Recommend
7 
 Thumb up
 Hide
133 Posts
[1]  Prev «  1 , 2 , 3 , 4 , 5  Next »  [6] | 

Abstract Games» Forums » General

Subject: Done to Depth rss

Your Tags: Add tags
Popular Tags: [View All]
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
And a related question: would Manifest Chesstiny be even better if the stones were done away with entirely, so that the turn rule became:

Quote:
On your turn, place any off-board piece onto any empty space on your back row, if possible, and then move any one of your Chess pieces on the board as normal.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nicholas Hjelmberg
Sweden
Saltsjö-Boo
flag msg tools
designer
Avatar
mbmbmbmbmb
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.

Sounds interesting. Thematically (if that really matters in an abstract game), you get to set up your army on the battle field as you wish. Gamewise, I get more time to prepare my opening/adjusting to my opponent's opening.

What you might lose with such a variant is the quick start of modern chess, an evolution that contributed to the game's popularity in Europe.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
nhjelmberg wrote:

What you might lose with such a variant is the quick start of modern chess, an evolution that contributed to the game's popularity in Europe.
My goal was to avoid this. Note you start moving your pieces from the very first turn, just like in FIDE or Fischer Random.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nicholas Hjelmberg
Sweden
Saltsjö-Boo
flag msg tools
designer
Avatar
mbmbmbmbmb
milomilo122 wrote:
nhjelmberg wrote:

What you might lose with such a variant is the quick start of modern chess, an evolution that contributed to the game's popularity in Europe.
My goal was to avoid this. Note you start moving your pieces from the very first turn, just like in FIDE or Fischer Random.

Good point! I read it but somehow didn't absorb it. Then I have no immediate arguments against the variant. It sounds like a variant that adds new paths without removing what makes chess special. (Perhaps I'll miss the strategic late castlings but they are probably redundant in this variant anyway.)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
milomilo122 wrote:
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
And a related question: would Manifest Chesstiny be even better if the stones were done away with entirely, so that the turn rule became:

Quote:
On your turn, place any off-board piece onto any empty space on your back row, if possible, and then move any one of your Chess pieces on the board as normal.
In that case the difference is that you can capture a stone but can't capture an empty square. So I like the second option better.

And it's clever, more so than Fischer's 'solution'.
Another solution to what many consider to not yet be a problem, is to test and find a new initial set up. One without the need for castling unless of course one finds it crucial for Chess to have an option to make a weird move each, once a game. It would once again become a Book with all the advantages and disadvantages.
Or one could turn to Grand Chess and have both a Book and no chance of ever finishing it from a human point of view.

There are many solutions but as yet there's also a broad concensus that there's no real problem. We're even catching up faster on climate change.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Peter Ward
msg tools
mb
Throwing this out there: what if there was a version which was sort of a hybrid of Sittuyin and Hive using Western pieces?

The board would start with the Sittuyin pawn-line (although I prefer the variant with 5 pawns on rank 3, 3 on rank 4 because it leaves the centre 2x2 square empty). Like Hive the other pieces start in reserve and you can plop them down anywhere behind your pawn line whenever you want, but the King has to be in play on or before your 4th turn.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
christianF wrote:
milomilo122 wrote:
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
And a related question: would Manifest Chesstiny be even better if the stones were done away with entirely, so that the turn rule became:

Quote:
On your turn, place any off-board piece onto any empty space on your back row, if possible, and then move any one of your Chess pieces on the board as normal.
In that case the difference is that you can capture a stone but can't capture an empty square. So I like the second option better.
I'm leaning that way too. I'll change it until I find something wrong with it.

Quote:
Another solution to what many consider to not yet be a problem, is to test and find a new initial set up. One without the need for castling unless of course one finds it crucial for Chess to have an option to make a weird move each, once a game. It would once again become a Book with all the advantages and disadvantages.
AlphaZero could be a great assistant in finding such a setup.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
Cortez527 wrote:
Throwing this out there: what if there was a version which was sort of a hybrid of Sittuyin and Hive using Western pieces?

