Statistics versus Individual Score

I believe the concept of an individual "score" should be removed and replaced with a robust statistics engine. Sorta like the stats engine that shipped with Unreal Tournament, although it doesn't have to be quite so detailed.

The problem with a single integrated "score" is that it forces the game designers to basically interpret what is important and not important in the game. In original base T2, on maps like Sanctuary the scores of the turret farmers often exceeded the other teammates scores by 2x or more. This is an implicit value judgement. Rather than have the designers make that judgement, it should simply be reported in some sort of grid or matrix that player "Foo" had X number of Turret Kills.

This was the critical flaw when Dave G decided that grabs were worth a point. This was also an implicit judgement, and it was a crappy one. Grabs should be statified separately from Caps, which should be statified separately from various kills, etc.

Granting a "Gen Defense" stat based on proximity to the generators is okay, same with a "Flag Carrier Defense" bonus or "Flag carrier Kill" or "Flag Return" bonus. They just need to be tracked distinctly, rather than putting them in a big blender and producing a single number.

There is virtually nothing interesting that a single score number can represent, unless you're talking about a game with relatively little depth of play, like Arena.

I like pizza mod for T2, it gives stats like that, along with the actual score.

It has things like, Fastest Grab, fastest cap, most grabs, most caps, most flag carrier kills, most MAs(for various weapons) and longest MAs, most kills(for each weapon), most shrike rams, most TKs.

Bunch of shit like that. And obviously it shows the name and the number.

I aggree tho, something with stats would be better than a scoring system.

As far as flag caps go, I do think 1 cap = 1 point, and matches shouldnt be won by grabs, obviously....but, I do like being able to see how many grabs each team has.

I agree with this but I'm one of those hardcore gamers who plays for th sake of playing games, and not a regular joe who just wants to shoot some folks.

The problem with taking out points is you lose the instant gratification, plus some of the incentive to play your best...games are about gratification, challenege and reward.

Civ2 is fun because it constantly rewards you...new buildings, new units, the throne room...you get a new advance for taking over every city...it's all very rewarding. If scores are taken out something would have to replace them to give players that 'hells yeah' feeling.

Anyhow I 100% agree with you, but I could see why for the casual player this might make the game less fun.

If you aren't playing for score, why the fuck are you playing? If you are in a ladder match then yes, personal score doesnt matter that much because there is a reward for your team winning. When not in competition NOBODY cares how well their team does. If you say otherwise you are stupid. Ever notice how very few people play siege when not in competition?

If you aren't playing for score, why the fuck are you playing? If you are in a ladder match then yes, personal score doesnt matter that much because there is a reward for your team winning. When not in competition NOBODY cares how well their team does. If you say otherwise you are stupid. Ever notice how very few people play siege when not in competition?

Generally i play for fun. for some people, point whoring is fun. for other's it's OOB beacon repairing, and for even others, it's playing like a team.

Generally i play for fun. for some people, point whoring is fun. for other's it's OOB beacon repairing, and for even others, it's playing like a team.

This's true.

Tribes is not the kind of game to point whore in.

If you aren't playing for score, why the fuck are you playing? If you are in a ladder match then yes, personal score doesnt matter that much because there is a reward for your team winning. When not in competition NOBODY cares how well their team does. If you say otherwise you are stupid. Ever notice how very few people play siege when not in competition?

Most tribes players gave up player for score a long time ago. Whether in a pub or in competition, it doesn't matter what your score is at the end of a map, it matters what your team's score is.

When not in competition NOBODY cares how well their team does. If you say otherwise you are stupid.

Why do you think that what I said means that I'm only going to look at my team's score? That's not what I said, or maybe you're responding to someone else's post.

The point is this. If I'm a newbie and I discover that I'm actually pretty decent at deploying stuff, or piloting a bomber, or something like that, then I ought to be incentived to do that, even if it doesn't show up in the final score. In a single-number score, it might not show up and I'm going to feel completely unrewarded for my efforts. If the stats show that I was the most prolific bombadier or the best shrike-rammer or whatever, then I have distinguished myself even though I don't have a lot of points.

