Why are the bots easier to beat than real players?

At the weekend I played the notorious Nelson bot (rated 1300) about 10 times.  Most games ended in stalemate or I won (just, but it's a war of attrition), he beat me on 3 occasions. 

I can consistently beat Emir (1000) and fair well against Sven (1100). I can even last 30 odd moves against some of the 2000+ bots. 

But playing real players around my own level seems to be more difficult. 

Is it that the bots are programmed to miss the odd capture which would generally not be missed by a real player (even a lower rated one)?

Or is it that I'm taking more risks playing the bots, perhaps being more aggressive due to 1) No time pressure* 2) No ratings at stake and 3) I know it's not a real person at other end so I care less how the bot might move.

*Despite there being no time pressure against the bots I probably make my moves faster than in normal play. 

What are your experiences of playing the bots compared to real people? 

Depending on the bot you play, they are designed to make mistakes. Your job is to exploit those errors. But beware they are sharp with tactics.Enjoy.

These bots don't grow. While you maybe playing people at your level, they are probably moving upwards in ratings and you are just one of their stepping stones.

The rating of bots is not a real rating. It does not adjust with the games they play. It is just a number that the chess.com guys assigned to the bots at the beginning, based on gut feeling. They did not do a very good job.

Online chess ratings are a joke anyway, but these ratings are a bad joke.

I don’t know. It seems even the Beth Harmon bots are programmed to drop a piece every few moves.

Hi! My name is Lauren Goodkind and I’m a chess coach  based in California: www.ChessByLauren.com.  

  In my own experience, bots play a lot differently than people.  Bots tend to hang pieces and move strangely at the lower levels.  

Perhaps you can ask the programmers who programmed the bots to see what they think.  

I'm doomed

This. This exactly.

I've tried playing a few of the bots one day and of all the things that bothered me, the fact they make weird uncompromising moves that very few people would play in that position messes me up because I'll have to reroute.

Nowadays there's one engine that I play against and analyze with on the Arena interface, I tweaked it so it favors piece activity mostly. I tweaked a few other settings to match my preferred playstyle a bit and it seems to work out quite handy, it makes a lot of moves I'd expect a human opponent to play when I do a practice sesh. Practically, of the moves that I'm surprised by, most of them are the type of romantic wildness that I adore. So either way it works out excellent, maybe even better than playing hoomans.

I see you're playing 30 minute games. This is a good choice as it gives you plenty of time to consider your moves. At your level (or anyone below 1000) the person who makes the fewest mistakes and blunders should win. Though that also applies to higher levels too.

Concentrate on opening principals, pawns to the middle, knights out and then bishops. You'll probably see a lot of players performing wayward queen and fried liver attacks at this level. Know how to counter those while developing your pieces and you should start to see your rating steadily climb. 

Play puzzles, take the lessons and search YT for beginners chess. There is a lot of good free content on there that can really help. 

Just go to the play menu and click on "computer". That should take you to the bots. 

Bot's require time to "think"... The engine itself is probably accurate, but the computer has a time limit for moving, so it's already at a disadvantage.

Bots usually have uncharacteristically odd understandings of the game, especially those below top-level. It's hard to emulate what a 1200 rated human would actually know about chess in a bot. Therefore, there will be some obvious loopholes in the bot's thinking process that make it somewhat easier to beat than a similarly-rated opponent. This is all to get the bot's rating level to the level that you're aiming for.

That's the main logic behind why a bot with drastically overhauled parameters plays better than a bot downgraded to play a certain level.

Beginners usually only look for piece activity and king safety, and don't always find tactics. Tune down the importance of material equality and you already have a much better bot. I guess chess.com has been doing sort of the same but it's not working out quite as good as what I use for practice.

if you beat those,then you can play the harder ones (i think)

try beating all of them maybe?

This may have been true in the past but I believe the developers are testing upgrading the chess engine used in the bots. I believe a few weeks ago the bots on the Android app got an upgrade to the new Komodo Monte Carlo engine that favors moves by win probability resulting in a more human positional style of play. You can read about it here ( https://www.chess.com/terms/komodo-chess-engine ) if you're interested.

May you beat my son

That would be one step closer, but you'd still be in a mess for after the opening since there's only so much data on played human games available.

That's interesting, I wasn't aware of that but thank you for letting me know!