Solution for stacked flagging freqs.

[espo]

I think we could at least try this for a probationary period and see how it goes.

 

When a player enters he either goes on 90 or 91 depending on which is smaller. If a player doesnt want to flag he can go to a priv.(Maybe bring back ?createteam) Lengthen the time on spectating a player for lack of flagging for fucks sake. Its way too short! If I get killed twice in center as lanc trying to get to the portal gate I get specced only to ?flag onto the losing freq.

 

This modification wont stop hoppers but it would keep flagging freqs bigger and more even and would also reduce confusion for newblettes who want to flag but don't understand ?flag and when they do figure it out, wonder why they got specced 30 sec. later. It's a real turn off for new players.

 

It almost seems as if someone was trying to re-invent the wheel when the current system was implemented. Simple Smart > Complicated Stupid.

[Weebles Wobble]

Agree

[Cheese]

Lengthen the time on spectating a player for lack of flagging for fucks sake. Its way too short! If I get killed twice in center as lanc trying to get to the portal gate I get specced only to ?flag onto the losing freq.

 

its also lulzy when it immediospecs u instant last flag touches ht

[Emit]

get better

[espo]

Yes Hyperspace, get better.

[P Nut]

<_< Edited November 19, 2010 by P Nut

[espo]

Not sure what that emoticon implies. PNUT, don't you agree with my idea? If not, please state your opinion on the matter. I am always interested in your thoughts about matters that concern HS flagging.

[Dav]

a nice way to decide where a player goes could be using exp rather then numbers. Would be complex but doable.

 

Perhaps the zone could take the average exp for a team.

 

The team with the lowest average exp will get players where exp > avg and the other can have exp < avg

 

If this was done with a condition that the difference in players was not larger then 2 or 3 it would be a nice way to do it as well.

 

Stopping hoppers could probably be done as well, although it may be harder then this sounds.

 

The zone keeps a record of who is on which team. If they leave ?flag and attempt to rejoin within a set time (say 2 mins) then it puts them back to the last known flagging freq.

 

Essentially you would get the following matrix.

 

 

1) Has player been on a flagging frequency in the last 2 minutes?

 

If yes - put playeron previous team

 

If no - continue

 

 

2) Is the difference between the players on frequencies 90 and 91 > 2?

 

If yes - put player on smallest team

 

if no - continue

 

3) Determine players exp level of player and compare to average exp all flagging players

 

If player exp is > avg - put player on low exp team

 

If player exp is < avg - put player on high exp team

[espo]

I like your input Dav.

 

Now if we could get some of these ideas in action.

[spidernl]

What is the advantage of dumping players right on freq 90/91 on entry?

That's pretty much the only new thing in this thread as far as I can see.

Edited November 23, 2010 by spidernl

[ArcticxWolf]

I don't understand how this will stop stacked flagging freqs.

[»SD>Big]

what about kill/death ratios? isn't that a better measure of skill?

[P Nut]

what about kill/death ratios? isn't that a better measure of skill?

No. A person can be rushing, being the most effective person on the team, and get a horrible k/d ratio because of it.

Another person could be camping gunline, getting an amazing ratio, without putting in much effort at all.

[ArcticxWolf]

What about basing it on overall percentage of games won, similar to league?

[Dav]

exp is a reliable measure of ability, you have to remember that a very skilled player may has a very basic ship and thus cannot be very effective.

 

@ ArctixWolf The key I wanted to point out in my post is that one measure of eveness is not enough. You need to keep numbers around even and stop hopping as well as balance the skill.

[Unix]

After a certain point, experience becomes meaningless. Around 10-13k is my own opinion.

[Dav]

this is true, then havving a difference of two when teams are full of 10k + players will make the teams less even. Make things more complex still but an "If exp < 10000" would fix it.

 

Unfortunatly i suspect that this method may be a bit too complex to impliment though.

[spidernl]

Man, this exp-based system is such a new idea! I haven't even seen the 10-13k exp 'argument' before!

 

Protip: this discussion isn't new, and no matter how 1337 your ideas are there still needs to be someone

with the time to actually develop (a) module(s) that implement 'em.

 

Also, if anything the square root or something of exp could be used. First couple 1000 points will have

far more impact than 10k+, as they should.

 

Edit: as someone with a 'little bit' of experience with ASSS modules I can assure you things like these

don't get too complex to implement that fast.

Edited November 25, 2010 by spidernl

[Dav]

its mostly processing time that might be an issue. I have no idea how much load a multi-step procedure like the one I described upon ?flag would take though.

[spidernl]

its mostly processing time that might be an issue. I have no idea how much load a multi-step procedure like the one I described upon ?flag would take though.

 

Then again there are some unneccessary steps in that procedure: Why take the average and not just add up, look which has least total exp and dump on that?

While taking the square root (or something similar, as I suggested) takes a bit of power, it can't be so significant that it'd noticeably slow down ?flag.

 

Edit: also, with the 'square root' method you'd probably keep two variables: freq 90's total "square rooted" exp and freq 91's "square rooted" exp. All you'd

have to do then is add or remove one player's "square rooted" exp on entry/leave and do a comparison every time a player uses ?flag.

Edited November 26, 2010 by spidernl

[Samapico]

its mostly processing time that might be an issue. I have no idea how much load a multi-step procedure like the one I described upon ?flag would take though.

That might have been true a decade ago... A dozen square roots won't kill any computer

[Dav]

im not saying it cant do it, just not sure if it will be laggy

[Patman1]

Why even bother square rooting, just adding up the exp should be enough. This would stop stacking.

This kind of module wouldn't be too complicated, it just sounds like a bunch of else if statements slapped together (or whatever the equivalent statement is in Continuum programming language).

 

As for hopping, I am bringing up another idea that I read from my own suggestion post, have a module to read a player's IP address and then lock him to the frequency he is on.

[Dr Brain]

Perhaps you guys are unaware, but the teams are already limited as to how far apart they can get in exp.