3rd Apolyton Civ3 Tournament : 17/Dec/2001-7/Jan/2002 (results are in!)

[Aeson]

Scoring

The problem with scoring isn't so much that early conquest and "bloating" can score so high, its that other ways can't compete. This is because the Bonus awarded for early victory is figured the same regardless of victory type. Of course a space launch/diplomatic/cultural victory isn't going to have a chance to occur in the BC's or early AD's, so the conquest (which usually can be had by then) bonus is usually the biggest. Because the Bonus for winning after about 1000AD becomes negligible, that means that the most points if conquest hasnt been achieved by then will be 2050 scores that have longer to pull up the average (which started at 0 in 4000BC).

The current Bonus formula is (2050 - Date) * Difficulty = Bonus. Difficulty is 1-6, corresponding to Cheiftain-Deity. There needs to be a modifier for each victory condition something like (2050 - Date + VictoryCondition) * Difficulty = Bonus. The number for each victory condition would be the date at which that victory would be considered "well done". If 1500 is used for the space launch VictoryCondition, then a 1500AD Space launch on Regent would get a bonus of 6150, instead of the 1650 bonus it would get with the current scoring system, which is a much more deserving bonus IMO. Map types and Number of Civs would affect the dates that "well done" victories would be achieved, but just general values for VictoryConditions would still improve the scoring system greatly.

Possible values for VictoryCondition:

Space Launch 1500
Domination 500
Conquest 0
Diplomatic 1400
Cultural 1600 (1800 for the 20,000 one city option)

If Modifiers for Map type and Number of Civs needed to be added, they could be added to the formula. Larger maps take longer to complete conquest or domination, and allow for more expansion, thus earlier dates for the other victory types. Smaller maps are just the opposite. Usually water levels and land mass types seem to even themselves out. Archipelago for instance, slows conquest, but also has less room for the other victory types to build on, and keeps Civs from gaining contact with each other as soon, so wouldnt need to be included. The number of Civs usually makes conquest take longer, and makes tech advancement faster (decreasing the time needed for space and diplomatic victories) but makes little difference in Cultural victories. The modifications would just need to be applied to each VictoryCondition, and the formula could remain the same.

VictoryConditions:

Space Launch = 1500 - NumOfCivs*10 - Landmass
- more civs mean faster overall research
- more landmass means more possible population/territory and faster research

Domination = 500 + Landmass
- the bigger the landmass, the more land needed to trigger domination

Conquest = 0 + NumOfCivs*10 + Landmass
- more landmass, more civs, more to conquer

Diplomatic = 1400 - NumOfCivs*10 - Landmass
- more civs, more land, faster fission can be researched

Cultural(100,000 total) = 1600 - NumOfCivs*2 - Landmass
- more land to build culture cities on
- more civs, faster research for new cultural buildings though less chance at cultural wonders

Cultural(20,000 city) = 1800 + NumOfCivs*10
- more civs, less chance at cultural from wonders

Landmass:

Huge = 300
Large = 100
Standard = 0
Small = -100
Tiny = -300

Of course, most of this is just off the top of my head, especially the landmass values. Certainly all the values could use a lot of fine tuning... just a scoring system that takes all of this into account (not just score by date and territory/population) would be nice. A side bonus for each wonder built (I usually dont build any, because it lowers scoring efficiency, best to just conquer them as it is now), turns of peace, and future techs could also be added. I know future techs are already in the scoring system, but with there being 540 turns in a game, and usually future techs are only going to count for 100 of them at most, they get "averaged" into insignificance. A seperate, non-averaged bonus would give them some meaning, and reward those who go the scientific route, rather than just doing enough to get the advances needed for a specific victory.

[SonicMeiyu]

While I agree in principle with the idea behind what Aeson is saying, I don't think Firaxis is going to do any of this, and I doubt we can modify the code to do it ourselves.

So for now I'm going to base my judgements of the quality of another players game relative to mine by the victory date and their method of victory.

Also, is anyone other than me a little confused by the "last automatic save game" method for submitting your game? When I attain a diplomatic victory in say 1705 AD (as I could have in this game) my last save is for 1700. The same thing happens with my space race victory (1796 AD but the last save is 1794 - see earlier posting for the most amazing game in the history of civilization! - please? someone download my zip?) Mark, do you want the last save before (so you can see the triggering event and watch the show) or after (so you can see the score with no big delay - thus forcing me to "play a couple more turns")?

Finally, where are the summary lists for the previous tournament results? Are they up yet or not? (I apologize in advance if it takes a whole lot of work to do this).

