Angeldust

YouTube Twitch Facebook Instagram Discord Twitter
iTunes App Store Google Play Store Steam (Windows, macOS, Linux) Windows

Hero battle (PvP) ratings

in English in Forums





Post 21–30 of 33

Previous « 1 2 3 4 » Next


# 21

by

Firefly

4 months ago


@Hummm: fair enough; I think there's a point to be made for both philosophies.

A) One one hand I think it's fair that shooting another player gives you a well-deserved share in their future performance, since you did manage to shoot them. If they start doing better than before, you're rewarded. If they turn out to get worse, you're punished a bit. In the long run you get a balanced reward for the actual performance of your target.

B) On the other hand I also think your 'in the moment' scoring can be fair. Which would also lead to a standard, ever-increasing score for all players since you'd immediately get some amount of points for shooting someone and no score would ever change because of the other players' performances.

System A shuts down cheating using fake accounts more naturally than system B. It's also a novelty system I think.

B is more conventional and has the strong point of scores being absolutely set in stone for each player exhaustion.

More feedback is welcome. I'd be happy to run with either system A (which is ready to go) or system B which'd need a bit more thought and development.


# 22

by

obi2002

4 months ago


Time based system would be more fair score wise, but system A is much more benefitial to everyone. It's a hard choice, maybe we should test System A first.


# 23

by

Hummm

4 months ago


I'm just saying that system A rewards people for dominating noobie players that later become good (with no effort on your part). There should be little reward for killing a noob (which is in there) but no reward later as they improve. If you want a big reward for killing them as an experienced player, you have to actually do some work after they've learned how to play.


# 24

by

Firefly

4 months ago


Suppose we'd go for a system B approach, then what is a good score algorithm to determine how many points you get for exhausting someone?

First question: does it matter what your own, current score is? Or should a target always give the same amount of points on exhaustion, no matter who shot them?

Second: I think it's important to prevent fake accounts from giving (many) points. But if everyone's score starts at zero, you can't use a fraction-of-target approach since everyone'd remain at zero. What is fair?

One approach I can think of is mixing system A's scoring with system B's temporal limitation.

Something like: every player starts at score 1. On exhaustion, a player receives [ Target's score / ( Target's EX'eds + 1) ] score points. For the fake accounts, the first EX would give you 1/2 point. The second EX 1/3 point, and ever diminishing from thereon out. I should probably do a test run using this system to see how it pans out.

Other ideas are welcome too!


# 25

by

obi2002

4 months ago


If you EX someone with 100 score and no nemeses, you get 1/2 of it so 50? For the 2nd 25, then 12,5 and so on? I like this a lot more, since other people EX the person you EX don't affect you. I vote for temporal A :)


# 26

by

Hummm

4 months ago


... just a brain storm...

every one starts with 10 points.
every time you EX someone you get 2% of their score (possibly with a hard cap)
every time you get EXed you lose 1% of your score (again, possibly with a hard cap)
scores can't go below 10.
if you EX someone below 20, you can't get more points from them until either:
a) they EX someone (anyone)
b) you or they leave The Zone (and return, which resets session stats)


# 27

by

Firefly

4 months ago


Thanks for the ideas and feedback. I'm going to experiment using a different rating system right now. Will post again once I have results.


# 28

by

Firefly

4 months ago


Going with a scoring system where each exhaustion counts for (target score / target EX'eds) at the time it takes place, I get the following ranking and scores:

1. obi2002—815
2. Firefly—513
3. Alex makarov—441
4. space is green—351
5. Tailsmancion—251
6. LIL XXARDD—170
7. ExcaliburGaming—146
8. ZOOYHS—144
9. WertuXAN—132
10. trieuthien—87
11. Hunter86—85
12. Hummm—84

Observations and advantages:
- scores only ever increase and don't decrease in your absence
- being exhausted doesn't affect your score at all
- you can calculate how many points you get for shooting someone using [ their score / ( their nemeses + 1 ) ]
- scores can be updated in near real-time

Overall it seems like a good idea to run with this scheme. Although I'm a bit concerned about the big score differences at the top. What do you all think?

(Additional note: players not doing PvP will have a score of 0. Once you get involved with the shooting, your score becomes 1.)


# 29

by

obi2002

4 months ago


The big score differences apear cuz everyone up there(Firefly, Alex, space and I) have been dominating every time a large enough group to gather some notable points is present...livestreams.

Once the game gets more players and possibly large groups outside of livestreams, I presume scores will sky rocket into the thousands(with this system).


# 30

by

Firefly

4 months ago


@obi: yes, in a way it's entirely reasonable to have these scores. I'm just not yet convinced if the system is entirely fair. For instance, a high 'K/D' ratio isn't really rewarded so Alex doesn't get a lot of benefit other than bragging rights.

Other than that, scores being available in real-time is a big advantage and greatly improves the interactivity and fun feel of scoring in PvP. For that reason alone I'd choose this new system over the previous, 'iterative graph' model.


Post 21–30 of 33

Previous « 1 2 3 4 » Next