Comments/Ratings for a Single Item

~~Earlier~~ ⇧Reverse Order⇩ ~~Later~~

The Black Ghost. Black gets a teleporting Ghost piece that can not capture to balance White's first move advantage. (8x8, Cells: 64) [All Comments] [Add Comment or Rating]

Anonymous wrote on Sun, Sep 28, 2003 07:53 PM UTC:Good ★★★★

I like the idea of using the 'Black Ghost' to balance the first-move
advantage of White. However, I would have preferred a different name for
this piece as I don't really like the term 'Black Ghost'. This name
just doesn't seem to fit with the rest of the chess pieces. May I
suggest
the Latin term 'Servus' instead (which means serf and therefore fits
into the medieval settings of chess. Also, 'pawn', which means 'foot
soldier', is also a Latin term), as this piece is obviously of the
non-military sort as it cannot capture the opponent's pieces, but it can
still be used as a blockader or a sacrifice.
Another issue is that this new piece may just over-balance the first-move
advantage of White. I think the benefits of having this piece may just
exceed the benefits of moving first.

Breadman wrote on Wed, Sep 1, 2004 07:34 PM UTC:Good ★★★★

I have attempted to analyze this piece using Zillions, but the program values it almost as much as a Queen, neglecting the fact that it can't capture. Thus, a human player can win either side easily. Rook vs. Ghost seems like it should be a draw with perfect play, but I have managed to win with it.

Anonymous wrote on Sat, Sep 4, 2004 07:35 AM UTC:

Zillions rates pieces based on mobility; even though this piece can not capture, the fact that it is very mobile makes zillions thing 'Wow. Look at all the squares this piece can move to! It must be extremely valuable!'

George Duke wrote on Sat, Sep 29, 2007 04:10 PM UTC:Good ★★★★

On subject lately of switching, swapping and Betza's Wand One of Teleportation, RBetza's Black Ghost(1996) introduces a weak teleporting piece here. To balance the first-move advantage for White is the object. There are GC Presets for Black Ghost because we played one. Black Ghost is a disrupter able to make two-move sequences unexpectedly out of straightforward attack. As boards now go to 8x10 and 10x10(9x9,9x10,12x12) primarily larger, it may be interesting to give Black Ghost more power. In 1990's Betza had serious reforms for standard 8x8 like this, whereas by 2001-2003 his style was more free-form, fanciful, and sarcastic. How weak can a piece get? Weaker, because as Jeremy Good has started to catalogue, there are pieces of negative value. This Comment leads into pieces of negative value because Betza himself has some. Besides, there are other very weak pieces in Betza we may take up first.

Jeremy Good wrote on Sat, Sep 29, 2007 06:52 PM UTC:

Among most of those with whom I've playtested this game, there is an impression verging on consensus that the Black Ghost is too powerful, that it more than overcomes the first move priority. The idea though is still a very nice one.

Gary Gifford wrote on Sun, Sep 30, 2007 10:02 AM UTC:

The way the current Black Ghost game is, barring a blunder by Black, I believe his Black Ghost side will always win.

Antoine Fourrière wrote on Sun, Sep 30, 2007 03:09 PM UTC:

I added a version with a White Ghost on c3, d3, e3 or f3 and a Black Ghost on hand in this zrf.

gnohmon wrote on Mon, Mar 5, 2012 12:46 PM UTC:

I strongly support the idea of having a White Ghost on a central square and a Black Ghost on hand. Maybe the original Black Ghost variant is not as good as I had thought...

H. G. Muller wrote on Mon, Mar 5, 2012 03:20 PM UTC:

Well, I wouldn't know about that, and I also doubt the statement that this game would be 'always won for black'. (That would make the Ghost significantly more valuable than a Pawn, as Pawn odds at worst reduced your score against an equal opponent to ~30%.)

I can comment on the puzzles coined on this page, however:

KNNKG seems a general draw. At first glance this is a bit surprising: KNNK under rules where black can pass his turn is generally won. The trick seems to be that the Ghost can hinder the Knights in going where they have to go to drive the black King to the corner. It throws itself into the path of the Knights like a kamikaze, knowing that it cannot be captured because KNN is always draw. It needs a certain agility to be able to hold off the mate; a non-capturing Dragon Horse (RF) is not strong enough to prevent a general win for the Knights, but a non-capturing Queen is enough to let the number of wtm wins drop to below 30%.

KRKG is in general a draw; when the black King is already driven back to the 7th or 8th rank, however, it seems winnable.

Joe Joyce wrote on Tue, Mar 6, 2012 03:58 PM UTC:

I played Black Ghost with Jeremy Good, and we found that the ghost was far more powerful than it looked. The point, I think, is that the ghost can frustrate action by white that took some moves to prep, and thus black gains tempo. In the hands of a good player, this is often decisive, especially if white is left out of position.

H. G. Muller wrote on Tue, Mar 6, 2012 04:28 PM UTC:

Well, I will see if I can play-test it with Fairy-Max. It seems a very expensive piece, in terms of branching ratio. I guess that to not explode the search tree, I would need to apply extra search-depth reduction to Ghost moves.

I see that there was a proposal to rename the piece. The name 'Journalist' suggests itself. Going everywhere, and getting in everybody's way! :-)

Joe Joyce wrote on Tue, Mar 6, 2012 05:34 PM UTC:

H.G., 'Journalist' is an excellent name, exactly as you've described it! ;-)

H. G. Muller wrote on Sun, Oct 23, 2022 10:56 AM UTC:

To get a reasonable value for the gost I had to adapt the way the Diagram calculates piece values; the usual formula M*(33+0.7*M) for a piece with M moves values it more than a Queen even when I do count non-captures only for 1/3 (and captures for 2/3) as I did. But it turns out that (0.11*N+0.89*C)*(22+0.7*N), with N the number of non-captures and C the number of captures, gives exactly the same result when C = N (as it is for non-divergent pieces), and puts more emphasis on captures. And it still predicts the same devaluation when a piece with 24 moves gets one captures or non-captures removed, from which I concluded that captures were worth twice as much as non-captures.

This still predicts a value of about a Rook, though. I guess the problem here is that non-captures contribute a lot through future mobility, where you can go after 2, 3, 4... moves. But for a Universal Leaper this provides no extra value, as it can get anywhere in a single move anyway. I now made the hack to discount the number of non-captures above 1/3 of the board area by a factor 10. That still makes the Ghost worth more than a Pawn. But some comments below claim it is worth more than Betza thought, so perhaps this is correct. In any case the Diagram would not sacrifice a minor for the Ghost now.

13 comments displayed

~~Earlier~~ ⇧Reverse Order⇩ ~~Later~~

Permalink to the exact comments currently displayed.