Scoreboard Proposal

This is a proposal to change the scoreboard display to provide a rating instead of just a killcount.  Based on the objectives and the gameplay of Tremulous, a killcount is an inadequate way to judge one’s worth in a match.  The killcounter gives people this idea that we’re playing team deathmatch.  I’m not saying the proposed system will be perfect, but rather it will be substantially less flawed than kills.  If you’ve played on a server with the Lakitu7 QVM, then you’ve probably seen the /mystats function in console.  This provides a wealth of knowledge that only the player can see.  We’re not here to debate why/why not that is the case, but rather, I bring it up because there is good information already there that can worked into a scoring system.  The point being that I know these functions work and it will ease the amount of coding that needs to be done on the project.  My other ideas were made without an understanding of the feasibility of implantation.  Thus, I submit a practical, implementable way to rate a player on the way Tremulous should be played. 

Basically, the amount of damage done to structures (friendly or otherwise) and players (friendly or otherwise) is recorded and yields a bulk number.  The general idea is to have a formula where the basic equation looks something like this:

Score = {A*(damage to players) + B*(damage to structures) – C*(damage to friendly structures) – D*(damage to friendly players)}/E

Now, to answer the concepts and the variables (A,B,C,D,E) and I will answer the building questions in a later section.  It is possible to record all of the damage done to the above mentioned items.  Now, instead of just taking kills, we have all this all data on damage but I’m adding modifiers.  The reason being, Tremulous is a game where you’re supposed to destroy the spawns and the last remaining players of the opposing team.  A base exists to protect the spawns and aid the opposing team.  The base and structures are the important things to destroy, so modifier B would be a multiplier greater than modifier A.  This would be done to encourage people to attack structures instead of padding their killcounter while still rewarding a player for getting kills.  It will also reward human players for killing bigger aliens since bigger aliens have more hp and can take more damage.  Also, since points can be subtracted from the score and the modifier can be made large, it will discourage team damage, especially on structures.  Now, the reason I’m using letters for modifiers is well, I don’t know what they’re going to be.  I’d like to know what the tremulous community thinks before I pour lots of time into establishing these.  Although, I do know what kind of modifier E will be, it will be a power of 10.  The point being, the score displayed will be a number from -100 to 100 with 0 being the reference point.  Say player Z has a numerator equivalent to 100,000.  His score will be divided by 1000 and he will have 100 for a score.  E will not be different for each player.  This divider is determined by the player with the highest score.  If Z has 100,000 and Y only has 100, Z’s score will be 100 and Y’s will be 0.1.  Y’s contribution is relative to Z’s but it is still on an absolute scale. 

This explains how the first part of the score will be made, now for part two which includes what I intend to do with builders.

Well, to put it bluntly, trying to score building fucks up everything.  It is extremely difficult to make building comparable to a score that is ever inflating.  How do you award points for a good decon when you help in moving a base?  How do you even code that?  You can record time for ckit held or time granger, but just because you’re a granger or have a ckit doesn’t mean you’re doing anything.  Humans can repair restructures, grangers cannot.  Does that make the alien builder less deserving of a score?  What do you do about that guy that sucks at building; his putting up structures doesn’t solve anything?  What about someone who would just decon and build a turret all game?  These are all epically problematic and I’m not going to be the guy who solves that.  So the best I can offer our tremulous builders is just recognition on the scoreboard.  If you build the most structures (in the case of human, build/repair), you will get a B next to your score.  I am talking to people to find out if it’s possible to hide that B until the end of the match, you don’t want the opposing team to know who’s building after all.  Also, secondary builders will also be rewarded with a B if they did F% as much construction as the primary builder.  Once again, F is used because I do not know what to make this yet.  I’ll be the first to day, that’s not perfect, but tough shit.  At least builders will finally be recognized with something.  Also, for those who say “but Medic, why should I repair or build if it won’t show up on my score because I need F% for it to?”  Well, hopefully you’ll decide that repairing/building when necessary to save your team is worth picking up a ckit or being a granger for a minute or two. 

