Source-
github.com/mr-yt12/lichess-gui-use-code-example/blob/master/Why%20KB%20is%20fair.mdpeople have been using mouses to play ultrabullet, or tablets, or phones, and since KB is so different and much faster, it gives an unfair advantage and therefore isn't fair.
Refutation:
ultrabullet or bullet weren't created or meant to be a mouse skill game. It is only by accident, due to limitations of our input methods. It just so happened that mouses have been the most common input device. Playing with a mouse is not a rule, nor is it a part of the game, it is most common simply due to the fact that there were no alternatives. If a sport is created with a rule, the rule becomes a part of the game, i.e. you cannot use hands in soccer, even though it would be more efficient. Fast chess weren't "created" with such rules, they only inherited the rules of chess, input methods are totally optional.
P.S. This argument is not about the time when the rule is / isn't created. It's about whether the game should have a certain rule, based on its purpose. I.e. it's not that the rules cannot be changed once the game has already been created, it's about whether the game's purpose presupposes a certain rule. The reasoning behind why fast chess shouldn't be considered a mouse skill game is discussed in the next paragraph.
Argument:
let's imagine someone created a big chess board with heavy chess pieces (the size of a person), and called it "Chess Lifting". Stronger people would now have an advantage. If someone used a forklift to move pieces, that would be unfair and cheating, since anyone with smaller muscles could now beat a bodybuilder. Using KB in ultrabullet is the same, since anyone using KB can easily flag a mouse player.
Refutation:
Shorter time controls and heavier pieces are different things. In the case of heavier pieces we can only see it being introduced for a single purpose: to make weightlifting a part of the game. Therefore it becomes a thing that has been defined by a rule, by agreement. Trying to invent a different method of weightlifting (using exoskeleton, for example) would be breaking rules of the game in question. Unless everyone used exoskeletons; but this would make it a different game, or a subdivision of another game. The previous game (where some other methods of lifting weights are used) would remain, even if everyone started playing the newly created game. The rule (weightlifting as a part of the game) was introduced intentionally and anything that would give an unfair advantage there would be cheating. Same as in football, where one cannot touch the ball with their hands. It's a part of the game, it's been introduced as a rule.
In the case of shorter time controls, it can be argued that a shorter time control has been introduced not to make the game depend on mouse skills, but to make the game depend on faster thinking. Both are important to win in fast chess, but the former could be considered a side effect, stemming from the limitations of our current technology. It is also consistent with the history of chess game duration. Before the internet was invented, almost no one played bullet. When digital input methods were introduced, people started playing online chess and some of the players — shorter time controls. New input methods allowed shortening of the chess game.
So it could be argued that ultrabullet was added as the next step in reducing the duration of the game and to create a game where fast thinking is more important, similar to how it is with sprinting and distance running. It doesn't diminish longer chess, just creates a new type of game. But, since our input methods can't yet catch up with the speed of thought, the game partially turns into a game of faster input skills, as a side effect. So, by improving our input methods, or choosing completely different input methods that don't calculate moves themselves, we get closer to competing only in faster thinking skills. Once we can just think of the move and it's instantly played, we could say that we reached the point where the speed of thought and input methods have been equalized.
Argument:
when using other input methods, you have to decide on your exact move, for example by dragging from e2 to e4, or typying e4. With the KB extension you can, for example, swipe along the fourth horizontal while pressing w, therefore you don't decide on what your move will be, but the move is randomly chosen (it can be either e4, d4, c4, and you can't know which). Therefore it becomes a Random Mover BOT and therefore KB is unfair.
