Talk:ChessInc

This is my response to the comments seen in my post about ChessInc

I rearranged some of the below discussion with Pokes, so that things are in a slightly different order

pokes: Welcome! It’s great that you’ve brought this energy here - often new players will end up not writing any proposals, and this is totally on the other end of the spectrum. Here are my thoughts reading through it:

FamilyMan: Thank you pokes!

pokes: As written, this game could easily take something like five years for us to play. Rules will usually be friendly to players checking in every day or two (twelve hours for spinning the wheel in this dynasty is unusually fast), and so things tend toward mechanics that are asynchronous. This makes chess almost a worst case scenario for a game to play here: Even if it were just two players not making moves through an LLC, we’d probably want to give them a day or two to move, making the game take a month or two, already the length of an average dynasty. But this has four games being played indirectly through subgames of nomic.

Family Man: First, note that while the corporate rule making (procedural) mechanism might be synchronous, the substantive chess mechanisms would in no way need to be synchronous. The two substantive chess mechanisms, that i can think of, would be

1. Changing your player’s quoted prices at which they would buy and sell each type of share, and how many they would buy. We barely have talked about designing the game state page that keeps track of these actions, but getting this right will be essential
2. taking a few turns a day

FamilyMan: How can the turn taking mechanism be asynchronous, you may ask? Here is one way it can be arranged to simulate the asynchronous experience: someone from the corporation volunteers to login every two nights at the same time as their counterpart (someone from the opposing side/corporation), and they continue the game for a couple of turns I envision this player making the choices recommended by the grand-master-level chess engine (the link to the chessboard in the original post has a recommendation system that can be toggled on. The only departures i envision would maybe favoring a move that captures pieces , especially if a side is losing and is going to be checkmated.

FamilyMan: What is interesting is the steady change in net present value. (more thoughts on NPV at the very bottom— scroll down)

FamilyMan: Second, note that while the procedural/ rule making within the corporation’s mini-Nomic would be synchronous, this will not be a problem. This synchronous mechanic would work, where other synchronous elements of blognomic gameplay (both procedural and substantive) may have failed for the following reasons: # I envision that the number of shareholders per corporation will usually and often be a very small number of people — much fewer then the number of players in the BlogNomic/MacroNomic/SuperNomic. All shares of some corporation will probably even be owned by only one player at some points. # The synchronous turns in a subnomic don’t need to be at a constant pace, and can be taken few and far between. Within a turn, there is no reason anyone has to vote within too small a window of time to make it not possible to start out with unanimity rulemaking procedure, as in vanilla nomic, but impossible in Blognomic. it might be someone’s turn. Even though there are no pressing corporate statutes to create, they want the ten points of having it adopted, so they propose something mundane. It might be 6 days before everyone has voted and that single turn has concluded. And that would be ok, because it is an available and useful tool of gameplay, but not a mechanism that would spend much time in the spotlight. The game could still be vibrant with activity, even if this particular mechanism is typically neglected. # For example, there might the first day be an agreement that the majority shareholder has to log in every two days at a set time (worked out with the opposing corporation) to continue the game, and to just play whatever the chess engine recommends for a couple of moves # Further refinements might soon after be made, that will make things more streamlined and/or open the floodgates of “the paradox of self-amendment” that fascinates me about Nomic. For example: # Example 1: a majority can vote for anything that demonstrably increases net present value, but only once a day # Example 2: a binding agreement for majority shareholder to buy everyone’s shares so that they can play once a day and vote for their own amendments every day, generating ten points every day, in exchange for giving them a cut

pokes: Usually initial proposals don’t have nearly enough content to be a fully playable game, but instead are broad strokes of mechanics that players can start acting upon. Having a victory condition would be unusual, because once one appears, the phase of the outer nomic game that is semi-cooperative inner-game construction ends and most of the outer game becomes maneuvering towards victory.

