Evaluation System

[TittenIgnition]

i like a system that has points like this:

first: 1
second: 2
third: 3
fourth: 4
fifth: 5

and the person with the least points wins. seems fair

[<hk>]

Negative Point Accumulation System (NPAS):
The best score on a map is counted as 0. The diffecence in milliseconds to rank1 is defined as the point value. Points add up over maps. The player with the least points, wins.

Example:

Code: Select all

23:000, rank1, 0 Points
23:080, rank2, 80 Points
...

Pro:

• Still Simple
• Players with close times almost get the same Point value
• No arbitrary made up values.
• The point system scales well. This means rank100 still gets the amount of points relative to his time, not 0.
• You cannot only tell, who is better but "how much better".

Con:

• Players can get more points on very very long maps because the time differences between ranks are usually bigger. If that long map happens to be a weakness it's hard to iron that out. (fixed in ANPAS)

This system pretty ok. It fixes most of the problems easily without getting more complex than it has to be. If all maps are pretty much the same length it's as fair as it can get.
This would be my choice.

[<hk>]

Normalised Negative Point Accumulation System (NNPAS):
The best score on a map is counted as 0. The diffecence in milliseconds to rank1 is defined as the point value. Points add up over maps. The player with the least points, wins.
To Iron out the very long map problem all point's are normalised from the best time on.

Example:

Code: Select all

23:000, rank1, 0 Points
23:080, rank2, 80 Points
...

Pro:

• Players with close times almost get the same Point value
• No arbitrary made up values.
• The point system scales well. This means rank100 still gets the amount of points relative to his time, not 0.
• You cannot only tell, who is better but "how much better".

Con:

• Pretty Complex and hard to teach.

This system pretty OK if you run it in a computer program in the background all the time. It's too abstract for the DFWC, though.

[<hk>]

Weird Adding Everything System (WAES):
This system combines the DRS and the (N)PAS.

I was a bit unclear, but it's points from rank + points from relative time, added together.
The first number would give simple points based on their rank (like 1st = +20, 2nd = +15, 3rd = +12 etc.)
The second number would give a point (20 - x) based on how far their time was from the rank 1 time. Some kind of exponential function would probably be good.

Pro:

• A little bit of each

Con:

• Complex
• Unprecise at low ranks
• long map problem (see above)

This system successfully combines the disadvantages of PAS with the disadvantages of DRS. Just adding stuff from different systems doesn't make it better. You don't get more precision because you still have the gaps in points per rank.

This is the worst Idea yet.



By far.



Seriously. This is so stupid I can't even begin to think of a way how this seemed sane to you when you wrote it. You couldn't have done a worse job.

[<hk>]

i like a system that has points like this:

first: 1
second: 2
third: 4
fifth: 5

and the person with the least points wins. seems fair

I guess this is just a joke.

But in case it's not.... :

What about rank3? Why does it get 4 points? I would suspect 3.
Why is there no rank4?

In case you just made a typo there, I have to say that this suggestion is like DRS.
You have a linear buildup of points relative to rank delta. It's just negative linear decline, just as DRS. OK, you did choose the values in a way so you get a normalization effect at rank1. That fixes half of one of many problems.

Please guys. Give your suggestions more thought before you post them. I'm serious.

[TittenIgnition]

i like a system that has points like this:

first: 1
second: 2
third: 4
fifth: 5

and the person with the least points wins. seems fair

I guess this is just a joke.

But in case it's not.... :

What about rank3? Why does it get 4 points? I would suspect 3.
Why is there no rank4?

In case you just made a typo there, I have to say that this suggestion is like DRS.
You have a linear buildup of points relative to rank delta. It's just negative linear decline, just as DRS. OK, you did choose the values in a way so you get a normalization effect at rank1. That fixes half of one of many problems.

Please guys. Give your suggestions more thought before you post them. I'm serious.

yes, it was a typo. it seemed like it might work. ill just leave everything up to people who know what theyre talking about then

[flu]

[quote="<hk>"][/quote]

If you choose the system, whatever you want, I could help you, if you need.

[Arcaon]

Weird Adding Everything System (WAES):
This system combines the DRS and the (N)PAS.

I was a bit unclear, but it's points from rank + points from relative time, added together.
The first number would give simple points based on their rank (like 1st = +20, 2nd = +15, 3rd = +12 etc.)
The second number would give a point (20 - x) based on how far their time was from the rank 1 time. Some kind of exponential function would probably be good.

Pro:

• A little bit of each

Con:

• Complex
• Unprecise at low ranks
• long map problem (see above)

This system successfully combines the disadvantages of PAS with the disadvantages of DRS. Just adding stuff from different systems doesn't make it better. You don't get more precision because you still have the gaps in points per rank.

This is the worst Idea yet.



By far.



Seriously. This is so stupid I can't even begin to think of a way how this seemed sane to you when you wrote it. You couldn't have done a worse job.

An exponential function would still give a proportional amount of points to low ranks, while still taking account to competitiveness with "high rank bonuses". An exponential function (from percentage, not number of frames of course) would not make difference for map length. I suppose complexity could be a problem for ya.

[<)Ghost(>]

hmm doesnt seem like ur going to find a great way to rank, so ur best bet is to take the best ranking system out of the ones u mentioned.

[Leaf]

In my opinion, I think the points should be based on position (1st or whatever) rather than time (2 seconds faster).

As an example, if there are seven maps, Person X could be the skillz and beat Person Y by 0.1sec on 6 maps. But let's say the Person Y found some wacky route on the last map, and his time was 2 seconds better than Person X's. If you distribute points by time, then Person Y would win DFWC because of some weird route, despite Person X owning the rest of the time.

I think that's unfair, as Person X is clearly better but Person Y wins.