why get engines?

Hi this article is about engines. Everyone says engines are great for improving and that you should definitely consider buying one. I do agree but take stockfish, its main feature is the analysis function but you can access good analysis even here on chess.com so why would I download stockfish when I can simply use it from a website. I've heard that stockfish also comes with a database, same thing, chess.com has about 3,000,000 games in their database and what about chessbase with 8,000,000. If anyone has any suggestions on engines with features that I can't already access or if there is something that I'm completely missing then I would like to hear them.

So many think they need the absolute highest rated engine at all times.  You have players that barely know how the pieces move, have no clue how to use an engine, and have no idea what engine analysis means.  And yet!  They have to have something rated 3800...because 3600 isnt good enough.

well then what are good examples of GUI? I am new to them are they worth it? 

You can just download a free version of Stockfish. The online analysis is limited. If you use Stockfish on your local computer, you'll be able to tweak the resource usage and everything else to your liking. You can set the search depth, the number of threads, the size of the hash tables, the endgame tablebases.

Regarding databases: These have nothing to do with engines. It's just something entirely different. You can use any engine with any database with any user interface, some of them free, some not (like Chessbase, which is arguably the best, but not free at all).

Try Arena (GUI) + Stockfish (engine), both free.

thanks for the advice! ( although I still think I have plenty of access to engines even though I would like to be persuaded .)

The engines calculate faster. You can get to crucial positions instead of wondering what the best lines are.

Certainly for many tasks your online analysis is perfectly fine. But sometimes it can be nice to see what happens if you let Stockfish calculate 60ply with plenty of resources.

 A valid point.  But let me share a story.  One night after chess club, a bunch of us went to a Dennys to go over a game someone had played.  They were sure they had a win but couldnt find it.  Someone pops up with: "Just run it through an engine..."  Sure that would be easier, but easier isnt always the best way to go about something.  So 6 of us sat there talking, analyzing, bouncing ideas off of each others, etc.  Here are 6 players of varying strength from USCF C class to Master.  After about 2 hours we all decided on what we thought was the correct continuation.  It was the some continuation an engine suggested.  So why put ourselves through all that?  To learn...to not just have the answer given to us.  To learn, and actually understand "why" it worked.

Is this a true statement?

Maybe sometimes. Is it always, more often than not, less often, or about the same?

Context is related to the OP and using analysis on chessdotcom...

By analysis, I'm also referring to the game reports. Because after all, those of us not pros, would like to improve.

I'll edit this post with a link to the original forum soon.

Take the reference game and look at the report for the move 9.Re1. The report says this is "best". What does that even mean? Are there no better (only equal/"alternative") moves?

When I saw this, I wanted to know if I was missing something. I started making a list of my candidate moves. 9.Re1 was one of about 5 or 6 I considered, but I couldn't for the life of me come up with anything to suggest it was even a top 2 or 3 move.

We know the depth of analysis (report) is shallow (18-20 iirc). Perhaps more depth is needed? I tried that. More lines? I tried 12 (yeah 12). I only got to depth of 34 (hey, I don't have forever). 9.Re1 was 9th on the list at that point. It had been higher/lower before, but that's where I stopped.

Here's that screenshot:

So, how useful is it actually and realistically?

My candidate moves were 9.d5 (hoping to push the e pawn down his throat, but the a7-g1 diagonal has me a little skeptical), 9.Bf4 (I ended up with 3 pawn islands and tripled pawns on the g-file in 1 version), 9.e5 (with ideas of trading LSBs, but giving up the d-file easily). I guess I turned to the engine too soon, instead of looking at it myself. I didn't like White playing the LSB back to d3 instead of be, kind of confused me (don't play this opening).

What are your candidate moves/ideas/plans? What does your engine "say"?

Edit: link as promised

https://www.chess.com/forum/view/site-feedback/accuracy-report-doesnt-understand-math

Of course, the difference between the engine's 9th and 5th choice is about 0.1, you could argue it made top 5 and I have no clue if 0.1 even matters.

So you don't have to flip back and forth...

After 8...Nf6, what do you consider playing as White?

@PerpetuallyPinned, engines are good for blunder checking and online analysis usually is sufficient here. For strategic advice, I'd not ask the engine. It's more fun than useful.

The best piece of advice youre going to get.

If i show you a position where white is +.4.  Would you be bale to know "why" white is better?

Of course, if you don't just look at the evaluation, the engine line might still offer a positional idea you just missed.

That's why I posted everything previously here.

I was looking at another game someone posted the other day. It was asking about the endgame, but I usually look at the entire game. Another endgame version I looked at I was pretty sure gave good chances for a draw. 1 move I wasn't sure about, checked with engine and finally (like depth 50 something) found a mate. Mate in 34 iirc, I let run longer to see if a quicker mate could be found. Guess what? Mate 35 popped up next, then mate in 36.

Has anyone seen that before?

Here's that thread:

https://www.chess.com/forum/view/endgames/what-endgame-principles-apply-in-this-endgame

I need to find the position I came up with again

Since i am unable to find a mate in 3anything, it doesnt matter. 

This does remind me of that game Carlsen and Caruana played in the WC match, where some neural network found a mate in 39? i believe it was.  Not one single person commented.  But once Svidler mentioned it in chat, then suddenly you get all these people typing: "OHHHHH...I cant believe he missed the mate..."

Here it is: