Game High Scores

[modman]

Please post an image of the scoring table.  I will keep track of the top five in this post below.
I'm going to keep track of the kills minus deaths you have, just to keep anyone from cheating.
1)
2)
3)
4)
5)

//EDIT it's going to be (kills - deaths)*(kills*deaths).  This will account for your efficiancy rate and sheer numbers.

[prehistoricman]

shouldn't you divide by deaths?  I would personally rank a player who had 200 kills and 100 deaths higher than a player who had 1000 kills and 900 deaths as the first used his units far more efficiently.

[@kukac@]

It's easier to count rate with kills/deaths (if its 1, then the kills and deaths are the same, if 2, then you killed twice as much as died, if it's 0.5, then you killed half of deaths, etc.)

[Omega]

Hmph, all my best games I have forgotten the score for... though I like the format, and agree it should be kills/deaths. Oh, and perhaps the list should have rules:
1) Must be on a map that comes with glest, to ensure you don't have a custom map that is cheated
2) No lying!

[modman]

Yes but about having to count rate, why would someone that has 1 kill no deaths have a higher score than someone with 999 kills 100 deaths?

I thought of what you said @kukac@ and above is the reasoning why not.

& prehistorian lets plug in the numbers.  You get (200-100)*(200*100) which is higher than the other one I estimate (to lazy right now to plug it in)

I will probably change the formula to divide the entire thing by (deaths times 2)
I'll write a program to figure it out faster for me. :]

THE SMILIES ARE BUSTED AGAIN!!!

[Omega]

You know, why not just post the number of kills and the number of deaths and leave it from there?
Yeah, I know, I'm a genius, but I try not to flaunt it.

[@kukac@]

Yes but about having to count rate, why would someone that has 1 kill no deaths have a higher score than someone with 999 kills 100 deaths?

Really? Have you ever divided any number by 0? What was the result?

It's a problem when someone kills 1231623513 and has zero deaths (heal sux!), but I get what you are saying.

Also you want scores, and not rates. How about resources*(k/d)*(1 when normal CPU, 2 when Ultra, 3 when an another player?)*(enemy number)?

[modman]

Yes, I realized dividing by zero would be a problem.  So I wrote a program that would compute the scores for me in C++.  I'm not that good, but can download it here.  It will be in a zipped folder.  Included will be the source code, so you can look at it and compile it yourself if you don't trust me not to infect your computer with a virus.

[Omega]

I'm going to trust you not to inflict my computer with a virus, but only because I don't think you're skilled enough to make a virus!
Ok, I'm just gonna compile it myself because I want to be able to see if it can be improved, and I always did like studying code to see how people made certain programs... Downloading...

//EDIT: Sorry, but its no good. After testing it, I enter things like 150 deaths, 62 kills and other random numbers. The process is horrible. It needs to be improved. I really like what @kukac@ suggested. I made a new Program that takes in opponent difficulty, number of opponents, and resources harvested. It can be downloaded here.

The only slight problem is that it is affected more by reducing the deaths by 4 than increasing the kills by 10...
It also has a completely different format than modman's program. It is in a zip program with the cpp file to allow changes and improvement.

PS: Modman, your source was missing

Code: [Select]

#include <string>You cannot use strings without including the header. Don't know how you compiled your own program, but...

[modman]

The process is horrible

What do you mean by that?  I tested it too, and I think there are no bugs at all, except maybe a questionable algorithm to some...

OK about the algorythm, how about multiplying rate and gain on kills together and then dividing that sum by something in the realm of resources harvested or resources harvested divided by 2 or 10 or something?  That way it would test to see how efficiant you were regarding your military.

I would like to take a look at your "superior" source...(not trying to pick a fight, not even saying you said it)

\\EDIT
A fine program.  Where do you learn your C++ if you have no book?  Internet?  (Not always the best source since it tends to be limited and not always complete, since if it's worth saying, someone will make money doing it.)   And I use bloodshed Dev-C++ and it said nothing about using strings without the string header.  I would assume that its functionality is included in iostream.  Plus the string function works fine without it.  Ask daniel.santos.  He probably knows.

For another program, I have previously written a geometry helper program that knows about 10 formulas, some more helpful than others.  It does cin, cout simple stuff and manipulating and plugging it in formulas.  You can download it here.  Again it includes the source as well as the compiled app.

[@kukac@]

Why do you need a program for it? Isn't the raw peasant intelligence enough for such a simple maths problem?  :lol:

Glest also has score system, wouldn't be much easier to use that?

[modman]

It counts resources way too much.
I have to disagree that a person could do it faster...Unless you have a 1 KHz processor, then it would be faster mentally!

It would be nice to have a "submit to glest.org" button on the score page.  Then it could be automated and no image editing could occur.  Also it would be set up to only accept scores on "official" maps.

[@kukac@]

Only official maps? Then what would be the reason of the community made maps?

[Omega]

Yes. I think any maps that has been uploaded and verified as 'fair' is good. Although it's easy to change the official maps anyway, so there'd be no sense.

[modman]

If you're going to go that far to cheat...
Someone who is going to cheat probably isn't going to know about the map editor.

[Omega]

Simple idea: have a future version of Glest include high scores, and if you're connected to the internet and you leave when the game says the match is over, glest will connect to a database and post the highscore. Dunno how to do that, but it'd be nice, and then Martino can set the parameters as he sees fit.