Evolution of the rules

I haven't spent much time researching this but I am sure chess and it's rules have evolved over time.   Tiny rule changes through the centuries, survival of the fittest, or most balanced rules.   It's interesting that chess, like an organism has evolved to what it is today.   If that's true, my question is, will it stop evolving?   I wonder what the next tiny rule change will be?   Any predictions?   Or maybe a better question is what rule would you change, or introduce, to make the game even better.  Or is it perfect as is?

time control have change.

In my own lifetime and tournament experience time controls have changed a LOT. Adjournments have been done away with and sudden death time controls have become more and more popular. Increments and time delays have become common and neither existed when I started in 1973.  The majority of changes I have seen though tend to be more in the interests of the organizers than the players. 

I think that 50-Move rules should expand a bit, for that some really complex ending mate(like 2kinght v.s. pawn) takes a lot of the moves(almost 50).

The US 5 round swiss used to be played rounds 1-3 on Sat and rounds 4-5 on Sunday then everyone went home. Now most have the option of playing round 1 fri night or Sat am .  However, if you choose the fri night option you will most likely have to spend 2 nights in a hotel. If you choose to play round 1 Sat am then you will have to play your first round at a more rapid time control , an SD time control like G/75 or G/90.  In either case the swiss pairing system gets messed up because some will choose to play fri night while others choose to play sat am, to add to the mess/complications many events also now allow those who lose round 1 on fri night to " re-enter " , which I very much oppose. Now, since hotel prices are now more than twice what they used to be I prefer to spend only one night in the hotel. This eliminates fri night as an option for me. I also dont want to play at the faster time control on Sat am and the pairing confusion created by having the two options for round 1 AND possible re-entries. The best solution for me then is to take a bye in round one and avoid the madness and extra expense . This would also mean I wouldnt have to play until Sat after lunch which would allow me ample time to drive to the event that morning. This would also save the expense of a couple of meals out . 

It was like that for a while (in the 80s I think there was a 75-move rule exception for the 2N v pawn endgame) until more tablebases were developed and it was found that there are many, many endgames that are won but where the winning line is over 50 moves. Do we really want a huge encyclopedia of ending positions with their own exception to the 50-move rule? It was decided that the 50-move rule would stand without exceptions.

http://en.wikipedia.org/wiki/History_of_chess

You can't expect there to really be a single path along which Chess evolved, as during most of history there  was not  a single  world-wide organization like FIDE enforcing the same set of rules everywhere. So there must have been many local variants, some catching on and spreading, some being canibalized by others by taking part of their rules (those that made them attractive), but not the entire game. When visiting Italy I was told that they even used different rules there before WW-II. (Something like some Pawns pushes being prescribed, but I forgot the details.)

Roughly the development went from Shatranj (Arabic / Persian Chess), with Elephants (Alfil) in stead of Bishops and a General (Ferz) in stead of a Queen, through a game with Bishops and a Queen that moved as King in addition to these on a 12x8 board (Courier Chess)  back to 8x8 where the (quite boring) Shatranj pieces were dropped (so with Bishopsand a Queen that moved as King). Then they invented the distand Queen moves ('Mad Queen variant'), which is very much like what we use  today. Courier Chess had already some prescribed double-pushes of Pawns, and these were probably generalized together with e.p. capture as a compromise to appease opponents. (Not sure whether this happened before or after the Queen got its distant moves, and if that happened in  the same order everywhere.)

Also not sure when castling was invented. My guess is after e.p. capture, as the rule that the King may not pass an attacked square does not strike me as something you would think of if you hadn't heard of e.p. capture first. Noteworthy, though, is that the Indian variant Chaturanga, which even predates Shatranj, already had a rule that the King could make a Knight jump as its first move, as long as you had not been checked yet (speeding up it reaching g2 to let out the Rook)! So the realization that there is a problem for which castling is onepossible solution is very old indeed.

There's always Kasimdzhanov's idea about abolishing draws to promote fighting chess. Instead of sharing the point the players are supposed to switch colours and keep playing with shorter time controls until one of them wins. I can't say I'm too fond of that idea myself. Today all six games in the women's GP were decisive, as were all four games in the latest round in the Russian Superfinal. I don't see how the existence of draws should be such a problem, and the result would probably be that few long games would be played since the players would need to save energy if they risked having to play many games every day.

http://www.chessbase.com/newsdetail.asp?newsid=7387

But what about the En Passant rule?  I wonder when that popped into existence.  I still don't understand why it's needed.  If I were the king of chess I'd drop it.   And while I'm at it, I'd change the pawn promotion rule:  A pawn should only be able to rescue a captured piece, not allow 2 queens.

In practice it matters very little to what you are allowed to promote. I once tried handicapping one side in computer self-play games by restricting all promotions to a 'short Rook', i.e. a Rook that can move at most 2 squares away. This piece has approximately the value of a Knight. But it had virtually no effect on the result; the thus handicapped side still scored 50%. So for simplicity the any-number-of-Queens rule is probably a good choice.

In Shatranj there was no promotion choice, but you could promote to as many Generals as you wanted, even though you only have one General in the opening psition. (Promotion to worthless pieces like General is a bad idea, though, and the main reason why Shatranj is so drawish.) It seems that in the even older Chaturanga the piece you get on promotion was determined by the file you promoted on (e.g. on a- and h-file you get a Rook, etc.)

The importance of the e.p. rule is that certain Pawn endings that in absense of the double-push would be Drawn, become won when you allow a double push. (E.g. white: Ph3,g2, black: Ph4, with the Kings tied up elsewhere; g2-g4 provides an easy win without e.p.) As  the double-push was mainly motivated to speed up openings, I can imagine it ran into a lot of opposition from people that would see all their end-game knowledge invalidated by it. So e.p. was invented as a compromise to prevent double-pushes opening up the possibility to sneak past a Pawn guarding the square in front of you. So with this modification the double-push became acceptable to enough people to become standard. Furthermore, this only makes sense when promotions are really important. If it was just to a Shatranj General, no one would have bothered to amend the rules to make it more or less difficult... Unfortunately the promotion rules for Courier Chess are not known.