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

I've been playing a lot against the computer here. I started playing it on Easy since most engines I've played in the past just absolutely cream me. But either I'm getting better and/or the computer is really easy on Easy, so I can just blitz wins. Now I play it a bunch on Hard and have been getting some good games. My performance rating is probably about 1900 USCF, give or take 50 points or so, just not quite Expert, I think.

I'm batting like 400 against the computer on Hard, so I was wondering if anyone knows the actual play-strength of the computer? It opens very well. When I post-mortem, its openings are almost always book/Master-level. But it is regularly dropping games to a USCF Class A/Expert, so its overall rating can't be Master strength.

FAQ entry.

     your good,

     what is the is the air-speed velocity of an unladen swallow?

chess.com says that it is 2000 on "hard" (see here), but I don't think that's accurate. I've played against it many times before, and I'm guessing it's more like 1600 FIDE "hard."

I've posted a thread (see here) listing several online engines of varying strengths. Jester is probably close to 2000 strength, and it easily outplays Little Chess Partner. The dumbed-down free online version of Shredder is probably master-level on "hard," and it absolutely destroys Little Chess Partner.

African or European?

SukerPuncher333

Thanks for the input. I'm really not sure how accurate how the estimated strengths are of Little ChessPartner.

First, I think "Easy" is stronger than just your average 1200 player. Since I think I'm about 1800-2000 USCF, I know I can crush "Easy" mode, but I still lost some games to it back when I was like 1500 USCF. This to me suggests that it's better than 1200 whatever. It still opens strong, which is something a 1200-rated player can't do.

1600 FIDE is probably too conservative for the "Hard" because this would only be 1700 USCF, which I know I'm stronger than. 1700-1900 FIDE would be more accurate.

TadDude

Thanks for pointing us to the FAQ. However, it only gives the approximate CHESS.COM RATING. This thread is asking for the approximate USCF/FIDE RATING. RTFQ! Still, thanks though.

Quite clever of chess.com to determine chess.com ratings before anyone played on the site. 

I may be under-estimating because I've played LCP too much and got used to its style, but I feel it's no stronger than 1700-1800 CFC rated human players that I know (CFC is roughly same as USCF). Maybe stronger than them at blitz because humans blunder tactically, but certainly not at slow time controls. Two of its major weaknesses:

1) way too materialistic. you can see its evaluation of each position throughout the game, and the evaluation is 99.999% based on material. even a single pawn outweighs the most extreme positional factors. doesn't care about doubled pawns, backward pawns, control of diagonals/files, central control, king safety -- you name it.

2) plays with no plan. after it runs out of book, it's "goals" are simply: 1) castle, 2) develop pieces, 3) push pawns as far as possible, and 4) shuffle pieces around or exchange pieces. I'll post up a game later that shows an extreme example where the computer has no plan and makes really silly moves

btw, what did you mean by "batting like 400" against it?

Here is an EXTREME example of Little Chess Partner (on the "hard" setting) making bad moves. It's game plan can be summarized as consisting of 3 parts:

1) develop piece and castle (which is good, but this is the only good part)

2) push pawns as far as possible

3) shuffle pieces around and wait to get mated

This was a blitz game and my attack must be full of inaccuracies. Any human would've defended better (or at least bothered to defend at all), but the computer just waited and allowed me all the time in the world to prepare the attack. This is why my estimate of LCP's playing strength is so low. I admit this type of position doesn't favour a computer's style, but still, come on...! Just look at its moves, lol. I'd give it 1700 FIDE max.

"btw, what did you mean by "batting like 400" against it?"

You're obviously not a golfer.  400 is 40% of 1000.

Batting 400 is a baseball term lol.  *sigh*  :)

ah I see. scoring 40%

SukerPuncher333

Yes, it is a baseball term, and I'm glad that confusion is cleared up.

bigpoison

You do know that you cannot be considered an American if you don't like/know anything about baseball!

Check out my avatar, dude.

SukerPuncher333

While it's true that you did beat "Hard" very handily in the game you demonstrated, what is your rating/the rating of the engine you used, in USCF or FIDE, when you played that game? You may be right that its strength may be no more than 1800 USCF, but I wonder if its not just a little stronger, like 1850-1950 USCF.

For one, you seem to notice a pattern in the way it plays, and you take advantage of the way it seems to calculate positions. When I play the computer, I don't look at the analysis it gives of the game or what it feels its winning chances are; I just try to play the position in front of me.

