Talk:Fischer random chess/Archive 1

Archive 1

White advantage

The text compared that it is "possible" in 960 for White to attack a Black Pawn in one move, to that it is "typical" in standard chess that an attack takes three moves. First, it only takes two moves for White, not three, to attack. Second, to compare what is "possible" in 960 to what is "typical" in standard chess, is inherently unfair and unhelpful comparison. (So I changed the text so the comparison is between what is possible in both games. The alternative was to compare what is "typical" in both games ... but that is a darker forest to tread.) Ihardlythinkso (talk) —Preceding undated comment added 08:32, 14 September 2010 (UTC).

"Much" is a cloudy compare. And for consistency, in the same paragraph later, the advantage is referred to simply as "greater."

A "defensive position" is not necessarily the same as or the result of making a move to defend a Pawn! —Preceding unsigned comment added by Ihardlythinkso (talk • contribs) 10:43, 14 September 2010 (UTC)

Neoliminal's Proposal

(The following is a proposal by Neoliminal)

Castling in Chess960; an appeal for simplicity.

Bobby Fischer said, "That's basically it. You can learn the rules in two minutes. It's a great game, and can become the standard for chess." In 2003 David A. Wheeler contacted many active in Fischer Random Chess to determine the exact castling rules, including Eric van Reem, Hans-Walter Schmitt, and R. Scharnagl. All agreed that there must be vacant squares between the King and his destination except for the participating Rook, clarifying the castling rules for Chess960.

The intention of this article below is to promote the idea that the castling rules as originally presented by Bobby Fischer for Chess960 are flawed from a game design standpoint and that changing these rules would benefit the advancement and acceptance of Chess960 by both the Standard Chess and non-chess playing communities. Obviously there will be resistance by some in the current Chess960 community, as any change might be seen as suspect.

First let's review caslting rules for Fischer Random and Standard Chess.

Fischer Random castling rules. (quoted from http://www.chessvariants.org/diffsetup.dir/fischerh.html)

"Of necessity, in Fischer Random Chess the castling rule is somewhat modified and broadened to allow for the possibility of each player castling either on or into his or her left side or on or into his or her right side of the board from all of these 960 starting positions. However, after "a"-side castling, the King and Rook find themselves on the usual squares: King on c1 (c8) and Rook on d1 (d8), after "h"-side castling : King on g1 (g8) , Rook on f1 (f8). Sometimes castling looks odd in Fischer Random Chess: e.g when your King is on e1 and a Rook is on f1, you only have to move your King to g1 ("King-move-only" castling)."

A much longer explaination can be found at http://home.att.ne.jp/moon/fischer/list/p_20/20_0.htm, from the original Fischer Random text.

FIDE rules for castling. (quote from http://www.fide.com/official/handbook.asp?level=EE101)

"Castling This is a move of the King and either Rook of the same colour on the same rank, counting as a single move of the King and executed as follows: the King is transferred from its original square two squares towards the Rook, then that Rook is transferred to the square the King has just crossed."

Please note that for brevity I have excluded from both the conditions under which castling is not allowed. I will include these for the offered variation at the end of this article.

I think it's important to review a history of castling to understand part of the impetus for this change. Before the 1500s, in order to speed up the game, the 'King's Leap' allowed the King to move two spaces as his first move (jumping one square). Castling become a single move of a rather common two move opening where the Rook was moved next to the King and the King would use his King's Leap to move on the other side of the Rook on the following turn.

There are two points to be made here. The first is the King's ability to 'King's Leap' two spaces as his first move was changed to castling as a single move. This is a balancing of power from a game design point of view, because the King can no longer move forward or diagonally two squares on the 'King's Leap', but must instead interact with a Rook in order to gain the advantage of moving two squares as his first move. The second point is that the Rook always ends up in the square leaped over by the King. (further research available at http://cunnan.sca.org.au/wiki/Chess)

In examining Bobby Fischer's rules for Chess960's castling we see that it totally ignores the origins of castling in favour of appealing to chess players who are already familiar with the castling positions that result from Standard Chess. Unfortunately the system is confusing to newer players who aren't as familiar with Standard Chess. The final resting squares of the King and Rook feel arbitrary to new player who may not have played enough Standard Chess to intuitively remember them.

It seems that simplification of the castling rules for Chess960 could help promote the game for beginners, streamline the rules and reconnect the game with it's historical roots. To that end I suggest the following castling rules for Chess960:

Castling This is a move of the King and either Rook of the same colour on the same rank, counting as a single move of the King and executed as follows: the King is transferred from its original square two squares towards (or over) the Rook, then that Rook is transferred to the square the King has just crossed (if it is not already there). If the King and Rook are adjacent in a corner and the King can not move two spaces over the Rook, then the King and Rook exchange squares.

(1) The right for castling has been lost:

  1.  if the King has already moved, or
  2.  with a Rook that has already moved 

(2) Castling is prevented temporarily

  1.  if the square on which the King stands, or the square which it must cross, or the square which it is to occupy, is attacked by one or more of the opponent`s pieces.
  2.  if there is any piece between the King and the Rook with which castling is to be effected. 

This rule is much cleaner than the current set of rules and is the most obvious to new players who are introduced to Chess960 as beggining Standard Chess players.

As an excersize for the reader, I suggest finding a new Standard Chess player, explaining the rules of Chess960 to them without revealing how castling works, and then ask them to guess how castling would work in the variant you have just shown them. neoliminal 10:00, 18 September 2005 (UTC)

I don't see the relevance of this long proposal, and the text above, to an encyclopedia article on "Fischer Random Chess". You don't like the rules of Fischer Random Chess, and that's fine, I think any game could be improved! But this article is about "Fischer Random Chess", not about what you think the rules should be. The rules of Fischer Random Chess, aka Chess960, are as they are described in the article: after castling, the pieces always end up in the standard positions of an orthodox position. Fischer has stated that repeatedly; there's no question or controversy over this. The rules you propose, while they sound very workable (and have a historical continuity), are not the same, because they would cause the king and rook to end up at different positions in most starting positions. Wikipedia describes things as they are, not what you wish they were. Major groups have already agreed upon the rules as stated in the article, and international masters have already been playing the game using the rules described in the article, which are not what you propose. Send an email to Mr. Fischer if you want the rules to change; he's currently in Iceland. I doubt he'll change them for you. Until someone like Fischer or a Chess960 organization changes the rules, we have to keep the article accurate. I'm particularly sensitive to this topic; the initial rulesets for Fischer Random Chess, although all agreed that the final positions were always the same as orthodox chess, were maddeningly ambiguous about the necessary conditions for castling. I spent a fair amount of time tracking down people to find out the "consensus position" on the rules, including IMs, and then documented the result here on the Wikipedia. We cannot just go "change the rules" because we don't like them. -- Dwheeler 06:22, 19 September 2005 (UTC)

