GOTM Difficulty Levels

[Cactus Pete]

EEO . . .

Why did you choose to use best scores in your formula? If you used average scores, you would utilize a great deal more information.

I do like your concept, but I would hope we could add to it in order to reduce the need for milking for all victory conditions (which was the original issue up for discussion).

 

[Obormot]

Then to be competitive you'll need to do exactly the same as you need now AND win by culture. The same old stuff, only even more time consuming. If you really want to make non-milked game competitive, you need to redesign the scoring system completely, using totally different formulas for different VCs. It will be extremely difficult to balance it, and even if you succeed at this, I think that the epthatlon and fast finishes will still be more interesting. But that is just me, so you can try.

I can suggest how to construct such a scoring system if anybody is interested, but balancing it will require a lot of work, and it will still not work perfectly. (i.e. in some games it will be shifted in favour of one VC, and in onther games in favour of a different VC).

 

[EEO]

Why did you choose to use best scores in your formula? If you used average scores, you would utilize a great deal more information.

... I had no time to do better. The point was not to find out the right numbers, but to express a simple idea that makes the choice of the best victory condition more difficult (for max score). In fact, for a given game, it is clear that an intermediary value (let's say between 1 and 10) exists for each multipler in such a way that the choice is not obvious.

I do like your concept, but I would hope we could add to it in order to reduce the need for milking for all victory conditions (which was the original issue up for discussion).

Certainly ... We are far from the end. I see (personally) the priorities for a new formula as follows:

1) Give each victory condition an equal chance to go for max score (at least for "standard" settings)
2) Reduce the need for crazy milking
3) Reward better a "civilized" play

If we accept an empirical approach, -and thus improving progressively the formula- 1) can progress quickly.

However, 2) could be more complex, because we have to foresee the subtle consequences of our changes. So, by addressing one way of milking, we can open the door to another way ...
A radical way to obtain it, is to favor fastest finish above all other considerations. This is already done by DaveMcW.
Another approach is to reduce the importance of population in score formula, because the current milking technique is based on it. Remove completly the population from the formula as DynamicSpirit suggests is interesting in terms of real-world matching. In terms of gameplay, I fear it will lead to simply favor fastest finish again, but we have to test it ... For the moment, I am inclined toward a solution that reduces the importance of the population factor in order to avoid milking , but keeps it to counter-balance extreme techniques used for fastest finish.

Finally, my point 3) is probably the easiest technically, but the hardest in terms of possible consensus. The success of civilization has probably much to do with the warmonger aspect of the game. It is certainly true for the one million people that bought the game (first expansion being "Warlords" is a sign ...) . I guess it's also partially true for the one thousand that play GOTMs (a poll would be interesting). Therefore, proposing more reward for a civilized gameplay could be interpreted as favoring a boring gameplay.

Well, maybe are we going too far in this thread. Do you think there is interest among the players for a new thread about civ score ? Is GOTM staff interested by such a question or is it too early ? Is some people already working on it and far beyond these considerations ?

 

[Obormot]

1) Give each victory condition an equal chance to go for max score (at least for "standard" settings)
2) Reduce the need for crazy milking

First of all we need to agree, what should be rewarded for each victory condition, select appropriate variables vi and construct a base score function B(v1, v2, ... vn) for each VC. This is a relatively easy thing to do.

The difficult part is comparing different VCs and comparing milked games to fast games. For that purpose we need to construct benchmark functions B0(t) which would represent how B(v(t)) increases as the time goes in a "perfect" game with current settings. It should depend on the type of the map, difficulty level, leader traits, UU, climate settings, AI rivals, and the quality of land near the start, and maybe something else that I forgot. This makes balancing incredibly difficult.

Then the final score at victory date t* would be S(t*,v) = const*B(v)/B0(t*).

3) Reward better a "civilized" play

That can be done by selecting appropriate vi. However I am strongly against doing this, especially in military victories.

 

[Shillen]

How did a difficulty level thread turn into a scoring thread?

 

[Cactus Pete]

You're right, Shillen, we've gotten a little carried away.

I like EEO's suggestion to see if the majority of active gotm players are interested in changing the scoring system to equalize for victory conditions and minimize milking. If so, then moderator action and presumably a new thread would be appropriate.

 

[DynamicSpirit]

I have now started a separate thread (and a poll) for this discussion, at http://forums.civfanatics.com/showthread.php?t=186101

 

[Cactus Pete]

How does someone not reading this thread find the poll?

 

[DynamicSpirit]

How does someone not reading this thread find the poll?

It's listed as one of the threads in the main GOTM forum. It's some way down the list because there's currently quite a few sticky threads.

 

[AlanH]

How does someone not reading this thread find the poll?

If you subscribe to this forum (Forum Tools menu at the top of the forum listing) you can get an email each time a new thread is posted here, or if you prefer not to have email notification you can spot from its icon color in your usercp that there are unread posts in it.

