QSC Results - Games 21 to date

[AlanH]

Once the Pyramids are done their biggest benefit is the granaries in every city. So I think the theory is that once you have the wonder you get the points for its benefits. In the case of the Pyramids you get points for all those granaries, plus the extra cash from no Granary maintenance.

 

[ainwood]

Is everthing worth the same number of points in every game?
Why is a wonder worth so much less after completion? In COTM4 I have Colossus, Mausoleum, and Forbidden Palace, for just 300 points total, while they cost double the shields to build. But the Pyramids which is not quite done is nearly 400 points.

Presumably when you say "almost done" you are using Dianthus' stats program to get the score?

For uncompleted improvements, you get 1 point for every shield accumulated. When the improvement is finished, you get points for the improvements, and wonders are worth 100. But as AlanH said - for the pyramids you get a granary in every city.

I have summarised the scoring here, but it may not be 100% correct (but close enough, fo now).

 

[Randy]

I may have missed this some where so if it has been asked before, sorry.

The last column (Jason QSC) what is this? In COTM 05 at rank 12, I have 1.38 there.
Xarin @ rank 1 has 1.09, SirPleb @ rank 13 has 1.64, and diophantus @ rank 22 has 0.32. It does not seem linear.

 

[AlanH]

Someone asked me to add a column containing the ratio of final Jason score to QSC score. The column heading is "Jason over QSC", an attempt to minimise the column width.

I did it as requested, but I'm thinking of removing it as no one has even commented on it until now, and I can't see that it throws any great light on the data.

 

[Randy]

OK, I get it now QSC score * last column = Jason score (sort of)

So if the last column = 1.0 you played the same thru the whole game.

If last column > 1.0 you improved and the higher the number the more improvement.

If last column < 1.0 then the opposite is true.

Cool!

Edit: leave it in, it's a quick check to see how you did after the QSC. If I would have saved a 1000 BC from GOTM 35 where I warmongered instead of building structure I may have had a 2.0+

 

[AlanH]

So if the last column = 1.0 you played the same thru the whole game.

You are correct in saying that higher ratios mean stronger late games or weaker early games. But I don't think 1.0 has any special significance, as the two scoring systems are completely different, and are based on completely different parameters.

 

[Randy]

1.0 would = 100%

So a qsc score of 10000 and a jason score of 10000 would be the same.

but 1.1 = 110%
qsc score 10000 and a jason score 11000.

and .9 would be 90%.
qsc 10000 and jason 9000.

 

[AlanH]

So a qsc score of 10000 and a jason score of 10000 would be the same.

But a QSC score of 10000 has no special meaning. A Jason score of 10000 means you hit the target, because the scores in all games are normalised to 10000 to even out variations in food level, difficulty and so on. There is no normalised "target" in the QSC, the score is just the sum of a lot of factors. As it happens the top players sometimes make it to QSC scores of 10K and beyond in specific games. But no normalisation is applied to ensure that this is the consistent optimum score across all games.

BTW. Dianthus spotted an error in the scoring process that meant we gave KillerLoop a lower score than he deserved in the COTM 5 results. We didn't give credit for all his Pyramid-shaped granaries, as we used an out of date version of the scoring utility. It's now fixed, so you should all go and marvel at KillerLoop's new high score.

 

[Dianthus]

Someone asked me to add a column containing the ratio of final Jason score to QSC score. The column heading is "Jason over QSC", an attempt to minimise the column width.

I did it as requested, but I'm thinking of removing it as no one has even commented on it until now, and I can't see that it throws any great light on the data.

Hey Alan, that "Jason over QSC" column is really cool .

I can see what you're saying about it not meaning anything in that you can't compare it across different games. It can be compared with other players for the same game though. How about normalizing it so it can be compared across games? I.e (100 * PlayerJason * MaxQSC) / (PlayerQSC * MaxJason)

 

[Tone]

I'd like to add my support to that idea.

 

[AlanH]

I've normalised the ratio column. Enjoy

 

[Mistfit]

As per usual great work Alan -

One question: How did zamint3 play cotm05 in vin civ?

sorry if this has been addressed

 

[AlanH]

.... because he had a return on the end of his version, but it's all better now.

Well, you did ask