how to prevent the end of the online chess game due to cheating?

@victorchess30 I think you might be over estimating your ability to detect cheaters.

What's background do you have that would give you the skills to detect cheating?

If your opponent takes more than ten seconds to play the first move, you abandon the game??? And you assume that person is cheating? What's the idea there?
It's not unusual for me to look up at the screen, say oh I've got a game, and settle myself in to concentrate before the first move.
I truly don't understand why you would see that as indicating cheating. You think they are using a computer on move 1?

In your comment #17, dear @ddwildin, you did not mention that you have already talked to him and have tried to convince him to stop blunder-checking his own moves with an engine during the game.
It was a wild speculation, it was wrong from me to have assumed that.
But please understand that it is your obligation to report your friend in case they didnt listen. :<
I think it is the responsibility everyone should take to help the administration of any chess site to keep the enviroment clean.
As long as you are silently tolerating this behaviour it is - sorry to repeat this again - "co-cheating".

there should be a way to detect cheaters before they commit the crime and prevent these people from making accounts

@victorchess30

This is really bad luck for you. Are you a magnet for cheaters or what??

I have played more than 5000 games on lichess.org and rarely played against a cheater (maybe 10 times or so that I can remember).

How can you be so sure that your opponent was cheating ? Maybe he just played a great game and you didn't (no offense intended).

Having "incredible elo level evolutions" can occur. There is a global pandemic and people are stuck inside their homes, so this is a great opportunity to learn chess and improve.
And differences in results at different cadences is very common for young or beginner players (young players play usually better at blitz because they don't want to wait a long time, and beginner players are better at slower time controls because they have trouble seeing tactics and so in rapid play).

Of course I don't deny that cheating exists here but I don't think it is quite common as you say.
Saying "I practically meet more cheaters than human players." That is a massive exageration.

A few words to the idea of a warning for cheat suspicion: I dislike it.
While the cheater could react to such a warning, the innocent has no chance to do anything against it. He remains feeling unsettled, helpless and angry. (And yes, for the same reason I also do not like the warnings against clock sitters)

thank you all for your interest in this topic.

I will clarify just one point, I am well aware that my reactions are excessive, but this is precisely the main consequence of cheating, that there is always a doubt.

as for the number of cheaters, I think like @belgianraf it's in the small levels that there are the most. at the higher levels, people have worked, it eliminates certain psychological profiles I think.

for those who don't know, there are free, undetectable softwares that can either play for you or just offer you a move, and whose strength can be set. you can play moves yourself or let it play at the right time. And this is valid for all speeds.

@Sarg0n Faced with the impossibility of controlling the problem, I suggest that a politely worded warning should deter at least some cheaters.

@tpr I completely agree with you : Denial is as bad as paranoia.

@h2b2 I am well aware that I don't have the level to estimate by myself if there is cheating. But I look at the analysis, the time taken to play, the complexity of the positions, and the profile of the player. I get an idea with all this. necessarily imperfect.

But the most important thing is that I've been playing on the internet for years, and I've been able to see the evolution. Before with the same means I almost never suspected cheating.

the problem has only gotten worse, (and I hope it's not just my mental health!)

@sparowe14 I know it's excessive, but I do it from experience.

@Sequoia_giganteum As said before, I think that at your level there are fewer cheaters. Cheating is a childish attitude. It happens more for beginners who can't stand the difficulty of progressing in this game.

@sheckley666 it seems simple to me to formulate a message that is not an accusation.

Well, please formulate.

@victorchess30 may I tell you that cheaters are indeed rare (I would say that 1 out 100-200 opponent is a "possible cheater"). Also, many people think that cheating involves playing only slow games and having a nice rating in classical/rapid. If you check on some fide profiles, you'll see that some people are way stronger at classical than in rapid or blitz. Also some people are obvious cheats (like having a rating of 2300 in classical but playing regular blitz games and having a rating of 1500), I think considereing people as non-cheats from the very beginning (even though you're losing) will help you feel less depressed. Many people think that every opponents who won a nice game against them is a cheat (even if the game lastezd like 10 moves and the opponent didn't have the time to do a single inacurracy). However, I'm not saying that cheaters don't exist. There's quite a bunch of cheaters but less than in other sites. But I think i agree wiht you: cheaters should receive a temporary banning for like 24h or 48h to prevent them playing against players.

@victorchess30
"I am well aware that I don't have the level to estimate by myself if there is cheating. But I look at the analysis, the time taken to play, the complexity of the positions, and the profile of the player. I get an idea with all this. necessarily imperfect."

It still takes a lot of chess knowledge (which we both don't have) to determine if the top move suggested by analysis is a natural human move or a move no human would make. If you played a title player, every move they play would be a top analysis move, but they would also be a natural move for a human to play.

Without that chess knowledge it's probably still possible use a computer to estimate the likelihood of computer assistance, however that requires different software that isn't built into lichess, and usually looks at many games.

Basically, I don't believe it's possible for you and me to accurately determine if someone cheated using the builtin analysis, and the fact you suspect a lot more people of cheating than most, I think you may have been getting it wrong.