Well, if you look over metaverse ranking like I do every single day, playing easy levels don't really change rankings in the metaverse. It's matter of how often submit very high score is the key in metaverse.

Now, I don't think playing easyer difficulty is a cheep way of getting metaverse rankings. But I do think playing bigger maps gives you better scores is not fair. I got 600 mhz pentium 3(overclocked) cpu, with win 98 with riva tnt vanta(or something) graphics card. I cannot play anything bigger than large map. Therefore there is almost no way for me to get 20000+ scores. Not everybody's rich and able to afford a very good computer. So why do they get better scores when they play gigantic map? Meaning you have better computer then you get better score?

That's not actually the strategy, of course. Many games are played in the GCCG style that do not result in huge early treasuries; I suspect it relates to the relative starting positions and PQ qualities of planets colonised by player and AIs. Thus, not all games are recorded as anomalous. The player in this style seeks to move through the wonders/TG/techs as smoothly as possible; money is only a means to an end and I have never worried if I'm up or down as I take the last planet.



Any strategy worth writing down falls under this definition. If it wasn't any good, do you think anyone would bother? The point is you are defining what constitutes a 'correct' way of getting high scores, as opposed to what you look down upon as 'cheese'.



That's a frankly outrageous claim, and I find it offensive. I was ranked 40th, IIRC, before I discussed using his strategy with EBZ. I've played hard and long to muster enough points to make a decent contribution to my team's score and advance myself to the top ten. I have won at normal, tough, challenging and even painful. What do I have to prove to you? I doubt I have the skill to win at maso, and I dont enjoy the size of gigantic maps. I chose to invest my gaming time in a different way: to flog the game as hard as possible to advance my team's score. I'm not trying to be a better person, just help my team. EVERY game I play is legal - no mods, no nothing. Every game Ray and Pontius play at gigantic higher levels scores 2.5 times the score I do. You set the scoring system; you made the rules. I play according to them. Everyone else is free to play the same, or differently - the top 25 is made up of all sorts of styles. I reject the notion that the GCCG style 'infringes on other people's enjoyment', because there is simply no evidence that is true. Everyone seems perfectly happy to press on in their own way, and many have seen the GCCG as an interesting experiment. Not all who tried it have stuck with it.



Would you care to define the 'correct' long term strategy as in the World According To Stardock? Really, you've just said as much. I thought strategy games were about developing strategies to win the game/Metaverse, not about figuring out THE strategy the developer really wants you to do.




I'll be even plainer: try fixing the scoring system in the game first. If you want maso/gig games to be worth ten times beg/small games rather than 2.5 times, then code the game that way. I dont mind if that's the rule. Don't come trashing players yelling 'cheese' when they actually discover a way to beat the game. I'm a customer, I paid the money, I'm not hacking, cheating, or modding the game. I have played many strat games and nver quite been good enough to play at the highest difficulty level in Civ3, etc. but my commitment to play LOTS counts for my team, here. Re-adjust downwards if you feel it proves your point.

Perhaps on the Stardock registration form you should also ask for my: religion, colour, creed, political alignment. That way I could be certain I was playing correctly.

I should really delete that last paragraph, but I'm just too annoyed. I love the game, and the boards. I just paid the money to upgrade to Drengin net last night, would you believe? Pity if I end up walking away after that, although from the enjoyment I've had over the last two months, the extra $US30 was still worth it.

Trade tech's (unless it involves a tg or wonder you haven't built yet) and wonders on long term contracts (I usually do 100 smackers for as many turns as I can get) They usually go into the red ie bankrupt. You probably have done it and never thought about it. EBZ just consciously does it more.




well in the top 20 or so, I have the fewest games in. So even though I know exactly how they do it, I can't duplicate it all that well. My urge to change this, that, the other just to see what it will do is too great.

Berserk,

My PC trembles on Large, too. But all other things being equal, you can play more games on smaller maps in the time it takes to play one game on a big map.

From the preview art on the Downloads page:

"Play in galaxies small enough to be finished in an hour or play in massive galaxies with thousands of planets and ships that take months to complete."

So, why is scoring based on map size a problem? If the Metaverse rewards dedication (and based on this thread, it certainly does), what's wrong with rewarding time spent on a big game?

Ultimately, my job is to  help the community as a whole.  To that end, the integrity of the Metaverse as a whole.

I am not as interested in the details of a particular exploit than whether a particular something passes the smell test.

When someone looks at the top players on the metaverse, they expect those top players to be winning at the hardest levels, not cranking out immensely high scores as the lowest level.

And bankrupting the AI can work to a degree at many levels, but at beginner, the AI is completely gullible. After all, if the tactic worked so well at higher levels, people would do it en masse and save a lot of time in beefing up their scores.

Our interest is preserving the general integrity of the metaverse as a whole. We're not going to worry too much about individual cases unless they go out of their way to make it apparent (such as making a post complaining about anomolous scores and then making it pretty apparent that they are cranking out "beginner" scores to inflate their metaverse ranking).

