Why is an unconvential move cheating?

Question

I heard about Brandon Jacobson, a GM, having been banned from chess.com for cheating.

It looks that the reason was that

[their anonymous account] reached an elo above 3100, but there’s a catch – they sacrificed their Rook on the second move of every game and still won.

Some [chess players, and chess.com] argue that his actions were clearly cheating, as sacrificing a Rook in every game goes against basic chess principles.

My knowledge of chess is absolutely basic (I play with my children from time to time) so I cannot understand the controversy: how playing a legitimate move is cheating? Even if you play it every time (I would argue that this puts you in an disadvantage because you are somehow predictable at the beginning).

Where is the cheating here?

Answer 1

The "main line" of this "opening" is, I suppose, 1.a4 e5 2.Ra3 Bxa3 3.Nxa3. White is down by an exchange, say 1 1/2 pawns. He has developed his QN but to a bad square. This cannot possibly be good. However, there are some mitigating factors. White does have the two Bishops, and the dark square Bishop is unopposed. A polite way to describe the state of Whites development is that it "remains flexible". A strong, fast, player who has put themselves frequently into this situation knows what they have to do, even if some of it is just figuring out what to do with that Knight.

For Black the situation is not difficult in principle. They should find a way to exert pressure that is enough to make White concede to exchanges, and then win the endgame, which means playing a long game. But this is risky against a fast player at a short time control. Black will need to spend much time strategizing an unfamiliar position, instead of engaging in a rapid flurry of traditional manoeuvres. For Jacobsen, this worked and there has been a brief flurry of interest in this "banned opening" even extending to imagining that it may have some objective merit. If the fad persists, people will prepare for it, and it will soon fade.

For chess.com, their anti-cheating measures involve assuming that unusually favorable statistics imply cheating, and it is hard to see what else they can do. It is however inevitable that even the best statistical methods can only be right most of the time, and must be wrong some of the time. Unusual statistics may simply imply unusual circumstances, and likelihood estimates usually ignore these. Chess.com relied on statistics to discredit and ban Hans Niemann in the past, but eventually backed down. They may back down on this.

To furnish the OP with an answer, playing a legitimate move is certainly not cheating (although deliberately playing a bad move might be seen as insulting and thereby applying psychological pressure). Although there is some statistical evidence that might suggest some other form of cheating, there is no direct evidence, and there is a plausible explanation without cheating. Magnus Carlsen supported this by playing the "banned opening" in a titled Tuesday event with the predictable outcome that he placed lower that he normally would, but still finished high in the ranking.

Answer 2

Let me give you an analogy: the world record for running a marathon (~42km) is at about 2 hours. The world class marathon runners consistently come close to that time.

Now, suppose I start to "run" marathons not by starting to run, like the others, but by spending the first hour sitting down, having a coffee and reading a book. After that one hour, while the others have already covered roughly half the distance, you see me finally starting to run and the next thing you see is that at the finish I am head to head with the other world class athletes.

Now, ask yourself: might I have just come up with an "unconventional way of running"? Or might it be that I used some method to cheat - hop into a car, use a bicycle, take a short-cut - in the time between start and finish?