The board would start with the Sittuyin pawn-line (although I prefer the variant with 5 pawns on rank 3, 3 on rank 4 because it leaves the centre 2x2 square empty). Like Hive the other pieces start in reserve and you can plop them down anywhere behind your pawn line whenever you want, but the King has to be in play on or before your 4th turn.
This is very similar to the game I just linked to, Manifest Chesstiny.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Peter Ward
msg tools
mb
milomilo122 wrote:
Cortez527 wrote:
Throwing this out there: what if there was a version which was sort of a hybrid of Sittuyin and Hive using Western pieces?

The board would start with the Sittuyin pawn-line (although I prefer the variant with 5 pawns on rank 3, 3 on rank 4 because it leaves the centre 2x2 square empty). Like Hive the other pieces start in reserve and you can plop them down anywhere behind your pawn line whenever you want, but the King has to be in play on or before your 4th turn.
This is very similar to the game I just linked to, Manifest Chesstiny.

Yep yep. My only difference was removing the stones, changing the starting pawn structure, and making placement a full turn. Mostly I just like the Sittuyin pawns and wish it was used in more variants
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
milomilo122 wrote:
I'm leaning that way too. I'll change it until I find something wrong with it.
I don't think there is. You may have early long range attacks on the pawn structure, but being able to move and/or cover a pawn are such that I can't on the face of it find a problem.

milomilo122 wrote:
Quote:
Another solution to what many consider to not yet be a problem, is to test and find a new initial set up. One without the need for castling unless of course one finds it crucial for Chess to have an option to make a weird move each, once a game. It would once again become a Book with all the advantages and disadvantages.
AlphaZero could be a great assistant in finding such a setup.
Good point.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
It occurs to me there are two different ways to constrain the problem of "fixing Chess":

1) make the best solution possible
2) make the best solution that Chess players might reasonably adopt

These lead to different kinds of solutions. By aiming for 1), you might end up with something like Grand Chess; by aiming for 2), you end up at more conservative solutions, like Sierawan Chess

But which kind of solution *should* one aim for, if the goal is to serve Chess players?

Thinking this way raises the possibility that solving the problem doesn't end with the invention of an improved game. The full solution is an improved game + sufficiently effective outreach/persuasion to drive adoption.


 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Corey Clark
Canada
flag msg tools
designer
mbmb
christianF wrote:

Permanency as such doesn't show up in the tree, you have to look at the game. A weird notion here is that Corey sees some things opposite of my view. I think Hex allows deep planning because it is all permanency. Othello to me is more tactical because for long stretches there's no permanency at all. It starts radiating from the corners, the explicitly permanent spots.

I will just say any designer or perennial Othello player owes it to himself to watch the videos on this channel.

https://www.youtube.com/user/Othellolessons/videos
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Richard Moxham
United Kingdom
flag msg tools
designer
Avatar
mb
christianF wrote:
I have nothing against it except its architecture.
Magnificently expressed.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmb
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
Looks at least like an interesting fun Chess variant. Have you intentionally not added it to the BGG database?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb
CoreyClark wrote:
christianF wrote:

Permanency as such doesn't show up in the tree, you have to look at the game. A weird notion here is that Corey sees some things opposite of my view. I think Hex allows deep planning because it is all permanency. Othello to me is more tactical because for long stretches there's no permanency at all. It starts radiating from the corners, the explicitly permanent spots.

I will just say any designer or perennial Othello player owes it to himself to watch the videos on this channel.

https://www.youtube.com/user/Othellolessons/videos
That's a very good introduction indeed. Note that I said 'to me'. I'll gladly admit that practice makes perfect but to me the game remains, how did I put it, an organism trapped in a poorly fitting exoskeleton.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron Browne
Netherlands
Maastricht
Limburg
flag msg tools
designer
Check out my Digital Ludeme Project: http://www.ludeme.eu
badge
Researcher in AI and automated game design.
Avatar
mbmbmbmbmb
Hi Nathan,

NJames wrote:
In order for a player to experience conceptual depth in a game, the game must submit to a series of increasingly difficult insights.
I saw a paper last year by Lantz et al. called "Depth in Strategic Games" (http://julian.togelius.com/Lantz2017Depth.pdf) that includes the following diagram. I think this summarises what you describe:


The dots indicate comprehensible strategies that players can learn.