Heh. I was just about to write something on this talk page to the effect of "I have this great idea for new castling rules, but wikipedia doesn't allow original research, what should I do?", but someone (Neoliminal, who I don't know) beat me to it with the exact same idea. N.B. to clarify the idea: the new suggestion is that castling is defined as moving the kings two steps towards a rook, and moving the rook to the square the king crossed. (This is consistent with FIDE chess, but inconsistent with current Chess960 rules which state that there are two different "correct" castled positions, Kg1/Rf1 and Kc1/Rd1 which one has to memorize.) Again, I'm aware that wikipedia is not the place for original research, but where should we go with this idea? Surely it's allowed to suggest improvements to Chess960, somewhere. Where can this idea be submitted to the general Chess960 community for consideration? Direct communication with Fischer is not an option. --Sunnan

Possible design explanation of Fischer's castling rule

In most (standard) chess books that I read, the explanation of castling is as follows: 1. To tuck the king safely (though not always) in the wing.

2. To centralize the rooks - allow the rooks to control the central files.

All of the above objective are met in this 1 move. e.g. "The Right way to play chess", by D. Brine Pritchard.

The object of the move is two-fold: it brings a rook into play in the center of the board, and it gives greater security to the king.

If the alternative, 2 king move castling was the norm for chess 960, then none of the above two would be satisfied in some positions or when castling on one side. So in effect there is no point to castling in certain situations. The reason for the new supposedly convoluted rule of castling is to fulfill the above aims. Thus, this new rule is actually much more logical from a design standpoint than the proposed one only adheres to the "king two moves to left or right principle".

In summary, I believe that Fischer's castling rules completely adhere to the main objective of the original castling move in standard chess which is to centralize the rooks and to move the king to safety (away from the center).

Perhaps, someone can elaborate further on above and include it on main page.

If you want to create a chess variation that has the rules above, that's fine! But it has to have another name, because "Fischer Random Chess" and "Chess960" are taken. Trying to call a different variation "Fischer Random Chess" seems in poor taste; someone who names a game after themselves should get to write its rules. I propose that you create a new name for your new chess variant, and get it into the usual websites, etc., for chess variants. If it starts to get used, then a different Wikipedia article (linked to from here) would be entirely appropriate. Feel free to name the variation after yourself; obviously Fischer did it :-). -- Dwheeler 05:52, 19 September 2005 (UTC)

Now, if you just want to simplify the explanation of the rules, so that they're clearer but have the same effect, that sounds like a great idea!! But once the rules above are recast to describe Fischer Random Chess as it is, I doubt that the rules above are an improvement over what we have. The proposed rules for when castling is allowed have two sets of two rules (4 rules total); the current description has 3 rules total, and 3 is shorter than 4. The vacant-square rule is simpler, but only because it changed the meaning of the rules, which is NOT okay; this article has to describe Fischer Random Chess. But if you can rewrite the rules to NOT change, but be clearer, that sounds like a wonderful idea and I'd love to see it. -- Dwheeler 06:09, 19 September 2005 (UTC)

Hi David. I was hoping to bump into you at some point and it's odd that it happens to be here in Wikipedia. Quoting you "We cannot just go "change the rules" because we don't like them. -- Dwheeler 06:22, 19 September" Of crouse we can! Fischer did. All the changes in chess happened because someone just "changed the rules". They just changed the name when they moved it from Fischer Random to Chess960. I purposely put this in the discussion area because I knew it didn't belong in the actual article. However there is a growing movement which finds these rules acceptible, simple, and elegant. Currently there are TWO names for one game variant. Fischer Random Chess and Chess960. My suggestion is to change Chess960 to use these new rules. Unfortunately Chess960 is linked to this article. Obviously Fischer is set in his ways and likes the rules as written. So let me be very clear. I want to change the rules to Chess960 (not the rules to Fischer Random on which they are based). In effect Chess960 becomes the variant. Fischer Random Chess remains as state in the current article. BTW, you can contact me at jklewis@umich.edu for more direct contact on this idea and where it seems to be heading. neoliminal 15:06, 19 September 2005 (UTC)

Hi, it's a delight to meet you here too! I see where you're going. The problem is that when the phrase "Chess960" was created, it was INTENTIONALLY defined as being "the same as Fischer Random Chess, but with a name that doesn't include any Grandmaster's name". I don't know of anyone who accepts the term "Chess960" as meaning the castling rules you propose. Here's a note about the Chess960 name, and here are castling rules clearly showing that the end is always the same as orthodox chess. They've even had monied tournaments with the castling rules as currently stated. I would argue that Chess960 and Fischer Random Chess are clearly defined by the world at large as being "the same thing", and saying otherwise in Wikipedia wouldn't be honest (which I know is your goal too). Instead, I propose that you work to create a NEW NAME for your chess variant, carefully write down its rules, and place them on some web page, for comment. Post a few games with commentary (to show that it really HAS been played, and also to help make sure that the game is playable and that the rules are clear). Then submit that to Chess Variants so you get wider distribution of the idea. Once it's on the web, both your site and another site like Chess Variants, it's no longer original research. Instead, it's a published chess variant, and you can add a new article specifically about it, with cross-links between this and Fischer Random Chess. I guess the question is at what point is something "popular enough" that it deserves its own article. I tend to be more of an inclusionist, so as long as there are multiple people who play it, and published rules for it, I'd be happy. -- Dwheeler 18:19, 19 September 2005 (UTC)

To that end, I've created a new section that lists and links to similar games. If you can name the new version ("Chess960.2" or some other name), and get it posted elsewhere, then there's now a place to put it. -- Dwheeler 18:44, 19 September 2005 (UTC)

I'll get SchemindMind.com to use the castling rule in a Chess960 variant and then once it's established I'll link it in the similar games section. neoliminal 21:35, 19 September 2005 (UTC)

I've decided to call the castling rule "Orthodox Castling" and to call the variant "Orthodox Chess960". Rules to be posted to the variants chess pages shortly. neoliminal 16:17, 20 September 2005 (UTC)

Excellent. I see you have a future career in marketing, given the name you chose :-). I saw the change you made to the text here, and I tweaked it slightly. I don't think the article should say "Wikipedia article coming shortly", so I removed that. Instead, for now this article notes its existence, and that's enough. Once it's published externally, and there's evidence that it's been played, then just create the article and link to it. -- Dwheeler 21:50, 21 September 2005 (UTC)

This variant is now called Chess480. It is currently played at www.schemingmind.com and the rules use standard FRC with the addition of Orthodox Castling.

