AGA modification, would it work?

[tiger314]

Reading about button Go and granularity, I had this idea. What if the AGA ruleset was modified to komi 7 and first to pass wins in case of a tie. Would such a ruleset be applicable?

[HermanHiddema]

I prefer "last to pass wins ties" (which requires removing the "white must pass last" rule, of course)

[oren]

Reading about button Go and granularity, I had this idea. What if the AGA ruleset was modified to komi 7 and first to pass wins in case of a tie. Would such a ruleset be applicable?

Do you have a reason for this change?

It seems a change for the sake of change without a benefit as I read it.

[RobertJasiek]

The texts are a bit ambiguous, so I am not sure if I understand their intention correctly. If they want to create a ruleset with pass stones and without an equal number of black and white moves (territory scoring?), this might result in pass-fights and therefore such a rule design does not work. However, maybe a clarification avoids such?

[Bill Spight]

Reading about button Go and granularity, I had this idea. What if the AGA ruleset was modified to komi 7 and first to pass wins in case of a tie. Would such a ruleset be applicable?

This is like button go with 7 point komi, since the only time the 1/2 point button matters to winning or losing is if there is a tie otherwise.

[tiger314]

The point of this theoretical ruleset is that during endgame, the players are encouraged to play more carefully, since not only the area controled matters, but also the player to play the last worthwhile move. Normally in case of an even number of dame points, the player to move can reinforce for free. This is avoided by this simple rule addition.

To make it more clear, all usual AGA rules including pass stones and white to pass last still apply. The only difference is the komi value and the additional rule regarding draws.

I am not proposing this as a ruleset to be adopted, I am just curious if it would work and what would be the consequences.

[HermanHiddema]

The texts are a bit ambiguous, so I am not sure if I understand their intention correctly. If they want to create a ruleset with pass stones and without an equal number of black and white moves (territory scoring?), this might result in pass-fights and therefore such a rule design does not work. However, maybe a clarification avoids such?

I'm not sure if you're referring to my post too, but let me respond:

If there are pass stones, no rule that white must play last, an integer komi, and a rule that the last player to pass wins ties, then there are no pass fights. This is equivalent to the rule I dubbed "reverse button go" on SL (except I worded it: last player to pass gets 1/2 a point)

Explanation, assuming two consecutive passes end the game:

Case 1. Suppose the last dame has just been played by black, and there is a tie (including the integer komi). White passes. Now black can pass and win, due to the "last pass wins ties" (both have given one pass stone)

Case 2. Suppose the last dame has just been played by black, and white is one point ahead (including the integer komi). White passes. Now if black passes, he loses. If he starts a pass fight (B ko-threat, W response, B pass, W pass) he gains one point (due to the extra pass stone). But since now it is white who passed last, white still wins.

Note that you can change the number of passes to end the game to whatever number you like, without causing pass fights.

This rule can basically serve as a drop in replacement for the "white passes last" rule, with the advantage that you only have to think about it in case of ties, which are rare.

[RobertJasiek]

Case 1. Suppose the last dame has just been played by black, and there is a tie (including the integer komi). White passes. Now black can pass and win, due to the "last pass wins ties"

Since Black wins in the manner, White tries something better than passing: White tenuki, Black pass (pass stone), White pass (pass stone and tie and White wins the tie as the last passing player).

Therefore Black must also improve his strategy: White tenuki, Black tenuki,... and we have a pass-fight.

***

How do you solve played out removals?

[tiger314]

Case 1. Suppose the last dame has just been played by black, and there is a tie (including the integer komi). White passes. Now black can pass and win, due to the "last pass wins ties"

Since Black wins in the manner, White tries something better than passing: White tenuki, Black pass (pass stone), White pass (pass stone and tie and White wins the tie as the last passing player).

Therefore Black must also improve his strategy: White tenuki, Black tenuki,... and we have a pass-fight.

***

How do you solve played out removals?

Could you please also comment on the original proposal: all rules from the AGA ruleset apply, except for an odd integer komi (7) and first to pass wins ties?

[HermanHiddema]

Case 1. Suppose the last dame has just been played by black, and there is a tie (including the integer komi). White passes. Now black can pass and win, due to the "last pass wins ties"

Since Black wins in the manner, White tries something better than passing: White tenuki, Black pass (pass stone), White pass (pass stone and tie and White wins the tie as the last passing player).

Therefore Black must also improve his strategy: White tenuki, Black tenuki,... and we have a pass-fight.

***

How do you solve played out removals?

Territory counting is used, so tenuki costs a point.

[RobertJasiek]

Territory counting is used, so tenuki costs a point.

Territory counting with the implication of territory scoring, I see.

However, a rigorous proof is missing why tenukis played in the opponent's territory combined with approach plays to achieve removals can never be an advantage for a tenuki player.

We only have Berlekamp's / my proposition that, for a static territory region, an equal number of black and white plays in it do not affect the region's score. So how do you prove that, in each territory region, there is the equal number? Without equal number, a player's tenuki(s) might swap the last passing player, and we might have a pass-fight.

Could you please also comment on the original proposal: all rules from the AGA ruleset apply, except for an odd integer komi (7) and first to pass wins ties?

Please define "to win a tie".

[DrStraw]

Case 1. Suppose the last dame has just been played by black, and there is a tie (including the integer komi). White passes. Now black can pass and win, due to the "last pass wins ties"

Since Black wins in the manner, White tries something better than passing: White tenuki, Black pass (pass stone), White pass (pass stone and tie and White wins the tie as the last passing player).

Therefore Black must also improve his strategy: White tenuki, Black tenuki,... and we have a pass-fight.

***

How do you solve played out removals?

Territory counting is used, so tenuki costs a point.

Tenuki doesn't cost anything, but passing could.

[Bill Spight]

Please define "to win a tie".

If the net score is 0, a certain player wins the game.

[RobertJasiek]

I expect "the first to pass wins in case of a tie" can lead to pass-fights. However, matters are worse because an early pass could become a tedomari in anticipation of a later tie.

[tiger314]

I don't think pass fight can occur because only the first pass of the entire game matters. As for an early pass: under any area ruleset any early passing costs points, so it is better to play dame (or some other point gaining move) than to pass in anticipation of a tie.