If you have an RSS news feed reader you can also subscribe it to this forum and receive notifications of new threads being posted. The RSS link is http://forums.civfanatics.com/external.php?forumids=165&type=rss2.

 

[Piscator]

Let's consider the numbers for GOTM8 (no time to go further):

f(domination)= 1.00
f(diplo)= 1.57
f(conquest)= 1.60
f(space)= 3.29
f(culture)= 6.57

Using a modifier for different victory conditions cannot work. It opens up exploits and more opportunities for score-milking (instead of reducing). For example in GOTM3, I had a cultural victory with 120K score (also 60% territory). This would give 780k score...

My point. Even if you change those modifiers people will play towards that victory condition only, making the game less versatile.

Of course, the current system is far from perfect. However, don't forget that if you want to milk a game, you have to own it first. And that's what the best players do anyway. So even if you change the scoring system, I wouldn't expect much of a landslide in how players perform, but just removing some sharp edges it has now.

If you want to balance scoring, the only way to me seems a scoring-mod. And that just seems a bit unprobable, but always good to think about. I'll put some ideas in the other thread.

 

[Déja]

If you want to balance scoring, the only way to me seems a scoring-mod. And that just seems a bit unprobable, but always good to think about. I'll put some ideas in the other thread.

I'm sure I'm not alone in both playing GotMs and working on mods for Civ. Besides, putting a GotM score in alongside the CivIV score would be extremely trivial. The hard part would be figuring out an adequate algorithm to calculate that score.

 

[Alraun]

A Warlord level game would be fine if it were made sufficiently difficult. Say putting us on a small island with crappy resources over the ocean away from contacting everyone else who are on a continent all together.

 

[Déja]

A Warlord level game would be fine if it were made sufficiently difficult. Say putting us on a small island with crappy resources over the ocean away from contacting everyone else who are on a continent all together.

What's the point, then? Why have an easy difficulty level (beneficial to the less experienced players) and make it artificially harder (non-beneficial to the less experienced players)?

I think that I would have to agree with those who have spoken before me that Warlord plays too differently than Noble+. It wouldn't be beneficial to anyone to play on such a lower level. (I'm not claiming to be very good, myself, as I have only recently been able to win more than 50% of my starts on Noble)

 

[Piscator]

I'm sure I'm not alone in both playing GotMs and working on mods for Civ. Besides, putting a GotM score in alongside the CivIV score would be extremely trivial. The hard part would be figuring out an adequate algorithm to calculate that score.

Of course, I know what the relative simple part is and what the hard part would be...

I mean unprobable because
1) The reluctance to make mods compulsary.
2) To get a scoring system that would work better.
3) To get a general acceptance/consensus of another scoring system.

 

[Déja]

Of course, I know what the relative simple part is and what the hard part would be...

I mean unprobable because
1) The reluctance to make mods compulsary.
2) To get a scoring system that would work better.
3) To get a general acceptance/consensus of another scoring system.

As far as mod compulsion, I doubt people would mind. Such a mod would be a couple kilobytes to download, and if it's done right, it would be trivial to install. Opening the save game automatically triggers the mod, etc... Besides, take a look at the success of HoF.

 

[Piscator]

As far as mod compulsion, I doubt people would mind. Such a mod would be a couple kilobytes to download, and if it's done right, it would be trivial to install. Opening the save game automatically triggers the mod, etc... Besides, take a look at the success of HoF.

That it's kilobytes to install is not much of a point. And I'm pretty sure that requiring a mod makes the step in participating in a gotm a bit higher for new players and the less computer savy.

The HOF mod caused quite some discussion prior to use, even optional use or not. While this mod isn't supposed to change the game, it still does (whip issues). How many people did use the mod in the last gotms? I can't seem to find some definite numbers about that.

 

[DynamicSpirit]

That it's kilobytes to install is not much of a point. And I'm pretty sure that requiring a mod makes the step in participating in a gotm a bit higher for new players and the less computer savy.

The HOF mod caused quite some discussion prior to use, even optional use or not. While this mod isn't supposed to change the game, it still does (whip issues). How many people did use the mod in the last gotms? I can't seem to find some definite numbers about that.

No one used the HOF mod prior to GOTM10 because it wasn't supported. GOTM10 was the first Civ 4 in which using HOF became a possibility. And GOTM10 is still running so - obviously - no figures on how many have used it yet.

And people already have to install the 1.61 patch before they can play the GOTM - it doesn't look to me like installing the mod is any harder than that.

 

[Alraun]

What's the point, then?

Uhhhhhh.... variety, the same as everything else?

 

[AlanH]

Moderator Action: Please focus the scoring debate in the thread started by DynamicSpirit for that purpose. This thread is about difficulty levels.