Okay. But why the weird name? "Chess 960" is for the number of initial starting positions... and that's not any different when the only difference is the different castling rules. -- Dwheeler 00:35, 30 September 2005 (UTC)

Someone pointed out that the only reason there are 960 different starting positions in FRC is bacause mirrored positions caslted differently. In essense there are now only 480 starting positions because half are mirrors where the new castling rule makes it tactically similar to it's mirror. That doesn't mean there aren't 960 starting positions in Chess480, just that tactically there are really only 480 positions because the mirrored positions aren't technically different, (aestestics aside). neoliminal 15:25, 8 October 2005 (UTC)

This is discussion for a reason.. I totally agree with the original poster. When playing chess960, all new players choose to use the neoliminal proposal, and when I suggest it to players already familiar with the game, more than half will usually agree to implement the new rule. It is pure logic. Of course in official tournaments they won't budge until enough people stir up a fuss to change the rule, and since the game is in its infancy of gaining popularity, it is safe to say this won't happen for a while. —Preceding unsigned comment added by 207.210.51.222 (talk) 07:20, 29 January 2008 (UTC)

Micky Adams reference correct?

The stuff about castling here seems very confused. For example, we have:

The king may not move through any square (towards its destination) that is under attack. Although it is not entirely clear, this probably should be interpreted as including occupation of the square by an opposing piece.

In other words, it isn't clear whether squares between the king and its destination can be occupied. But immediately after that it says:

All squares between the king and rook, and their final squares, must not be occupied by any other pieces. Castling cannot capture pieces.

In other words, squares between the king and its destination definitely can't be occupied. Also, squares between the rook and its destination can't be occupied. But then in the section on "FullChess" (a term I've never before come across, incidentally), it says that:

Later versions of the Fischer Random Chess rules clarified that indeed, Fischer intended to allow castling to "jump" other pieces.

In other words, squares between the king or rook and their destinations definitely can be occupied. So which is it? Is there, in fact, a single authority on Fischer Random rules (if so, what?) or is the game played with a number of little variations in castling rules?

On a completely different point, Micky Adams is described here as "world champion in blitz (rapid) chess" - is that really true? I don't know of any blitz world championship (I might just be forgetting something, of course). There's a blitz rating list, I think - does it perhaps mean that he was number one on that list (I seem to remember that he was at some point - maybe he still is)? --Camembert

That's a quote from one of the references (I've forgotten which one, but it shouldn't be hard to find). I haven't tried to track down a second source or to falsify the original source. Please do try to track that down to a second source or falsify the claim, if you can. -- Dwheeler 02:37, 27 September 2005 (UTC)

Older discussion on rules

When Fischer first came out with his rules, his castling rules were maddeningly ambiguous, which is part of the problem. "FullChess" was an attempt to give clear rules on the point. There is, finally, a single authority on Fischer Random chess rules - the official rules. The only version of the official rules that seems to be online is the Spanish version (see the reference list), so be prepared to use Babelfish. The rules are very clear that the spaces CAN be occupied between the king and its final position by your own pieces, as long as the squares between the king and rook initial positions ARE NOT occupied. In orthodox chess, those are one and the same, but not so in many Fischer Random positions. One weird thing is, what happens if there's an opposing piece that the king "jumps over" - is that okay? The official rules don't seem to be clear at all about this point (at least via Babelfish), unfortunately.

I'm making an inter-library loan for the book on Fischerandom chess. When I get that book, I'll see if it clarifies the castling rules further.

I've had a look at all the sites linked to from the article and read around the subject a bit. As far as I can see, those Spanish rules are just a translation of the rules at http://www.frankfurtwest.de/ChessClassic/cc03/e/c960/rules.htm - these are the rules used at the Mainz Chess Classic (where Leko and Adams had their match).

I'm not very sure what the confusion is over the king being able to leap over enemy pieces - what makes you think that enemy pieces might be treated any differently to friendly pieces? There's nothing in the rules to suggest they should be treated differently, and a piece doesn't attack the square on which it stands, so I don't understand what the problem is.

However, being able to leap an enemy piece is dependent on being able to leap a piece at all, which I must admit, I'm still not convinced is allowed. You say above "The rules are very clear that the spaces CAN be occupied between the king and its final position by your own pieces". I'm sorry, but I can't see anything that clearly says this is allowed - could you point out the relevant part to me? FICS, which purports to use the same rules as were used at Mainz, doesn't allow castling in that situation. ICC explicitly states that jumping over pieces is *not* allowed; their rules for Fischer Random say ". . . squares the king passes over or onto cannot be attacked by the opponent or occupied by pieces, squares the rook passes over or onto cannot be occupied . . .". Maybe these servers are just implementing the rules incorrectly, but I think this shows at the very least that the rules are not absolutely clear on this point. (Incidentally, in none of the games in the Leko-Adams match did this situation arise - I don't know about the 2002 Mainz event - I've not yet managed to track down the games played there).

Also, I'm not very sure we should spill so many words on "FullChess" - as far as I can tell the name hasn't spread very far beyond the website linked to from the article. I'm going to leave all this as it is for now, but I may well give it a rewrite at some point in the future. --Camembert

Early versions of the Fischer Random chess rules didn't give enough details on the rules for castling, which is why there are so many different interpretations. As you note, the ICC doesn't allow the king to jump over pieces, nor does FullChess. However, the writeups on the "Chess Variants" don't forbid this. The ruleset by Eric van Reem certainly doesn't either. I found one ruleset that clearly stated that some squares could be filled, but now I can't find it (naturally!). I'm going to try to get the Fischerandom book and see if that clarifies things. FullChess is sometimes claimed to be the same as Fischer Random, but it seems to have different rules, so I think some text clarifying the differences is the right thing to do. --Dwheeler 22:15 9 Jul 2003 (UTC)