FamilyMan: I totally get your point. But I am intrigued by the thought of not having nomic be this sort of two-segment game you are describing: I see Nomic as a game that relies more on using rulemaking as the instrument to maneuver towards victory rather then creating the rules that we think will be the best game and then playing the game we designed by committee. I do not have your experience with Nomic (online or otherwise) so I don’t have the perspective you have acquired.. But i want to understand why what you are describing is better, rather than just taking your word for it. Can you point out any dynasties that were not the way you described and show why they were lame?

pokes: I didn't make a value judgement about either kind of system, just noting what is common in BlogNomic specifically.

FamilyMan: I think also there is an implication that the more developed the initial prompt, the less opportunity to develop the game. But it isn’t obvious to me why that would be. Can you show any examples of that being true?

pokes: By analogy, a machine with more moving parts is harder to modify in-place. But also, tweaking a game is fun, but players also like to be part of the building of the game.

pokes: It’s also a lot to take in. It would take a lot of effort to figure out how to start playing this game, and frankly if I were thrown into it as a whole game instead of figuring it out piece-by-piece as dynasties here go, I would probably idle out.

FamilyMan: This is a good point. The first thing that comes to mind to tighten it up in terms of length is: How about if i omitted the modified sub-nomic ruleset from the dynastic rules and plan to introduce a vote once the game is underway— Would you advise that sort of thing? Would it still be too much to take in? My thinking, when including the details of the subnomic, was: every Nomic player has probably read the vanilla 1982 ruleset and if there are only like five changes to that original ruleset (indeed, some of them were just removing rules), it is not that much to actually take in. But perhaps i could just show the rules that are different, and then give a link to the 1982 ruleset, rather then copying the text for most of the rules verbatim.

pokes: I've read the 1982 ruleset, but have almost no experience playing it.

FamilyMan: Perhaps, also, i could propose the payout system in a votable post soon after the inaugural address, rather then in the inaugural address. Would you advise that sort of thing?

pokes: Yes. It's worth noting that all of the rules have to be added by proposal. The Ascension Address itself can't add dynastic rules to start out with.

pokes: Please, throw some proposals our way! You’ll get a lot more  s though if it’s in much smaller chunks.

FamilyMan: I already wrote the huge chunks above. whoops. Read at your own leisure— no rush

Kevan: HE/HIM 07-12-2020 10:17:35 UTC Kevan: Good to see some enthusiasm. A chess dynasty would be fun to try - there’s no reason why the game described here would take years, when it’s an amendable game of Nomic. We could agree for individual chess games to be played out between two particular human avatars, in real time.

FamilyMan: Thanks for the support. I agree it wouldn’t take five years, however my vision is a little different. If by “in real time” you mean each game takes half an hour, that leaves no opportunities to buy and sell as conditions slowly change. As i have mentioned, What is interesting is the steady change in net present value

Kevan: And as Nomic players I’m sure we’d want to tinker with the rules of chess and the types of pieces that were used, if nothing else because not all of us will be great chess players.

FamilyMan: Good idea. Lichess (the website i linked that would be hosting these games) has six supported variants. i am open to trying any of them. I have worked with the lichess code, but i don’t think i have time to implement and host a lichess server with a new variant . But i am open to additional variants, if someone can find a way to host the games online. I think standard chess would be best, but i am open to variants.

Josh HE/HIM 07-12-2020 11:02:39 UTC josh: Looks fun. A couple of issues but we’d expect those to be ironed out in real play. Family Man: Thanks! Josh: Now all you have to do is win a dynasty and you’ll be free to implement it :) FamilyMan: Gee thanks (lol)

FamilyMan: I mentioned before that What is interesting is the steady change in net present value. Here are a few examples of how NPVs might change: At first, shares of the second, third, and fourth corporation are valued low, lower, and lowest, because those games might not even happen! But as it becomes clear that someone will win by checkmate in, say, the third game, shares of the fourth game will be huge, and there will be a huge payout. (If you missed it in the original post, the condition of the fourth board being played is a decisive win by checkmate in the third game) Once one side (say it is white in the first game) has a clear way to force checkmate, they might make a deal with the opposing side so that besides winning by checkmate, they also take every single piece on the opponent’s side (in other words, they make a deal for the other side to lose dramatically) : owners of shares of white then get a huge payout and shareholders in the other side are left with nothing but debt, to which they declare “Our liability is limited, we don’t owe anyone anything, this corporation is now liquidated” Once this deal above takes place, the NPV of the chess board ownership corporation becomes merely what is left on the board in white’s army

FamilyMan: This is a follow up comment -- I am saying this a day or so since writing my comments above -- I have made the first three chessboards human vs computer, so that should quell any concerns about the gameplay being too staggered and taking years, once and for all!