Most tribes players gave up player for score a long time ago. Whether in a pub or in competition, it doesn't matter what your score is at the end of a map, it matters what your team's score is.

That is BS and you know it. Don't kid yourself. Nobody cares about teamscore on a pub.

Why do you think that what I said means that I'm only going to look at my team's score? That's not what I said, or maybe you're responding to someone else's post.

The point is this. If I'm a newbie and I discover that I'm actually pretty decent at deploying stuff, or piloting a bomber, or something like that, then I ought to be incentived to do that, even if it doesn't show up in the final score. In a single-number score, it might not show up and I'm going to feel completely unrewarded for my efforts. If the stats show that I was the most prolific bombadier or the best shrike-rammer or whatever, then I have distinguished myself even though I don't have a lot of points.

We aren't arguing the same thing here. I didn't say stats are bad. In fact i think stats are great. My point was that taking out personal scores would be stupid.

Fair enough. I'm okay with eliminating personal score completely, but I don't mind leaving it in as long as the stats are there.

That is BS and you know it. Don't kid yourself. Nobody cares about teamscore on a pub.

That's a perfect example of a shitty pub. If you honestly care more about your player score than your team score in a pub, you're probably ruining the pub for everyone else.

I made a ridiculously long post on this issue in the "Fixing Pubs" thread. Rather than retype everything I have to say about stats, I'll just quote it in case anyone has a few hours to kill :P


I think having admins giving out trophies and mvp votes would lead to a huge favorites system that would be highly abused.


I think the problem with the points system we see in tribes1 and tribes2 is it takes objective information and uses it to draw a subjective rank (their point value.) Perhapes what we need to do is have the computer report just the objective data and let people draw their own conclusion. The rankings would then be driven by a group mentality bases on the objective reports of the server.


Now the problem comes in reporting that objective material in a way that can be read quickly, so I'll dedicate the rest of my post to analyzing such a system (please realize it's all just brainstorming -- I hate sounding like I'm giving out orders by saying something like "this should be this.")

Ok first of all, most of would agree that most roles are divided up into offense and defense. But really there are some gray areas, such as someone playing midfield who would chase cappers, esscort carriers, or just pick off strays. Or those who would fly a vehicle around the map ending up on either side of the map. I can really think of no fair way to distinguish a role for these people since half the time they're on D and the other half of the time they're on O. The most fair method I can come up with then is to record the activites of both (objective :))
If we use the flagstand way points to distinguish which zone belons to Offense, and which zone belongs to Defense we can draw a line down the median to determine where the player's current activity is taking place. Alternetly we could use base waypoints and raw a radius around those base waypoints were any overlap or exclusion could be considerd nuetral. I don't really like adding extra zones as it will make things harder to read in the future, but I think I may like the radius system better. You could even add an OOB zone, but once again, readability is a factor. Flooding people with information isn't always the best way to go.

Ok, let's go with O/D/N (offense, defense, neutral) using the radius system just for the sake of foundation and start comming up with important role defining stats.

First of all, I think the first thing to display on our player roster is where everyone is spending their time in game. The quickest way to come up with comparitive results for this is tiny bar graphs. Just pick 3 colors (I'll use Red for O, Green for D, and Gray for N) and make a bar graph that represents the time each person has spent in which zone of the map. It's quick, it's easy, and it gives at a quick glance an idea where people are spending their time on the map. Very objective, ok let's move on so we can determine a way for people to distingusih if that's time well spent, or if they're only jerking off.

Now it gets complicated. There are a billion things people could be doing, and like SoulJa said, it's hard for a computer to determine what's important. So let's stay focused are list objective things.