It's true that the rules don't explicitly forbid jumping, but nor do they explicitly allow it, and my feeling is that allowing the king to jump over pieces while castling is so counter-intuitive that there would be an explicit note in the rules allowing it if it was meant to be allowed. Anyway, hopefully the Gligoric book will clear this up. --Camembert

Ahah!! I found it! It was in van Reem's description, see its "Remark" section. Squares can be filled that would ALWAYS be empty in orthodox chess, depnding on initial position. These were the rules used by grandmasters, which is about as official as we're likely to get. I removed the stuff about the opponent's piece, nothing seems to complain about that. I'm still going to try to get the Gligoric book (anyone else have a copy?), but until I get it, this seems to be the best available information. --Dwheeler 04:02 10 Jul 2003 (UTC)

Hm, I'm not sure that's what the "Remark" is supposed to be about. It's true, as it says, that squares can be filled that would always be vacant in orthodox chess, but that's not the same thing as saying that pieces can be jumped over. For instance, consider a starting position with the white king on d1, and white rook on c1 - it is possible to castle a-side even if a1, b1 and e1 are occupied, which would be impossible in orthodox chess; however, no pieces are leapt over in this case. I think that's the sort of position the comment is meant to refer to. Or, to look at it from the other side, consider a starting position with the king on b1 and rooks on a1 and c1 - is white allowed to castle h-side if d1 is occupied? Whether the "Remark" is meant to cover jumping over pieces or not, it doesn't cover this situation (only the squres e1 and h1 are mentioned in h-side castling, not d1).

I'll see if my library can get hold of the Gligoric as well. In the meantime, I'm going to email the press guy from the Mainz Chess Classic and see if he (or somebody else at Mainz) can clear this up - whatever the rule is meant to be, it certainly could be expressed more clearly than it is. --Camembert

That'd be great, thanks for trying to get hold of those folks. -- Dwheeler 16:35 10 Jul 2003 (UTC)

I'm now talking via email with Reinhard Scharnagl, Eric van Reem, and some others. I hope to get this worked out. -- Dwheeler 23:02 14 Jul 2003 (UTC)

Jolly good. I emailed a couple of the people from Mainz the other day (though not van Reem himself), but have yet to receive a reply. I hope this can be sorted out as well. --Camembert

van Reem's initial reaction is that the literal rules are not what he intended. But, what "the chess community" or Bobby Fischer really want isn't yet clear. In the past there hadn't been much discussion about the castling rules. Apparantly we've started such discussions, and they will try to work out clear, unambiguous rules. What exactly the rules will be is unclear yet. I think for the moment we leave the text as is, listing published alternatives. I'm in discussions with van Reem and others to make sure there's a simple, clear definition. Once there is one, we can label it as the official rules according to X and then note the other options later as historical information (if we want it in at all). Hmmm, writing an encyclopedia article causes a change in what it describes. I'm not sure what to call that process. -- Dwheeler 23:31 15 Jul 2003 (UTC)

There seems to be universal agreement that "jumping" pieces was not intended, even by van Reem. So, I've rewritten the text here to clarify the rules as intended. --Dwheeler 19:17 23 Jul 2003 (UTC)

Thanks very much for sorting all this out. We've got a better article and a better set of "official" rules in the offing. Maybe you should write to people about the orthodox chess world championship and sort that little mess out next ;) --Camembert

Curious naming comment

Bridge players would probably suggest it be called "duplicate chess".

Why? Even after scanning the article for bridge, I don't get the quip. (I'm assuming it's a quip). Would someone care to enlighten this bridge-nitwit? --- JRM 11:38, 2004 Sep 1 (UTC)

It's a reference to duplicate bridge, of course, but I don't get the point either. I'm going to remove that sentence from the article pending some sort of explanation. --Camembert 01:04, 12 August 2005 (UTC)

"Duplicate bride" uses the same "deal" for each player, so random chance isn't involved in the hands. It's connection to Chess960 is spurious at best.66.65.152.2 02:18, 4 January 2006 (UTC)

Another way to produce a 960 position

The Chronos chess clock has a "random position" feature. In fact, it has two options. You can generate a random position where white's bishops need not be on oppposite colors, and the king need not be between the rooks. This could be called a random 5040 position. Or, choose a position where white's bishops must be on opposite colors, and the king need not be between the rooks. This could be called a random 2880 position. You can create a random 960 position from a random 2880 position by examining the location of the king. If it does not lie between the rooks, then swap the king's position with the rook nearest to it.

I hope this does not constitute an advertisement. The Chronos is one of the more expensive clocks. I certainly would not advise purchase of one based solely on this feature. Maybe someone more ambitious than myself could add an entry in the Wikipedia for "chess clock."

--Twixter 18:14, 25 July 2005 (UTC)

Diagram

"The starting position for Fischer random chess must meet the following rules: .... The white king is placed somewhere between the two white rooks. .... Note that the king never starts on file a or h, because there has to be room for a rook."

The diagram shows the white king starting on h1. WTF? TheMadBaron 22:08, 11 August 2005 (UTC)

The diagram shows the white king on f1. The piece on h1 is the queen. --Camembert 01:01, 12 August 2005 (UTC)

DOH! TheMadBaron 07:42, 12 August 2005 (UTC)

Useless trivia: One of the very first "Fischer Random" games was held when the rules for the starting position were very unclear. If I recall correctly, the "standard" starting position had the king NOT between the rooks. After that oops, the rules for the starting position were clarified to what was intended.

Flaws in the initial set up systems provided.

There are serious flaws in the set up systems provided on this page. They will not result in an equal distribution of all 960 positions. The restrictions provided for in the initial set up need to be followed strictly. For example, you should never place the Queen before placing the Rooks and then the King as these pieces have more restrictions on how they may be placed. The only ideal solution (until I can think about this further) is to select a number randomly from 1-960 and consult a chart. neoliminal 14:59, 23 September 2005 (UTC)

Rolling a d1000 would work! Just reroll results between 961 and 1000.--Sonjaaa 06:14, 24 September 2005 (UTC)

Some of the setups that have been proposed aren't of equal distribution, but I did an analysis earlier, and the Queen-before-Rooks placement for the dice-rolling approach is okay. The trick is that you never place just the rooks at all; you place all the other pieces , until there are 3 holes left. Then you place Rook-King-Rook in the 3 holes, in the only way allowed. Since there's no rolling to place the rooks -- it's always a specific way -- it can't have an uneven distribution. Am I missing something? -- Dwheeler 02:42, 27 September 2005 (UTC)

No, you have it right. The Rook/King/Rook must be the last. I was wrong about the Queen being placed before the Rooks. Any system where the King or Rooks are placed before other pieces may result in a problem, as the King and Rooks share a bell curve in terms of placement.