For people who jsut want to play a mindless progress quest system, here's where they shoudl go:

http://www.progressquest.com/

 

 

Because there is a way to beat the gigantic game in one day. Not one month. I do not know but people do it. That don't mean that they are wrong, because they have spend their time to figure it out which I didn't.

I think I'll follow MadProphet's lead, the winds seem to be blowing towards this thread. copy-paste:

---

The way I see it (and this could be wrong for some people - if so, I apologize to them), people don't use the cranking guide cause they enjoy bankrupting AIs, or they only have 20 minutes a day to play. They use it to get a better score, period. So I'd call it cheese. But you know what? My definition of cheese is pretty harsh. "Anything you do to maximize your metaverse score." So if I'm playing a maso game and see fifty frigates and transports approaching my last defenceless planet, I don't bother playing the game out to the end and submitting to reduce my metaverse score. I don't see many losses submitted around here, so I guess we're all a bit cheesy. Admit it. (Well, nearly all - I'm sure some people submit every loss and play every game through to the end - kudos to them.)

This isn't a problem with the difficulty - beginner is supposed to be easy to win, that's the point, and this way is just as easy as the "by the book" methods. If the the devs want to change anything, it should be the scoring, but that'll just open Pandora's box again...
whatever happens, any recalc shouldn't be retrospective IMHO.

On a side note, about the scoring system: I won't get into the actual algorithms used for working out game score and player score. But how about simply not counting losses in the score? How about an option to have them show up on the record or be hidden? If the devs really want all the saved games they can get their hands on, this would let them see all the games where the AI kicked major butt.

BTW, the suggestion that Stardock should focus development resources to fight a never ending war against cheese tactics is absurd. If someone came up with a trainer, is it our job to keep patching GalCiv to foil trainers? No. Instead we can police the metaverse.

We're not about to do that. Especially when we can much more easily state that the principle of the metaverse is to balance dedication and skill. It will never be perfect. 

If you don't like how we run the Metaverse, there is nothing forcing anyone to use it.

If we find that the metaverse rankings don't make sense, we'll adjust the scores retroactively by changing the formulas on scoring. Each submitted game contains the data that was used to calculate the score.

Restrospectively using ver 1.05 I'm so glad I did this experiment . . . I think here's the crux of the problem pure and simple IMHO. Sorry I could never quite get a Maso to add to the List but this has gotta say something


8/3/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 10960
Difficulty: Cakewalk Win / Lost: Win: Military Conquest


8/4/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 10910
Difficulty: Easy Win / Lost: Win: Military Conquest


8/5/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 11300
Difficulty: Simple Win / Lost: Win: Military Conquest


8/6/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 11880
Difficulty: Beginner Win / Lost: Win: Military Conquest


8/7/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 12020
Difficulty: Normal Win / Lost: Win: Military Conquest


8/8/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 11310
Difficulty: Challenging Win / Lost: Win: Military Conquest


8/8/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 11850
Difficulty: Tough Win / Lost: Win: Military Conquest


8/9/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 12580
Difficulty: Painful Win / Lost: Win: Military Conquest


8/10/2003 (1.050000) Morality: Pure Good Galaxy Size: Large 12650
Difficulty: Crippling Win / Lost: Win: Military Conquest


Note - I got all wonders and Trade goods in ALL these games except on Crippling where I missed the Galactic Stock Exchange and that Historical Preserve . . . ?

So wheres the incentive not to bend the rules? I play painful mainly and sometimes crippling when I'm zoned in but as you can see it would make little differnce considering the current scoring system as it is.

To be blunt I haven't brought into the beta and don't plan to until I can see it rewards different difficulty levels appropreately . . . :notsure:

What we'll likely do is after GalCiv 1.1 comes out scores at normal difficulty will be that score to the 1.0 power. Scores below normal (liek beginner) will be taken to the .9 power and scores above normal will probably be shifted to the 1.05 power.

Tomorrow's beta will have a much more significant difference in scoring between the difficulty levels.

That make sense to me. And if you might consider same thing for the map sizes as well? Getting 2.5 times score points in gigantic map is way too much bonus. If playing gigantic map takes weeks to play, then I wouldn't complain, but some people can play two gigantic maps a day. My point is what is so dedicated about playing gigantic game when you can beat it in one day?

None of the gigantic game subssions I've seen have finsihed up in a single day.

Heck, it takes hours just to move a ship to the other side of the galaxy.

Gigantic in GalCiv 1.1 will also be 28x28 sectors (that's hundreds of more sectors).

Allow me to suggest that a spread of 0.9 to 1.05 is too limited as compared with the actual difficulty in play involved. Perhaps 0.6 to 1.4? or even higher...

Wait, wait. Will there still be an old gigantic size choice?