And in terms of "castling, pushing pawns, and regrouping pieces", I think that's the standard strategies of almost all chess players, mostly class players, but Masters still use these basic guidelines as well. It also has a lot to do with the opening that you choose. 1. d4 with the computer going into a Benoni and you advancing your pawn. Of course it's a slower game, so a large part of most 1. d4 games involve shuffling of pieces in order to achieve a desirable position. Gaining space via pawn pushing is also a huge part of the strategy of these openings. It also doesn't hurt that you're opening like a Master, giving the computer little chance to seize the initiative in the opening.

When I play, I often play random openings. Sometimes I REALLY surprise myself and open like a Master in for the first dozen moves in one of the games I posted. In other games, I resign after like 15 moves because I blew some random opening/opening variation.

Back to your game, the computer seemed to be playing pretty well. It puts the question to your Bishop, and you decide to trade dark Bishop for his Knight, thus not loosing tempo and doubling his pawns because he can't recapture with his pinned Knight. However, this is a personal preference, as classical players prefer not to trade away Bishops for Knights in the opening. But with the advent of computers and engines, many more players are willing to do so. I've never seen so many players play the Ruy Lopez Exchange Variation in my life more so than now, even when I was a little kid and where all of us thought that doubled pawns = death in the opening lol.

Pushing the h-pawn doesn't seem like such a bad idea, especially if you like attacking and need to play for the win. However, I suppose the computer didn't see that you wanted a Knight outpost on f5, and thus, his isolated h-pawn was soon without a defender. And any Queen's Gambit type of game will involve piece shuffling due to the delaying of tactics, it's okay to waste tempos via moving the same piece multiple times in order to achieve a better position on a closed board.

The computer is playing pretty good. But you're just playing better, at least USCF Expert or even Master level (about 1900-2000 FIDE).

So does any staff know the answer to this question?

brandonQDSH, something I forgot to ask: are you blitzing through your games against LCP? If so then it's a different story. I do think LCP is around 2000 by blitz standards. I was only referring to long games. In fact just after that game, I played it again and lost badly after getting my queen trapped.

I'm only about 1800 CFC (or USCF), maybe a little higher, but like you said this game is somewhat unfair because I've played LCP so much that its style has become predictable. Anyways, my point was to illustrate what I think is LCP's poor "game plan," which limits its strength (this poor plan is the main reason I believe LCP is more like 1700 than 2000, in slow games): 1) develop/castle, 2) push pawns, 3) shuffle pieces. I see this pattern over and over again. Part 1 is definitely good, no doubt. Part 2, pushing pawns, is also necessary, but LCP was doing it almost randomly. This game isn't the best example, but I swear LCP will push all its pawns if given the chance! Shuffling pieces is also necessary especially in a closed position, but come on, it was doing Rd8-Rb8-Rd8-Rb8 not to mention Kf8-Kg8-Kf8-Kg8, lol. The computer played good at the opening (developed, castled), but after that stage all it was really doing is pushing pawns, shuffling pieces, and waiting for the human to make a mistake. Think back to cases where you lost -- it's because you played too carelessly or attacked too recklessly, and over-looked a tactic dropping some material, correct? Rarely does LCP actually outplay you positionally or strategically. Therefore, in blitz LCP could well be 2000. But in a slow game we make less tactical blunders, and so LCP's strength goes down.

By the way, here's an online engine that I think does play at 2000 strength in slow games (and even stronger in blitz)

http://www.ludochess.com/jester_eng/jester_eng.html

Set it to level 10, though so far I haven't noticed much difference between 1 and 10. Unlike LCP, Jester actually plays positional chess. You can really feel how much stronger it is.

For an even stronger engine, try: http://www.shredderchess.com/play-chess-online.html

This is about master strength, but below GM level. If you compare these computers, it's hard to imagine LCP being 2000 strength. If LCP is 2000, then Jester is what...2300? And Shredder Online must be 2600?

SukerPuncher333

Thanks for recommending some strong engines to try out. I will do so in the near future. I play on slower time controls when I play Little ChessParter, around G/45 or G/60.

Some of my losses stem from playing an unfamiliar opening. Since I'm not a Master, I can't play EVERY opening at Master strength. The computer, however, does, and sometimes it can get a great middlegame position on me. I try too hard to equalize, and I end up dropping a pawn. In some rare cases, I even drop a piece.

And yeah, you're right. Some of my losses stem from me attacking too recklessly, with the computer able to patiently and calmly find the correct moves on the defense.

I can play positionally, and I do love me some Queen's Gambit Declined, and I can play it pretty well. But at heart, I'm a tactical player. I guess I should try playing the computer with 1. d4, but its 1. e4 e5 play is pretty sound.

I asked much the same question not too long ago.  Here's the thread:

http://www.chess.com/forum/view/community/how-strong-is-little-chess-partner