---

Fairness:

Is the game fair? Isn't there an advantage for first move? During an analysis of the intial setup of orthodox chess, that seems to be the superior setup. This means that any Chess960 setup, except the one identical to orthodox is inferior and can be exploited by the first player. Does anyone have more information on this?

From play it appears that Chess960 has the same first move advantages enjoyed in standard chess. 66.65.152.2 02:22, 4 January 2006 (UTC)

Just briefly studying different setups aside from the regular chess setup show either abuse of first move by well positioned pieces or weak control of center as opposed to the regular setup. Any other opinions/research on this matter? 70.111.224.85 03:25, 6 January 2006 (UTC)

This is an ugly problem with traditional "chess1" as a proper sport. However, this unfairness need not and should not exist in chess960. The SOLUTION is simple. The problem persists only if we let it. Player AA makes the first move of a White piece. Next, now and only now, player BB has the one time option to decide whether BB wants to play as White or Black. This compels AA to open with a move that gives both colors equal chances. In my chess960 book I explain the way to avoid players having to switch chairs or rotate the board. This "Fair First Move" rule works in chess960 because there is such an enormous pool of reasonable first moves. Plus, the move 1. e24 is a different move in different initial setups. In contrast, decent opening moves are in too short supply in chess1. One reason is the initial setup never varies, and another is that its one initial setup is symmetrical on the two wings 'a' and 'h'. It is a straight jacket. This reminds of the mother who tells her son to cut the last cupcake in half, and tells her daughter to choose her preferred half. (Gene Milener)

---

I think it's important that we clearly note that most of the methods listed here do not produce truly random outcomes. I tried doing it with a single note at the top of the "other methods" section, but user 70.33.210.218 seems not to have liked that, so instead I have put notes at the top of each method. It'd be unfair to make readers read halfway through each method before finding some vaguely phrased comment about it not being random.

Actually I'm wondering how useful it is to list all these faulty methods, when the method with ordinary six-sided dice, listed early in the article, is so easy and produces a random result. And the article is "too long" per Wikipedia already. I won't delete them but I will suggest that someone else think about doing so. DanielCristofani 00:22, 19 November 2005 (UTC)

Daniel C. is right, there are too many setup determination methods listed in this wiki article. People should use one 6-sided die and be done with it. (Gene Milener)

---

If Chess960 is ever to be played in all of the venues and under all of the circumstances that standard chess is played, a convenient way to set up the pieces is necessary. Furthermore, the faster the pieces are set up, the better.

A convenient way is necessary, yes; it is desirable that the randomization method not take longer than twenty seconds or so, yes. I don't see that a method taking eight seconds is much better than a method taking twelve.

Perfect randomness is an ideal which is not met by any ordinary physical apparatus. I think the main question to be faced is whether a particular departure from randomness, even theoretical randomness, is so bad as to compromise the main goal of Chess960. In other words, could someone with knowledge of this departure prepare openings enough to gain a measurable advantage over someone who does not take advantage of this departure?

Whether the bias will significantly compromise gameplay is only one question. Another is whether using a known biased method is actually allowed by Fischer or by the accepted rules of Chess960; remember, the "main question to be faced" for Wikipedia is not what this game should be, but what it is (as was made clear in the discussion about simplifying the castling rule). A third question is whether the biased methods actually take any less time or are any easier than the unbiased ones. Granted that the bias will not significantly compromise gameplay, I would still like to see some information on the second and third questions.

Isn't rather early to pass judgement on what the balance of opinion between convenience and theoretical randomness will be several years from now?

(Again, it has not been shown that the known biased methods are appreciably more convenient than the theoretically random ones. Some, like the put-the-pieces-in-a-group-and-flatten-them-into-a-line method, are certainly more convenient, and thus should be listed, assuming Chess960 rules allow the use of biased methods. For others it is less apparent. Granted, we don't know how people will decide to make the tradeoff; still, a method should only be included in the list if it has SOME advantage over any other given method in the list. But again I will leave it to other people to decide if all of these methods meet that criterion.)

To this end I propose that a variety of methods for setting up the pieces be readily available. Those that give random results in theory certainly have greater aesthetic appeal, and should be labelled as such. For the others, it would be desirable, if convenient, to provide information about the nature of the departure from randomness. Since the text of these individual methods can be cut and pasted anywhere, it makes sense that each one of these carry its own statement about randomness or lack thereof.

Fair enough.

On the skepticism front, it seems appropriate that, for now, random position generators built into clocks, computer software, or other apparatus, be given a free ride. But, when high stakes depend on the outcome, some certification of good behavior should be expected.

One other remark. In some religious circles, and in some parts of the world, dice are considered to be gambling apparatus, and are proscribed.

So we should include at least one method that does not use dice. DanielCristofani 08:34, 27 November 2005 (UTC)

There are alternatives to dice that are probably well known in such circles too. Common items like coins or a pencil as a rolling log can be used. Doesn't belong in this article. — Preceding unsigned comment added by 88.112.76.119 (talk) 19:42, 26 June 2012 (UTC)

                  Edward Northam     23 Nov 2005

You think you have the castling rules nailed down?

I don't see anything in this page that would imply that a-side castling must be done with the rook that is closer to the a-side than the king is. The rules have changed enough that we can't take that for granted. DanielCristofani 22:48, 15 November 2005 (UTC)

Let me revise that a little; I see things in this page that would imply that, but they are labeled as consequences or side effects of the rules, whereas the rules do not imply it. DanielCristofani 22:51, 15 November 2005 (UTC)

Stolen sections

A few lines in this article appear to have been taken wholesale from other sources. For example: "...all of the study necessary to play conventional chess made it hard work, and he had gotten into chess in order to avoid work!" is not only non-encyclopedic and POV, it also gets several Googles hits. [1] --LostLeviathan 04:13, 25 December 2005 (UTC)

New Chess960 book to add to list

NEW CHESS960 BOOK TO ADD TO LIST:

We need more chess960 books to be published.

In Jan 2006 I (Gene Milener) published a new chess960 book, in English. It is available on Amazon.com and Amazon.co.uk. It is entitled "Play Stronger Chess by Examining Chess960: Usable Strategies of Fischer Random Chess Discovered" (abbrev. PSCbyEC960). ISBN 0-9774521-0-7.

