 |
|  |
 |
|
nbarclay
|
 |
Huntsville, Alabama
Dec 1999 time: 23:32
|
|
quote: Originally posted by Moonbars
The AI is quite dumb when it decides what to build, therefore it seems to build Cols regardless of if it needs one, or if it would be better building something else. If you have an under-powered building, you are hurting the AI. |
The main reason why human players don't make colosseums a higher priority is that we use the luxury slider instead and, at some point, very possibly get enough luxuries that we do not need either the luxury slider or colosseums. AIs, in contrast, don't use the luxury slider under normal circumstances, and can easily find themselves without much in the way of luxuries (especially if a human controls a lot of luxuries and is reluctant to trade). So even though colosseums aren't viewed as especially valuable from a human perspective, I would expect that they're almost always worthwhile for AIs.
|
|
|  |
 |
|  |
 |
|
alexman
|
 |
Hunt Valley, MD
Jan 1970 time: 00:32
|
|
quote: Originally posted by nbarclay
The effects of courthouses and factories are easy to see just by comparing cities before and after the improvement is built.
|
Yes, but in what order is it better to build them? You can't know that without playing it out twice. It depends on the level of corruption in the city. This is just one example where most players make a decision based on intuition rather than on calculations, it's one of the most fun aspects of a strategy game for me.
quote:
As I recall, in our original C3C discussion of the colosseum, you came down very strongly in favor of making the colosseum-or-cathedral question more interesting.
|
Sure, the colosseum-or-cathedral decision is more interesting than the colosseum-or-not decision. That's why I voted for the change. But the current proposal is even more interesting to early-game strategy, where colosseums used to be a no-brainer no-build.
quote:
To me, the quesiton is entirely one of whether the benefits outweigh the disadvantages enough to justify the amount of deviation from the stock rules. And my view is that it is not.
|
We have already deviated from the stock rules in the current version. That deviation affects strategy, so it's a considerable deviation. The proposed idea affects strategy in a more interesting way, so I think it's a better proposal.
|
|
|  |
 |
|
punkbass2000
|
 |
Waterloo, ON, Canada
Feb 2002 time: 00:32
|
|
"Yes, but in what order is it better to build them? You can't know that without playing it out twice. It depends on the level of corruption in the city. This is just one example where most players make a decision based on intuition rather than on calculations, it's one of the most fun aspects of a strategy game for me."
I don't think this is true. It would be a rare situation where the choice is diffcult. Either the city is prodcucing a lot of waste and little shields, in which case why would you even consider a Factory, or it's producing a lot of shields and little waste, in which case, why would you even consider a Courthouse. If it produces little of both, then you probably don't really want either and certainly not a 240-shield, 4gpt(?) Factory. Only when a city produces, say 8 spt and 4 waste would the question be even remotely difficult and the Courthouse probably wins just because it's cheaper. 10 turns for that Courthouse will probably net you 2spt, while it would take 30 for that Factory to net you 4spt. Build the Courthouse first; it will take 34 turns in total. Build the factory first; it will take 38. Yes, these are simplistic models/examples, but I really don't think there are a whole lot of semi-corrupt cities producing amounts of waste and shield that would make this an interesting decision.
|
|
|  |
 |
|
punkbass2000
|
 |
Waterloo, ON, Canada
Feb 2002 time: 00:32
|
|
I would think most AU players fall into that category, however. They're not particularly difficult calculations. The hardest part would be the Courthouse calcualtion, I was just presuming 2spt gained, as that would be the case often enough. Could easily only be 1spt. But that fact that it's difficult to predict, to me, is a bit of a flaw with the game; it should be dependable. IT's less of a bug deal for Factories, Marketplaces, etc., and once you play the game a bit it's easy to tell how many shields/commerce you're going to get, but it would be nice if you had a concept of what exactly the effects of an improvement would, even if that forecast came with the assumption that the city wouldn't change in anyway in the meantime.