[Aeson]

You're probably right that they won't do anything, but oh well. It would only take a few minutes to modify the scoring code with the most basic of changes to the formula. There shouldn't be any extra debugging with the code needed either. The only real time consuming part would be in balancing the landmass and number of civ variables if they chose to include them. Of course even unbalanced, it wouldnt be any worse than the current system

Judging based on Date for each victory condition is a good plan. The only problem is there is no real way to compare different victory conditions. Also for Diplomatic victories, sometimes date is meaningless, as quite often the only requirements are that you build the UN and not make anyone mad at you. Letting the AI do all the work (research) on difficulties higher than Regent means that you can get to Fission faster than building an empire and researching it yourself.

[SonicMeiyu]

I don't think it makes sense to compare scores from different maps, and since everyone has the same map in these tournament games all you need to know to evaluate them (at least qualitatively), is the victory condition, date, and score. When comparing within a victory type both date and score are good indicators of the quality of the game.

I think the victory conditions come in at least two classes that can be easily compared with each other by score:

1) diplomatic, space race, and cultural
2) domination and conquest

I'm kind of guessing with the cultural condition because I've never won a game this way. Basically, class 1 rewards the quiet, peaceful, and technologically advanced civ, while class 2 rewards the aggressive, warlike civ. Class 1 also requires late game victory, while class 2 can occur at any time.

Ideally, I would alter the calculation of score by reducing the score during war time - perhaps a 50% penalty (analagous to the 0 score one gets during anarchy). A peaceful civ that controls 40% of the world should not be considered inferior to a war mongering horde.

Pragmatically, to intercompare the two classes I would suggest a 25% bonus to the class 1 victory type. But the size of the bonus is highly subjective and so it only lets you see if two games are "in the same ballpark" (to use a North American phrase - would you say something like "on the same pitch" in Europe?).

[Chemical Ollie]

I am sorry if I have led this discussion somewhat out of topic, but I think that a discussion about the scoring system makes sense in a tournament aftermath. Anyway I have not found any separate thread about scoring.

Sonics' suggestion is worth considering, to have two "classes", one for peaceful players and one for aggressive. But on the other hand, you can win peacefully after a long time of aggression if you are careful enough not to over expand your land mass.

My strategy for CIV2 was to expand like crazy in the early game, then do peaceful city development and research in the mid game. I used to conquer the world in the modern age and left just a few cities of the last civ when I made peace. Then I went for population optimisation, future tech and a full spaceship that was launched just a few turns short of time-out. That strategy normally gave me 400-600 % on the huge earth map under emperor or deity, depending on how lazy I had been in the boring micro management. I know now that I have to do something completely different to get a good high-score on CIV3.

One thing in CIV2 to encourage peaceful playing was that you got score points for the numbers of years of peace. I miss that feature in CIV3. Once you understand how the scoring system works, people can do crazy things that never would happen in real world just to optimise the score. I saw in this thread that some guy had sold off his culture improvements to delay the vic, so he got more time to optimise the population. That was very creative, but not realistic. I would not call it cheating, but it decreases my admiration for the game designers that such acts could by nessessary to improve the score.

[SonicMeiyu]

Originally posted by Olaf Hårfagre
Sonics' suggestion is worth considering, to have two "classes", one for peaceful players and one for aggressive. But on the other hand, you can win peacefully after a long time of aggression if you are careful enough not to over expand your land mass.

You can basically tell how aggressive a player was by their score, and there are two types of agression: early settler agression, and military agression. To get a score over 2500 you need at least one of these, to getup to 4000+ you need both (with early military aggression).

In my old civ2 days I liked to develop a medium sized high quality empire, large enough to lead in tech and money, but not so large that each turn took forever. Mayber 15 or so cities. Initially I'd ignore other civ's attacks and just defend my territory until late game when I had modern units against their late middle ages to early industrial units. The result was a very high score because only the last turn counted. So the old system rewarded late aggression while the new one rewards early aggression.

If you want to "win" the tounament you'll have to adapt your playing style. But I'd rather submit my game, tell a little story of my experiences, and let the players who like to conquor the world with horesmen have their glory. Personally, I can't stand playing that way. I don't have the patience to send an army of 100 horsemen, archers, and catapults one little square at a time.

[philler]

that's the spirit Sonic!

as I said once, who gives a fudge about the score?
having fun playing and sharing experiences is what matters IMO.

[JLPOWELL]

Domination Victory 5200 pts 1240 AD

3rd Apolyton Civ3 Tournament Domination Victory 5200 pts 1240 AD

Well this game is just about as addictive as Civ 2. I haven not played but one Monarch level game before, and it is nice to not be playing catch-up all the time. I usually play at Emperor level, and would like to see a few scenarios at that level. I have tried Deity a few times, and been crushed flat each time. (Not even road kill more like bug on windshield.)