Armor Prefernce [vehicle prefernce?]
Kills [contains about 3094203234 sub stats]
Destruction [blowing up turrents, gens, etc]
Deploying/Repairing
Capping


First is easy, Armor/Vehicle prefernce. We can get a fairly good idea what kind of role someone is trying to fill by what armor they change into most often between spawns. It should be too dificult to come up with a fair algorithm to determine a favorite, but I can't decide if a favorite armor (which may be too subjective) should take the place the current armor. On the one hand, people can change roles in the middle of the map giving a flase status in results. On the other hand people looking at the roster will undoubtably see people midchange which could lead to a false impression of their role.
I don't want to clutter up the screen by showing both. Really what's important is what's going on at any current time, thus it's not very important to factor what someone was doing at the beginning of the map (I know, it contradicts the idea of my zone time logger from before) Ok, here's a solution (and this can work for the zone time logger too): Only display the favorite armor of the last 2 minutes (1,3,5 whatever, I picked 2). As you can see this is getting pretty complicated. I finish up by saying armor prefernce should (imo) bedivided into: naked(for those who do not visit an inv at all), light, medium, heavy, and possibly vehicle (if you spend more time in a vehicle than anything else -- getting subjective again, but it's still based on simple math and not so much opinion influenced algorithm. Of course you'd specify which vehicle) I think that the can best, and most compact way to display this imformation is with a mini icon or letter system. I would color code it for even easier readability as it might be hard to tell the medium form a skinny or a fatty, and it could be difficult to display a vehicle with a 40x40 icon. You coulc even go as simple as a letter, but that could lead to confusion (what does the N mean?) Icon with a letter overlay may be nice. Whatever, I'm sure irrational has better plans for the GUI than me.

W00t! We can move on :) (jesus fucking christ this is long...Thank you if you actually ended up reading this far -- I think we're half way there :D) Ok...Kills. This is one of the biggest deals and probably a big soft spot for a lot of people. Which kills are more important? Well, that's why we're objective about it :] It'd be outright stupid to report every single sub stat in kills, we can simplify by only reporting according to zones, but then our reports aaren't accurate enough to be role defining. Ok, since I'm brainstorming I'm going to create another list of the important factors that go into kills.
--
Zone Location of Hunter (this is you)
Zone Location of Pray (this is the bad guy)
-->Locations could be as follows: O, D, N or OOB
--
Kills of flag carrier
Kills of flag esscort
Kills of flag chaser
Kills of HoF
--> these are all important as they have to do with the flag
That leaves the other shit you can kill, which includes farmers, repair monkeys, the LD, the HD, piolets, HOs, and custom roles.

Now LD are usually scatterd all over the enemy zone
People hanging around assets like turrets and gens are usually guarding (when it comes to gens) or reapairing them. They're usually just fodder as the attackers main objective is the gen, turret, or inv station.
When it comes to the flag, you are almost always out to kill the bitch.
As for those playing D, anything that comes into your base's zone needs to die.

So then, what do we need to keep track of reguarding kills? I thin we need to look at location (O/D/N) and assets (gens, turrets, invs, flag). Ok, time for me to list.

Flag O Kills (clearing, esscorting)
Flag D Kills (kills made by those who kill enemies with the flag or around the flag)
Assest D Kills (Gen/Inv Guards who kills would be saboteurs)
General D (Defenders who kill enemies in their general zone (ho's, esscorts & cappers not around the flag, etc)
General O Kills (those who go O and just blow the enemies up. Gen guards, farmers, light D, etc)
OOB kills (these people are generally worthless, but who knows.)
N kills (only will happen when there's no flag around. Probably someone killing an incomming Ho...or someone tag chasing)
Turret kills (might as well keep track of them here)
Teamkills

Ok, that's a fiar ammount of kill stats. I'm sure I left out plenty of roles, but they'll fall into general O or general D. What's important is that these kills aren't given any special point vaules to weigh aginst each other. What will happen is a person's kill proficiency will be displayed and nothing else. It's a little tricky to determine a kill proficiency, and maybe even a little subjective. But we'll take the most obvious and just say which ever kill stat is the highest will be displayed as the profeceny. There after we can (maybe important, maybe not) display another bar graph of the next top 5 (4,2 whatever) as a percentage of the top proficiency. This would be subject to some testing. General D kills may unbalance Flag D kills for example. I'm not going to go into details, b/c I'm sure many people would have many differnt ways to organize it, but I hope you get the general idea.