Refutation:
you still decide on your every move. Let's look at how the mouse can be used in ultrabullet. Let's say that you don't care which move / premove to play: b4, c4, d4, e4, f4. So you start moving them in turn. You drag from b2 to b4, c3 (you have a pawn on c3) to c4 and so on. You drag pawns one after the other. What will decide your move in this case? We could say it's decided by circumstances. For example, at some point when c4 is premoved, if your opponent makes a move, c4 will be played. But why? If your opponent had taken a bit longer to play their move, you would have already premoved d4, and it would have been played instead. Have you decided on your move then? Or has it been circumstances? In this case you wanted either of your pawns to move to the forth diagonal, and it wasn't important for you which one would. You had a set of moves in mind and you tried all of them as quickly as you could, you have decided on each of them, but only one of them has been played, because of the circumstances.
When you swipe with your cursor along the fourth horizontal while pressing w, you also decide on your moves, however you do it much quicker. Your moves are still sequential. You don't choose the whole horizontal and let the computer randomly choose a square. Your cursor is hovering over different squares at each point in time. What move will be executed depends on the circumstances, the move is not chosen randomly by a computer algorithm. What move will be played depends on circumstances both when using the mouse and KB. At each point in time, only one move is chosen. Moreover, you have control on when you start and finish pressing w.
Now let's say it's your move, and you need to move some pawn to the fourth diagonal. You swipe across the forth diagonal, and one of your pawns moves there. Let's say some pawns couldn't move, because those moves would be illegal. Could it be said that the computer has chosen a legal move for you? If you play with the mouse and make an illegal move, what will happen? The move won't be played. Now let's say you have very little time and you don't see that one of your pawns can't move, but anyway just in case you decide to quickly move two pawns sequentially, because you understand that if you move one of them, and it doesn't move, having to react to this event and try another move will take more time than trying to move more than one pawn right away. Has your move been randomly chosen by a computer algorithm? Lichess board is programmed to discard illegal moves. They simply don't get played. It doesn't constitute a penalty for you. When you swipe with KB, same as how it is with the mouse, the moves are not chosen for you. Lichess board doesn't decide on a move for you, it simply discards all illegal moves until you input a legal one. When using KB, you don't command to play either of a set of moves. You play them continuously and deliberately (you always have control of when to start and stop swiping, or pressing the key, or the direction of your cursor). Illegal moves are simply discarded (that's how Lichess is functioning) and once a legal move has been chosen by you, it will be played.
One might think it's impossible to decide on the moves so quickly, and they would therefore think that this argument is fallacious because it uses technicalities to construct a loophole in the common sense to justify KB usage. The counterargument to that is that you certainly can decide on your moves so quickly and here's how. Imagine you need to type "12345" quickly as a password. You've done it many times and you need these digits to be typed in that exact order. You cannot just press them all at once for the computer to chose the order randomly. You have to actually make sure "1" is pressed first, "2" -- second, and so on. How quickly can you do it? Imagine using both hands, or one hand. Try practicing. After enough practice you will be able to get it right almost each time and extremely quickly.
One might counterargue that a chess move consists of two inputs: the initial and the destination square". And therefore you have to use at least two letters. But that is not a rule, it's just a necessity to be able to communicate moves in some cases. You can use "Be3", "e5" notation to communicate moves, and it is as discrete and as legit as UCI. Using KB is similar to using the algebraic chess notation. You choose the piece on your keyboard (B -- bishop) and you choose the square with your cursor (e3). When there are two pieces that can go to the same square, you can choose which one will do so, because twin pieces (knights and rooks) are assigned two keys and one of the pieces is marked with a green square.
Argument:
king walks (circling your king in a time scramble) are unfair because it's not the user who generates the moves, but the KB extension.
Refutation:
it's still the user who generates the moves. Imagine a device that could read your thoughts and transfer them directly into the computer. Now imagine thinking about the king moving up. Then imagine the king moving up and down quickly. Then imagine the king going in circles, imagine quickly thinking about the king going to all nearby squares. You don't have to name the moves, direction, or say the words "king", "square". Just your slightest thought of the king going somewhere will actually move the king there, if the move is legal. Using KB is similar, the difference is that you first have to transfer your thoughts to your hand muscles. So moving your hand whatever way is the same thing you could achieve with your thoughts. The fact that it might be reflex and that the moves are not always deliberate doesn't invalidate the legitimacy of such moves, because your thoughts can be also reflex, and there is nothing invalid about producing an action in a game reflexively. When playing with a mouse, certain quick moves are also produced reflexively.