Excerpts of the book are available at http://CastleLong.com/.

I am hoping mention of PSCbyEC960 can be added to the mentions Wikipedia already lists for the only other two chess960 books yet published...

In 2004 Reinhard Scharnagl published a book about chess960. It is in German, and it is available on Amazon.co.uk. Scharnagl's book is listed at the bottom of the Wikipedia chess960 article.

Also listed is Svetozar Gligoric's book on "Fischerandom" chess, published in 2002.

Thank you. Gene Milener

Gene M 06:26, 9 February 2006 (UTC)

WP:NOADS, WP:COI. Please read.Jasper Deng (talk) 06:12, 11 February 2011 (UTC)

Propriety of "Non-Random setup"

It doesn't seem appropriate for the author of that particular section to make a personal appeal to tournament directors, "In the Author's viewpoint it would be desirable that the play&start mode could be applied to every chess event immediately ". Am I wrong on that?

No, you're right. Roberto Rovida (not a known chess authority) added big chunks of original research to this page four days ago, deleting some useful content at the same time. Of course, most of the randomization methods on here are original research, and superfluous, but given that I've said I'm not going to be the one to delete them, I am going to try to rephrase Rovida's method (and restore the wiped content) rather than just revert the article. DanielCristofani 09:19, 11 March 2006 (UTC)

opening traps?

Are there any known opening traps in some chess960 starting setups? These tend to be well known by experienced players in standard chess, but maybe there's some surprises in some chess960 initial setups.--Sonjaaa 06:46, 17 March 2006 (UTC)

move page to Chess960

This game seems to be called Chess960 way more often than Fischer Random Chess at the professional level. I think we should move it. Here's a Google frequency test, for example:

• 70,500 for chess960
• 28,900 for "fischer random chess"

--Sonjaaa 06:46, 17 March 2006 (UTC)

Hm. I get different numbers:

chess960: "about 78,600", "In order to blah blah blah, very similar to the 316 already displayed.", "about 82,000".

"fischer random chess": "about 65,000", "In order to blah blah blah, very similar to the 483 already displayed.", "about 67,400".

"fischerrandom chess": "about 162", "In order to blah blah blah, very similar to the 65 already displayed.", "about 161".

This means "fischer random chess" has a larger number of substantially different pages using it. And the numbers are not off by a factor of 2, anyway. So Google is not decisive.

I want to be clear: the page move sounds like a good idea, offhand, and if there is evidence or consensus that professional opinion favors the name "Chess960" I will support it unreservedly. DanielCristofani 23:11, 23 March 2006 (UTC)

It seems Sonjaaa is right. I've moved it. —Nightstallion (?) ^{Seen this already?} 08:29, 28 March 2006 (UTC)

Thanks, good idea. I contributed to the early versions of this article, and at the time "Fischer Random Chess" was more common. I think you're right, for a variety of reasons Chess960 has become more common. -- Dwheeler 04:39, 25 May 2006 (UTC)

I just did a bunch of edits of the text itself, so that it usually refers to the game as Chess960. Exceptions are at the beginning (where it notes the older name), the discussion on the name, and the discussion about early tournaments. I think the move is complete. -- Dwheeler 20:44, 22 June 2006 (UTC)

I wish I had joined the debate earlier. Google results and professional opinion or consensus does not give anyone the right to take it upon themselves to just change the name of a patented game does it? I would assume that if Fischer wished, he could probably turn to the courts if tournaments were held using that name. Millions of games have someone's name on them - Rubik's Cube anyone? Wikipedia should have remained on the objective side and kept the name as Fischer Random Chess, not succumbing to popular opinion - in my most humble opinion. TheChessPlayer 7 July 2006

960?

This means that there are 4×4×6×5×4 = 1920 possible positions if the two knights were different in some way. However, the two knights are indistinguishable during play; if they were swapped, there would be no difference. I'm confused about something. Doesn't the same hold true for the two bishops? If you swapped the two bishops, there'd be no difference, so, shouldn't it be 480 possible positions? I'm sure I must be overlooking something, but what? Nik42 04:34, 3 April 2006 (UTC)

Never mind, I just realized what I was overlooking. The bishops are determined independantly of each other, and the first bishop's position does not restrict the second one's Nik42 04:37, 3 April 2006 (UTC)

There are 960 possible starting positions (PSPs), not, as the main article states, "960 possible (non-mirrored) positions." The king cannot start on a or h, since it must have a rook on either side and the king's occupying a or h would prevent this. Exactly half of all PSPs involve the king at b, c, or d.

With king at b, left rook is at a and there are 6 possibilities for right rook. With king at c, left rook is at a or b and there are 5 possibilities for right rook. With king at d, left rook is at a, b, or c and there are 4 possibilities for right rook. So the total PSPs, just for the king and rooks = (1x7)+(2x6)+(3x5) = 28.

Of these 28, only 4 combinations involve the occupation of 3 dark or 3 light squares; the other 24 occupy either either 1 dark and 2 light squares, or 2 dark and 1 light square. This is significant because the next step is the placement of bishops, which is square-color sensitive.

Each of the 4 monochromatic-squares occupation of the king and rooks allows 4 combinations for the 2 bishops. (Remember, the bishops must be located on opposite-color squares.) Each of the 24 non-monochromatic-squares occupation of the king and rooks allows 6 combinations for the 2 bishops. So the total PSPs, just for the king, the rooks, and the bishops = (4x4)+(24x6) = 160.

Each of the 160 combinations leaves 3 spaces for the queen and the knights. If the queen is placed next, she has 3 possibilities. The remaining knights collectively have only one possibility at this point. This is because the knights, unlike the rooks and the bishops, have no set-up rules that distinguish one of a kind from the other. So the total PSPs for all pieces = 160x3 = 480.

But remember, the forgoing discussion considers the king only occupying b,c, and d. Each of the 480 has a mirror alternate that includes king occupying e,f, or g. So the total PSPs, if we include the mirror alternates = 480x2 = 960.