Three games are shown. The game on the left (white dots) is too simple. There are a couple of simple strategies that players quickly learn then they have mastered the game (this would apply to Tic-Tac-Toe).

The game on the right (white dots) is too hard. Even to learn the most basic strategy players would have to play for a long time, and until then are essentially making random moves until they work out what's going on. Very unsatisfying - these types of games do not survive.

The game in the middle (black dots) is juuuuuust right. There are a number of strategies that players can accumulate linearly, starting with simple strategies that become apparent soon after playing, to more complex and powerful strategies that emerge as the player learns more about the game. This is the "minute to learn, lifetime to master" type of game.

I think this concept of the strategy ladder is a clear and powerful - and correct! - way to describe strategic depth. Though where the author has "Computational Resources" on the X axis I'd put something like "Experience" or "Mental Effort". And it suggests that perfect play is generally achievable, which I don't think is the case for most games.

NJames wrote:
We can probably identify some things as necessary for conceptual depth, and also some things that will probably signal conceptual depth, but I doubt we can ever guarantee conceptual depth.
I'm exploring this very issue in my current project (http://ludeme.eu/index.html). I believe that it is possible to automatically learn such strategies and to measure games for their strategic potential in this way.

NJames wrote:
As proof of the last point, I note that there is always the possibility of discovering an insight that breaks the game.
I think this is measurable too, although we'll never be able to prove the non-existence of winning strategies that have not been discovered.

NJames wrote:
Since it is human-attainable insight that matters, even solving a game by supercomputer wouldn't permit a guarantee.
Absolutely! I totally agree. For this reason we are aiming for human-level or "plausible AI" that plays games well rather than AlphaGo style AI that defeats all human players. That would sort of defeat the purpose, since we are trying to recreate and measure the "human" experience of playing the game. An added bonus is that this means that we don't have to spend so much time and effort developing overpowered AIs.

Regards,
Cameron
10 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
christian freeling
Netherlands
flag msg tools
designer
Avatar
mbmb

Thanks Cameron, that"s a good post. The stepping stone diagram represents my own thinking about depth and the issue wheher a game might be considered a strategy game or a tactical game.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
TPoG
Denmark
flag msg tools
ISBN: 1-85723-146-5
badge
Hmm...
Avatar
mbmbmbmbmb
Great post Cameron!
I was trying to remember the referenced paper. Nick has mentioned it earlier in several other treads (or maybe at his homepage).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nathan James
United States
Covington
Ohio
flag msg tools
designer
mbmbmb
Cameron, thank you for that contribution!

The graph is an excellent illustration of what I have in mind. I recall seeing that graph, but I had forgotten it. From what I remember, the graph doesn't show actual games, but rather illustrates the concept that some games might be too hard, others too easy, and some just right.

While a strategy game can be sabotaged by any step on the ladder being to difficult to reach, (that is, too great a distance on the horizontal axis,) I don't think the same holds for the effectiveness of the step, (too large a distance on the vertical axis,) until it is the last step and achieves optimal play. I see no reason why an initial, obvious strategy that took you halfway to perfection would be a problem, provided the rest of the journey consists of steps that are reasonably attainable (not so hard as to frustrate or remain forever a mystery), and reasonably effective (at least enough so as to be noticed).

The total length of the journey is important also, and if we are talking about a Great Game, like Chess, the journey must be very, very long. On the other hand, I can see a use for games which could be fully explored within a single lifetime or even in shorter periods. These would be attractive for those who enjoy variety, but also it might be possible to create games that have an excellent learning slope, but which are too short to stand the test of centuries.

