11 thoughts on “Sandpit

  1. My idée fixe is that wage thieves should face jail time. The civil system is broken and the Fair Work Ombudsman is mostly useless and surprisingly incompetent. I have found their phone and on line advice service unreliable.

    It is absurd that an employee who steals $50k from the boss will get a criminal record whereas a boss who steals $50k from an employee will likely get away with it or pay the worker a fraction of what they are owed and silence them with a non-disclosure agreement.

  2. Hugo, I agree but I don’t think we will see the necessary changes under this system, namely global capitalism. Following the ideas of Jonathan Nitzan and Shimshon Bichler (Capital as Power) and the ideas of Wolfgang Streeck (How Will Capitalism End), I would see the current system as both unsustainable and unreformable. The only hope I can find is that the system must undergo the collapse and new and hopefully wiser societies (made wise by the travails of the collapse) might arise. Grim hope, I know.

  3. A General Theory of Virtual Modelling?
    (Some Conclusions from Array Games)


    Computer virtual modelling, as dynamic system simulation, presents the best hope of usefully modelling the complexity of socio-economic systems including their interactions with real physical and ecological systems. My experiments with array games strongly suggest a general theory of virtual modelling is possible. Common array games present one-, two- and three-dimensional arrays which model spatial dimensions and provide arenas for action according to game rules. Action occurs as digital time steps through successive array states. Backgammon features a 1-D array, despite the 2-D u-shaped path the checkers follow around the board. Chess features a 2-D array. Computer Real Time Strategy games (RTS) are virtually modelled and feature 3-D arrays. Movement rules implemented as algorithms govern the placing of “counters” or “men” on the array.

    The time dimension is mandatory for any game. A game requires duration and events. A succession of events (as game states in a series) must occur. In-game time in array games is digital. In RTS computer games, in-game time is simulated analogue, and effectively analogue at the level of human perception, whilst still being strictly digital at the computational level. An RTS game proceeds visually by frame rates (still frames per second) and computationally/algorithmically by a closed loop computational cycle. There is a control loop within which all input and output data must be transferred (to and from I/O devices) and also within which a full internal computational cycle must occur.

    Complex systems style computer games, as virtual models, ought to be especially useful from an experimental perspective as they can be constructed as;
    1. “Cooperative-competitive” games. – (Market systems are also cooperative-competitive).
    2. Malleable virtual models – They can be iteratively modified and retested quickly.
    3. Dynamic models (not static) – They model changes over time.
    4. Chaotic systems – Are dynamic systems highly sensitive to initial conditions.
    5. Potentially non-deterministic systems – This applies to models which integrate human competitors as agents via I/O if human brains themselves are non-deterministic.

    The best way into this topic is to look at the “laws” and “rules” of competitive array games, including RTS computer games. In all cooperative-competitive games, people come together in some manner and agree to abide by certain rules, more or less, for the sake of “fair competition”. As stated above, we should not forget that market economies show many characteristics of being “cooperative-competitive” games. Market dynamics are also time-dependent and could be modelled by calculating successive array states. Strictly speaking however, markets should never be modelled without the concomitant modelling of real economy and real world inputs and outputs. Ignoring real world laws, particularly physical and biological laws, will lead to highly distorted models.
    Cooperative Competitive Contests have Laws as well as Rules

    It is important to clarify definitions. It is standard to talk of the Laws of Rugby or the Laws of Chess. However, with proper definitions of “laws”, “rules” and “parameters”, the official guides governing the playing of such games should be called rules, as in the Rules of Rugby or the Rules of Chess. The legal laws and regulations of ownership and competition in extant socio-economic systems are also “rules” under this definition. Rules are made by humans and can be re-made otherwise.

    We need to reserve the word “Laws” for a more rigorous use. Rules are by nature arbitrary and can be changed, although perhaps not too extensively or the essential character of a game will eventually be changed. Rules set the parameters of games; the bounds and factors involved. Laws, however, describe invariable relationships. Even in a modelled system there are Laws which are surprisingly independent of rules, at least until changes to the rule set become so extensive that they change the entire character (the genre) of a game or competition. Laws, correctly derived, are unchallengeable. The use of the word “Laws” in this context is meant to imply the same rigour as is implied in the phrase the “Laws of Physics”.

    Complex RTS game system behaviour can be described by the discoverable Laws of RTS in the very same way that real world physical behaviour can be described by the discovered Laws of physics. At first glance, this might seem an untenable assertion. The physical universe is a real and objective “given” to science. It is pre-existent and its known Laws are the dependable and inviolable Laws we have so far discovered through scientific investigation and experiment. An RTS engine is a mathematically and algorithmically modelled system created by humans and presumably fully alterable at the will of the designers. It would seem to be an arbitrary system with rules only and ones which we can change at will. However, this is not the case. It still has Laws. We can change the rules and parameters of RTS in its virtual world but we cannot change its Laws.

