Life In 19x19

Would anyone be interested in the idea of a L19 Hahn system tournament?

I've long been an advocate of the point values not mattering, the result being the only thing, but I'm being steadily convinced otherwise by watching the success of tournaments like the Go to Innovation tournament in Germany, which has had some top players playing some really interesting games.

The idea of having points actually making a difference does mean that the game rarely reaches a point where playing sub-optimally is worth it, as every 10 points on the board makes a difference.

Some fun games involving seriously strong players (including Hwang In-Seong losing by only 9.5 points to a 3d, despite giving 9 handicap stones!) in this format are as follows:

Hwang In-Seong (7d) vs Choi Young-Sik (3d) - http://eidogo.com/#2JU3QMzS - B+9.5 with 9 handi stones
Cho Seok-Bin (7d) vs Harald Lange (2d) - http://eidogo.com/#2ulj0CtE - W+26.5 with 5 handi stones

What is this? Bangneki without money? In my opinion it turns go into a psychological game like Poker and it discourages the elegant behavior of resigning when the game is close but you don't think you can catch up.

I think it would be interesting. I might be willing to play.

gowan wrote:What is this? Bangneki without money? In my opinion it turns go into a psychological game like Poker and it discourages the elegant behavior of resigning when the game is close but you don't think you can catch up.

I agree on the discouragement of resignation thing. We could have a rule where if the players agree that {x} is about 15 points ahead it can count as a result of that size at that point if the players agree in the chat?

I don't understand why you think it will make the game psychological. It's already pretty psychological isn't it? (just look at magicwand's games )

I might be interested. Never tried it before.

Although an interesting idea, I don't really like stepped nature of the system. It can be extremely important to fight the last ko to differentiate between a 19.5 or 20.5 point loss, but it can also be completely irrelevant to give away three stones in the endgame and change the score from 11.5 to 17.5. If you're going to play for the points, I think any one of them should count, no matter how little.

How about

• Use a whole value komi (say 7).
• For a score difference of X:
  
  • the winner receives 5 + 2log(X+1) points
  • the loser receives 5 - 2log(X+1) points
• any loss by more than 30 points or by resignation is counted as X = 31

That means that:

• getting a tie scores 5 - 5
• win by 1 point scores 6 - 4
• win by 2 points scores 6.585 - 3.415
• win by 3 points scores 7 - 3
• ...
• win by 31+ or by resign is 10-0.

Any point on the board (up to 31) counts towards your score, but the effect gets increasingly smaller.

HermanHiddema wrote:Although an interesting idea, I don't really like stepped nature of the system. It can be extremely important to fight the last ko to differentiate between a 19.5 or 20.5 point loss, but it can also be completely irrelevant to give away three stones in the endgame and change the score from 11.5 to 17.5. If you're going to play for the points, I think any one of them should count, no matter how little.

How about

• Use a whole value komi (say 7).
• For a score difference of X:
  
  • the winner receives 5 + 2log(X+1) points
  • the loser receives 5 - 2log(X+1) points
• any loss by more than 30 points or by resignation is counted as X = 31

That means that:

• getting a tie scores 5 - 5
• win by 1 point scores 6 - 4
• win by 2 points scores 6.585 - 3.415
• win by 3 points scores 7 - 3
• ...
• win by 31+ or by resign is 10-0.

Any point on the board (up to 31) counts towards your score, but the effect gets increasingly smaller.

That's an idea - of course, the han system seems to move 10 points for every 10 points on the board - so it would be easier I guess just to say 6 points komi (or 0 with handicap), and even is 50-50, and 1 point per point - so winning by 2 is 52-48 etc?

topazg wrote:That's an idea - of course, the han system seems to move 10 points for every 10 points on the board - so it would be easier I guess just to say 6 points komi (or 0 with handicap), and even is 50-50, and 1 point per point - so winning by 2 is 52-48 etc?

Yeah, that's a lot easier.

I like a logarithmic scale because it makes the difference between winning and losing more important, and also makes close games more exciting (points won/lost in a close game are more significant). Going from -1 to +1 gains you 2 points (6-4 instead of 4-6), while going from -1 to -3 loses you only 1 point (3-7 instead of 4-6). It means that taking risks that might improve your score but might also reduce it are generally worthwhile

Just a personal preference though

Maybe it would be a good idea to create a poll.

LovroKlc wrote:Maybe it would be a good idea to create a poll.

Yeah, I think so - we'll get some interest up first. currently we have me, Kirby, Daniel, and possibly Herman?

What level of players are you looking for?

Anyone and everyone. If we get a big enough interest, I'll come up with a tournament structure

I'd be interested in playing especially if there is some more interest from the lower SDK range of players.

Possibly interested, but it would depend where it was played. Cool idea though.

Make it a Tygem (English) client tournament maybe. I think one could get some sponsoring from the staff right now as they want to promote Tygem to the west. Prizes like Tygem-betting points or teaching games.