Jason Scoring

[brennan]

When the staff calculate the best dates for spaceship, 100K, etc, they have to take into consideration the possibility that a player can reach the domination limit.

Finally, the idea presented elsewhere of giving score bonus for playing "peaceful" or "fair" games is absurd to me. You might as well argue to pay a reduced price for the game in the store because you're not going to use the game's full potential. However, I hope in future versions that the penalties for treacherous behaviour will be potentially much more severe than they are today. The AI should be able to recognize liars and scoundrels and gang up against us. Occasionally, trade embargos are a nuisance, but they seem to sign those against each other to the same extent.

Right, so lets take away the scoring benefit of reaching the domination limit no matter what the victory condition you are aiming for. Let's encourage people to use the game to its full potential by levelling the playing field for other victory conditions: If you play for a diplo win, warmongering behaviour should count against you. If you play for a culture win likewise. The way things stand at present, almost all high scores are generated by the same strategy, regardless of the final victory aim.

 

[MeteorPunch]

Right, so lets take away the scoring benefit of reaching the domination limit no matter what the victory condition you are aiming for. Let's encourage people to use the game to its full potential by levelling the playing field for other victory conditions: If you play for a diplo win, warmongering behaviour should count against you. If you play for a culture win likewise. The way things stand at present, almost all high scores are generated by the same strategy, regardless of the final victory aim.

You're proposing a different scoring system for a different game. In Civ 3, it helps with every victory condition to have as much land as possible, plus luxuries and happiness are super-important as well. Based on the design of the game, the scoring system is pretty great, and Jason is amazing.

I would expect Civ 4, if it has more variety of ways to win/dominate the game by diplomacy, culture, etc., then hopefully they will have a scoring system more to your liking. The fact that we have to face is that Civ 3 is mostly a war strategy game, and warring often, benefits most, and the score reflects that.

 

[brennan]

I would contend that the scoring system is over-simplistic and rather poor. It only rewards you for having a large empire. It takes no account of your Culture - which is supposed to be a crucial game mechanic. Just an example.

I agree that I hope for better things from Civ4. A better role for a navy, for example. All I am trying to push for here is that we find a way to score GOTMs that does not automatically make the Dominate-Milk strategy the best every time *yawn*.

 

[MeteorPunch]

Could you please elaborate a little on how you think scoring should be? Points for culture isn't a bad idea, but then people would find ways to get to the domination limit, and milk culture, and get the fastest space race victory.

 

[brennan]

Scoring would have to be based on your victory condition, rather than on a set baseline, since the baseline rewards milking under all VCs.

Non-military wins could award bonuses based upon the average AI opinion of you and penalise you for declaring war, capturing/razing cities etc, (In Civ1 you actually scored additional points for every turn you were at peace), specifically to prevent use of domination limit milking. The point being to encourage a second play style, similar to #CC play, to counterpoint the all-out warfare style that always seems to dominate the high-score tables. But 'normalized' so that it appears on the same tables. We would, of course, still have good old Histographic scores to compare with, since that is a valid VC.

It all boils down to a game balance issue: the bigger your civ the better. And the introduction of new victory conditions has not changed this fact. I find this unfortunate.

 

[Dianthus]

Scoring would have to be based on your victory condition, rather than on a set baseline, since the baseline rewards milking under all VCs.

...

But 'normalized' so that it appears on the same tables.

You've just hit the nail on the head with the "normalized" bit. That's the hard bit. The problem is that it doesn't just have to be normalized for the victory conditions but also based on other factors such as map type/size and civ traits. If the set of games were large enough each month for GOTM then this normalization could be automatic assuming that the best game for each victory condition was the best possible. That's not an assumption that can be made though, so this normalization is difficult. Any ideas for how to achieve this normalization?

 

[occam]

Hey, I read *some* of that thread you linked and nowhere did it seem to address the "standard deviation" issue in Jason scoring.... to wit: milking tends to look the same at the end scoring but slight differences in conquest date make huge scoring differences.

Am I crazy?

 

[solenoozerec]

If you play for a diplo win, warmongering behaviour should count against you.

Why? Do you think that. War is the most effective diplomacy.

If you play for a culture win likewise.

How can you achive 100K peacefuly?

The way things stand at present, almost all high scores are generated by the same strategy, regardless of the final victory aim.

High scores - yes. Not awards, though.

It takes no account of your Culture

It does. The more culture you have, the more territory you have, read - score.

But in general, I agree that Firaxis score is somewhat primitive score ad Jason score is a simple add-on to correct it a little bit.

However, I think the more proper place for this discussion should be in civ4 forum.
Maybe what we have in GOTM is not perfect, but it is certainly good and work for most people.

 

[leopalas]

Hi,

After reading all of the other thread something really stroke me. The maths, statistics and econometrics involved in trying to present a reasonable good score formula are beyond my understanding. So, i have a great respect for guys like Aeson that at least tried to get some logic out of the dubious Firaxis system. The thing is that the creation of this new system, because its based on the Firaxis system, implied that in the end, the absolutes would be the same. Territory and Population. Thats something that we cant escape from unless an entire new system was made from scratch. Megalou referred real life comom sense to say that happiness is not controversial as to access the merits of a civilization. I tottaly agree with him. A strategy game should try to approach real-life circunstances as well as it cans. What about territory? That is the main part of the score in the early part of the game thus having an exponential effect on the final score. Territory is definitely not a good measurement of a Civ achievements. Some of the great civs the world has known werent that big. Some eventually died because of its over-size. Even today apart from the US and Canada ( with a small population, i must say), im not seeing any large nations doing that well on overall development, as opposed to small countries like most of the European countries. The problem is that culture, and the ability a country has to "impose" its culture to other nations should in fact be a factor of development and how well a civ has done. Greece was the perfect example in the past ( low territory, low population). All of the western civilization is still based on their early achievments. Science is also an important factor. These are simply not contemplated in the Firaxis system. In a very ignorant way, i was just suggesting that some kind of extra-ponderation would be given to the games that present that edge, either in culture or science. Those games are precisely the ones that score low on the Firaxis system: 20K/100K/Diplo/Space. The argument that a good milking in those kinds of game will achieve the same results as fast Dom/Conq is a fact, but no one will do that kind of thing unless he is going for the Cow. So, this is a sylogism, and in the end the fact is that we have 2/3 of victory conditions that just arent pursued. Its an amputation to the spirit of the game and to comon-sense.
At least we could try to find a way to include extra points in the GPR for those who get the fast finish awards. Again im on shaky ground here as i dont know how it is calculated but this would be another option that i think most would agree with.

P.S. BTW Megalou, its not false modesty when i say i do not consider myself a top player. When i started playing GOTM i thought i was... Till i saw guys like SirPleb, Dynamic, Klarius and others. Wich doesnt mean im not improving...

 

[Gato Loco]

Isn't this whole thread a bit beside the pont? The only reason we care so much about the Jason score is because the list of results is ordered by Jason score. If it were ordered by time spent playing, we'd all be complaining about how the system's unfair to people who don't do OCC. You can still compare games by other things (e.g. the fastest 20k, the highest culture spaceship, or the most points per hour played.) if you want to. In short, Jason score is a way of ranking domination victories, and there's no reason to care about it if you aren't going for domination or semi-domination. If you wan't recognition for something else, just advertise your feats in the write-up.

 

[Dianthus]

If you wan't recognition for something else, just advertise your feats in the write-up.

Well said Gato Loco . One of the games that impressed me most in GOTM was one that scored just 334 jason points, but was a great game with a superb writeup. It was Zwingli's 2 city Conquest challenge from GOTM18. Just take a look at these: (well worth the read)
http://forums.civfanatics.com/showpost.php?p=900578&postcount=60
http://forums.civfanatics.com/showpost.php?p=907388&postcount=15
http://forums.civfanatics.com/showpost.php?p=907391&postcount=16
http://forums.civfanatics.com/showpost.php?p=935088&postcount=60
http://forums.civfanatics.com/showpost.php?p=938068&postcount=68
http://forums.civfanatics.com/showpost.php?p=946048&postcount=83

 

[Megalou]

Megalou referred real life comom sense to say that happiness is not controversial as to access the merits of a civilization. I tottaly agree with him. A strategy game should try to approach real-life circunstances as well as it cans. What about territory? That is the main part of the score in the early part of the game thus having an exponential effect on the final score. Territory is definitely not a good measurement of a Civ achievements...

You gave me some credit and I should give some to you because I agree that the mere acquisition of territory is somehow overrated. Well put!

BTW Megalou, its not false modesty when i say i do not consider myself a top player. When i started playing GOTM i thought i was... Till i saw guys like SirPleb, Dynamic, Klarius and others. Wich doesnt mean im not improving...

I don't think it was false modesty, but all the same: where the method is the most effective, the competition is the most stiff. As for the aestetics, domination is a bit sluggish, but conquest is the purest victory - forgive the cynicism - and I for one have been impressed by many more domination games than diplomatic games, for example. The first time I saw one of the top players you mentioned get Fission from the ToE wonder on a rather high difficulty level and win the game the same turn, I was truly impressed, but that was at least a year ago and I wonder if there is really as much variation in diplomatic games as there is in the military ones. And I say this because diplomatic victory is one of the victory conditions that is usually missing in the medal roll. I could be wrong, but there is something very predictable about amassing beakers (and culture points in your core cities) as opposed to deploying an army.

Edit: I take back the last part about variation. To be truly successful in the science department you need good military skills too, both in PTW where you need a second core and in C3C where you need scientists in corrupt (conquered) cities. I guess I was a bit "over-intimidated" by those posters who think that some victory conditions don't get enough credit. Like others have just pointed out, the credit is in the eye of the beholder.

 

[AlanH]

GOTM provides one form of player-player competition based on Civ3. There are others - SGOTM, PBEM, HoF. The Jason score used in GOTM reflects the scoring system as defined by the designers of the game. It tries to make it less sensitive to the victory condition chosen, but the basic scoring parameters are still the same.

The Jason scoring system leads to military victories taking most of the high score awards, mainly because it rewards rampant territorial growth and that's the usual prerequisite for a military victory. There's no compromise required for domination players and little required for conquest. For a high scoring later victory you have to maximise territory and population even though these are not prime drivers for your goal.

Asking a single scoring system to measure the achievements of people who want to play the game in different ways is a bit like asking for a way to compare runners, walkers, hurdlers, wheelchair athletes and cyclists in a single Olympic race. They could all use the same race track, they all rely on muscle power to get from the start to the finish, and they all measure their performance against the clock. But you can't create a scoring system to compare them all directly and absolutely.

Bragging rights are available for fastest finishers, and for anyone else who wants to declare a personal goal and set out to achieve it. If it's a popular goal then others will wish to emulate or better it, and a new competition will be born.

 

[MeteorPunch]

Well said, Alan.

I like the system now, but if any of you geniuses out there come up with another scoring system besides Jason and fastest finish, please propose it here. Something like an expanded QSC-type score. I think it might be a cool idea

 

[denyd]

Just a quick question for all the number crunchers out there.

Is it better to maintain a larger number of happy faces and finish your game 5 turns later or to pop-rush / draft citizens to accelerate your victory, yet have a lower score due to the lessened number of happy/total faces?

 

[AlanH]

That sounds like another of those string-measuring questions. Winning faster improves your score only if you are on the right part of the Jason curve.

However, given that the happy faces are averaged over the whole game and you are only talking about a small reduction for a few turns, I doubt if the loss of score would be significant wherever you are on the curve. Whereas winning faster *can* have a big effect on score if the date is right.

 

[Xerol]

Well said, Alan.

I like the system now, but if any of you geniuses out there come up with another scoring system besides Jason and fastest finish, please propose it here. Something like an expanded QSC-type score. I think it might be a cool idea

Read my idea above about weighted victory dates, population, and territory scores. I think I'll actually start a thread where people can post their GOTM43 'results' and I'll try to work out the right percentage weights for the game. The conquest and domination victories shouldn't move much, but 20k, Diplo, and Space should move a lot towards the higher end, but not for everyone obviously. Under this system, for a 20k, if you give up a few hundred years to expand to the domination limit, you'll actually lose points.

If anyone wants to go back and "pull up" their results for the last 3-4 GOTMs, I'd really appreciate it so I have some starting data to work out how this scoring system could work. Basically, all you need to do is load up your final save(the one you submitted), go to the F8 screen, and write down the numbers for Territory, Happy Citizens, Content/Specialist Citizens, and Finish Date. If you have any points for Future Tech(I doubt anyone does, but it could fit into it) write those down too. Compile as many as you can from the last year and PM them to me, I'll start a thread about it in a week or two after I get enough data compiled.