The live pairings on this site are completely stupid and need to be changed

How is it that you play a bunch of under 1000 players, they beat the crap out of you, and then you get a slew of dumb dumbs. The ratings don't reflect any level of common sense play.

They use adaptive bots. You play well, and then all of a sudden the opponent plays like a GM. You play poorly and then the opponent becomes a retard. We aren't playing humans.

Your reading comprehension needs some work. I was referring to moves in a game, not the result of one game to the next.

One gets series of out of shape players and series of accute players all the time. Same thing happens in real life, regarding OTB competition. There is no point for "Elo regulating bots". It would bring zero benefits and chess players of all sorts (tired, drunk, distracted, motivated, etc) are plenty.

A conspiracy, such as the famous "Bots conspiracy" weak players often imagine because they understand close to nothing of chess and know chess history even less, need a reason. Here: no reason. At all.

I don't think they are bots, it's harder to get a match in a less common time control. But many indeed keep noticing that ratings are not accurate, it's almost like ratings tell you nothing about what to expect from the opponent. He could be suspiciously strong or he could be very weak. Apart from the fair play issues, level of concentration could vary among players. That's why I'm promoting full cessation of PvP online chess. I believe people should only play bots. Against humans - OTB only (with increased security measures). Stop playing chess with humans online. This only leads to frustration. Bots are fun and are actually more human-like than many humans.

There are at least 2 different style bots. Adaptive bots and static bots. The static bots are at set levels and make mistakes at a certain interval. That is what makes them not fun. The human can play surprising good moves, but they tend to go back to their true level. The adaptive bots improve while you play well and mess up as you mess up.

That's a myth, there's no certain interval. Some bots are humanized very well using neural networks. Dumb bot logic is in the past.

For bots that you call "static", Komodo engine (used for chess.com bots) decreases skill by reducing search depth, reducing opening book knowledge, altering piece values for evaluation and increasing variety of move choice. There's also dynamism (amount of risk) setting, it can lower the strength a bit too. But there's no "scheduled blunder" like many think.

Quote about adaptive bots:

Auto Skill has Dragon automatically adjust its internal Skill level as you play against it. It is a fun feature. Once you set your approximate Skill level, with Auto Skill on, it adjusts Skill based on the evaluation of the game position, giving you better chances of winning if you fall behind. If you are winning, Auto Skill will raise its level. If you are losing, it lowers the level. It is biased to increase your winning chances. On average it will shoot for a level 1-2 lower than you. Auto Skill can lower the Skill level up to 5 points, and raise it up to 3 points from your current Skill setting while you play.

That's -625 ... +375 Elo strength variation. On chess.com such bots are presented as "Adaptive". They're great for learning to walk you through all stages of the game.

Another setting that applies to both "static" and adaptive bots is Personality. Quote:

The Personalities are:

Default: This is the default and strongest personality. You have full control over Contempt settings. Most of the other personalities will play roughly one skill level weaker than Default.

Aggressive: The Aggressive personality will attack relentlessly with no concern for their own safety. They prefer active pieces, and are biased toward playing with the Queen.

Defensive: The Defensive personality emphasizes king safety and a solid position above all else.

Active: The Active personality tends toward open positions and well-placed pieces.

Positional: The Positional personality tends toward solid play, maneuvering, and more closed positions.

Endgame: The Endgame personality prefers playing through to the end to win by promoting a pawn.

Beginner: The Beginner personality doesn't understand the fundamentals of chess, especially endgame principles, and is mostly looking to check and capture. But please note that it will drop in strength between one and two levels, so you still have to set the skill level appropriately if you hope to win!

Human: The Human personality is optimized to play like and to beat strong human players, playing aggressively, avoiding pointless simplification and repetition, aiming for open positions, and emphasizing static advantages. If you are a fairly strong player and want the experience of playing against a simulation of an even stronger human player than yourself, we suggest using this mode and finding a skill level at which you can score 25% or so at your preferred rate of play.

This is from Komodo website, we don't have access to these settings on chess.com but chess.com probably uses all these settings to create various bots.

Are you that one guy who made the "This site is rigged" forum? Your names are similar.

If yes, stop whining and play.

If no, my bad but just speculation.

But im pretty sure i'm correct. You're talking about the same "adaptive bot thing".

TryToMoveThatWay is the account you were on.

again, my bad if i got it mixed up just speculation.

However, chess.com has better things to do than play around and mess up elo ratings. You think a website, used by 1b+ people will do such things? They have to make sure no fair play policy is violated or if the servers are running fine or if they can keep it financially stable. They simply wont go hahahaha lets mess up elo.

True.

Some people get overconfident when you play bad and they make mistakes.

Some people get motivated when they are losing.

I kinda disagree. Bots use stockfish and if they are like uhhh lets say 1600 then they will play the third or second best move using stockfish. They are very un - bot - like and most of the time the third or second best move doesn't make sense for humans to play.