Generally speaking, build order is style-based and also strategic, but again, it generally starts to become easier to tell what should be built where when you play a bit. This would not be true with the proposed Colosseum change. One of the most annoying things I find with the idea is the simple fact that it's based on 'years' as opposed to 'turns'. Trying to predict how many turns before the bonus is difficult for just about anyone, I would think. You'd pretty well need a list of which turn corressponds with which year, or at least the number of turns for each unit of years (ie. 50 years/turn until 2750) and even then, the calculations would not be straight-forward. If I build a Colosseum in 660BC, how may turns until it pays for itself? I doubt many of us could do it without a fair bit of homework. What if I build it in 660AD? Another round of intense calculation.
Then again, OTOH, this is certainly different from most decisions. I doubt anyone ever builds a Wonder for the Tourism bonus. An improvement with it does at least make for interesting decisions, even if not very predictable, and the learning curve will not be friendly as it will be so hard to line-up cause and effect with each other.
But, OTOH, if it is such a divergence from anything stock even knows, perhaps it is not a change for the AU mod.
|
|
|  |
 |
|
The pirate
|
|
quote: Originally posted by alexman
Of course, it would. All you need is one player to do the calculation once, and then a rule of thumb becomes available for all. For example: "A Colosseum never pays for itself if you build it after 500 AD" (you would still build it if you need the happiness though).
|
Does also depend on when you anticipate the game ending. Based on your approx calc in the 'a neat bug thread', a colesseum built in 500BC pays approx 1gpt return by 1500 AD. I often still have games going at this stage, but my impression is many players have won by this stage, is this right?
|
|
|  |
 |
|
nbarclay
|
 |
Huntsville, Alabama
Dec 1999 time: 23:32
|
|
quote: Originally posted by alexman
Even if there is one city where the build order matters (and there are plenty of cases where it does), you have a situation where you rely on your gut, not on a calculator. |
Your gut, based on large numbers of cases in the past where you've built courthouses and seen their effects and where you've built factories and seen their effects. You even know the formula for factories, although a bit more guesswork is involved in figuring out the effect of courthouses.
But seeing what the effect of a tourism bonus really is is a whole lot more difficult because the effect doesn't kick in at all for a long time and then it increases with time in several steps when it does kick in. Changes in the ratio of turns to years make the situation even more complicated, and so do questions of when the game is likely to end - especially if a player hasn't made up his mind what kind of victory to pursue or what kind of strategy to pursue along the way.
In a later post, you suggest the idea of rules of thumb. But rules of thumb would require plugging in guesses about when the game is likely to end, which is something that varies from game to game and from player to player and that players may or may not be able to predict with any reliability. Players would either have to look up the appropriate rule of thumb for their best guess at when the game is likely to end or apply a mathematical formula to the rule of thumb. That's a vastly bigger complication than just reducing a building's maintenance cost as we do now.
Further, with normal build decisions, the question is, "Which building will benefit me more during the time in between when I finish the first one and when I finish the second one?" For example, if your science slider is set low prioritizing gold over science and/or you need extra happy faces from luxuries, you build a marketplace before a library, but if your science slider is set high and happiness isn't an issue, it's better to build the library first. Similarly, with the courthouse/factory question, if you have lots of waste, a courthouse first makes more sense, but if you have little, a factory first makes more sense. And if you make an imperfect choice, once the city has both improvements, you're getting the same benefits no matter which order they were built in. The period of time directly affected by imperfect choices of order is itself generally very brief.
Granted, the doubling of culture generation values after a thousand years can make things a little more complex if culture plays a significant role in players' build choices. But even that is really not much more than a little extra "oomph" for the easy-to-understand concept that building cultural buildings earlier gives you more culture.
