Predictability: good or bad?

[DrSpike]

So there are a lot of things in civ which should be random, but, due to the sterling work of spirited individuals, it is clear they aren't.

The question: Is this a good thing, since you have to really know the game to be good, or does knowing the position of all the huts and specials by 3500 or even earlier remove the pleasure of discovery for you?

[rah]

Based on the know patterns, the ability to predict where all the huts was always available. It's just with this new method, you need only find one vs 2-3-4 the old way. So it just happens a few turns earlier than it used to.

[ramses II.]

Using his own brain is a fair method compared with AI cheating ...

[DrSpike]

Rah:

My point is not whether it is worse given very recent information, but whether or not the predictability we have understood for a few years now is desirable. I mentioned huts and specials just as an example. You didn't answer the question I posed.

[samson]

When predictability is low, you have a game of chance; when predictability is high, you have a game of management.

The aim of strategy is to increase your opponent's predictability and reduce your own. Strategy must always account for randomness, but the best way of doing this is to reduce its occurrence and thus its impact on your plans.

In Civ, you play against the game as well as opponents. When the game (via random or poorly understood events) dictates your play style, you become predictable and vulnerable. When your knowledge of the game allows you to reduce this randomness, the opposite happens.

[DrSpike]

I respectfully disagree. For instance all the skill required to play civ well is still required if you remove the 'predictable' hut/special placement, oedo years, how the maps are created etc.......it hardly becomes a game of chance in these circumstances.

There are of course things that should be known to maximise propensity for skill, but the things above are not in that category IMO.

[rah]

I agree that skill is still required. But the predictability is a great advantage if the other player is not aware of the predictability.

Alas, I wish there was no hut pattern, no special pattern, no oedo year, etc, because I really like the exploring part of the early game and some of these diminish the degree of discovery. But it's a game, and they're part of it, so to be competitive it is important that we learn what we can, because if we don't, someone else will.

Rich

[solo]

Predictability is good, as is any increase in knowledge.

Individual players still have the choice whether or not they want to use this knowledge in their games, so it comes down to a matter of playing preference.

Highly competitive players are always seeking a strategic edge over opponents, and will use predictability to increase their advantage.

Casual players, who do not care to much about "winning", probably enjoy the surprise of bumping into a hut they didn't know was there, or being paid a huge commodity bonus they had no idea was coming to them.

Some people, including myself, like to play both ways, depending on their mood. In Civ II, you have the choice, and this is what makes it such a great game.

[DrSpike]

Originally posted by rah


Alas, I wish there was no hut pattern, no special pattern, no oedo year, etc, because I really like the exploring part of the early game and some of these diminish the degree of discovery

Rich

Now we're cooking. Of course in reality you have to play the game as it is (in MP anyway, in SP you have more flexibility), but the quoted remark is what I was after.

I didn't want to tip my hand immediately but I agree with this statement entirely. And my reply to Samson adds a further element, in that it is only the sorts of predictability (in factors that should be random) that is due to laziness on the part of the programmers that I object to.

[samson]

Originally posted by DrSpike
I respectfully disagree. For instance all the skill required to play civ well is still required if you remove the 'predictable' hut/special placement, oedo years, how the maps are created etc.......it hardly becomes a game of chance in these circumstances.

There are of course things that should be known to maximise propensity for skill, but the things above are not in that category IMO.

What "should be" predictable and what shouldn't is a matter of opinion. What is predictable is a matter of fact. The location of huts and specials, revolution years, even the supply and demand of cities, are predictable once the factors governing them are known.

When a person first starts playing civ, cities going into disorder seems to be a random/unpredictable event. As he acquires skill in the game, he finds that it isn't. Skill is an acquired knowledge of the game's workings. This applies to any aspect of the game.

Chance vs. Managability isn't an absolute, it's a spectrum. The outcome of huts is unpredictable (for now) and this introduces an element of chance into the game. The location of huts is predictable and this provides an opportunity for planned action for those who wish to take advantage of it. Skill isn't an absolute either, it's a broad spectrum with many aspects of specialization.

[DrSpike]

Originally posted by samson


What "should be" predictable and what shouldn't is a matter of opinion.

Sure. That's why I asked for opinions.

My opinion is that there are elements of predictability that are contra to the spirit of civ. Rah seems to agree. Yourself and Solo seem to always prefer predictability, with the choice always being present to 'ignore' the information you could use.

[DaveV]

Originally posted by rah
Alas, I wish there was no hut pattern, no special pattern, no oedo year, etc, because I really like the exploring part of the early game and some of these diminish the degree of discovery. But it's a game, and they're part of it, so to be competitive it is important that we learn what we can, because if we don't, someone else will.

What really killed MP Civ II for me was black clicking and city triangulation. I think both of those tricks go very much against the exploration spirit of the game, but apparently it's another of those "everybody does it" things.

[DrSpike]

Originally posted by DaveV

What really killed MP Civ II for me was black clicking and city triangulation.