So, here’s an example:

Say that as a goon I kill 10 people and 3 turrets.  For now, assume every person and turret had full hp; also for right now, say that B is equal to 2 and A is equal to 1.  Which means that I'm saying killing structures is 2x more important than killing people, so my score would be 2,080.  This would be divided by 100 to give me a score of 20.8.  Say a teammate of mine killed 1 turret and 5 people.  He'd get 360 for the turret and 500 for the people which would equal 860.  Well, the divider is 100 so his score would be 8.6.  But, say this teammate decided to put up the Overmind and a few eggs and the rest of the base.  Well, he’ll get a score of 8.6B to show his efforts as a builder. 

A few notes for editing purposes if I did not cover or something is suggested: 
Note 1:  Score is separate for each team. 

That actually sounds really nice to me. I'd like to see Lakitu/someone make that so we can test it out on the 1.2 server. It really sounds like the most promising formula I've heard thus far.

Only complaint I can think of so far: "Whaaat? I got a B? I'm an A+ builder!" (Ie, leaves some room for confusion)

Also, we'll have to _REALLY_ advertise the modifier values so that people know what to go for. Them trying for a high score is what we want, because it means that they will help out in a bunch of areas.

Another small issue: those "pro" noobs will complain that their score is low and they can't figure out why. Then you try to explain and they will be "oh well that's gay." However, I think that for the most part, we should just ignore these kinds of people.

Yeah, that's what I was thinking.  And while I understand your lettering issue, I figured B for buildier makes the most sense.

I think that's a great idea medic.
Hopefully killwhoring will subside a bit after these changes

How about we keep the "Kills" column, to keep the "pro" noobs happy, but sort the list according to Medic's score value. Also kills should be awarded to the person who did the most damage, not the person who put the last shot in. Lastly, should buildings count as kills? And also let the OM/egg/RC/node evos/credits be shared across all the attackers, not just the owner of the last bullet.

I'd still like to phase out kills, even it if means just renaming it to "score." That should be the least we can do.

While I agree to some of Amanieu's suggestions (about splitting up evos and creds), I do think that we should get rid of kills altogether. I'm not too worried about keeping the "pro noobs" happy -- a lot of them have already sworn off 1.2 as it is.

You could give points for repairing and building structures; though the problem is people may build/decon to increase scores, etc.

First off I think this is really great work.  You've put substantial quality thought into this and it shows. 

I do want to chime in on a couple of ideas.  You could reward builders for the amount of time their structures stay up.  Though this could be modified if their team is exceptional, in general it would be more reliable.  I think that the incentive of a win is enough to encourage good building, but maybe I'm too much of an optimist.

I really like how you've weighted things.  Should FF be weighted too?

@Amanieu
Kills are viewable via the mystats so you can still view kills.
I'm not trying to change the way kills are awarded, you'd have to take that up with Norf and Timbo.  Kills don't mean anything to me.
The building damage is one of the main factors in score.  Whether a building is a kill or not doesn't matter to me much. 

@Gustavas
Repairs are going to be factored in as part of rewarding someone with a B.  Rewarding for a structure staying up... well I don't really see how that factors into anything (the phrasing is dickish, sorry, but I just don't see why one would get points for a turret staying up because there's so many variables that determine how long that ret stays up), not to mention it would be hard to assign that a value that complements the score.  Yes, a base move is a great thing but does a guy that put an RC up at 10 seconds into the game deserve to reap benefits all game long?  I realize the RC is staying up, but it's a fact that the RC is needed.  Maybe if he had to stay next to the RC to keep it up and sacrifice his round, but otherwise, I just don't see why it should be worked in. 
Friendly fire is a part of the equation.  Team damage subtracts from your score. 

@Tyranis
Yeah, that's one part of the many reasons building isn't factored into the score formula. 