    A subtle difficulty arises at this point. Mathematically and algorithmically modelled virtual systems as formal systems internally demonstrate axiomatic behaviours. Yet where these axiomatic internal behaviours reflect real-world Laws, I will be terming these behaviours or relationships as “Laws” inside the modelled virtual system. The reason they are termed “Laws” is because they are congruent or isomorphic (referring to the correspondence theory of truth) with real world Laws (like some of the Laws of Physics). This congruence is enforced by the unavoidable requirement to make virtual models which are pragmatically recognisable and useable by humans (specifically by human brains and/or minds). Biological laws of perception and human internal cognition modelling support this statement of enforced necessity.

    At their most fundamental level, RTS games partake of some of the essential Law-bound nature of the real physical world. There are clear reasons for this. The physical and biological Laws of the real world impose both “hard” and “elastic” constraints on the game world. In other words, the Laws of the real world push into and impose themselves on the game world or modelled world. The central example is dimensionality. We are familiar with the notion that we live in a world of four obvious dimensions, meaning apparently obvious to our senses. These are height, width and depth (the x, y and z axes of 3D space) plus time. How much of our conception of dimensionality is derived from the senses homo sapiens developed through the process of evolution in order to survive in the world at the “meso” scale at which we exist? We do not exist at the quantum or atomic scale. We do not exist at the cosmic scale. Our senses have evolved to be functional at the scale at which we exist and to detect threats and opportunities at this level. At the same time, how much do our notions of 3D space plus time owe to our enculturation and education? These more ontological questions hint at some of the complex issues behind the scene.

    A game with a geometric or array basis, like chess, need not model all the space dimensions but it must model some (two in the case of chess). A chess board models two space dimensions and is a finite two-dimensional array (8 x 8). This arbitrary array size (or other arbitrary array sizes which also could have been chosen), combined with the arbitrary movement rules of chess (or other arbitrary movement rules which also could have been chosen) generate “in-game” or internal Laws of chess which will consistently reflect some aspects of the physical Laws of the real world. It has already been mentioned that the time dimension is mandatory for a game or any dynamic model.

    An example of an in-game “Law” in chess is that a piece will “control” more squares from the centre of the board than it does from the corner of the board. This is mathematically and geometrically axiomatic, that is to say it is an axiomatic result of chess being an array game. This “control” is a movement potential dependent on the movement rules of chess and modified by obstacles imposed by other pieces on the board and the edges of the board (of the array). This Law is a direct analogue (or isomorph) of the space dimensionality of the real world where, for example, I can sweep more horizontal area with my arms while standing in the centre of the room than I can while standing in a corner of the room. To reiterate, when I speak of arbitrary rules like the movement rules of chess I will use the word “rules” un-capitalised. When I speak of in-game “Laws” I will use the word capitalised. The crucial thing about these Laws is that they are mathematically and geometrically axiomatic in-game and their direct analogues are empirically true in the real world.

    The Laws of RTS genre computer games (or other array games like chess) are of a combined mathematical and geometrical nature. If we change the rules and parameters of a game, whilst still keeping the essential and specific array nature of the game, this does not change the Laws of that game. The modern queen did not exist in pre-modern variants of chess like Indian chess (Chaturanga) or Persian chess. The mere invention of the queen (which combines the moves of rook and modern bishop) did not change the Law that a piece “controls” more squares from the centre of the board than from a corner. Changing the rules and parameters (bounds or factors) of chess, for example changing the board to a 10 x 10 array and adding two new pieces and two more pawns, still would not change the Law that a piece “controls” more squares from the centre of the board than from a corner. Such Laws are inherent to the game genre, in this case the genre of games with geometric arrays, digital time or turn based progress (discrete countable moves) and rule-governed array states.

    Thus, asserting that all array games, including RTS games as arrays of tiles or pixels, have the same reliable basic Laws is really asserting no more than the following. To use the example of RTS, RTS games are all of a structure which models the three space dimensions plus time as the arena of action and then models and places simulacra of materials, energies and units (agents as machine or human representations) within that 3D, time-dependent world. All these features can be, and indeed must be, quantified. They are governed by mathematical-geometric Laws analogous to the classical Laws of physics of the real world. These Laws are made explicitly analogous in those modern games which use Newtonian physics engines.

