyone has around at least 50 wins. This will not guarantee zero cheating, but it surely would have less cheaters than what we currently have.

I would also like to set my game search filter, aside from color and rating, also set # of wins greater than or equal to 50, and account is greater than or equal to one month old.