Therefore, it is incorrect for the main article to state that there are "960 possible (non-mirrored) positions." It should simply state that there are "960 possible positions." 71.61.5.211 (talk) 07:02, 12 December 2011 (UTC)

Yes, the phrase "non-mirrored" was meant to cover that mirrored setups are still unique regarding play due to castling making otherwise mirrored setups different. But the phrase is unfortunate and confusing, as you point out. What if the text is changed to read: "... full castling options are retained in all starting positions, resulting in 960 unique positions."? Ihardlythinkso (talk) 07:48, 12 December 2011 (UTC)

Or even: "... resulting in 960 different positions."? Ihardlythinkso (talk) 22:00, 20 December 2011 (UTC)

Serious Complaint

Bobby Fischer did not invent anything. Shuffle Chess was around a long time before Bobby announced that everyone had ruined the game and that we'd have to use his special method to make it real again. Bobby merely made a few trivial additions to the ruleset. This is common knowledge, and I'm really appalled this article exists without anyone having ever properly checked up on the history of this variant. Krabbe ZincAtari

• Hello there; it seems right - I urge you to make the necessary changes for this rather big mistake. And it would be much better if you can then provide more references supporting the older origin of the invention of this chess-variant. Thank you, __ Maysara 17:36, 19 May 2006 (UTC) -- It turned out that shuffle chess is not synonymous with Chess960 but is a variety of it. Then all is fine with this article. Best, __ Maysara 13:02, 22 May 2006 (UTC)

• "Shuffle Chess" is used as the general name for any chess ruleset where the starting position is random, as well as a particular version that allows arbitrary locations of the pieces in the back row. Chess960 / Fischer Random Chess is an example of the general class of shuffle chess variants, but it is a specific version created by Fischer... I find no evidence that this particular ruleset was published before. You may consider the additions to the ruleset trivial, but generally different rulesets are given different names so that it's easy to agree on the ruleset used. So I think the article is correct, Fischer created this set. -- Dwheeler 04:37, 25 May 2006 (UTC)

Position systems removed?

Dividing up the article is fine, but it's odd that the position systems were removed entirely to the startup. It means that the "#177" marker at the top isn't explained anywhere in the article. Should we move the position system text back here?

I think we need to remove #177 from there. As far as I understand different Chess960 software will generate different position given the number "177", there is no widely accepted standard. The article is already long, I would propose to keep less essential stuff in a separate article. Andreas Kaufmann 23:35, 14 September 2006 (UTC)

No, all correctly working Chess960 software will generate the same position given the same number. See Chess960 starting position and Chess960 Enumbering Scheme for more information. --ZeroOne 15:45, 15 September 2006 (UTC)

In Chess960 starting position you can read: ...Late in 2005, the program Fritz9 became available. It has a Chess960 option, but, for some unexplained reason, it assigns idns to SPs in a different way. Andreas Kaufmann 07:19, 16 September 2006 (UTC)

Nice

This is a good article on a complicated subject. Good work.

Preposterous Claim

This sentence:

...it makes computer chess programs much weaker, as they depend on the opening book to beat humans.

is preposterous. The exact opposite is true. A computer will be able to "memorize" every single possible opening in any of the 960 starting positions. All it takes is someone diligent enough to feed all this data into the computer. One of the two major advantages computers have over humans is their practically limitless memory (the other is speed). All you have to do is buy more hard drives and connect them to the computer. Remember, the whole point of Chess960 is to diminish the dependence on memorization. Therefore, with time, computers will rule Chess960 and even the best humans will have a very hard time keeping up. I'm removing this ridiculous statement within a few days. Let me know if you have an objection.

Say the computer memorizes every possible sequence for (say) the first six moves on each side. Forget having someone "diligent" feed it in; the computer can just generate these sequences. Now, take the billions of positions at the end of these sequences. Which of these positions are best for the computer? The computer really can't tell; estimating the value of a position is what computers are worst at. And no human is going to look at these billions of positions and tell the computer which are best, either. Not knowing which outcomes are best, the computer also cannot know which moves are best. With regular chess, we know which moves are good in the opening from many decades of study, plus we have massive databases of master chess games, so we can say that people who play move X in the opening go on to lose the game 81% of the time, or whatever. Based on this we can tell the computer what specific move to play in many opening positions, and that's its "opening book". There is no quick way for a human, a computer, or a human-computer team to learn the same amount of opening theory for the other 959 starting positions. In essence, Chess960 drops the computer (and the human) straight into midgame-style play, where positions are assessed using limited lookahead and a lot of guesstimation and intuitive thinking. Again, this is the part the computer is worst at. So that preposterous claim is correct. I wonder if we can rephase it so it's clearer. DanielCristofani 00:47, 1 December 2006 (UTC)

I did not mean that the computer be fed all possible openings. I meant it will be fed all openings from past games so it could analyze them. This is what is done with all chess programs today.

With regular chess, we know which moves are good in the opening from many decades of study, plus we have massive databases of master chess games, so we can say that people who play move X in the opening go on to lose the game 81% of the time, or whatever.

So, in 100 years, when computers will have had decades of "studying" tens of thousands of games and situations, who's going to be on top? No human will be able to beat a computer then, since every time you start a game, the human starts from scratch whereas the computer has hundreds of past games to look at. This claim is therefore true today, but as time goes by, will become less and less true.

In essence, Chess960 drops the computer (and the human) straight into midgame-style play, where positions are assessed using limited lookahead and a lot of guesstimation and intuitive thinking.

Again, that's true today. What about the year 2113? Will it be true then?

In 2113, people will start having named openings and systems in Chess960, assuming of course it substitutes standard chess. Computers will just be as good as humans. This is what I believe. -Sibahi_talk 23:26, 1 December 2006 (UTC)

There's no way a human can sit to a game in 2113, look at a starting position that was just created by a die-toss, and say, "oh, yeah, I remember that, I've seen several games that stated this way and I remember how they developed". There are just too many different starting positions and openings for each of these positions. The computer will pop all of the previous games out of its hard drive in a millisecond, and for the first few moves of the game the computer will have a huge advantage. After that, it's back to common sense, and the human has the advantage again.