The game ended unexpectedly in 1240 with Greece at peace. I had planed on taking a few turns to finish some wonders, get military tradition and smite Rome, they being nearest. I was not even considering anything but a purely military solution. This left 4 wonders in various stages of construction at the end, but more victory points than expected. I usually don’t pay much attention to the points.

So how, you ask was this heroic wondrous feat of military genius accomplished? Well as Dilbert said in response to Ratbert’s prediction of “Edge after six and a half rotations followed by chicken sounds” That was luck luck luck luck.

Luck because this scenario fit my style of play perfectly, I always fight, I just do not have much interest in a cultural or space victory. I might go for diplomacy, but I keep Clausewitz too close to heart so my diplomacy is usually at bayonet point. Besides the computer players seem all to be trained by Machiavelli and cooperate little with the human. But then again they are oh so gullible; they never seem to recognize that where that knife in the back came from.

Luck to start next to my nemesis the Zulu. I like to get horsemen and use their hit and run tactics, but the impi are a real pain if you try that against them. So I took a risk and went after them early with a small swarm of archers.

Luck again to defeat the Zulu with fairly narrow odds. The Spartains are still upset from all the forced labor there.

And luck again as it turns out I would not have had any horses until I dealt with the Zulu anyway.

How to? Well not too subtle. War is simple, find the enemy and kill him. That which works for Schwarzkopf works for Patton and works for Conan as well. As always the devil is in the details. Unfortunately the AI is a tactical moron. Don’t get me wrong, I think Firaxis has done a pretty good job with the AI overall, but in a fluid situation with mobile units, it breaks down pretty badly. A relatively small force of horsemen can wreak havoc taking very few losses. And then of course you can upgrade them all to knights, which REALLY leave the hoof prints on the enemy. And don’t get me started on cavalry, which are way too good.

Some more good luck (some of which you can make for yourself) was getting leaders when they were needed. Being able to build a Forbidden Palace at just when needed was useful and I used one to rush a wonder as well. The Heroic Epic is definitely worth building. A squadron of elite horsemen each delivering a coup de grace nearly every turn is bound to cause a few leaders to emerge over time.

At the end I had 31 knights most of which started off life as horsemen. Also there were 8 musket men plus a surviving archer and a couple of swordsmen. And oh yes the first warrior I built. Didn’t have heart to retire the old guy, don’t think he ever fought a battle.

Speaking of time wow. Still can’t believe how much I have spent on this game.

Thanks to Apolyton for the scenario. Playing a scenario is better than a random start, but you can’t really write one for yourself. Spoils the surprise

BTW getting rid of the huts was a good idea it removed a factor, which was very chance related. AS the barbarians were set to raging there was plenty of chance as it was. I got sacked pretty badly by a group of over TWENTY horsemen which at the time was bigger than my entire army. It’s a good thing they disappear once they sack a city, as it is they creamed Cornith from size 8 or so to size 1 and cost over $1000. If they had hit Athens instead they could have caused some really serious damage.

[Chemical Ollie]

Originally posted by philler
that's the spirit Sonic!

as I said once, who gives a fudge about the score?
having fun playing and sharing experiences is what matters IMO.

I partly agree, but my other brain half can't drop the competition instinct. I simply want to be best, not only beat my own high score, but all others. However, right now I am still at the low end of the learning curve for CIV3 and I took part in the tournament just too learn more, but once I have learned how certain things work, I will try to beat you all.

In this tournament game I did a lot of save/reload "cheating" just to see how I could improve things if I did them differently, and I learned a lot from that. As I submitted no game file it does not matter that I cheated anyway, so don't be angry. I will share some of my findings here. Some of you might regard my advise as newbie stuff, but then just skip to the next posting:

- Chop down all forest and put mines on all grassland in the early game. Even if it is unlogic in real world, the game is designed to encourage it. The AI does it all the time. It was the only way I was able to beat the AI in a wonder race I had to reload over and over before I found the solution.

- Start building Pyramids, HG or GL after only 3-4 cities. Use the other cities for further expansion. The AI is fast on those early wonders.

