is it all smurfs or what?
Sort:
Im stuck at 6-700 rating but 90% of my opponents play better than 1300+ lvl bots. I find it hard to believe that these people are as beginner as I am when they turn all my tactics against me and see every trap while hardly using any of their clock. Idk if the bots just need work to bring them up to the rating they claimed to be or if I’m just getting constantly smurfed. But its frustrating to try and get better when im losing consistently at half the rating that I can beat the bots.
The reason why you are losing is that you are not making good moves. You aren't castling, you are placing your knight on the rim (Nh6 for example in one game). It's like pushing your chair away from the table and saying you can't reach the fruit on the table. It makes it harder to gain an advantage this way.
Play with the basics which include:
1) Pawn moves in center
2) Knights toward center
3) Castle early
4) Don't give up the bishop pair
5) Don't move the same piece twice
There are more, but that is a start.
ryanriffe wrote:
Im stuck at 6-700 rating but 90% of my opponents play better than 1300+ lvl bots. I find it hard to believe that these people are as beginner as I am when they turn all my tactics against me and see every trap while hardly using any of their clock. Idk if the bots just need work to bring them up to the rating they claimed to be or if I’m just getting constantly smurfed. But its frustrating to try and get better when im losing consistently at half the rating that I can beat the bots.
Comparing players to bots really isn't that useful. Most of the bots are overrated. Turns out it's pretty hard to dumb down an engine to accurately reflect a certain rating level.
Martin_Stahl wrote:
ryanriffe wrote:
Im stuck at 6-700 rating but 90% of my opponents play better than 1300+ lvl bots. I find it hard to believe that these people are as beginner as I am when they turn all my tactics against me and see every trap while hardly using any of their clock. Idk if the bots just need work to bring them up to the rating they claimed to be or if I’m just getting constantly smurfed. But its frustrating to try and get better when im losing consistently at half the rating that I can beat the bots.
Comparing players to bots really isn't that useful. Most of the bots are overrated. Turns out it's pretty hard to dumb down an engine to accurately reflect a certain rating level.
Wouldn't it be the other way? The 1500 bots are playing like a 1200. They are already "dumbed down".
Fiendamental wrote:
Martin_Stahl wrote:
ryanriffe wrote:
Im stuck at 6-700 rating but 90% of my opponents play better than 1300+ lvl bots. I find it hard to believe that these people are as beginner as I am when they turn all my tactics against me and see every trap while hardly using any of their clock. Idk if the bots just need work to bring them up to the rating they claimed to be or if I’m just getting constantly smurfed. But its frustrating to try and get better when im losing consistently at half the rating that I can beat the bots.
Comparing players to bots really isn't that useful. Most of the bots are overrated. Turns out it's pretty hard to dumb down an engine to accurately reflect a certain rating level.
Wouldn't it be the other way? The 1500 bots are playing like a 1200. They are already "dumbed down".
The bot is a 3000+ bot with tweaks and settings to try and make it play more like lower rated players. It's an estimated strength and is essentially guesswork.
The site could probably track everyone that plays a bot and get a better estimate, which used to be done when the bots could be played in Live, but I don't know if that was used to set estimates or what. Also, before the bots played server side and now I'm not sure how the code runs, so that can have an impact on strength too.
the bots are very overrated, you shouldn't compare the two
“Turns out its pretty hard to dumb down an engine to accurately reflect a certain rating level”
That’s kinda funny.
Uhm also just to add, imo I find that bots are surprisingly pretty good estimates of their stated rating level considering it’s not the easiest task to do I think the designers did a pretty cool job of it. I’d like to see how they coded it or designed it and have them post a YouTube video series or something about them.
Except for the adaptive bots which are designed to allow you the advantage. The 2000 rated adaptive bot (natasha) is worlds easier for me to play against than the 2000 or 2200 (nora bot modeled after magnus carlsens style supposedly). I’ve noticed some of the bots who aren’t adaptive will let you get an opening advantage and then start playing better and reclaim equality once the middle game begins. (Jonas for example I think).
That’s kinda funny.
Uhm also just to add, imo I find that bots are surprisingly pretty good estimates of their stated rating level considering it’s not the easiest task to do I think the designers did a pretty cool job of it. I’d like to see how they coded it or designed it and have them post a YouTube video series or something about them.
Except for the adaptive bots which are designed to allow you the advantage. The 2000 rated adaptive bot (natasha) is worlds easier for me to play against than the 2000 or 2200 (nora bot modeled after magnus carlsens style supposedly). I’ve noticed some of the bots who aren’t adaptive will let you get an opening advantage and then start playing better and reclaim equality once the middle game begins. (Jonas for example I think).
zix3i7im_V3 wrote:
“Turns out its pretty hard to dumb down an engine to accurately reflect a certain rating level”
That’s kinda funny.
Uhm also just to add, imo I find that bots are surprisingly pretty good estimates of their stated rating level considering it’s not the easiest task to do I think the designers did a pretty cool job of it. I’d like to see how they coded it or designed it and have them post a YouTube video series or something about them.
Except for the adaptive bots which are designed to allow you the advantage. The 2000 rated adaptive bot (natasha) is worlds easier for me to play against than the 2000 or 2200 (nora bot modeled after magnus carlsens style supposedly). I’ve noticed some of the bots who aren’t adaptive will let you get an opening advantage and then start playing better and reclaim equality once the middle game begins. (Jonas for example I think).
That’s kinda funny.
Uhm also just to add, imo I find that bots are surprisingly pretty good estimates of their stated rating level considering it’s not the easiest task to do I think the designers did a pretty cool job of it. I’d like to see how they coded it or designed it and have them post a YouTube video series or something about them.
Except for the adaptive bots which are designed to allow you the advantage. The 2000 rated adaptive bot (natasha) is worlds easier for me to play against than the 2000 or 2200 (nora bot modeled after magnus carlsens style supposedly). I’ve noticed some of the bots who aren’t adaptive will let you get an opening advantage and then start playing better and reclaim equality once the middle game begins. (Jonas for example I think).
The site uses Dragon/Komodo and the engine options for the bots. There's probably some other tweaks, such as bot specific opening books that also impact things.
Fiendamental wrote:
The reason why you are losing is that you are not making good moves. You aren't castling, you are placing your knight on the rim (Nh6 for example in one game). It's like pushing your chair away from the table and saying you can't reach the fruit on the table. It makes it harder to gain an advantage this way.
Play with the basics which include:
1) Pawn moves in center
2) Knights toward center
3) Castle early
4) Don't give up the bishop pair
5) Don't move the same piece twice
There are more, but that is a start.
I'm with you that I'm a pretty weak player. I'm just questioning why a 1300 bot lets me get away with mistakes that even 700s punish me for. Either way I'd say I'm getting better and having fun doing it. Just a minor frustration really.