    Summary of Implications

    RTS computer games are mathematically, geometrically and algorithmically governed systems models of the real world and as such they have invariable relationships both internally and with regard to the real world which they model. These relationships can be described by Laws. There are Laws intrinsic to the real world and Axioms intrinsic to the virtual model world. These sets overlap (have to overlap as demonstrated above) and thus there is a subset of Laws shared by both the real and virtual (modelled) world.

    Laws describe the invariable relationships in a system. Hence Laws (correctly derived) are unchallengeable. We cannot model without concomitantly and necessarily importing some Laws from the real world into the model. We also “import” or commit some matter and energy, or at least some energy, to set up the model. The replication of real-world Laws in games and models makes them pragmatically playable or pragmatically useable. A cooperative-competitive game in particular cannot be played if it is a system which the brain finds totally unrecognisable due to a complete absence of recognisable Laws and rules. Indeed, such a system would be a non-simulacrum or non-model and it would be indecipherable nonsense to the human brain.

    Many games and many forms of art (literature, painting, cinema) model the world as a 3D time-advancing space where events (processes) occur and which is “populated” with materials, energies and agents be these agents biological entities or automata. Complex Systems Models as virtual models must embody systems of some essential Laws which are isomorphic in relation to the real-world systems possessing these Laws. This is so if they (the Complex Systems Models) are to present human brains with models of the real world which are recognisable and pragmatically useable.

    Time is Special

    We must consider the time and space scaling problems of modelling. We quickly discover that time is special in that it presents special problems for RTS modelling precisely because RTS is real time modelling. More generally, time creates a special problem for any virtual modelling, perhaps in all cases, or at least certainly when it is real time modelling. Real time virtual modelling essentially means that humans share time with and interact with the model in some way via I/O (input/output procedures). Humans then are interacting in real time with the model. The human player obviously does not share space dimensions with the virtual game. The player is in a real 3D space and the game is in a mathematically modelled virtual space. However, interactively the player inputting commands and acting on outputted information does share real time with the game. If overall game-time and/or specific tasks in-game are to proceed at a different pace to real time then game time must be distorted compared to human perception and reaction times. This last refers to the pace at which humans have experiences perceptually and cognitively. Time is special in this case as it is the one dimension humans share with the game.

    It is tricky to grasp and express this idea of “time specialness” for virtual real-time modelling. Only the space dimensions can be proportionally scaled for modelling purposes. As well as proportionally scaling space dimensions, we can un-proportionally scale or distort space dimensions in models if we wish. However, we cannot proportionally scale time when real-time virtual modelling. We can only distort time in the in-game world. Under conditions survivable by humans, the 3D space dimensions and the time dimension are pragmatically and apparently separate and non-interactive; meaning at the Newtonian physics level. This is not the case in the 4-dimensional spacetime model of Relativity but everyday human survival, perceptions and actions occur at an apparent Newtonian level.

    A tabular summary might help to explain this.
    Modelling 3D Space dimensions:
    • Proportional scaling means dimensions of 1x, 1y, 1z may become for example x/2, y/2, z/2 in a half scale model.
    • A distorted scaling example is that 1x, 1y, 1z becomes 1x/2, 1y/2, z/4.
    • Another distorted scaling example is that humans become 1x/2, 1y/2, 1z/2 but buildings become 1x/4, 1y/4, 1z/4. (This is what one might call category distortion.)

    The time dimension stands alone in Newtonian modelling. It has nothing with which to be relative under Newtonian physics assumptions, at least not in the way that space dimensions have other space dimensions with which to be relative, as in the examples above. The Newtonian model assumes absolute time just as it assumes absolute space. However, absolute space has three dimensions which can be relativized and thus be scaled one to the other in both the “Newtonian-like” real-world and in the virtual model and from the real world to the modelled world.

    Time is a solo dimension which is of a unique type under Newtonian assumptions and under closely Newtonian-like real conditions. It can be “scaled” (compressed or stretched) in the modelling process going from real world to virtual model but this will lead to different results for kinetic energy and momentum calculations. The makers of a rigorous Newtonian engine virtual model will not want to distort kinetic energy and momentum values in their model. Otherwise, a scaled human figure on scaled landscape on a computer screen will scurry like an ant; behaving as if it has the momentum effects and power to weight ratio of an ant. Thus, the only ratio of time that a rigorous Newtonian model can use from the real world to the modelled world is 1:1.