Argument:
multipremoves are unfair because Lichess is a one premove website.
Refutation:
Just because a feature doesn't exist, doesn't make it unfair to obtain it via different means, unless there is a rule that specifically restricts the use of this feature. But Lichess is a one premove website not by reason of a rule that restricts the amount of premoves, but because people who develop Lichess might not think it's worth adding multiple premoves, as one premove + hover is better and simpler.
What are some uses of multiple premoves:
Premoving a checkmate when the opponent leaves in hope to return later and claim the victory.
Time scrambles: being able to effectively promote pawns, being able to deliever checkmates with little time.
Fun — multipremoving a checkmate or a beautiful sacrifice during a stream can be very impressive.
The original argument was that multiple premoves are too complex to implement, and since the use cases are not ubiqutous enough it's not worth it. This might be partially true, since Lichess often listens to requests and implements features that are frequently requested and are going to be used by a significant amount of people. Since most people are satisfied with only one premove, and adding multiple premoves is not an easy task (considering all the bugs that can occur, design questions and many things that may come up which aren't obvious at first), it's therefore probably not worth it for now to add it for the Lichess team. Here's a request for mutltiple premoves from a few years ago:
github.com/ornicar/lila/issues/3890Argument:
KB is unfair because it's significantly faster than other input methods.
Refutation:
the first point applies. Speed of the input method isn't in the rules of ultrabullet. It wasn't made to be a mouse skill game, it's only due to limitations of our technology that is has been. The speed with which one can input moves, produced by the brain or reflexes, is irrelevant, since any existing limits are there not due to restrictions imposed by a rule, but due to limitations of our technology.
Argument:
using more than one key at a time is unfair.
Refutation:
the point two applies. Our brain is capable of quickly imagining two or more pieces going to a certain square, or multiple squares almost simultaneously. If our brain can create such a signal, that can be clearly interpreted by a machine and doesn't include any additional knowledge generated by the machine, then our muscle reflexes (largely controlled by the brain) producing similar input falls into the legal area.
Argument:
what if we create a slider with all legal moves, and just move this slider back and forth to produce random moves? This would be the same as KB and would be a random move generator in essense and therefore unfair.
Refutation:
this wouldn't be the same as KB. First, KB doesn't show you the legal moves. Only legal moves are played, because Lichess interface would always reject illegal moves. (There is an assumption in KB code however that the user would always play legal premoves [pseudo-legal moves]. Such an assumtion might be controversial, but there are similar incidents: Lichess shows legal moves when a piece is clicked. Can this be considered computer assistance? It can, but we assume it doesn't significantly impact the game. This should still remain a subject for discussion.) So we arrive at a slider with all generally possible moves [pseudo-legal moves], which the user can move however they want and it would produce a move. Such a slider is not equivalent to the KB extension. It would be a different game, and there is no proof that the user moving the slider can translate the moves, written on it, into an image of a board with pieces. When the user is playing with KB, they can do similar things they would be able to do with a brain-computer interface (BCI). And not only those things would be similar in speed, but also in nature. With a slider, it would be impossible for a human brain to immediately translate all the moves written on the slider into an actual board with pieces. The problem with this concept is that there are no strict borders to where human abilty begins and ends. We can only speculate approximately. But in the case of the slider, the amount of operations the brain would have to perform to actually translate it into an image of the board with pieces is so great that it's definitely not an appropriate KB counter argument.
Argument:
KB is illegal because it's not a standard Lichess feature.
Refutation:
mouses are also not provided by Lichess, one has to buy one themselves. Mouses use drivers and code that are not part of Lichess. There is no requirement for a feature to be natively implemented in Lichess to be legal.
Argument:
you have to install a third-party extension to use KB, so it's not fair.
Refutation:
you also have to buy a third-party mouse (Lichess doesn't produce their own mouses).