You have a point, but you need to re-estimate the time frame. For a computer to learn the Chess960 opening as well as it knows the chess opening, using the same methods, people will have to have played 960 times as many Chess960 games as they have played regular chess games. Maybe that will never happen; and there's no WAY it will happen within 5,000 years. At which point we can think of a new solution. Assuming our civilization has not collapsed by then. DanielCristofani 02:42, 2 December 2006 (UTC)

Hold on here. This is highly dependent on context. By "computers" do you mean specialized chess-playing computers like Deep Blue, which are pitted against world champions, or general chess programs, which are pitted against the average chess player when his boss isn't watching? Computers that are designed specifically to play chess will find it harder to tweak their moves (Deep Blue for example was tailored specifically against Kasparov, so one can only imagine how much work would be needed to tweak for different start positions). In that sense, yes, "computers" will lose much of the advantage that they currently have. However, if you are talking about general chess programs, they rely mainly on general heuristics -- and these heuristics aren't likely to change because of different start positions. So these programs, which are already quite tough vs average/casual players, aren't going to lose much if anything. 24.6.99.30 06:21, 17 January 2007 (UTC)

Who really cares if the computer masters the game.. the point is that there's no way for HUMANS to do so which is all that really matters, and because of this it'll never be bland, nor would you play the 'same' game over, over, and over... -thefairman

While it is interesting to debate such issues, the criterion to decide whether the comment is in place or not is not whether we can come up with logical arguments to support it, but whether we can find references to support it. 188.169.229.30 (talk) 13:12, 22 April 2012 (UTC)

Grammatical Error

Using the word "die" sounds horribly confusing. I don´t know where you guys are from but, I think using the word "dice" would be a lot more clear for a lot more people, especially those with limited knowledge of the english languange.

One die , two dice. Die is singular, dice is plural. I don't see it's confusing at all. Besides, the context is not confusing at all, even if you have no idea what the word 'die' means. -Sibahi 09:23, 23 June 2007 (UTC)

We're from English speaking regions of the world. Any further questions? —Preceding unsigned comment added by 207.210.51.222 (talk) 07:07, 29 January 2008 (UTC)

POV and citation needed

This quote from the opening: "The random setup makes creativity and talent more important than memorization of opening moves." is stated as though it is a fact and not an opinion and sounds distinctly POV. And it is not cited.

It seems to me like it could be improved by giving citations for the view and changing it to something like "Proponents such as Mr. X, Mr. Y and Ms. Z believe the random setup makes creativity and talent more important than memorization of opening moves (which they believe to be a weakness in standard chess)" 75.166.41.38 (talk) —Preceding undated comment added 03:27, 24 January 2011 (UTC).

Shuffle chess

"Shuffle chess" redirects here, but according to the article Chess960 is a specific variant of shuffle chess. This is not so clear. It will be good if somebody who knows what shuffle chess means exactly writes about it in short. Perhaps the redirection should also be cancelled, with a short article about shuffle chess mentioning Chess960 - as well as other variants - as special cases. 188.169.229.30 (talk) 13:23, 22 April 2012 (UTC)

The reason for the redirect is that shuffle chess never really caught on, while Chess960 became very popular. A lot of that is because of Bobby Fischer, but if you play a few games of each you will find that shuffle chess is less chesslike. If you end up with two white bishops and your opponent has two black bishops, you end up putting most of your men on white while your opponent favors black squares. That requires a different strategy than chess. Also, nobody ever defined shuffle chess as well as Fischer defined Chess960, so you have to decide whether you can castle and under what conditions. I really don't think a separate article on shuffle chess would survive deletion for being non-notable. --Guy Macon (talk) 18:34, 22 April 2012 (UTC)

Castling question

In the section titled "Castling," in the paragraph beginning "When castling using a computer interface..." there's a bit that says "...he king is moving to his at least two steps..."

Question: Moving to his what? Left? Right? Side? Wjl2 (talk) 14:59, 16 May 2012 (UTC)

It must mean at least two squares to the left or right. But that is a very bad paragraph (as are some others in that section). I think that either they were written by someone who is not a native English speaker or it has been vandalized. Bubba73 ^{You talkin' to me?} 19:33, 16 May 2012 (UTC)

Unprotected pawns

As the article mentions, an unprotected pawn in the starting position may strongly shift the balance toward White. It might seem as a good idea to restrict the starting positions to those that have no unprotected pawns, as is the case in normal chess. Has this ever been suggested (by a notable person) as another chess variant, and has the number of such possible positions been calculated? (The starting position given in the infobox has an unprotected pawn (a), so it might seem that unprotected pawns are rather normal in Chess960.) --Roentgenium111 (talk) 23:51, 24 May 2012 (UTC)

Reinhard Scharnagl did that kind of work for Capablanca Random Chess. For 960 I don't know. (Am not convinced either, that a 960 position w/ unprotected pawn "strongly shifts the balance toward White". Why should it anymore than in standard chess where White often has first chance of attack, e.g. 1. e4 e5 2. Nf3 ?) Ihardlythinkso (talk) 03:42, 25 May 2012 (UTC)

Thanks for the link! I've manually found about 90 positions in Chess960 without unprotected pawns, but would still be interested in the total number. (I do see a difference in White's advantage: e.g. in the situation you give Black voluntarily put his pawn in an unprotected position, whereas with an unprotected starting position, this is unvoluntary and Black has to sacrifice a move to protect the pawn. But I'm no good chess player, so maybe you should discuss this with the editors who added that assertion to the article.)--Roentgenium111 (talk) 11:45, 12 June 2012 (UTC)

Here's an interesting article from 2009 saying 360 positions have all pawns defended.

I'm sure there's lots other research on this too. Regarding voluntary vs. involuntary, does it matter? (In both cases Black needs to react to the threat. Is there compulsion for Black to defend the pawn? White's pawn is equally vulnerable, as in Petrov's 1. e4 e5 2. Nf3 Nf6.) Am sure computer analysis has begun on the 600 positions w/ at least one pawn undefended, so what we really need is a ref when one is available. (Does the 2006 Milener book contain anything in this regard?) Ok, Ihardlythinkso (talk) 13:42, 12 June 2012 (UTC)

Advantage

Is there any way to determine whether the randomization results in even or uneven initial position? --174.44.93.242 (talk) 23:18, 26 May 2012 (UTC)

Well, you could put all 960 starting positions into a good chess engine. Bubba73 ^{You talkin' to me?} 00:07, 27 May 2012 (UTC)

It depends on what you mean by "even". It's hard to see how white could be worse in any of the starting positions. So he must have better practical chances of winning. If you mean to ask if it's a draw by theoretical best play, it would require solving chess. — Preceding unsigned comment added by 88.112.76.119 (talk) 02:28, 26 June 2012 (UTC)

Why shouldn't White be worse off in some starting positions?

@OP: This blog post gives statistical data on which position seems to favor Black or White, with large differences (but probably too small sample sizes). --Roentgenium111 (talk) 22:39, 1 August 2012 (UTC)