The problem with friendly fire is that sometimes it is necessary:
A) You have nodes but no arm and you desperately need someone to build, therefore you TK a naked rifle and ask him to build.
B) You have just moved base and you decon old structures by TKIng them while sitting in your new base.

Well, I think it's safe to ignore that argument, Amanieu.

1) The incident with the tk is rare, and players are willing to lose score now for that. If the situation is desperate, it'll still happen. It's too hard to figure out what is needed or not. I think the safest way is to just ignore this situation. Like I said, it's rare.

2) Mark decon is the default in 1.2. No need to tk old structures. They should be marked.

So, overall, I don't think its worth trying to figure out some complex way to solve problem 1) because it's so rare. Situation 2 shouldn't happen.

One more thing about kills though: You should get more points by killing a rant than by killing a dretch, and you should get more points killing a chainsuit than killing a naked rifle.

That sounds great, Medic, especially if the modifier for destruction of enemy buildings is relatively high compared to the modifier for enemy kills.  We'll see a lot more building rushes and less *I'm a battlesuit and I see a dretch so I'll put off my rush to finish it off even though it's no threat to me if I walk away from it and kill it later*  

For points for a structures's time up, you'll have builders build in crap locations out of the enemy's reach just to boost their score.  While I agree that the builder(s) should be rewarded, it'll hurt the gameplay more than it'll help.  We'll have builders trying to repeater/egg spam where they'll just be easy pray.

Also, the idea of keeping the kill counter should be kicked.  It goes against what we're trying to minimize with the score system, that is, kill whoring.  Sure, more weight could be put on score.  However, I'll remind everyone that we still have people that show off their kill score even when their team has lost.  It won't matter even if the player with the highest kill is at the bottom of the scoring; he/she will still gloat.  "Pro" noobs, as we've come to refer to them as, are in it for themselves.

As for advertising it: It doesn't need to be advertised heavily; It only needs to be documented somewhere for players to see.  Word will spread through word of mouth.  It does fine with friendly fire (servers that have it and not), share, etc.  Players will figure out how the scoring works and will spread it.  You'll have new people coming in wondering why someone's score just jumped or why it's so high and they'll ask.  Whether they take the answer to their the right way or not is up to them.  There are plenty of great deathmatch games for the "pro" noobs to killwhore in and those games are the ones they should play.

I do have a suggestion.  I believe we should factor deaths into the scoring equation since the scoring is how we're trying to determine the value of a player..  Feeding allows the enemy to rack up kills and grow in strength, thus, making the player that is feeding a less valuable player and points should be deducted.

Don't you already get more credits for killing a tyrant versus a dretch?

Make different buildings give different amounts of points:
OM>Egg>Booster>Hovel>Tube/Hive/Trapper
RC>Node>Arm/Medi/DC>Tesla/Turret
Double the points if the building is killed in SD

Agreed

First of all the number of credits is irrelevant in this thread.
Yesterday I join a s3 vs s3 game, aliens had 2 rants and lots of dretches. My team members kept killing dretches, but I lucied the 2 rant. (We lost shortly after) After that I was on the bottom of the list because I concentrated on killing the bigger aliens instead of killwhoring the small ones. Maybe do 1 point for each evo? Rant =5 adv goon =4 goon =3 etc
And for aliens 100 creds = 1 point

Also, should we be able to compare a human score to an alien score?

This is just what i suggested on pt.com.

I think I understand what you're getting at. What about points getting subtracted for staying near the RC for a certain length of time without gaining credits? This would dissuade camping, but allow defending. Of couse it wouldn't apply to ckits. And I say RC because the RC should always be camping grounds, while other structures may be a part of a forward base.

Also, if you're defending the base, but not gaining ANY creds, then you're better off leaving with a squad of teammates. This would encourage leaving the base, while at the same time encouraging teamwork, as people don't want to leave the base without back-up. All of this without touching the current balance.

Anybody see what I'm trying to say?