Argument:
KB or Multipremoves are unfair because not everyone knows about them.
Refutation:
This argument presumes that fairness of an input method is dependant on the decisions that people make about this input method. But this is the opposite of how it should be: the decisions should be made based on the fairness of the method itself. I.e. if someone implemented a KB extension, and only used it themselves, without releasing it to the public, this would make the situation unfair for others who don't know about and can't access this input method, but it wouldn't make the input method itself unfair. The decision of the person who hasn't shared the code of this input method would be unfair, but not the input method itself.
Since no unfair decisions have been made, and the code and the extensions are available to everyone equally (and if the extensions break or are difficult to install, it's happening to everyone in the same fasion), the situation is therefore fair and the propagation of awareness of the extension's existence is happening in its due time. I.e. it's not possible with the current circumstances for it to propagate faster. Since it's impossible to make everyone aware of something simultaneously, and the speed of the propagation depends on people's decisions, which are currently the best they can get, since no major figure is deliberately preventing this propagation, we can't therefore use this criteria as a criteria for fairness of the circumstances. Unless the feature is forced on everyone and becomes the only default, the propagation is going to happen gradually.
Argument:
KB is not unfair or illegal but it makes it not fun for many people.
Refutation:
KB does make it unfun for many people. But it also makes it fun for others. Many people play fast time control for speed, and KB increases speed, which is what many people seek. One solution is to create an oasis for non-KB players where they can only play each other and avoid KB players. But that's probably impossible, since it would require implementing a new infrastracture from Lichess. The other solution is to ban the use of KB. But then we come to another problem. Maybe in 10-15 years there will gradually appear new input methods which will use the person's brain to send commands directly to the computer. They will appear gradually, perhaps expensive at first, so not everyone will have access to them. These interfaces will be fast and can be similar in speed to KB. They also will improve and become faster. Then, if we ban the use of KB now, we'll just postpone the dilemma, until those interfaces appear, giving us the same dilemma again. If we then follow the same approach and ban the use of those interfaces, not only would we face the problem of people using them anyway (since there will be no way to properly distinguish between them and other input devices [such as mouses or tablets]), but the next question would arise. If we ban those interfaces that we deem too fast to be fair, why didn't we ban interfaces such as tablets, which are also different and allow the use of two fingers instead of a single cursor to input moves? Where is the limit that deems the interface too fast? And why would we even try to impose this limit, if fast chess's appeal is in speed anyway? If someone doesn't want the speed to increase, then why not play time controls with increament, which are not about speed? If someone wants to specifically use the mouse because of the skills assosiated with it, then they could create teams and organize tournaments for mouse only players, and also try other games where mouse specific skills are important.
Argument:
KB makes it unfun because it turns endgames into chaotic moves.
Refutation:
that is also the case with other input devices. With the mouse, when one tries to avoid checks and win or draw on time, they often chaotically move their king to the nearby squares, often trying multiple of them, since it can be hard to see which are occupied by checks. Another problem is it's difficult to move pieces to long distances, so the game often turns into trying to shuffle pieces in the same place, without having a plan to checkmate the opponent's king or win material. With KB one can more easily activate pieces and deliver checks or a checkmate, since it's easier to move pieces to long distances and it's easier to quickly switch between pieces. With multipremoves present it's also sometimes possible to quickly promote a pawn (which is very difficult with a mouse) and deliever a checkmate by using queens or a rook and a queen. Occasionally the moves will be chaotic, when the players don't see a possibility for a checkmate. However, there are many aspects to KB time scrambles. You can use pieces that are far away from your cursor to bring them closer to the opponent's king and restrict its movements, winning you more time. You can promote pawns more effectively and use it to deliver checks. You don't have to shuffle pieces in the same place, and you actually can move them more effectively across the board. There are many more aspects. Therefore KB and the mouse have their own upsides and downsides in time scrambles, and we cannot say that using KB in time scrambles makes the games more chaotic.
