So what goes into this power graph anyway?

[Gedrin]

From what I can see it's not all the items in the score and also includes other items.

Seems like it totals things like:
City count * city_factor
Pop Count * pop_factor
Something for improvements
something for units... perhaps that power points thing.

Anyone know?

Gedrin

[quinns]

Good question, as the Powergraph is the main measure of success (win/loss) in the new CtP PBEM Rating system.

Does anybody know?

[Gedrin]

Eegads.... It's not exactly a good one.

Tests have shown that my score hits 40% of the total score in an 8 player game while my power is rather middle of the road.

Why do they not use the score page?

Gedrin

[Martin the Dane]

Gerdin:
The 40% is not of the total score in the game, as I normaly end up with several hundred percent, no mater what the graph shows. The score is a percentage of a number set in 'DiffDB.txt' to ilustrate I have copied the line for diety in the pack version.

BASE_SCORE	4600	# base score from which to generate the civ score

So the number displayed on the points page is the only relaiable number for a rating system.

[Martinov]

I would like to know myself - from observation the powergraph seems to measure the total "weight" of your units and buildings (measured in cogs). It noticably jumps on turns when you build a wonder/several units/buildings. I think it also weights in tech. I don't think it measures people, I have often had large pop empires with few buildings that graphed at half the size of smaller opponents with more buildings.

[Gedrin]

Martin:

Thanks but I think you misunderstood what I was talking about. When I said '40% of the total score'

I was not speaking of:
myScore/maxScore = .4

I was refering to:
myScore/SUM(all players scores... including mine) = .4

However I don't include
diff;opponents bonuses in my calculation.
In an 8 player game with all players equal this number should be 1/8 = .125
So if I have .4 then my enemies average .6/7 = .086 each... which means I am beating them by about 500%. That's a pretty dominating lead!

[quinns]

I thought Martin the Dane was referring to "x" when he mentioned the "several hundred percentage points higher". (I still don't know what this was in reference to, then.)

So, just so I understand what you're saying, the "percentage" that's given in the "rankings" screen is calculated by the formula: 100 x (your score)/(all players score including your score). Is that right?

You bring up a good point, and I just want to know if I'm off base with using the "Power Graph" to determine the PBEM CTP Ratings.

Regards.
[This message has been edited by quinns (edited November 02, 2000).]

[Gedrin]

No, wow...

I think we are talking about two seperate things here.
My statement was:
Tests have shown that my score hits 40% of the total score in an 8 player game while my power is rather middle of the road.

I admit I was a little vague. By total score I mean the sum of the scores that appear on the 'points page' for all players minus the diff;opponents size bonuses.

There is no way to get these values that I am aware of for the computer players except via SLIC. I only way I know know that my score represents 40% of this total because I have a SLIC trigger print a message when it occurs.

Now when that event occurs my power does not appear to be 40% of the values represented in the 'power graph' (which incidentally appears to be unrelated to the 'score').

Martin's referance was to the % given on the points page which is
myScore/maxScore
where maxScore is a fixed number depending on the difficulty level.

This value also has nothing to do with the power chart or what I was speaking of when I said 40% of the total.

So this % on the points page does nothing to tell you how one player did vs another but how they did against a standard...

Gedrin

[quinns]

Alright. Thanks a lot for clarifying.

Well, I guess we will just use the power graph for determining who is ahead at the ten turn mark (for PBEM ratings purposes). I can't see any better way of measuring, other than "power".

It seems, from your statements, that the "civilization" score is based upon all aspects of the game, while the "power" graph is just that -- a measure of a nation's military and economic strength as compared to the most powerful nation.

Thanks again.

[quinns]

Thanks for that explanation. Then, that should work fine with the ratings -- (please see the "PBEM Ladders" topic under CTP Multiplayer if you don't know what I'm talking about regarding PBEM ratings).

In general, regarding PBEM ratings, all we care about is which human players are ahead of which other human players. I think that we can pull all of that information from the Power Graph.

The only thing I'm still confused about is when you said, "...I usually have several hundred percentage points...". According to your formula: (my total)/(sum of all players' totals including my total) -- it would be impossible to have a percentage higher than 100% (or 1.0000). I must have missed something.

Thanks again and best regards.

[This message has been edited by quinns (edited November 02, 2000).]

[Gedrin]

WEll if
x = myScore/SUM(all players scores... including mine)

then x represents how much of the totalScore is mine... on average the others will have equal portions of what's left. IE

y = (1-x)/(#players -1)

[If you took y * SUM(of all scores) you would get the average of everyones score but x, ie me. ]

if you take x/y you get
(myScore/totalScore) / (aveEnemyScore/totalScore)
= myScore/aveEnemyScore

so while I never know myScore and enemyScores explicity it does not matter. I base all my score based triggers off score ratios.

Gedrin

[Martin the Dane]

OK, just so everyone knows it: Gerdin is right,

quote: Martin's referance was to the % given on the points page which is myScore/maxScore where maxScore is a fixed number depending on the difficulty level.

But of cause I should have figured that Gerdin had cooked up a way of getting the total score of the game.

[quinns]

After doing some testing on this POWER GRAPH (NOT the Civ Score, which is different than the Power Graph), here's what I've come up with (for those of you interested).

1. The powergraph looks the SAME for everyone throughout the turn. That is, it doesn't change as it moves from player to player in the same turn.

2. The powergraph line for each player is made up of:
a) The total production actually produced and still functioning. For example, a Warrior takes 135 production points to produce. You wont get any PowerGraph credit for this, until "construction complete" on the Warrior, then you'll get full credit (135 power graph points). The same holds true for city Improvements and Wonders. E.g. when a Granary is completed, you'll get full credit for the Granary (540 points). This does not seem to hold true for Advances or Tile Improvements though.
b) You seem to get about 50 points for each pop unit available to work. Each city counts as equal to [(city population factor plus 1) times 50] (or there abouts).

3. There is exactly a 2 turn lag on the power graph. That is, if it is turn 5, the power graph reflects the above "in-field production" as of the end of turn 3.

So much for my power graph "test" and analysis. Comments welcome.


[This message has been edited by quinns (edited November 30, 2000).]

[Keygen]

The score calculates the polution level also.
You can suffer a severe loss of points for great pollution.
A PBEM game in which I am involved is a good example.
It is a custom scenario made by Shavlor.
Deathmatch, in a small map, 5 players, 6 cities each with many modern improvements wich produce much pollution.
Here comes the strange and the negative thing on using the score instead of the power graph for measuring the players.
I am on the top of the charts, I have conquered all of the cities of another player, my power is generaly almost twofold than the other's but my score is -303 points. So my score is 0% .
I produce the most pollution from anyone else so I probably have the worst score while I am the most powerful nation undoubtedly.
Conclusion? Score including pollution penalty is way too out of measuring rated players!

[This message has been edited by Keygen (edited February 10, 2001).]