The entire nature of the tourist attraction bonus, with a cost paid for a benefit that won't even begin to kick in for a thousand years and with the value of the bonus depending so heavily on timing, is completely different from how build order decisions normally work. It introduces a complex new element to the game, and that's not something I think a conservative mod ought to do.
|
|
|  |
 |
|
punkbass2000
|
 |
Waterloo, ON, Canada
Feb 2002 time: 00:32
|
|
quote: Originally posted by alexman
Of course, it would. All you need is one player to do the calculation once, and then a rule of thumb becomes available for all. For example: "A Colosseum never pays for itself if you build it after 500 AD" (you would still build it if you need the happiness though).
Edit: This is not unlike other decisions like the location of your FP. The calculations are complex, otherwise you don't know if you've done the right thing until you've played the game out at least twice, but you rely on a rule of thumb to determine when and where to build it. |
There's no comparison. The FP is a wonder (a small one, albeit, but that kinda strengthens the point, IMO). You build it once per game and, until C3C, it's pretty easy to follow. And also, you just need to learn how rank corruption vs. distance to get a good idea of where to place it, and its effects are immediate.
Colosseums, on the flip-side, are city-improvements that can be built in any city. The proposed effect would not be easy to follow, even with a pre-made chronological graph of the cost/benefit analysis. Aside from dificulties in making that graph easy to read (charting both year and turn on one graph may not be straight-forward some, I know it isn't for me), this is decision that needs to be made many times throughout game and has no simple heuristic. On top of that, you need to determine ahead of time when the Colosseum will actually be built. Assuming at least ten turns of building time, it may not be so easy to [redict that the year with much precision. The FP, conversely, need only the basic understanding that it does not enable a second "core" and that getting it up faster is usually better than better location. Sure, for optimization it could be difficult, but a slight error one ring either way is not going to dramatically effect the cost/benefit of the FP. The Colosseum's construction can vary by one turn and easily effect ten turns of revenue collection at every potential Tourism increase.
This, BTW, comes into line with Nathan's point about culture: Culture only doubles once and is generally not a major consideration for building. Tourism increases up to six times throughout the game and I doubt anyone could argue that Culture comes even close to the importance of Commerce, even if a Culture Victory is your overall goal (only real exceptions would be HoF and maybe AU courses).
|
|
|  |
 |
|
Rommel2D
|
|
ACS Staff Member / Hosted Site Admin
|
 |
Upper Midwest
Feb 2003 time: 23:32
|
|
I'm not seeing the reason to worry about this 'complex calculation' for the tourism bonus. Do most players find the difference between 1, 2, 4 etc. gpt to be that significant that late in the game? Once I have Smith's and a bankroll going, anything under 100 gpt barely seems worth worrying about. Anybody who's played a few games past the industrial age and followed the date on the main GUI should have a rough feel for where they'll be in a thousand years, and that any colosseum built at the time will have an extra bonus somewhere down the line. I just can't imagine that much benefit will be gained by calculating individual gpt figures that late in the game, much less working it out hundreds of turns in advance.
I think the camparison to the FP formula is valid. My point being that the game designers put alot of effort and time into changing that formula to match the original vision, yet keep it playable. They came up with something they were satisfied with, and most players have accepted as an improvement to the game. It is what it is and players can work with it to various degrees to get their desired results: building a FP improves your civ, but it isn't critical to optimize the location. The same seems likely with colosseums.
If a few playtests shows this is a game breaker, then we'd know. Otherwise, I doubt squeezing the optimal amount of gold from colosseums would prove to be critical to a winning strategy. Just because the two formulas are too complex to keep in the front of your mind and make 100% correct micro-management decisions with doesn't take the fun out of the game. Just the opposite IMO.
