Dude, you’re the one who brought it up in the first place! I’m just pointing out why your analogy is faulty.
It can be described and accounted for, but not TOTALLY. No matter how precisely the rules specify the properties of a legal baseball diamond, it will always possess other properties that may potentially affect gameplay but remain unspecified. This is not the case with chessboards.
It is the case with satellites, however. Even if you use formal methods to prove that your simulation is complete and consistent, that doesn’t PROVE that the real-world satellite will behave in exactly the same manner. Because your simulation is based upon ideal components that are abstract representations of the real components in the real satellite. These ideal components are necessarily simplifications (otherwise it wouldn’t be a simulation, would it?). If you’ve done your job right the ideal matches the real in virtually all operational cases. But the map is not the territory, you know?
The rules of baseball are based upon the idea of an ideal ball that has certain properties. The behavior of this ideal ball matches the behavior of real balls in virtually all operational cases. But not all. And no matter how thoroughly you attempt to define the properties of the ideal ball, there will always be some elements of the real ball that are simplified away and have the potential to affect play.
Why the quotes around “game designer”?
I’m aware of what software verification is. (I also have a Masters in Computer Science, BTW.) But, no, I don’t do it for a living. Because it’s pretty much useless as a methodology for dealing with the sorts of rules systems that I work with on a day to day basis.
Yes, absolutely. There’s nothing magic about baseball.
Really? I don’t know, but I would guess that a sanctioning body that compiles a set of chess rules will describe it in terms of squares, not as a data structure.
Sure, we could play online and treat it as a data structure, but it is only conceptual. The game itself, to the extent it has written rules over the centuries, was surely physical in nature, no?
And that is relevant to the OP how? Not sure of your point.
I wonder, honestly, if your work shades your understanding of a game that happens in a continuous euclidean space instead of a discrete data structure?
Honestly, I wonder that. Is it possible?
Too lazy to do this I guess: game designer
So you design games that others implement, and they go out the door with no QA to check that your game design and other levels of implementation between the application and the hardware are all sufficiently connected to work well? Must be great to have a team that can write code that doesn’t need testing!
Then I am really wondering about if you think a game involving continuous elements can ever be self contained in the way the OP asks? How about roulette in your opinion? Do the rules there not cover all cases? Are there some cases that are covered by contradictory cases?
I will probably see about drumming up some help. If you are interested, let me know, and I will contact you with how I want to do it. I have a few ideas about how to divide the labor.
All that matters is that it be an 8x8 grid. It doesn’t matter if the board is made of wood, marble, electrons, or neurons, as long as it represents an 8x8 grid it is *functionally identical *to all other chess boards.
No. Expert chess players have always been able to play without a board. The gameplay in chess is completely decoupled from its physical representation. Chess on a computer is identical to chess on a physical board. It is exactly the same game.
The ball described in the rules is the map. The ball that players actually throw around is the territory. A formal proof about the rules for baseball is not a formal proof about the game of baseball.
Certainly. If you were talking about a baseball videogame, I’d absolutely agree that formal verification might be possible … although still largely useless.
Ha, no, of course we test. Not very rigorously though. It’s just not worth it. We don’t attempt any sort of formal verification.
A useful question is: Is a simulation of the game functionally identical to the game? Or does the simulation change the game by modifying or simplifying some element?
I don’t know enough about the rules of roulette to be sure, but I suspect that as long as you had access to a true random number source (not pseudorandom numbers) you could create a simulation that was functionally identical to the original. That suggests that roulette is a purely abstract game and could be analyzed by the methods described in the OP.
I understand the functional equivalence. But what do the rules of the sanctioning body/bodies say? do they specify a data structure or a board that might be made out of any materials?
You just said it was functionally identical and now you saw it is exactly the same. Those are related but not identical terms to me. What do they mean to you or the sanctioning bodies?
Maybe but if you skip back to the OP, you will see it was very clearly about the rules. It wasn’t me that took it into the realm of the operational game (if you will)
How is a videogame of basketball continuous? it is digital, not analog equipment, right? So at best it is a digital representation/approximation of a continuous realworld phenomena. OK, you can do it to whatever precision in time or calculation that hardware allows, but it is still digital and hence not continuous. Right? I am not a big gamer so I am not up to date on the latest hardware, but I would be surprised if you tell me there is a analog/continuous basketball game in common use that you are using as an example. Am I that out of date?
That sounds like a marketing decision more than a technical one. Can you share what games we are talking about? I can imagine lots of games I have seen do not need much testing because they are not really all that complex under the covers. Others, not so much.
Useful in some contexts, but not in others. I maintain it is not helpful in the context of the OP.