Yes, these are elements of predictability I could do without.

[rah]

I don't mind the complicated ones like supply/demand, optimal roading, etc. You have to program them some way. But things like huts and specials could have easily been done more random. (like civIII, and I hate admitting that, even though, two years from now, we'll probably have figured that out too )

In MP there is no choice to 'ignore' because if you do, you've given your opponent the advantage. While huts and special patterns are quite known, a lot of MP players do not use black clicking. They are at a disadvantage to those that do. But black clicking is a flaw not a pattern type thing. (edited because while posting, everyone jumped on black clicking )

RAH

[LDiCesare]

that is due to laziness on the part of the programmers that I object to.

Hi, sorry, but as a programmer I can't let that one pass.
It is not lazy programming, it is a design choice. If everything is random, instead of one seed in the save game, you need many locations. You also need much more resources at world generation and such. At least that is true for huts and specials. So these choices (considering the amount of them) may have been dictated by performance/size issues - you must fit the program on a CD for instance, along with save games. They didn't manage to crank a map editor with ToT for lack of space on the CD, so that point is probably valid. Now, it may not be, but don't accuse people of being lazy when in fact the algorithms they coded are probably more complicated than what they would have been with random distribution.

Example with a pseudo-text-based map save file:
Non random (size, seeds, terrain)
4 4 1 2 0 0 0 0 0 0 0 2 0 1 1 2 1 2 3 0
Random (size, huts, specials, terrain)
4 4 0 0 0 0 0 0 0 2 0 1 1 2 1 2 3 0
0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0
0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 1
Save game size tripled! Means you also will take three times as long to open/save games...

There are many kinds of different randomness:
Huts location (should be random IMO)
Huts content (should not be random IMO)
Specials pattern (good and evil, see below)
Oedo years (agreed for lazy programming here)
Map (we should have random and non random maps)
Available techs (think MOO to get the point - in civ, they probably should be non random)
Starting techs (why are they random? it is not really needed and there are strange stuff in it like the affinity of egyptians for masonry and the superiority of light blue civs...)

I am more concerned by the "key civ" stuff than with other non-random but funny stuff like oedo year. Key civ is a hassle to keep track of if you are interested in (check your position every turn, pray that no civ becomes extinct...). The same holds for starting techs: They are random but include strange factors which make it worthwhile to choose a color rather than another, and Chinese or Americans rather than Persians.

The patterns special is a point I want to discuss a bit. In most games, I'd like it to be random. In OCC, I don't. I couldn't play OCC (I always run random starts) if I had no clue where to build. The exploration phase would take longer and that would probably ruin the chances of winning. More importantly, hadn't we had a constant pattern, OCC might never have surfaced (or been as enjoyable).

[DrSpike]

Come on, you can't be telling me there is no better way (even under the restrictions you mentioned) of programming the artificially predictable factors under discussion?

[LDiCesare]

I mean that randomness is information and that the file for the map will have three times as much information in it, hence however you compress it it will still be three times as big.
Now, no computer program is random, so you always simulate randomness, and yes it could have been better done, but you would have had two kinds of patterns instead of one for instance, which is pain in the a** to code, and would still leave us unhappy because it wouldn't take that long to find out which pattern to choose.
And again, a game like MOM(edited, not MOO, but MOM, which is more like civ in terms of maps) had all the randomness you could want from its maps, huts and specials, so yes, it was feasible even at Civ2 time. I just wnat to stress it is not a question of being lazy, it is a question of designing things properly. So please blame the designers rather than the programmers here if you want to blame someone. Designers should get the programmers to do what they want, not the other way around - though don't tell people at Clash of Civs that I said that they would remind me of it when I don't want to code something -.

[solo]

How well a game is designed, makes a big difference. Sometimes predictability can spoil a game, when taking advantage of that predictability creates an insurmountable edge over the AI or human opponents. Games flawed in this way become boring and never last long.

Overall, Civ II's design is excellent, but in my opinion there are flaws, and I would agree that one of them was listing the continent # next to the grid coordinates that are displayed in the game. This allows black clicking, which does go against the spirit of the game, where the "unknown" should be left that way until actually explored by players. No ingenuity or skill was needed to notice or implement this piece of visible information.

On the other hand, I think a good design feature for a game is including little puzzles, which when unlocked, provide more predictability. This gives a game more complexity and depth and opens up the possibility for new or surprising strategies. One thing many of us have found lacking in Civ III, is the lack of depth and lack of hidden puzzles. For players like myself, and especially for Oedo and Samson, as much fun comes from the "eureka" moment of solving one of these puzzles, as comes from playing the game.

Many years ago, many chess players were complaining how the increase in knowledge was making that game too predictable, especially in the first part of the game, the "best" strategies were limited to a group of "standard openings". Some suggested switching the position of the some of the pieces in the set up, to give chess fresh start, i.e. lessen predictability and let pure skill prevail. But the design of the game and its depth prevailed. Some players started experimenting with "hypermodern" (well, it was to them) openings which ended up revolutionizing opening theory. Chess has such depth that players no longer lament important discoveries about the game. The more knowledge, the better.

Civ II does not have the depth of chess, but I do not think recent discoveries about the game have made it predictable enough to become boring. Instead, I think they have given the game new legs by opening up a whole new area of strategy. Although available to all, this new knowledge about trade will take some skill to implement well.

[LDiCesare]

I agree civ2's design is quite good, but the specials patterns could have been changed with little trouble, and is glaringly obvious. The trade part is non-random, but the formulas make sense. Well, they seem to. I think some random factors could have been added at the time and would have added to the fun. They were not either because the designers were too lazy to look at other games like MoM which used them or because programmers told them that would be too expensive to code, or because Sid likes bananas.

Anyway, the game IS good, even without much randomness, and the good point there is that you can use it both to make things simpler for you (cheap cheat), but also to make things more challenging (OCC), so it is a bit a double-edged sword. However, moddable randomness could even have been better (instead of no huts/huts, have no huts/regularly placed huts/randomly placed huts). This option would even cancel my arguments about save file size, as you could warn that some options will slow the game.

[solo]

I agree that really randomizing hut locations would have made Civ II a better game. Same goes for the location of specials and for revolution years. But I like being challenged by puzzles about such things as changing tech lists, key civs, and the complexities of the trading system.

[rah]

Or at least a pattern that was harder to visualize and had some variations.

[DrSpike]

It can't be hard to come up with a pattern that is to all practical intents and purposes indecipherable.

[Straybow]

Random (size, huts, specials, terrain)
…
Save game size tripled! Means you also will take three times as long to open/save games.

LDiC, the one reason why Civ is limited to 7 civs + barbs is map knowledge. With 8 total "players" you can use 8 bits/tile to keep track of who has explored what, right? But then you also have to keep track of who has seen terrain improvements, cities, and units as they change or move during the game. Oops, suddenly it isn't so simple; you've got 7 bits (or more) per tile feature to keep track of.

All considered, a couple extra bits per tile (which can be allocated in many efficient ways) dedicated to specials and huts will not appreciably change save file size or loading time.

If I could ever manage to get FreeCiv running I'd tell you how well their coding works, or not, as the case may be. I do know they have totally random special and hut locations.

[Scouse Gits]

I think I hold the middle ground here - I use Oedo, of course, but in SP (as opposed to Succession Games) ignore a lot of the 'pattern knowledge' that we now have - personally I hardly ever black click etc...

At the end of the day - this knowledge allows one to play even better, at the expense of much cerebration, this is only 'needed' when there are other humans involved.

Good or Bad - don't ask me - it just is

SG[1]

[Straybow]

I consult the Seers of Blackclick if I happen to be on a smallish continent and finding something within Trireme range is critical.

[rah]

Using no extra memory or counters they could have designed the special and hut pattern that could have been more unpredictable. Come on, add an overlapping pattern and it would take finding too many reference points (or at least considerable more) to figure out what pattern you were in. Besides think of building on a special and having another one in range. The possiblities are enless. You could get some real spectacular locations.

RAH

[DrSpike]

Yeah I think the guy saying it would make the game too big was full of ****, to be honest. It was lazy, nothing more. And I think (although all MPers agree you have to play the game as it is) most MPers would prefer to be able not to have black map clicking, triangulation, hut and special patterns. I know I would.

[LDiCesare]

Okay, performance were maybe not an issue in civ2 (although I think the code mimics civ1, in which it might have been). But I stick to the only point I wanted to make: It was lazy designing more than programming. Sure, it is easy to do some random stuff. They didn't want to because the DESIGNERS didn't want to. As I said, MoM was out between civ and civ2, so they could have done it if they had just looked at another game. Now, don't blame programmers please.It would have been easier for a designer to say "do the same as in mom" instead of "do the same as in civ1", so blame the designers who were too lazy to even look at other existing games (saying "do huts as in mom" for instance) instead of the programmers.

As for complicated patterns, the trade goods is quite complicated, but not undecipherable, and still not random. I think, however, this one has some merit not being random.

[DrSpike]

I must confess I don't understand the programmers/designers distinction. Wouldn't the programmers make the decision how to implement the 'random' hut pattern?

[LDiCesare]

In a first release (civ1) they would. In subsequent versions, unless told differently, they would stick with what theu used to do. So yes in civ1 they may have done better, but that was the first game of that kind, and you can't get it right all on the first time. Then, unless told otherwise, they would reuse the same concept exactly. I don't know how it works exactly at game companies, I know that when I applied to work for one of them, they clearly told me that, as a programmer, I wouldn't have any option to affect the design, (and so I don't work for them). That ment while designing such a game the people at that firm would have had the programmers either reuse existing code or develop one which did exactly the same thing (which is supposed to be way faster because you don't have effects on data structures etc.)...
Anyway, I think I agree with you on the fact that someone has been lazy. The who we can't know for sure, but that probably doesn't matter much.