    Implications for Realtime Virtual Modelling

    If a virtual game takes 1 hour of real world time to play out then it is a truism that it occupies 1 real world hour of the player’s time. People are time constrained and need games to play out in a reasonable time span. But speeding time up in a game with a Newtonian physics engine model ruins the calculated results and modelled outcomes of kinetic energy and momentum effects. In some real-time games like car races or first-person shooters keeping time scaling at 1:1 causes no problems. In RTS, which models long term economic-military struggle it does cause modelling problems. Motion in the model may be sped up a bit but not beyond the point where things become perceptibly ridiculous and action too fast to manage with merely human reactions. Beyond that point, the modeller has to use “category distortion”. That is to say buildings are constructed in an absurdly short times and even battles occur in absurdly short times but not so absurdly short as building construction times. Speeding up economic action time compared to military action time has extensive distorting effects in the “grand strategy” sense: the marshalling of economy and military for total war. These time problems might not concern a virtual society-market model (which can be run mathematically without graphical depictions of progress) except as follows.

    Firstly, if real human agents making decisions are to be incorporated into virtual test models via I/O, then the model can only be run time-proportionally at a pace where human decisions can be made (and sent through modelled communication infrastructure delays) at a normal real-world speed. The speed that the human agents can work at in practice will determine the time advance in the virtual model rather than its internal computational capacity which will be much greater. If rational agents can be automated then this will not hold true. However, one would want to test the model in some 1:1 time runs with human agents in order to “calibrate” and test the model with both human and then automated rational agents. If automated rational agents are behaviourally only a “little bit different” from human rational agents in even simplified decision contexts then this “little bit different” is going to qualify as a small starting condition difference at each iteration of the model’s time evolution. Small initial changes like this in chaotic systems can lead to greatly different outcomes. This chaos effect will likely distort findings from the model.

    Secondly, in a comprehensive economic model which looked at not just market processes but productive or constructive and destructive or consumption processes, the real time problem becomes fully manifest in simulations. Market operations must be time-synched at 1:1 with productive and constructive processes. Thus, to fully model a new firm entrant to a system, we cannot assume that the new firm appears instantly relative to the usual cycles of market trading. The new firm has construction times, setup times, shakedown times, operations windup times and so on. These must be time-synched to market trading cycles and even to pre-emptive responses of other firms gearing up to face new competition. At least, this is a summary of the issue I see with “time specialness” in relation to comprehensive economic models.

    Footnote: This is only about half of the essential points on this topic. There is a second half which deals with the Laws and Meta-Laws of virtual systems modelling. The Meta-Laws strand follows the line of thinking that there may be Laws of Laws. I will only post this second and rather more speculative part if there is any interest here in this theorising. If the above theorising to this point is found to be simplistic or non-sensical, I leave it to J.Q. to delete it if he deems that best.

  4. Ikonoclast, the Victorian ALP says it will pass laws to jail employers for up to 10 years’ for wage theft. The Queensland ALP made industrial manslaughter a crime last year, so progress does happen.

    I joined the ALP last year so I can exert my tiny little fly speck of influence. I understand cynicism and I’ve been there most of my life, but it is also a cop out. Cynicism gives us an excuse to do nothing. Doing nothing and being cynical is what the right want the lefties to do.

  5. Hugo, I agree with your first post. There are other biases in the system. For example for employees income is defined as wages for tax purposes (leaving aside any income from financial or physical assets).. For businesses, income is defined as profit for tax purposes. But wages are like revenue for business. Why can’t employees deduct the costs of generating the revenue? In the case of Sydney, $50,000 p.a. is a reasonable estimate for the cost of living to generate the revenue. But the tax free amount is only $18200. Run such a system for decades and there will be consequences.

  6. Political commentary all too often centers too much around economics. The opposite argument to neoliberal politics is anti-neoliberal politics. This gets us just so far. We might well come at this whole thing from another angle: ethics.

    Speaking as a confirmed atheist, I nevertheless present a moral critique of modern politics of all stripes – even while acknowledging that one side is doubtlessly more unethical than the other.

    Take that seemingly intractable problem of irregular migrants/ offshore detention.

    Two common arguments around this are: racist based fear of muslims (for detention), and economic cost (against detention).

    My argument is more fundamental: is it right or wrong? Ethically it is clearly wrong. Massively wrong. End of story, because in my book ethics trumps all other arguments.

  7. Ernestine,

    As the Fair Work Ombudsman (FWO) has pointed out, small companies with one or two owners often declare themselves bankrupt once workers threaten to take them to court for unpaid wages and unpaid superannuation. If you look at the unpaid wage cases taken to court by the FWO, well over half the companies are in liquidation by the trial date. The owners often set up a new company once the trial is over. This process is called “phoenixing” and some bosses have done it half a dozen or more times. Phoenixing is supposed to be illegal but the laws are weak and very rarely enforced.

  8. Ernestine, I am interested in your opinion on my long post above about virtual modelling. Be as harsh a critic as you wish. It’s better if I know my theorising is naive, wrong, old-hat, ponderous, tendentious, vague, preposterous, risible etc. I understand that not addressing it at all is also your prerogative. 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s