Hmm, here I thought you would tell me about the physics of the ball and the wheel and the bearings and the tray (I don’t know what it is called) and how in some weird case the ball, physics being what it is, could balance on one of the slot dividers, and hence you could never make consistent and complete rules for it.
Because that is essentially the argument that you (or someone, I am tired and forget who) made regarding baseball physics - I can’t make complete and consistent rules because there are too many mechanical physical things happening.
And similarly for tennis, but not for roulette. Hmm. I wonder where there dividing line is? What is the difference between tennis and roulette in this regard to you? I don’t see one myself - a ball travels around a well defined field, and physical things happen until the ball is in a certain position with respect to the boundaries, and then scoring is assessed. So what is the difference to you at that level of abstraction?
Here’s a link to the FIDE rules for chess. All they say is the board must consist of 8x8 squares of alternating black and white.
Functionally identical means identical in all ways related to the strategy and tactics of them game. A fancy board may be aesthetically pleasing, but it exerts no influence on the outcome of the game.
But analyzing the rules in isolation is purely an academic exercise. It doesn’t guarantee that the rules will cover all situations that arise in actual baseball games. So why bother doing it?
I’m not quite sure where you’re headed with your distinction between digital and continuous. Yes, of course a baseball videogame is digital.
Most testing is focused on crash bugs and play balance. There usually is no specific effort by the testing team to insure that the rules of the game are complete because formal correctness has little connection to fun. We’d rather deliver an entertaining experience with some small gaps and discontinuities, than an airtight set of rules that’s dull to interact with. It’s a different goal than in most software development where the focus is on creating a spec and acheiving it. A piece of game software can meet its original spec in every way and still be a total failure.
But the wheel in roulette is functionally just a random number generator. It can be replaced by any other fair mechanism for generating random numbers without changing the play of the game.
It’s a continuum. Purely abstract games like chess lie at one end. Sports like baseball lie at the other. In between you have games with great and greater reliance on the specific physical behavior of their components.
For example, Monopoly depends *slightly *upon the physical behavior of the dice. But you can replace the dice with a radically different means of generating random numbers – drawing slips of paper out of a cup, for example – without changing the gameplay in any meaningful way. You can’t say the same for tennis.
So while it is possible to represent some aspects of the game in software,in your head, if you don’t have squares, you are playing a variation of the rules, nt the sanctioned rules. Fine, that’s what I thought it would be. If your video game shows the squares, that is fine. If the game is between your mind and mine only, then it is not the same.
So you get to decide which part of the rules to ignore?
Are there already situations that arise that are not covered by the rules? We have heard of a few in this thread, it seems they are one-offs mostly, which is to be expected.
As I said, more interesting to me is the search for cases where the same situation might be covered by contradictory rules. I gave one such possible example.
Why do it? Why not? I don’t understand your objection - do you feel that the result would be a game different from the one you know and love now?
well, I asked about basketball, of course the answer is the same. I wanted to take it a step away from any emotions attached to the OP, but whatever.
The issue is that you and others have posited that the rules of bball can’t be complete because of the real world physics, which are necessarily continuous in nature.
But then t appears you say that a “functionally equivalent” computerized version can be made.
And that seems to me you ant to have it both ways - for you, you can define physics as a discrete simulation and for me, not so much.
That makes me wonder if we are even talking about the same thing. Of course you can make a video game of baseball or basketball that people will enjoy. You can even tweak the physics of the alternate world and that could be fun too. Such games exist for both sports.
But that has nothing to do with the OP, and my point is that the claim that because there are continuous metrics in measuring elements of the game (ball size, field levelness, heck, even 90 feet to the base), that is no obstruction whatsoever to the rules being internally consistent and complete.
I agree with all of that within the realm of video games.
The difference between games for fun, and MLB, is that MLB is a big business, and there are real risks to same if issues should arise. Simply doing an analysis to identify if potential issues of a certain type exist so one can decide what to do about it is prudent. I am sure MLB does that routinely in many areas of its business. My OP simply asked if it is known if they (or similar sanctioning bodies around the world) have ever looked at this one area.
But it’s not. You are assuming the distribution of the wheel, and you don’t know that really. sure, you could make a good guess, but it is not the same thing.
And see above about discrete simulations, physics, and the relationship to the OP.
What is the difference between roulette and tennis then? In both cases play is activated by human kinematics, and scoring relies upon the resulting location of the ball with respect to the rules-defined physical layout of the playing surface.
what on earth does this have to do with whether or not the rules can be specified completely, and in a non-contradictory manner?