More pokes commentary

If both sides are going to be run by a chess engine, but what the chess engine is trying to do is run by an LLC, it seems like the majority of the dynamics of chess boil down to the cases of

• both sides trying to win -> probably results in a stalemate
• one side trying to win; one side disorganized, or trying to lose -> blowout for the winning side
• both sides trying to lose, or delay the game -> game is delayed indefinitely

As such, it might be possible to capture most of the outer game by replace the chess games with a simplified model; perhaps instead it could be something like 5 rounds of each side submitting whether they're trying to win or lose, and scoring based on the end result of those rounds. This would be a more reasonable length, and be more transparent and modifiable in the outer game.

Some other more specific things:

• How are the decisions of the LLC coupled to the games of chess? Where are the states of the games stored, and how can they canonically be updated?
• What does "cumulative dividends" mean?

FamilyMan's response to 'More pokes commentary'

Pokes proposed: As such, it might be possible to capture most of the outer game by replace the chess games with a simplified model; perhaps instead it could be something like 5 rounds of each side submitting whether they're trying to win or lose, and scoring based on the end result of those rounds. This would be a more reasonable length, and be more transparent and modifiable in the outer game.

Answer: How about we add one little rule to the chess game so that the computer is not able to always understand the best move? No chess engine in the world would be able to give advice if we self-enforced the following rule (i made up this variation a while ago, it works quite nicely) :

1. let m = turn number

1. If m < 25 and m % 4 is 1 and the previous turn did not entail a capture by you or by opponent:
2. then:
3. do not move the same piece as in your last ply, except if there is no other way to get out of check.
4. else If m < 50 and If m % 4 is 2 and the previous two turns did not entail a capture by you or by opponent:
5. then:
6. do not move the same piece as in your last ply or the ply before that, except if there is no other way to get out of check.
7. else:
8. All moves that would be legal in standard chess are legal.

Pokes asked: How are the decisions of the LLC coupled to the games of chess? Answers For each game of chess, there are two sides: Black and White. For each game, each side is represented with a distinct corporate personality. As owners of that "side", they get to decide their own side's moves. Unless they are the black side in the first three games, in which case they dont get to decide the actions of their corporations.

pokes: But, how would I know that the RW LLC is authorized to make the move that it did? Is it only through making a rule in their subnomic? As an aside, it's an issue that the subnomic can apparently freely alter the outer gamestate (e.g. "players who vote against winning proposals shall receive ¤10 each"). If I had sufficient control over a subnomic, it would be easy to pass a subnomic rule stating that I've achieved victory in the outer dynasty. Or collude with another player to use the 10 reward for voting against a proposal to grind out infinite money.

Pokes Asked: Where are the states of the games stored, and how can they canonically be updated? Answer : If you view the ChessInc Intro, there is a sentence: In the middle of the lobby are four chess boards IF you click the link above, you will see how all changes to game state will be taken care of. Players will need to make lichess accounts corresponding to their blognomic names.

pokes: I think most people would want the state to be tracked on a wiki page that can be edited more flexibly. (say a BlogNomic proposal passes that moves a piece out-of-turn, or illegally. Then we can't track our state anymore on lichess.)

Pokes Asked: What does "cumulative dividends" mean? Answer: When you capture a piece, that is a dividend. When you mate your opponent, that is a dividend. As stated currently, all dividends are awarded to shareholders once net payout per share is determined (ie after the end of that game of chess. Thus, no "dividends as you go", but accumulated and paid out after mate.

pokes: I don't understand the difference then between this, and omitting dividends entirely and saying that the LLC shares each pay out as gold coins at the end of the game.