Check again Frogboy, I have submitted a few in my early days of playing, where I submitted a gigantic in 6 hours.

quote Frogboy -
- - - - - - - - - - - - - - - - - - - - - - - - - - -

Heck, it takes hours just to move a ship to the other side of the galaxy.

Gigantic in GalCiv 1.1 will also be 28x28 sectors (that's hundreds of more sectors).

- - - - - - - - - - - - - - - - - - - - - - - - - -

Hehe . . . me two . Any tweaks are welcome and appreciated (I've got my Credit Card in my hot little hand . . . !)

Wow! 28 x 28! You'll have to rename it from gigantic to 'realistic!' or mabye 'Leviathan' . . . a little joke there . . . hehe.

Thanks Frogboy

Almost hate to ask this. Does the scoring take into account the differences within each level? For example, my current game has all the AIs set at normal, except the Yor set at intelligent. I have a difficult game. Which is probably easier than setting all the AIs to bright, which also is a difficult game.

couslee,

I understand that the scoring is based on the individual intelligence level for each of the AIs.

~SDC~

phew seems to be winding down....

Beware . . . tis' just the eye of the storm.

Actually that would be me--ten games less than you, Pontius I haven't played in a couple weeks but I'll be back.

I have thoughts on this but I'll hold off until we see how the new beta goes, and I suggest that others do the same.

The thing here is that this is a valid strategy if you can do it - but the AIs need to also apply a little common sense and consider their cash flow, and also to disregard gifts in determining how much they'll spend per turn - one enhancement might be to allow a "mixed" cash deal - so much now and a different amount per turn (the latter based on their anticipated cashflow), or alternatively that they only pay what they can rather than going massively into debt.

My style is (was?) to go for as large a treasury as I could get in order to concentrate on research - as that seems gives the biggest dividend for score.

I don't particularly enjoy the "pure combat" side of the game personally - yet was being seriously "downgraded" in scores for alliance and culture victories - ie for using diplomacy rather than brute force. (I'd like to see the facility for diplomacy and espionage developed for GC2!)



This is also a testbed where people who are interested in the strategies can develop them, and then start applying them further up the system, or for testing mods. As an aside, it may be worth making clear what the criteria chosen for the score checking are so that mod builders and designers can design good mods that don't unbalance the system - at present, it is too nebulous to be safe to add a mod in.

It has been suggested that some sort of "diminishing returns" might be applied to the metaverse scoring system for games less than normal difficulty - ie where the scores of the first (or perhaps best!) n games at a particular level get full score and they are then progressively weighted downward. To me, that approach makes a lot of sense if you don't like the tactic and you want to encourage people to move up the difficulty ladder. This could be extended to bonuses for the first few games at levels above normal.

As I've said elsewhere, my main concern was (and is) to see about resolving what appeared to be a problem. It's extremely frustrating to spend 2-3 hours working a game through to have the score rejected - and if it was happening to me, it was happening to other people.



Thanks for that - it's appreciated - although I'm well over the initial annoyance now.

Alex


~SDC~

Tried playing a small beginner game last night. Don't have the patience to wade through the tech tree to get a high score, so I had to wipe out the other cultures and move on LOL

Regarding the 'bankrupting' discussion above - it seems to me that it is an AI issue rather than a Metaverse issue. Maybe it is time for Alien Econ 101.

I my opinion, the Metaverse and forums are one of the best features of the game. I think that the Metaverse should only accept unmodded games (so that it is a level playing field) and that games with a higher difficulty level should score more points than easy games. It's weird because some of my hardest games have generated pretty crappy scores (stuck in a corner bribing aliens into wars left and right while building an economy to support Dreadnaughts).

I don't think I like the idea that playing an unmodded version of the game can generate an anomolous score. Games with 'inappropriate' scores should be analyzed and used to figure out how to make the AIs smarter rather than being ignored. I know I have had to readjust my thinking in the past when the AIs got smarter about Trade Goods.

Thanks for the offers to join empires, I think the wonderful people of Krikket said something similar just before I wiped them out heh heh heh. Feel free to join me in my quest to be one of the top 10 Empires.

And finally, because it can not be said enough, thank you Brad, Cari, et al for creating a very entertaining game.

I have to go with Frogboy on several points here. First, and most importantly, the AI should be stupid on beginner level. Saying the AI shouldn't let itself go bankrupt strikes me as silly, since on intelligent and up, it won't (or at least not irrevocably so). I might drive the occasional minor into bankruptcy, but the majors certainly don't want to let me. Second, the .9 to 1.1 range sounds good to me. If you consider a 12K game (as in Cotal's example), this would be a difference between 4690 points and 30,698, which sounds eminently reasonable. Expanding the range even more would be ok, but this is fine. Personally, I'd like a finer gradation (maso should beat crippling by a good bit, for example). Finally, I want to add my voice to the idea that playing on beginner should be for, well, beginners. Minimizing the challenge of the game shouldn't be a dominant way of playing.

I'm just saying that on all levels the AI probably should be a little smarter about not getting into situations where it is accumulating massive debt.