Anyway, if you don’t like the roulette example, then how about either of these: 100 yard dash, and high jump. Both require the interaction of humans on a physical playing surface defined by rules, and scoring results from the rules defined degree of success compared to the competitors. Yet clearly there is complex motion with up to 6 degrees of freedom involved in scoring the results.
Are these simple games not able to have complete and internally consistent rules simply because they exist in a physical space with real world objects that humans attempt to navigate kinematically?
Uh, no. The representation of the board doesn’t matter. An imaginary board works exactly the same way as a real board.
Nothing in the rules says the board has to be fancy. I’m saying that a player might enjoy playing on a fancy board for aesthetic reasons, but the look of the board has no influence on the gameplay.
You’re assuming I love baseball … .
I assure you, I this isn’t an emotional discussion for me. I enjoy analyzing rule sets and the way that rule sets intersect with player experience and physical reality. You’re posing some interesting epistomological questions.
I’m not saying that at all. A computerized version of baseball is NOT functionally equivalent to a game played on a physical field. This is an important difference between baseball and chess.
A roulette wheel is merely a random number generator. The particular way it operates has no connection to its function within the rules of roulette. You could eliminate the wheel entirely and play the game by drawing slips of paper from a hat. The formal structure of roulette would be unchanged.
The function of the tennis ball is quite different. You can’t remove or replace it without changing gameplay in fundamental ways. The physical behavior of the roulette wheel is incidental to the game of roulette, while the physical behavior of a tennis ball is essential to the game of tennis.
Yes.
Here, look at the high school high jump rules, specifically the note to #6: “… if the landing pad hits the standard and causes the bar to be dislodged, the jump shall be ruled a fair jump.” Clearly the rules recognize that even if the bar is dislodged, the athlete might have had a fair jump. What if there was a freak gust of wind? An earthquake? A wayward pigeon? An explosion? A meteor? It’s easy to imagine other, more bizarre circumstances where the judge might rule that the jump was fair, even though the bar was dislodged. Even in something as simple as track and field sometimes the rules just have to say “Who knows? Let the ump decide.”
A functionally equivalent computerized version cannot be made of baseball, and I don’t think anyone has claimed that it can. At most, you could create a computerized baseball game that was very functionally similar to real, physical baseball, but there will always be things that can happen in a real baseball game that can’t happen in a computerized baseball game. A functionally equivalent computerized version can, however, be made of chess.
Put it this way: If I sit Gary Kasparov down in front of any computer chess game, from a 1970s mainframe to a modern supercomputer, he will play a very good game on it. But if I sit Barry Bonds down in front of the latest XBox baseball game, he’ll probably be lousy at it. Why is that?
Sure, the sanctioning body could say that in the rules if they want to, but apparently they don’t. they specifically (according to your quote) mention “squares”. Ask any 5 year old what a square is, if it got lost during your education in abstraction
I don’t deny that you get pretty much the same cognitive experience playing the game that way, it is an excellent practice method, a fine exhibition when done well, and an excellent ways to research game strategy.
But nothing stops the sanctioning body from making the equivalence part of the rules, and it appears they have chosen to not do so despite being well aware of the alternative representation.
Yes, the board does not have to be fancy. It just has to have sqaures. Not abstract representations of squares, just squares. When we pay computer chess, we agree to a conceit regarding the sanctioned rules. Not major in this case, but a conceit nonetheless.
You are assuming a binary definition of love not entered into evidence sir
yes, it is getting a bit more off topic, but at least it is a evolving to a coherent discussion. Not sure what its relatinship to the OP is, which means others who might participate won’t find it here, but oh well…
That misses the point. You can not know the actual distribution of the physical device, whatever it happens to be. You could, via statistical testing, estimate it, but you can’t KNOW it. In order to implement it, even with a discrete RNG (random number generator), you have to make assumptions that are not present in the real world.
And by the way, the same is true for a computerized chess PLAYER. It is only a simulation of a real human being, and a pretty poor one at that when it comes to all a player can do, the errors it might make due to its human-ness, the social interaction with the player, the body language that influences an opponent, etc. Which is in part what seems to be your claim that baseball can’t be the same experience in a computerized version as in the real world.
So, while computers are good in chess for testing strategies, they should not be mistaken for anything other than a limited simulation of human players on a physical board. It is no different than baseball in this regard, other than the number of players involved.
do you really believe this, or are you just trying to support the no longer defensible? I will give you a nice way out if you want.
why do you say in one game, you can change the equipment and it is still the same game, but in the other, you change the equipment and it is a different game? I would play roulette with a tennis ball and I could play tennis with a roulette ball, and if the rules allowed the variation, it is still the original game.