- Build your military early, even if you intend to do a peaceful game. If not, the AI civs will come for your ass, no mather how smoothly you go in the diplomacy. Sometimes it helps just to upgrade a unit or two, as the AI fears your strongest unit. It saved me from a German invasion once. They walked into my territory without asking for permission. If I asked them to leave, they declared war and rolled over my mediocre defences like rhinos. I reloaded and bribed them with some furs to keep them happy. They did not declare war but walked in anyway and I could do nothing about it. Even if I still was shipping them furs, they sneak-attacked my capital and took it. I reloaded once again and rushed for gunpowder. Lucky as I was, the saltpeter city in Zululand defected to me and I upgraded a few hoplites to muskets. When I did that, the German invasion army turned in its tracks and started to walk home again, no shots fired. Then I fended them off with some forts in the chokepoint so they could not walk in like that any more. Even if they had some former Zulu cities on my side of the chokepoint, those were probably too corrupt to build an invasion army. I started to focus on building my military instead of culture, which kept the Krauts from declaring war until I decided to terminate them with my merciless hordes of cavallery. Then the decision was in my hand, not theirs.

- Use artillery, ships and bombers to kill off as much of the enemy population as possible before you send your attack units in. Then starve the rest of the population down to town size 1 and let it regrow with your own population. Such cities hardly ever defect. I forgot to starve a couple of invaded cities in this game and they all defected after some turns, while the others did not. Strangely, starvation gives no happyness penalty, which would be the case in real world.

- Armies are great for defence. I built lots of armies and loaded them with 4 mech inf each. The AI will not even bother to waste attack units on that stack, even if there are 25 artillery to catch under it. Armies are great for city defence too.

- The AI has great spies and will almost always attack your weakest point. Don't give them any opportunities.

- Don't push a civ too hard if it has nukes. I tried to invade Bejing and put my stack of artillery on the hills just south of the town. Just before that, the Japs and Romans had (without my involvement) decided to join my cruzade against China, so the Chinks were fighting a 3-way war and were really down on their knees. When I started the barrage the bastards nuked my capital and one more city. Even if they could not go in with ground troups to take advantage of the nuking, it was a real mess to clean up. It also made plains of my grasslands, destroyed all roads, killed half the city pop and some city improvements. And I did not even get any warning or explaination about what happened and who did it. Just a big mushroom when you least expect it and BOOOOOM, there goes your capital. But it must have been the Chinks, as I was in war with no one else. I reloaded, and this time I took 2 other cities before I approached Bejing. The Japs and Romans did not join me this time, and there fell no nukes on me. China was finally willing to talk to me (after 40 turns of refusal and painful war unhappiness). Later they nuked Japan to extinction but that was not my problem, was it?

I now look forward to tournament #4! I think I have learned enough to try that without any reloading. And I will go aggressive conquest early to maximise score, (if the next game is on default rules).

[LitheDuke]

I agree with the brutal comments - again no victory here.
Please can we have another game soon - I don't mid what level, as long as there are no Zulu, Germans or Aztecs



Oh, and thanks for the lesson - I learned a lot, just not enough...

[Webber]

I need more tournaments!!!!

Please please please post the next tournament. This game is far too adicting. After giving up in this tourney (military style just isn't my thing) I am anxiously awaiting the 4th. Maybe it's just me, but building up a huge army and just rampaging throughout the world, city by city, isn't all that much fun. I for more enjoy trying my hand at real empire building and beating the opposition is all aspects of the game, not just military might. This might sound wierd coming from a Marine, but I just think it's more fun, challenging, and realistic playing this way.

If the 4th tourney has already been posted somewhere and I'm just not seeing it, could someone please clue me in? Thanks. Out here.


Webber

[MarkG]

Originally posted by SonicMeiyu
Mark, do you want the last save before (so you can see the triggering event and watch the show) or after (so you can see the score with no big delay - thus forcing me to "play a couple more turns")?

i dont think the game saves a new saved game anyway, when you win.

Finally, where are the summary lists for the previous tournament results? Are they up yet or not? (I apologize in advance if it takes a whole lot of work to do this).

with all the server trouble and the rest of apolyton work, i havent been able to do them, i've only done half of the 1st tournament. help is welcome....

[Peter Cross]

Great game!!! Looking forward to next one.....

[GhostQ]

diplomatic(!) 1868 ad

heres my game
i managed to achieve a diplomatic victory at 1868 just a few turns before space race. at the start of the game i was bound by zulus, was too lazy to drive them back and went for tech =). my relations were so good with everyone (3 being gracious!) i built the un and voted and won.

[Duncan Idaho]

Originally posted by MarkG
with all the server trouble and the rest of apolyton work, i havent been able to do them, i've only done half of the 1st tournament. help is welcome....

I'd be happy to help. I'm assuming this means:
1) download all saved games
2) check victory type and score
3) post results

Drop me a line or something and I'll get to it.