|
|
|  |
 |
|  |
 |
|
punkbass2000
|
 |
Waterloo, ON, Canada
Feb 2002 time: 00:32
|
|
"I'm not seeing the reason to worry about this 'complex calculation' for the tourism bonus. Do most players find the difference between 1, 2, 4 etc. gpt to be that significant that late in the game? "
That's the game-breaker, IMO. It's not necessarily that late in the game. If one beelined for Construction, they could could easily have a colosseum by 2000BC, generating 10gpt by 0AD, and 14gpt by 500AD. Yes, if you don't consider this sort of play, it's likely that the Colosseum will never become that valuable, but even a Colosseum in 1000BC which would not even need much particular focus can produce as much in 1000AD and 1500AD, respectiely, which is nothing to sneeze at. This is in addition, BTW, to the two Content Faces you're gaining in the mean time, allowing two extra citizens to grow/prosper and (likely) provide ~4fpt, 2spt and 2cpt, (plus modifiers on the cpt early in the game). Honestly, I would play the game completely differently. I would choose the Sumerians, who are not a bad civ as it is, Trade early for Masonry, researching IW in the mean time, then straight to Construction (afterMath ). With some pertinent pre-builds (I won't have trouble using the Palace, as well as Granaries/Temples can get me half way there. I would even consider poop-rushing due to associated rewards) I could probably have two at 2000BC and four by 1500BC, depending on numbers of factors like Difficulty, map size, etc. Still, at 500 BC you're making ~12gpt. Not bad. And it directly helps alleviate problems that Republic has, further strengthening that beast. Money and MP being some of the main problems with switching to Republic, which you're probably doing right around this time (and Colosseums still gain in value during Anarchy).
|
|
|  |
 |
|
Dominae
|
|
quote: Originally posted by nbarclay
Granted, the doubling of culture generation values after a thousand years can make things a little more complex if culture plays a significant role in players' build choices. But even that is really not much more than a little extra "oomph" for the easy-to-understand concept that building cultural buildings earlier gives you more culture.
|
Concept: Building Colosseums earlier gives you more Gold later on.
Q1: Do I want more Gold?
A1: Yes.
Q2: Will the game last until "later on"?
A2: Possibly. If yes, build Colosseums. If probably not, do not.
Please point out the extreme advantage a player-with-a-calculator derives beyond this rule-of-thumb analysis. I can understand pb2k's argument that Tourism might change the face of the late-game (and therefore affect the early-game more than we think), but your whole argument that there is something incomprehensibly complicated about Tourism I simply do not get.
As alexman points out, every turn you have to make choices in Civ3 that a-player-with-a-calculator is always better off making than you: one more Worker, or start on a Barracks? FP now, or FP later? FP where? Courthouse or Cathedral? Where do I place my cities? Should I trade this tech to this AI first, or that tech to that AI first? etc. etc. If you think you have the right answers to all these questions at all junctures (but somehow cannot figure out Tourism), why are you still playing Civ3? In a turn-based game like this, information is a very powerful thing; just look how look the Civ3 Strategy forum has remained alive! Yet 95% of players could not care less about most of this info, yet they succeed anyway. The Tourism mechanic is not something that is beyond their understanding.
Like I said, if Tourism Colosseums fundamentally change the face of the early-game (beeline for Construction at full speed, prebuild Colosseums everywhere, all the time), then I agree that the change does not belong in the AU mod. But if all it does is give players a new avenue to explore (that is, it makes Colosseums more interesting), then I believe it does.
I think some testing is what we need here. Can anyone "break" Tourism Colosseums? In other words, given a game that lasts well past 1500AD, does building Colosseums very early give you an unfair advantage versus the AI?
Last edited by Dominae on 09-01-2005 at 04:28
|
|
|  |
 |
|
Dominae
|
|
Picking up on what Moonbars just said:
quote: Originally posted by punkbass2000
That's the game-breaker, IMO. It's not necessarily that late in the game. If one beelined for Construction, they could could easily have a colosseum by 2000BC, generating 10gpt by 0AD, and 14gpt by 500AD.
|
How many turns after 2150BC is 2000BC? About 3-4. Please go back and check the various DAR threads in AU games and in the Strat forum that display players' progress at 2150BC: are they prepared to start building Colosseums en masse? Just building one Colosseum, maybe, but prioritizing them in every city (a total of 5, tops) would be akin to building Wonders while you are still REXing. Do you really think the Tourism bonus is worth that kind of sacrifice? Even at 1500-1000BC, you are still putting the finishing touches on your REX and have nothing more than a fledgeling military: would you really start poprushing a 100+ Shield improvement? Not to mention the fact that you missed the Philosophy beeline...(actually, this would be a nice side effect, IMO!).