Are you really sure you are not arguing that you get to pick and choose variations to support your arguments, but variations are verboten for the games I suggest?
You are really stretching.
For one thing, no one has proposed that a RULE that says let the ump decide is not a valid rule. It already exists in baseball as someone quoted upthread.
In principle, that could cover closing the set. As I said, I am more interested in conflicting rules in baseball, and jtgain provides us with some candidate examples other than the one I mentioned.
But for the discussion we are having now, let me restate my understanding in somewhat more technical terms.
Let’s stop focusing on computer-chess vs. real chess. Instead, let’s agree that chess rules are closed and consistent and not in conflict with each other. Agreed, right?
Same for checkers, tic-tac-toe.
I place these games in the category of discrete games, with a finite game space. Maybe large space, as in chess, but still finite.
So we agree that at least some finite games can have complete and consistent spaces.
I say some, because some might not. A crossword puzzle for instance might have multiple valid solutions if not carefully constructed. Generally it doesn’t, but it is certainly possible to construct one that does.
Sudoku is another example - in the published examples, there seems to be a unique solution. But I am not sure exactly how those are constructed, and if removing some of the “anchor” spots you are provided would still leave a unique solution or no solution at all. Well, I have a somewhat better understading than that, but I think you get my point.
OTOH, in games where human kinematics are involved, or devices that operate in 3 or more degrees of freedom (plus time!) are involved, the game space is necessarily not discrete, but continuous.
It appears to me that some in this thread might argue that ANY game space that is continuous can never have rules that are complete and consistent. They might even argue that you can’t have either one alone, let alone in combination, I am not sure of that.
I say balderdash to that. I have given many examples, I can give more. The argument against me seems to be that in a physical game, acts of god such as meteors might occur and rules can not anticipate that.
As though a meteor can not hit a chess board as well as it can land on a baseball field :dubious:
These events are simply edge cases, and do not require enumeration, only identification of need. We already agreed (right?) that chess is closed and consistent, yet our board can be knocked over by meteor or wind and if we don;'t have a record, we can’t continue. Similarly if our clock falls off the table and is shattered. We might know the remaining time, but (I don’t realy know if this is in the rules or not) we can’t simply postpone the game until we get a new clock because that gives us more time to analyze then the rules allowed at the beginning of the game. I highly doubt the rules allow changing the rules in the middle of the game, if they do, I retract my agreement that they are complete and consistent.
Finally, here is an example of a game that exists in a physical space, has one player, does not rely on random functionality, and keeps score itself according to rules configured (electronically or mechanically depending on the implementation). Physically, it is self contained. I introduce you to the old fashioned pinball machine.
As a non-discrete, continuous game, can the rules of an individual pinball game be both complete and consistent in the same way that chess is (knowing that both are subject to the same acts of god and that the rules can account for that)?
Why? because the baseball game is not a very good simulation I suppose. But that doesn’t mean a better simulation can’t be created. Pilots can walk back and forth between modern planes and their training simulators without feeling awkward for example. But that takes a bit more than an XBox
I am not sure what you mean when you say things can happen in a real game that can not happen in a simulation? Like what? Feeling wind? If that is important, I can build air control into a simulation. The haptic experience of sliding into second base and getting a strawberry? I could do that in principle. Maybe not today, but no reason it can’t be developed. The feel of the crack of the bat? the snap of a crisp throw in a glove? All possible given sufficient research.
Not that any of those examples have anything to do with the rules. So is there an example of something in the rules that can never be accounted for in a simulation?
Because in some games it makes a difference in how the game is played and in other games it doesn’t. It doesn’t matter if my pawns are made of wood or metal … chess is exactly the same game. But if you can’t play tennis with a wooden tennis ball.
The function of the wheel in roulette is to generate random numbers. You can replace it with another method of generating random numbers and as long as the distribution is identical, it doesn’t make any difference in how the game is played. You can’t make the same substitution with a tennis ball.
Yup. In fact, once you introduce that rule you don’t need any others.
You’re fine to limit the discussion to “Are there any contradictory rules in baseball?” But that’s rather more limited in scope than the OP.
So the experience of flying a training simulator is *identical *to the experience of flying a real plane? Not pretty close, but identical? Every possible failure mode of a modern jetliner is coded into simulators? Even failure modes that haven’t happened yet?
Rules are constraints, correct? They’re limitations on player action.
Are you willing to admit that the codified rules of baseball are only *approximations *of the actual constraints that apply on the ballfield? The rules assume that things like balls, bases, bats, etc. have certain fixed properties. But because the actual equipment are real things, not abstractions, they may have properties that differ in subtle but significant ways from the idealized versions that exist in the rules.