It sounds like you and I are largely in agreement on what type of game is desirable, that is, one which provides a human experience of continued learning. While I believe this type of depth is subjective, in that it can not be understood without reference to the person or persons who experience it, I do think that it is a real quality. "Subjective yet real" is hard to understand from a certain philosophical framework, nevertheless, I think it is an accurate description.

I wonder at your plan to automate and measure strategies of games, and I will review your site with interest. At this point I am trying to discover what I can about depth by way axioms and clear logic. (You quoted and agreed with a few of those axiomatic conclusions towards the end of your post.) One of my conclusions is that depth is ultimately not measurable, although I can imagine making some progress toward algorithmic assessments. In fact, that is what I imagine all would-be abstract strategy designers are doing.

If depth is ultimately not measurable, then attempts will be estimates rather than objective measurements. This is not to say they will be fruitless. In a way, they will be strategies for producing strategies, heuristics for heuristics, which is delightful even to think about.

As some support for the idea that depth is not ultimately measurable consider this axiom:
camb wrote:
... we'll never be able to prove the non-existence of winning strategies that have not been discovered.
And imagine for a moment that the graph you posted was not three different games but one game. Having discovered the center ladder, even to have plotted it all the way to perfection, would not mean that the leftmost ladder wasn't waiting to be discovered. And if it is waiting to be discovered, in what sense would measuring the qualities of either of the other two ladders represent a true measure of the game's depth?

Before I close, let me mention that I would heartily encourage relabeling the horizontal axis in that graph. Computational resources suggests that as we move to the right we eventually could be solving the game by brute force. I think what designers and players are interested in is instead beautiful intuitions that avoid the necessity for that kind of analysis. I don't know that piling up computational resources is a good proxy there.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
russ wrote:
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
Looks at least like an interesting fun Chess variant. Have you intentionally not added it to the BGG database?
I haven't subjected it to serious testing and I'm hesitant to add anything to the database without having done so.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Florent Becker
France
Orléans
flag msg tools
Avatar
mbmbmbmbmb
milomilo122 wrote:
russ wrote:
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
Looks at least like an interesting fun Chess variant. Have you intentionally not added it to the BGG database?
I haven't subjected it to serious testing and I'm hesitant to add anything to the database without having done so.

If you kickstart it, you don't need to.
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
galbolle wrote:
milomilo122 wrote:
russ wrote:
milomilo122 wrote:

Christian (or anyone else with considered opinions on this subject):

How about Manifest Chesstiny? Is that a good or bad solution? I haven't given it much thought but it feels like it could be an improvement over Fischer Random, as a tournament game.
Looks at least like an interesting fun Chess variant. Have you intentionally not added it to the BGG database?
I haven't subjected it to serious testing and I'm hesitant to add anything to the database without having done so.
If you kickstart it, you don't need to.
ZING!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
The Player of Games wrote:
Great post Cameron!
I was trying to remember the referenced paper. Nick has mentioned it earlier in several other treads (or maybe at his homepage).
Yeah, I think every discussion of depth should start from that paper, at the moment.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nathan James
United States
Covington
Ohio
flag msg tools
designer
mbmbmb
milomilo122 wrote:

Yeah, I think every discussion of depth should start from that paper, at the moment.
I was disappointed after reading the paper because I don't think the quality they describe, d, is equivalent to depth at all. The depth we are looking for is that of layered and nuanced strategies, where new concepts build on old ones. The model proposed in the paper doesn't reference the refinement of the strategy at all, it is only concerned with how the strategy functions when given a certain amount of computational resources.

Measuring d is equivalent to asking how many different world champions would we have if we played tournaments with every conceivable time control.

The method the paper describes has the advantage of promising an objective measurement. It has two major disadvantages. The first is that no device could ever produce the measurement, since it would require assessing every possible algorithm, not just reasonable or effective ones. The second disadvantage is that the measurement isn't depth, just a somewhat relevant approximation.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
[1]  Prev «  1 , 2 , 3 , 4 , 5  Next »  [6] |