|
|
|  |
 |
|
Modo44
|
 |
in western Poland
Jul 2004 time: 06:32
|
|
I checked some old games, and some new ones. Ranging from Monarch to Demigod difficulty, mostly with the AU mod. And I rarely find Colosseums built BC. They do happen if I have a very big city early on, but that's about it. Otherwise I'm too busy with my REX, basic infrastructure, and units, to go for a Colosseum, let alone need one.
With the current proposal, it would at least make sense for Militaristic civs to build Colosseums for happiness, once thay are available. But even they would have to get the tech first, and need the rather expensive happiness. I think chances are small for that to happen before 500 BC, or even later.
Yes, one could make an effort, and get a couple Colosseums up by 1000 BC, but, as Moonbars noted, this would require to hurt one's initial REX big time. Meaning, a couple relatively big cities very early. Definately something to look forward to... in 5CC games perhaps. 
|
|
|  |
 |
|
Dominae
|
|
quote: Originally posted by nbarclay
It's that "strategy" predicated on guesswork isn't all that strategic in my view. |
Every decision you make is based on a certain amount guesswork, that's why you are not able to beat Deity or Sid every time you play. While you might very well try to claim that every decision you make is the right one given the information available to you (har har), this is not true for every single player that plays at Demigod level and below (conservatively speaking, surely). Go back and read the DARs of that last few AU courses of Monarch and Regent-level players: they are explicitly stating that they are making guesses all the time! And you?
Strategic play is not perfect play. A player might incorporate a certain number of bombardement units in his/her offensive stack, only to later realise that this was probably not the best strategy due to the unexpected dominance of fast-movers in his/her campaign. Similarly, a player might incorporate Colosseums into his/her strategy on the mistaken assumption that the game might last until 1500AD. As players get better, they learn what works and what does not (or, they get better by learning what works and what does not). The AU mod's job is not to facilitate this process by making everything obvious. Rather, it is to make for an interesting/balanced game for veterans and newbies alike. By suddenly making Colosseums better, I do not think we are doing players a disservice by "forcing" them to think more. Was Firaxis in the wrong for "forcing" you to learn to use Settler-pumps?!
|
|
|  |
 |
|
nbarclay
|
 |
Huntsville, Alabama
Dec 1999 time: 23:32
|
|
Dominae, I'll certainly agree that in a game like Civ 3, there is a probabilistic element to strategy. You have to plan in terms of the various things that the AIs might do, and for the vaguries of the random number generator. Thus, choices are often made based largely on a player's best guess about what is likely to be happening elsewhere in the world or is likely to happen in the future. The need to learn how to play the probabilities while at the same time protecting against being harmed excessively if something unlikely happens is indeed a very important element of Civ 3 strategy.
And if all that were involved in the tourist attraction question were a need for players to guess at when a game is likely to end, with players having a clear picture of how the cost and benefits of a colosseum compare for any given guess at when the game will probably end, I would agree that that is a perfectly legitimate probabilistic element to expect players to deal with. Just the need to guess when the game is likely to end, in and of itself, is not a significant concern to me.
But when the guesswork results from a lack of understanding of complex game mechanics, I view that as a very different issue. Players are most capable of planning strategy when game mechanics are easily understood so the players have a clear understanding either of what will happen (if the effect of a choice is definite) or of what might happen (if random number generators or decisions by other players are involved). When a game mechanic is so complex that players don't have enough information to get a clear picture of what the results of an action will or are likely to be, that seriously undermines the ability to make good strategic choices.
|
|
|  |
 |
|  |
All times are GMT. The time now is 05:32. Apolyton Time is 00:32. |
top of page
|
|
|
Forum Rules:
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
|
HTML code is ON
vB code is ON
Smilies are ON
[IMG] code is ON
|
|
|
|
|
|