Author |
Topic: Essay by Christian Freeling on inventing games (Read 539986 times) |
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #855 on: Jan 4th, 2012, 3:11pm » |
|
on Jan 4th, 2012, 2:41pm, MarkSteere wrote: WHY would you want to admire Mark Steere's game in MY TOPIC ?? |
| It must be love, love, love ...
|
« Last Edit: Jan 4th, 2012, 5:05pm by christianF » |
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #856 on: Jan 5th, 2012, 2:03am » |
|
The rules of Rive are extraordinarily difficult for most, otherwise intelligent people to comprehend - conceptually simple though Rive is. Rive's stone placement rule, honed by an exchange of emails with RWS: "A group here is comprised of interconnected stones of both colors. If you can place a stone in isolation, you must do so. Otherwise your stone must be placed adjacent to up to three groups, of which the largest must be as small as possible." A common misinterpretation: "So you just have to connect to the smallest group. Right?"
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #857 on: Jan 5th, 2012, 11:11am » |
|
I've put up a couple of example games, the second one against Jos Dekker, a Penalty-6 game, among them. It gives a good impression of the 'crunching phase' at the end. We've finished it for the gallery, but the endscore difference was already clear to both at move 40. The cut at F12 (white 33) was probably decisive and should have been plugged by black at the previous turn. Marcel's program lost its first game, but now I'm up against a new version that has been improved on four points. I'm very interested in its development, or in any bot development regarding simple organic mechanisms for that matter.
|
|
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #858 on: Jan 5th, 2012, 12:31pm » |
|
on Jan 4th, 2012, 11:32am, NickBentley wrote: I've got no problem with the approach. |
| It's a multifaceted marketing blitz worthy of Kris Burm. And it's a big load of shibui.
|
|
IP Logged |
|
|
|
clyring
Forum Guru
Arimaa player #6218
Gender:
Posts: 362
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #859 on: Jan 5th, 2012, 1:09pm » |
|
on Jan 5th, 2012, 11:11am, christianF wrote:The cut at F12 (white 33) was probably decisive and should have been plugged by black at the previous turn.that matter. |
| White 33 was not fatal- If black had patched the hole on move 32, it would only net black 6 moku, not the 10 needed to win.
|
« Last Edit: Jan 5th, 2012, 1:10pm by clyring » |
IP Logged |
I administer the Endless Endgame Event (EEE). Players welcome!
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #860 on: Jan 5th, 2012, 1:14pm » |
|
Shibumi's ballyhooed "depth" will never materialize due to insufficient boardspace. I.e. board size, piece count, variety, complexity, etc.
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #861 on: Jan 5th, 2012, 1:42pm » |
|
on Jan 5th, 2012, 1:09pm, clyring wrote: White 33 was not fatal- If black had patched the hole on move 32, it would only net black 6 moku, not the 10 needed to win. |
| It's a bit more complicated than that. The cutting stone cost me 5 points (1 for the stone -6 for the penalty). It cost Jos 6 points (that the connection would have given). So far I win 1 point. But that's not the main issue. 33. F12 'saved' white a liberty for the crunch: I did not have to fill in any of the internal liberties of the bottomleft group. If 32. F2-S15-M1-I2 would instead have been F2-S15-M1-F12 (it's the same group) then he would have gained 6 for the connection (while I would have gained 6 for not having to place an isolated stone anymore), but more importantly I2 would have remained vacant for crunchtime! So plugging would have given him an extra liberty and taken one from me (because I would have had to grow at move 33). That's 9 against 8 liberties instead of the 10 against 7 we were now at, at move 40. That's only counting the two biggest groups of course, the rest are plugged in the proces. That would have meant only one forced black invasion instead of two, and may have cost black the game, I'm not entirely sure about one or two points there. game
|
« Last Edit: Jan 5th, 2012, 1:47pm by christianF » |
IP Logged |
|
|
|
clyring
Forum Guru
Arimaa player #6218
Gender:
Posts: 362
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #862 on: Jan 5th, 2012, 2:05pm » |
|
on Jan 5th, 2012, 1:42pm, christianF wrote: It's a bit more complicated than that. |
| It doesn't seem complicated to me. on Jan 5th, 2012, 1:42pm, christianF wrote: The cutting stone cost me 5 points (1 for the stone -6 for the penalty). It cost Jos 6 points (that the connection would have given). So far I win 1 point. But that's not the main issue. |
| It cost Jos 7 points to your 5 because not only is there one more group black group, there is one black stone fewer there; you won 2 points locally. (But this is indeed not the main issue.) on Jan 5th, 2012, 1:42pm, christianF wrote:33. F12 'saved' white a liberty for the crunch: I did not have to fill in any of the internal liberties of the bottomleft group. If 32. F2-S15-M1-I2 would instead have been F2-S15-M1-F12 (it's the same group) then he would have gained 6 for the connection (while I would have gained 6 for not having to place an isolated stone anymore), but more importantly I2 would have remained vacant for crunchtime! So plugging would have given him an extra liberty and taken one from me (because I would have had to grow at move 33). That's 9 against 8 liberties instead of the 10 against 7 we were now at, at move 40. That's only counting the two biggest groups of course, the rest are plugged in the proces. That would have meant only one forced black invasion instead of two, and may have cost black the game, I'm not entirely sure about one or two points there. game |
| Yes, this reasoning is correct- This means one forced black invasion instead of two. Each such invasion costs black 5 points and white 1 point. Playing one fewer invasion would have netted black 4 points in addition to the 2 points gained locally for a total of 6 points that blunder cost black. Assuming subsequent play to be accurate (I didn't check), white would have won by 3 points anyway.
|
|
IP Logged |
I administer the Endless Endgame Event (EEE). Players welcome!
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #863 on: Jan 5th, 2012, 2:16pm » |
|
on Jan 5th, 2012, 2:05pm, clyring wrote:Assuming subsequent play to be accurate (I didn't check), white would have won by 3 points anyway. |
| Subsequent play was only to finish the game for the gallery, we agreed on the outcome well before move 40. So I might have won after all, if black hadn't blundered, that's good to know. I like to play, but I hate to analyse, thanks.
|
|
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #864 on: Jan 6th, 2012, 1:30pm » |
|
In all the shibui philosophy discussion, there's been no mention of shibui economics. Ironic because it's obviously the primary consideration. The next larger (yet still puny) size is 83 percent more expensive to produce. How to manage turn order advantage on a tiny board 1. Massive recycling a. 3x5 Cephalopod b. 3x3x5 Rive 2. Initial, pseudo random phase a. 9x9 Flume (about half filled at resignation) b. Typically small Dots and Boxes
|
|
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #865 on: Jan 6th, 2012, 2:33pm » |
|
Luis programmed Flume for Zillions, btw. He left out the green stone border with my blessing.
|
|
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #866 on: Jan 6th, 2012, 2:43pm » |
|
I don't have hard finitude and soft finitude and al dente finitude. I have only finitude.
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #867 on: Jan 6th, 2012, 2:49pm » |
|
on Jan 6th, 2012, 2:43pm, MarkSteere wrote: I hope it's not terminal.
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #868 on: Jan 7th, 2012, 2:50pm » |
|
Symple at mindsports now offers players a choice between boards 11x11 to 19x19. I've won my second game aginst Jos Dekker, indicating that I'm not too bad for a beginner. We've started up another one that is already well on its way. So far so good, what about Marcel's bot? Well ... I lost the second game, so I'm glad to confess it does rather better than I had expected. First of all, it cares about 'initiative'.For argument's sake, the term 'initiative' is used here, in the early opening, as having the move in a position with the same number of groups. Since white begins, he has the initiative. It means that he can start growing, with black trailing at the same number of groups, or, if black grows first, follow suit with one group more. This gives one additional growing option every turn. The balancing rule makes that the 'initiative' can be traded for a couple of points + 'influence' by either player at any moment. You grow these very points, and the resulting 2-groups have more liberties than singles, hence the increased influence. Marcel's bot now acts on this intelligently. While in the first game it immediately took a 1-point compensation for white's initiative, it now values 'initiative' much higher. And here's how: White 4: white can grow his groups and have '3 plus influence' for leaving black the initiative. Or he can place a single and keep the initiative. He does the latter (F14). Black 4: black can grow his groups and place a single, and have '3 plus influence' for leaving white the initiative. Or he can place only a single and leave white the initiative, aiming for either getting the initiative at white-5 or else the option on '4 plus influence' should white decline to grow. I decided for the former (M3-M9-D13-M13), because offering white '4 plus influence' seemed an offer he wouldn't decline. Was it? I'll never know, but white's next move showed the bot did 'know' what it was given the compensation for: it didn't grow (which would have been bad: black would get the initiative for only one grown group less) but kept the initiative, which is the point of the whole exchange. The working of this sophisticated balancing mechanism, in which the stakes increase at every move and in which timing depends on boardsize, on the height of the penalty and on the actual position, is 'understood' by the bot in as far as it acts intelligently on it. That's more than the one human who doesn't understand it can claim. To excuse him in his own words: "There can be only One". And not only does the program take the balancing mechanism in its stride, it played a strong game throughout, from my point of view. Of course we both made mistakes, and I made one or two more, and I'm a beginner and all that, but this is a strong bot, especially (but not only) in the endgame tactics! I'm very glad to have an unemotional and constant leveled opponent to test strategical and tactical ideas against, and I'm learning a lot. Next to Havannah, with the Challenge and all pending, Symple will be my game for the forseeable future.
|
|
IP Logged |
|
|
|
omar
Forum Guru
Arimaa player #2
Gender:
Posts: 1003
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #869 on: Jan 7th, 2012, 11:06pm » |
|
Any news about when the Havannah challenge match is going to be? Have any dates been set yet?
|
|
IP Logged |
|
|
|
|