Does Anyone Know the Approximate USCF/FIDE Rating of Little ChessPartner On Strength: Hard? Erik?

jkpastorius

It seems like the staff doesn't even want to take a stab at this issue, which is really quite sad.

I suppose that it is fairly accurate to post a FAQ stating that it is approximately 2000 rated (Chess.com standards I suppose, but it really doesn't say).

SukerPuncher333

You know I was just thinking, just because the computer plays at a 2000 Chess.com rating, it doesn't necessarily mean its USCF/FIDE rating would be much lower, if at all. Of course, a human with a 2000 Chess.com rating, using tools like 3 day time controls, analysis board, opening database, books, etc. would have a much lower USCF/FIDE rating due to the inability to use these tools in an OTB match in a tournament setting.

However, if the computer can make these same moves in mere seconds, then it is will probably do well in an OTB tournament (for one, it would never run into time pressure!! lol). Imagine if you could play a 1900 FIDE, making the same strength of moves OTB as you do here on Chess.com. You'd probably be giving him or her a run for his or her money. Since you're borderline 2200 here, it stands to reason that you can bash a 2000 rated player.

That would be true if we are playing against LCP as if we were playing correspondence. But since we are playing it OTB style, we can't really use our CC rating to gauge its strength. Let's say my CC rating is 2200, my OTB rating 1800, and I score 50% against LCP -- that would give it a rating of 1800, not 2200 (since I'm playing it OTB style, not CC style).

Maybe you could post a game where you feel it was playing at 2000-ish strength? It could be just a matter of style (you simply like to steer the game into positions that favour LCP's style over yours).

SukerPuncher333

I guess what I'm trying to say is imagine a player who is rated 2000 on Chess.com; he or she may use databases, books, extended time controls, analysis board, "submit move" feature, and whatever tools available. Now imagine if that player could play in FIDE tournaments, but had the power to freeze time whenever it was his or her turn to move, and had like 3-7 days to make his or her move using all of the aforementioned tools. I'm sure a player like that would shred a lot of tournament players. His or her 2000 rating on Chess.com wouldn't be too far off from his or her FIDE rating.

Of course, without this mythical power to stop time in an OTB tournament, a 2000 Chess.com player would probably play much worse due to the inability to use all of the tools and advantages legally available in Correspondence Chess. But if this player did have this hypothetical power, then I doubt his or her Chess.com rating would be so "inflated".

But Little ChessPartner, in essence, has the ability to do this. So to me, I am making the moves in real time G/45 or G/60, but to the computer it is as if it has days to contemplate its move. I think the computer would actually do very well in a USCF or international tournament.

From your experience with the computer, I can see why you think its rating is inflated. But you do have a high Online Chess rating, so depending how you play, your OTB rating could also be quite high if you're not overly reliant on the various tools that Chess.com allows. Moreover, as you have stated before, since you are familiar with its playing style, and use its printed analysis of the game to your advantage, its as if you can read its mind, making it much easier for you to beat, even in standard time controls. I mean, if you went to a USCF tournament, and after every move your opponent wrote out an x-ply analysis of the position, and evaluated how many points he or she feels it is ahead/behind, then you'd have a huge advantage.

I see what you are saying. A 2000-rated chess.com player might only be 1800 OTB, but with the tools of CC (analysis board, opening book, days to think, etc) the moves he's making on chess.com are actually more like 2000 level, making these ratings quite accurate.

It could be true to some extent. I haven't thought of it that way. My estimate was based on human players that I know, mostly rated 1600-1800. I just can't imagine them scoring less than 50% against LCP on "hard" in serious games. Maybe in a casual short game, but not in a rated tournament game with classical time controls.

SukerPuncher333

If you take a look at my recent post where the computer chooses the Scandinavian/Center-Counter against 1. e4, you can see that the computer doesn't waste moves by just randomly pushing pawns or shuffling pieces around. It actually plays like I do, with minimal pawn movement and maximum piece activity.

Notice that after the initial 1. e4 d5 he doesn't make another pawns push until 11. ... e5, which is very logical in that it fights for central control and attacks my weak a-pawn. He does shuffle his pieces around, but in the way that you're supposed to, continually improving his position against my un-coordinated pieces until White can no longer maintain the defense. 

All in all, he plays a rather strong game against my failed experiment in the opening =/

Yeah you are right, in this case LCP suddenly felt very strong.

LCP's "original" plan is to develop, castle, and push pawns (and finally shuffle pieces if everything else is done), but your game reminds me: if at any time you have an undefended piece/pawn, LCP will usually attack it.

In your case, LCP made a bunch of moves attacking your undefended pieces. It might look like LCP was playing positionally, making crushing advances with its pieces, invading your squares, etc, but I think it was mainly driven by undefended pawns in your camp.

That's all generalization of course, so here's what I mean specifically: on move 6, perhaps Bxe2 would be better? giving faster development and keeping your c-pawn protected. I think this move started off a chain of events...

1) Since Nxe2 blocked off your f1 bishop, your c-pawn is left unguarded...

2) ...so you push b3 to guard it (spending a tempo), but now your b-pawn is unguarded...

3) ...so you play Rb1 to protect it (spending a tempo), but now your rook b1 is vulnerable to his Bf5...

4) ...so you play Rb2 to move it out of danger (spending a tempo), but now your a-pawn is unguarded...

5) ...so you play Nb1 to protect it (spending a tempo to un-develop!)

By now you've spent several tempi with no progress in development, while LCP has somehow gotten a mean-looking position, simply by attacking at whatever target you left undefended. So it was somewhat fortunate on LCP's part to have a chain of targets to "guide" its moves. But I think in a longer game, you would've calculated further ahead and had more success in breaking that "chain." I admit the game example I posted earlier is an extreme case of LCP being dumb, but I think your game is also a pretty extreme case of LCP being smart. I might also be underestimating LCP's strength after all, seeing as there's another thread where a 2100+ (Omicron) said he finally beat LCP on "hard."

By the way, move 16 by the computer looks like an amazing positional move, but I think LCP was only trying to attack your c-pawn. Unfortunately, whether or not LCP intended this, the positional impact of this move was crushing as it locked in your c1 bishop.

I'm rated 880 here,  and it creams me like any other engine even on easy!  (I have no USCF)

Jester Lv.10 on 10 seconds is not that strong. My rating is 2150, and with the Black pieces and 40 minutes of thought, Jester was crushed in the endgame. I would say he is around 2000 USCF. Check out this game: