Metaverse Scoring

I think I've written several theses worth of text on this topic in the past, but I did a site search from google and couldn't find anything of interest so I guess I have to try and replicate it.

First off no one is going to be able to answer this question to the level of detail that you ask. Precisely how you got 30K as opposed to 20K or 40K is unknowable. What is knowable is given that you got one score in one game what can you do in the next game to better that score.

Basically you score points for your population (society), your economy (income), your military might and your technology. These items are directly shown in the first four curves displayed under the timeline tab in your civilization manager. Basically the value of each of these four curves is divided by some function of the turn number and then accumulated and divided by the turn number again. This is done each turn and at the end of the game that is your total score.

The main point is that score is divided by the turn number so this causes early values to be worth more than later values. The other main point is that since these values are summed up turn by turn what you have is essentially a front end weighted intergral. In other words the score is proportional to the area under the curve. The consequence of this is that it doesn't do much good to have a high level of income for example for just a few turns. To have a noticeable effect you need high values of these curves held for long periods of time.

So to get a higher score all you need to do is get these curves higher earlier. Poplulation and income is obvious. Mititary is your military might that is calculated by sum of attack + defense + HP/10 of all ships. What ships you kill have nothing to do with it. Technology is simply the number of RP's spent and it doesn't matter once the tech tree is done RP's spent beyond this point continue to add to score.

A couple of quick points and that is that Technology is never a very big contribution to your score and any effort that you spend on maximizing your tech spending is probably a waste of time from a scoring point of view. Population is an important portion of your score. Nowadays with the morale limitations it's hard to get more than 13B to 20B per planet but failing to get these levels leveas points on the table.

The quickest and easiest source of big points for the inexperienced player is your military score where you build an array of 24 military starbases fill them with modules giving a 64 point bonus per ship per starbase (i.e. 1536 point per ship bonus total) and then build thousands of 1/1 attack fighters to put under the starbase array. This is not everyone's cup of tea but if your trying for top points it's a necessity.

Beyond that it's a matter of getting things done better and faster and how this is done is far more than can be simply discussed in a single thread.

Anyway that's pretty much the basics.

Interesting. You must have an engineering or math degree I'm glad I've taken calculus so I know what you're talking about.

In regard to the points for income, it's measured by total revenue not net income correct?

Also, with military, I'm guessing ships I build that are destroyed do not generate any points for me. Or is this also the integral area under the curve where these ships contribute to score for the length of their existence?

Thanks for the info

They give points while they are alive, as far as I can tell. Once they are destroyed, no more points.

This is the only part I can argue with. I assure you, someone(s) at Stardock know exactly what formula is used to calculate a score. So it is not entirely unknowable. It is, however, unknowable to us, as Stardock has decided this should not be public information. I can't say I really blame them, seeing the extremes some people will go through to maximize score - myself included.

Other than that, what Mumble has to say is as close to the actual score process as we'll ever know.

Correct. If you make one military ship, you will see your military curve jump up from 0. If that remains your only military ship, and it is destroyed, then your military curve drops back to 0. The integral under the curve is what matters (well, the integral under the curve divided by the amount of weeks that have gone by, or something like that; otherwise, just extending the game out longer and longer would allow people to boost their scores, regardless of whether or not their society improved).

With regard to Income, you are again correct: it is your revenue, not profit, that matters. Whether you're spending all your revenue each turn or you're spending 5 times your revenue per turn doesn't matter, your Econ score would be the same.

That's pretty much what I meant. I knew the information existed, I just had no idea if there was a wiki I couldn't find or something. And yes I do agree the exact formula shouldn't be divulged to maintain some fairness. And hot damn is calculus cool. Mumble, how did you come to the realization about the score being integral-related? Did you read it, or figure it out due to a sizeable command of math knowledge?

What I'm curious about too, but I don't think this is the case for Galciv2, is score is scaled based on galaxy size. This was the case in master of orion 2, but I don't think it is with galciv2.

I'm fairly certain that increased scores due to larger galaxies stems from having larger militaries, populations, and revenue due to the increased amount of planets.



Cool thanks

Yes total income not net.

Yes area under the curve for military just like everything else.

And yes someone at Stardock must know. In fact it could very well be simple with no constants or other complicating factors. Certainly there are units in which these value are measured which in effect are essentially constants in the equation but comment made by various Stardockians have implied that it's very straightforward.

By that I mean it seems to me that the units of population are in trillions whereas income, military and tech appear to be in millions. One thing that correlates that is the "wrap" that occurs in the graphs of population at around the 6.4 trillion level. A similar "wrap" occurs in your military rating at about 6 million and then again at 12 million.

This implies that the score you would get from an eventual population of 6 trillion should be about the same as the score you would get from an eventual military rating of 6 million assuming that the shape and start of the curves are similar.

This also points out why tech is so insignificant. With a whole lot of optimization I've been able to get tech production into the 200K range but that pales in comparison to the 6 million range of military and the equivalent 6 trillion range for population. Income also can never get close to the 6 million range but I usually get it into the 1.2 million bc range (in a abundant all gigantic game) and that does provide a noticeable contribution to the score.

You're right I am an engineer not a mathematician. But there's another interesting thing going on that I have a gut understanding of and have difficulty explaining but I'll give it a shot anyway.

I mentioned that each turn the values of the four components of score get divided by the turn number and the sum of these four components get divided by the turn number again and accumulated along with every turn’s contribution to score. The net effect of this is that each turn's contribution to total score is divided by a function of turn number squared.

Most folks would assume that this should cause scores to eventually start declining at some point in time but in fact they don't. One reason for this is that although each turns contribution is divided by the turn number squared, the net effect of summing all the individual turns is the same as multiplying by the number of turns and that in effect makes the total score inversely proportional to the number of turns *not* the number of turns squared. This actually has an important effect that required a major scoring change a couple of years ago.

But more simply the score never goes down because to do so the contributions of each new turn would have to go negative. It is true that because each turn's contribution to the total score is divided by the turn number that each new turn's contribution to total score gets smaller and smaller. But as long as there is dramatic growth in at least one of your curves it's useful to continue the game. However once all of your curves asymptotically stop growing there's not any point in continuing.

But going back to the situation where each turn’s contribution is divided by the turn number squared. This wasn’t always the case. Initially each turn’s contribution was only divided by turn number and it was thought that this was sufficient to keep an upper bound on score. However this ignored the fact the summing each individual turns score contribution was effectively the same as multiplying the average turn score by the total number of turns. The bottom line of this is that before the “extra” division by the turn number your score would simply never stop growing. This was demonstrated by someone that submitted a 210 *year* 4 million point game. This was before a lot of high score techniques like the military starbase array, among others, were developed. Anyway this caused a big fuss and the scoring change and recalculation of previous scores.

I include the above for completeness only. Generally this is nothing that someone new is interested in but at some point this understanding of how scoring works does in fact help those that are interested in achieving the highest scores possible, do so.

The first few times I explain this the typical response I get is, Huh? But I have had folks say they finally did get what it was I was trying to explain after about the 3rd or 4th time and that it did begin to make sense. If I was a mathematician then I'm sure I could explain it more rigorously but I’m not sure that would help all that much unless the reader was a mathematician as well.

Perhaps DA or Neilo could chime in a bit here as well because I think both have finally "got" what I'm talking about here and perhaps they can interpret.

BTW this is what I meant by writing a thesis. This must be the 4th or 5th time I've written pretty much this same explanation.

As someone who thinks playing a game on large is a big deal, I find having values that high mind-blowing. But I guess it makes sense on gigantic abundant galaxies. Just to see what happened, I boosted my population one game to hit the 1 trillion mark and that was a big deal to me hehe.

I also think considering how much money and effort goes into getting research, it should be weighted a bit more.



Actually, I think you explain it better because you're NOT a mathematician. I graduated from a university that had tons of engineering students. I struggled with math until I realized I could do well in it with hard work, and I learned calculus from an electrical engineer from India. I taught him how to drink, he taught me that lol. Tangents aside though, I get what you're saying.

The gist of the newer division by a squared denominator equates to diminishing returns with what you have vs how long you have it, like a limit approaching zero, which can then drop to the axis and become zero.

Sort of. But once your values stop growing you would have diminishing returns whether the denominator was either turn number squared or simply turn number. It's just that if the total effect was turn number squared it would diminish much quicker. But this isn't really the point.

Actually this whole thing about whether total score is inversely proportional to turn number or turn number squared would be totally academic except for one thing, and this is really how you can tell whether you're a good player or not.

When you look at your final score displayed during the end game summary you see a value for each of the four components of score (pop, income, military and tech) and you see the total score.

If you add up the four separate components of score they never add up to the total score. In some cases the sum of the four components of score will be far less than your reported total score in other cases the sum of the four components of score will be far greater than your reported total score.

What I think is going on is this. The individual running totals for each of the four components are calculated as they always were where each turns value divided by the turn number is added to the accumulating total for that component. My guess is that when they added the additional divide by turn number they implemented it at the point where each of the four running totals are summed into the accumulating total score.

Anyway this explains why the sum of the four components of score does not match the total score. But most importantly whether the sum of the four components of score is greater or less than the total score speaks volumes about your game. It basically gives you a measure of how successful you are taking advantage of the "front end weighted" nature of the calculation.

Basically the smaller the 4 components of score with respect to the total score the more this indicates that you are able to get high values on your curves early in the game. That's really where the big differences in peoples games lie.

Anyone can put up a military starbase array and build a bunch of ships to put under it, and failing to do so will cause your score to be less than someone that does. However getting your income, pop, military up earlier than the next guy is *really* where the major scoring differences lie.

When I was starting out but had been playing for awhile, I was disappointed at the scores I achieved compared to those of the experts. I had been around awhile and I knew I was doing the same things in my game that they were doing in their's. So why were they significantly outscoring me. The secret is that although they did exactly the same things that I did they were doing them sooner and faster in the game. That is they started doing them earlier in the game than I did and the rate at which they increased their values was faster than mine. So even though we'd both end up with the same total value of income or pop or military, they had much more "area under the curve".

This is the true secret of high score. Most people don't want to hear it. They like to think it's as simple as knowing a couple of "tricks" and while knowing a few "tricks" is essential it's really only the tip of the iceberg. Developing the knack of increasing your pop and income early in the game, of being able to conquer the AI sooner, etc is what's important and mostly this takes a long time of hard work improving your game in small steps and it's virtually impossible for someone to be able to simply just tell you how to do this.

Many thanks for the detailed explanation, Mumblefratz.

Sure, no problem. Hopefully it's helpful to at least some folks.

I just wanted to confirm that the technology score can pretty much be ignored.

I just finished my first game where I got over 1 million points. My technology score was only 20,000

When I started the game I did not turn off Technology Victory. I didn't realize this until I had finished the researching all of the Tech trees. At that point I had to stop researching completely.

So around year 3 my Tech spending went to 0 and my technology score never increased.

My military score was something like 675,000 so I was still able to clear 1 million points.

Interestingly, at the end of the game I had the option to end as either a Tech Victory or as a Military Conquest.

If I went with Military Conquest I would get over 1 million points.

If I went with Tech Victory my score was something like 350,000 points.

This was true for many stages of the game, the Tech Victory score seems to be capped. I am not sure if that is a hard cap, or if it was based on when I reached the final tech.

In any event, if you have the patience the tech score can be ignored.

- Livonya

Congratulations on being the first to achieve a 1 million point game. AFAIK you're the first to do so. I've come close (957K) but that was with DL. I know Magnumaniac has a 975K DA game which is now 2nd.

I'm still playing DL for a couple reasons one of which is that I'm still convinced I can reach 1 million points in DL. Anyway I think I have a good shot with my current game. I do have no doubt that I'll easily be able to get over 1 million once I do move on to DA though.

However the point of this reply is not just chest pounding. Primarily it's about the relationship between the various victory condition types.

I had a smaller game awhile back that I was able to win on about the same turn (i.e within two turns) in all four of the different victory conditions. The reported game length in all cases was 7 years so a two turn difference isn't very large, but it was a small galaxy game which makes the scale somewhat different.

Anyway based on that game I got the following ratios of scores between the various victory conditions.

Beyond Mortality = 1.0
Political Victory = 1.2
Cultural Conquest = 1.4
Military Conquest = 2.0

This data is detailed in the Score dependency on victory condition thread.

Basically I found that if you could do things pretty much the same a military conquest was worth twice that of a tech win.

Your results suggest closer to a 3 to 1 ratio although you mention a hard cap on the Tech win which at the score levels I was looking at I would never see.

I believe that the tech win score really has nothing magical to do with the Tech component of your total score. Basically I am pretty well convinced that the 4 components of score (income, population, military and tech) are treated identically in all 4 different victory conditions other than the overal multiplier to score that applies to the different victroy condition.

However an overall cap on the score of a tech win could be present and would account for the difference because based on everything else I've seen, even though your tech spending stopped your total score sould still increase due to increasing military even in a tech win.

So how many ships did you end up building and did you do a final upgrade of your 1/1 fighters to BHE/ZPA dreadnoughts?

Congrats on the record score Livonya!  :CONGRAT:  I almost didnt believe it when I saw it.

Sentient species taste better...

Mumblefratz -

When I was playing the game I would end occasionally to see where I was at score wise. At around year 4 if I went with Beyond Mortality then I would get about 350,000 and it never went up again. I was shocked by that. Military Conquest continued to grow, but I saw no growth at all with Beyond Mortality for the rest of the game.

There may be a hard cap, or it might just cap at some point based on when you start to research the Beyond Mortality line. I am not sure.

My final ship count was well over 50,000 ships.

I never did any sort of upgrade with my ships. I am not sure if the system could have done that. When you have 50,000 ships in one place it is nearly impossible to manipulate them without causing something akin to a game freeze.

I was producing the most powerful ship I could on each plant in the shortest amount of time, and then moving them to my military star base cluster (painfully slow and boring).

Since I couldn't research I was 100% military spending and my biggest planet was putting out something like 2,750 BC per turn and producing 1 ship per turn.

I always track all factors at the end of year 7 to compare games, and then play out the 8th year as fast I can... trying not to pull out my hair with impatience.

I will post my end year stats when I get home. It is interesting.

I think it is possible and faster to try to go research. In my next game I will see how close I can get to 1 million using research rather than military. I think I can get close, and it will be a LOT faster of a game.

The key of course is winning the game very, very fast. I had all but 1 planet under my control by the end of year 4, and if I could speed that up then I could get a higher score.

I also accidentally lowered my score because I was counting on doing research, and so I saved the Neutrality Learning Centers that I got from caputured planets, which was pointless once I had to stop doing research. If I would have turned them into stock markets immediately then my econ score would have been higher.

- Livonya

The answer to that one is yes, and yes   

I'll start by saying nice game  :CONGRAT:  I'll continue with a   :HOT:   for patience (50,000 ships  :SURPRISED:  :SURPRISED: )

Then I'll tell you that in the 975K DA game I managed, I had less than 1500 ships in total and military was the smallest score component. My boredom threshold is obviously much lower than yours   



Yes you could, with that level of patience - end of 2nd year, early in 3rd year is the target I aim for. The only thing that slows it down is the AIs ability to keep up the colonisation pace!

Did I say  :CONGRAT:  for getting the first legitimate million point game?

Actually, that wasn't correct. I just looked. I conquered the universe in August of 2228, which is much sooner than I stated.

I should have done it by the end of the 2nd year, but I spent way too much energy boosting my initial tech spending and trading. I should have just put everything into violence and destruction.

The next game is going to be a blood bath asap.

- Livonya

If you were willing to build 17K ships and upgrade then you'd be hitting 1.5M easy.   

See my reply in your Notes from a 1 million point game thread.

You could do better by a number of things I point out in that reply. In particular it sounds like you seriously shortchanged your income by having too much industry on your planets and certainly any reseach buildings, even on precursor libraries, are a waste as well. Basically in DL the optimum approach is 100% stockmarkets on every planet, no industry, no research and starports only on the 300 or so planets from which you rush buy a 1/1 fighter each and every turn.

In DA the economics of engines in particular makes this less of the obvious choice but at most I would warrent you want no more industry on any one planet then it takes to build a bare huge hull per turn (the rest being stockmarkets) and then use your cash to do a more cost efficient upgrade to your 1/1 fighter instead of a rush buy of the entire ship.

Generally controlling pop at 13B on PQ10 and below and 20B on PQ11+ is reasonable although for DA you might want to increase to PQ13+ for 20B unless a food or approval bonus tile is present.

Again see more detail in my reply to your Notes from a 1 million point game thread.

I'd be more likely to be hitting my head on the keyboard   



1 turn of doing that would put me off this game altogether - talk about RSI!

There are ways around that as well. This is a subject that's soon to come up for discussion in the ToE forums.

Awesome effort in breaking 1 million points Livonya! There should be a unique achievement medal for your efforts!
:CONGRAT:

Congrats Livonya.

Certainly congrats are in order.

BTW I see you're still unaffiliated. I don't know how long you're going to stick around this time but regardless you may as well join the Tyranny of Evil and help us defeat those furballs.

I see we think all the same lines Mumble, i have extended an invitation to you Livonya to join the Tyranny of Evil and aid us in our quest to de claw the kitty cats....

Hope to see you in the ToE soon....

I sent a PM asking him to join the Geckodons.