Bug or am I the next GM?

Hi.

First of all, this is great site to use, even for me as a chess novice.

Now directly to the topic.

As I am lingering around the lower third part of all site players, certainly no wonders can be expected from my play, at least at this point, yet.

So I am using in between real chess players also the site's computer-generated players as a pastime or training.

My favorite for cheering me up is Sven-bot, as I can beat him, if I really put some effort into my play.

Nelson-bot is the nut I still have to crack, but I think I can learn a lot from more advanced playing like his and I enjoy him always finnding a way to kick my a**.

As I am always curious, I thought why not try some more advanced bots like Antonio-bot and above as it's also very interesting to lose against them. Logically, I pushed to the limit, out of curiosity, and engaged today at lunch break the chess engine set to max. (3200), no hints allowed, no take-backs and random color choice chosen.

I was just looking for some minor distraction, as I did not expect anything worthwile to come out of it. I just wanted to get a feeling on how many moves it would take me with the bot players' increasing strength for me to be beaten.

Imagine my increasing astonishment, when I saw the engine react to my quite probably mediocre moves.

I finished up winning against the engine. At one point I even managed to get a second queen, something I haven't even managed to achieve with Sven-bot.

Do I get now some "GM h.c." title (or a Minor Master title) or is that chess  engine just buggy at higher ratings or is the engine ok, but my play a fluke in the matrix?

At least thanks very much for the overwhelming moment, when I won against the engine! I took a screenshot of it, as "proof".

Has something similar happened to other curious players, as well?

I still have to go through the game analysis.

Have all a nice day and a great weekend!

Meanwhile, I found other players ' forum entries, having had similar unexpectedly easy experiences with Komodo25(3200).
So if nobody can offer a better explanation, I assume it's a bug, not a feature.

Or feel free to try it yourself ... if your level is not too advanced.

Because maybe it's like the situation with the very advanced alien Asgard race in Stargate SG-1 coming to Earth and it's character Thor asking for help against a seemingly unbeatable adversary, the Replicators.
The reason is, as he puts it, roughly, that Asgard technology is simply too advanced and the Replicators are adapting too quickly to that.
Therefore he comes to Earth to seek help, as humans used more primitive methods to combat the Replicators (with a certain degree of success).

One human character (Daniel) then asks the Asgard:
 "Wait a minute, you're actually saying that you need someone dumber than you are?"

Before the Asgard can answer (or is willing to answer), character Colonel O'Neill comments on that "You may have come to the right place."

For me this is one of best lines of the series.

For those of you who want to relieve that hilarious dialogue:

http://www.stargate-sg1-solutions.com/wiki/4.01_%22Small_Victories_Part_2%22_Transcript

Again, I won quite easily against self-adjusted engine level 1700. The engine did quite basic errors and gave away important pieces like the queen way to easily. IMHO as a chess novice, I'd say the self-choosable engine levels are all (although I only tested 3200 and 1700) playing on 300 to 500 level. At least that's my work hypothesis. 

It seems to be just the engine, where the level can be chosen freely between 250 to 3200, which is not working properly. By not properly I understand that I expect a level way above my own to be basically unbeatable for me. If it's beatable then the engine's level is not as advertised, but misleading.

The computer characters like Sven and Nelson seem to play decently at around a level a human player would play (well, at least within +/- 200, I estimate) and usually do not loose important pieces too easily. As far as my little experience goes, they are not prone to the same failures the freely adjustable engine is.

Too bad the self-selectable engine level is then basically worthless.

Does anybody have different insights or insights corroborating my view?

I’ve had similar experiences with using a random endgame setup, winnable, but barely, and the engine couldn’t get a good defense. This was with Stockfish, not Komodo, but still. My guess is, the engine runs in the browser, and is programmed to respond in a second or so, which simply isn’t enough time on some iPad or phone to calculate to a meaningful depth.

Why the predefined bots are better, I don’t know. Maybe they run in the cloud.

I’m guessing, of course. But considering I’ve seen videos of GMs or NMs barely holding on or losing against 3200 chess.com settings, I’m guessing it must be different per where or on what system you run it.

Would be nice if they’d explain this somewhere.

Hi.

Thanks for that insight!

I wouldn't have guessed that the engine is running on my device. But if you are correct (and I remember that the engine reacted rather quickly), that could well explain that mishap. I'm playing on a mobile phone only, with often less than optimal connection speed (4G is a rarity).

Too bad, if they would have chosen to limit response time and engine performance would therefore be throttled on mobile devices. I  wouldn't care, if I'd have to wait 30s for a good result. After all, the emphasis is on playing level and not on speed.

Maybe chess.com could shed some light on that aspect. Play performance should not depend on the device it's being played on.