Cap off EXP to help balance teams

[»Ceiu]

Omni... do you mean like a jackpot system? No.

[Samapico]

The jackpot system was removed because it made people not want to flag for 90% of the time, while JP is raising.

And even then, winning a JP made people complain because some of them apparently never want to win, just raise JP.

[omni]

Not a jackpot system based on kills. A jackpot system based off time elapsed. This way ragequitters won't ruin game for the winning team.

[Samapico]

Same argument applies... people will just wait for a big jackpot.

Also, people will wake up in the middle of the night to come and win a big jackpot when no one's in pub

Based on kills or time, it's the same... Number of kills is pretty much proportional to the time elapsed... duh.

[omni]

What's wrong with having people playing in the middle of the night? There will still be two teams trying to win, unless it's like only one person trying to win.

 

Waiting for a big jackpot is fine. It gives the people who like to center more time to center. Eventually as the jackpot gets big, the focus will shift to flagging and people wont just try to win right away the way they do now. Now it's just grab flags and win continuously until the other team ragequits and people begin to center and pop declines.

 

As I told you kills and time elapsed isn't the same. If it was based on kills, the winning team wouldn't be able to get more money if the other team quits.

[Samapico]

Ok, so they'll have like 3 minutes worth of jackpot more, big deal. If it's based on kills, it's not like the jackpot drops because the other team quits.

 

Anyway, not gonna happen

[omni]

Why only 3 minutes? I dont get what you're saying lol

[Samapico]

Both teams playing for 15 minutes

Team A invades FR and get all flags

 

Team B counter attacks

 

Team A uses 50 fields to clear Team B

 

Team B ragequits

 

Team A has had all flags for at least a minute already... 2 minutes until they drop all, about 1 more minute after all-dropped to win

 

There you go. Both kills and time went up all this time, except after the ragequit, and during the win, which takes ~3-4 minutes

[Dr Brain]

Systems where the money accumulates over time, whether due to kills or due to elapsed time, will not be implemented in the zone because our aim is for relatively quick flag games, rather than drawn out annoying games.

[omni]

Well, just to clarify, for Samapico, the winning team would neut and let jackpot grow before they win. Why would they have to win right away?

[Aileron]

Three ideas, which relate to this topic:

 

For ragequitters: Simply run a check of the number of players playing every 15 seconds or so. If its higher than the last value, keep it, if not, throw it away. Base the jackpot off of the number of players in the flag match when it had the most people playing rather than the population at the end.

 

int players

int newplayers

 

(function to check players)

if newplayers > players (

players = newplayers; )

end

 

 

For the team evener: Before checking xp, even out the lanc pilots based upon the sell value of the lanc they own. Do this check regardless if the player is actually flying the lanc, and put a minimum on it to prevent ppl from messing up the system with a cheap lanc they don't actually use. Downside is that it may be harder for the experienced lanc pilots to be able to fly other ships

 

 

For the ?lanc command: Do a location check of where the lanc is, and check if that sector has flags in it. Mark the lancs who are in a sector with flags in it with a #.

[Acer]

http://i302.photobucket.com/albums/nn114/Acer1001/evener.png

 

^Exp Evener fail.

[Dr Brain]

It doesn't automatically even after a whole team quits.

[Samapico]

http://i302.photobucket.com/albums/nn114/Acer1001/evener.png

 

^Exp Evener fail.

team 1 fail

[Acer]

Well maybe if there is a big difference (3-5 people) the evener should pick a few players with ... uh not a lot of exp but not to little exp. Like someone with 5000 exp to go on the other team. =O

[Gannon8]

Well maybe if there is a big difference (3-5 people) the evener should pick a few players with ... uh not a lot of exp but not to little exp. Like someone with 5000 exp to go on the other team. =O

That is going to cause a LOT of people to quit.

What if the only people playing were those with >10,000exp, and one team quits? It's gonna take the player's prize money away, making them furious.

[Kilo]

Probably still wouldn't offset all of the money people made by hopping though. Too bad, so sad.

[spidernl]

The best way to make the evening system work is probably to turn exp into an amount of "points" using a formula which curves towards 8-9k exp and then 'slows down' again from that point, meaning the higher your exp past that point, the less influence that has on your "points".

 

Edit: Well, besides a full-blown evening system that actually checks what items you have, how well you usually do in a team. That'd require insane amounts of work to work well, most likely, though.

'Nother Edit: Ohsh- guess I wasn't quite the first person to think of it

Edited August 18, 2009 by spidernl

[Dr Brain]

The exp balancer relies on the fact that your "usefulness to the team" is roughly correlated to your exp, and linear (linear in the sense of super position, not in the sense of a line). It also assumes that the exp is an unbiased estimator of your usefulness. While it's not unbiased, and it's not linear, the negative impact of these deviations decreases with arena size.

 

So in an arena of 20, you would only notice a small difference in the "ideal" balancer, and the exp balancer. In an arena of 100, there would be no difference at all. Since balancing arena sizes of less than 20 isn't a primary concern (we mostly care that big games are balanced), I'm not going to devote hours and hours of dev time to creating the "ideal" balancer, since no one would notice anyway.

[spidernl]

I'm not going to "fix" the current balancing system if it's not broken.

 

Logarithmic is a terrible idea. It means that someone with 2x as much exp would count only slightly more (how much depends on the logarithm base). There's not a large enough range of exp to make logarithms behave the way you want.

 

Similarly, the gripe that exp stops being useful at 13000 has no bearing on relative player worth. Players that play more often typically fare better.

 

I can assure you that someone with 30k experience is not as valuable to a team as 3 with 10k experience. Honest.

[LeftsideChaos]

Maybe the max exp the balancer pays attention to should be 8k exp? 8k is enough for AD, as well as the main upgrades.

 

So if a person has 23k exp, they're only seen as 8k exp by the balancer.

[Dr Brain]

If the zone population was 2-4, yes, it might be a problem. As it is, it is not a problem.