Ok! (jesus christ this is long, I'll bet I could type whatever the fuck I wanted to right now and it wouldn't matter b/c you probably wouldn't read it anyways. I think I've been typing for 2 hours. Jesus. Ok I'll make up a tribe joke as an intermission if you're actually still with me. "Why do Heavy Armors make lousy medicine men? B/c thier mortars always explode!" :lol: ahhhhh....ok moving on :-|)
Destruction stats!!! This is easy to track. How much of what shit does one blow up? Time to list
Base Turrets
Base Sensors
Gens
Inv Stations
Vehicle Stations (if they have these)
Deployable Turrets
Deployable Invs
Other Deployable shit no one really gives a fuck about

This one's just like the kill stat tracking. It records how much of the ENEMY shit you blow up and then displays your proficieny (gens, invs, etc) and how many times you've blown it up. Then bar graph the next top five (color coded as always). I don't really need to explain it all that more. I'm sure there would be tweaking involved, but as always, just general brainstorming.



Deploying and Repairing. I grouped these to gether b/c it's all contructive.

First repair stats. List:
Gens
Invs
Turrets
Deployables
Players
Yea, I simlifyed it a bit. But it should give a good idea of what these monkeys repair the most, so you can tell them to stop fucking with deployables. I dunno, I'm even considering this a waste of bandwidth even. I guess we all agree that repair is the most suck job there is. But then there are deployables, that would overright any contructive stats (so farmers don't look like an ass). And I'd even consider having the constuction/detruction stat display one or the other. Anyways, turrters would get their kill records in the kill stat, but damge they inflict would go into an efficiency (points of damage induced vs. points of damage repaired? Invs would just have to work based on use. I have no idea how to compare it to the other construction area. Loadouts changed? Ammo resupplied? I have no idea. The whole constuction stat seems rather stupid, but if someone took the time to organize it useful information could be displayed.


Finally (holy shit, FIANLLY) there is capping. Nice and easy to track. What to record?
Caps
Grabs
Yea, do both. It should give an idea of who's at least trying I think. Well, that seems simple enough.



Ok, there you have it. Now here's an example of how the stat screen would look (only your team, I see no reason why you should know what the other team is doing).

http://www.phoenix-blaze.com/~sirlucius/tvstat.gif


Yeah, well, maybe that still looks a little hard to read. Better color coding or something.


Oh well, that concludes my idea.

Summary: I agree with mojo

me too.

points don't really matter, it's really just a ranking system...

if you got 200 points in a t1 map that'd be awesome...unless someone else got 210.

The whole point of online gaming is to be better than someone else (hence the whole Tribes as a sport catchphrase).

It's cool to see lots of statisticcs, but then you lose the sense of accomplishment for being ranked #1...because you might have had the fastest cap, but then lolnow had the most MA's and machinegunjoe had the most repairs, and...you get the idea...

That being said, having a FEW different categories would be cool like lucious' thoughts on:

Kills [contains about 3094203234 sub stats]
Destruction [blowing up turrents, gens, etc]
Deploying/Repairing
Capping

I think this would keep the fun of being the best in a category, without losing the sense of accomplishment by being the very best.

and just a thought...I remeber the first time I saw a +100 score (yeah almondjoy!)...I was pretty blown away, and I remember thinking that I was going to get as good as that guy...and that was what kept me playing T1 for 2 years. Too many stats and you'd lose that feeling of awe, and there would be no hook to the game IMHO.

btw that was a terrible joke lucious ;)

Any player that's worth a shit doesen't look at score to determine whether or not someone was skilled or who did the best. You can get plenty of points making stupid plays and cause your team to lose, whereas you can get a much lower score while making lots of smart plays and making your team win.

If you aren't playing for score, why the fuck are you playing? If you are in a ladder match then yes, personal score doesnt matter that much because there is a reward for your team winning. When not in competition NOBODY cares how well their team does. If you say otherwise you are stupid. Ever notice how very few people play siege when not in competition?

Honestly though, the best score system would be none at all. If you have no individual scores and only a team CTF score. People (esp newer people), will be more inclined to work towards that instead of looking at their score every 5 seconds. Good players would still stand out readily due to great plays (what matters). I know that a no individual score system will never ever happen though for a multitude of reasons.

ok.. name a successful game/gametype that does not include personal scores. When you can't think of one, why not try and figure out why.

ok.. name a successful game/gametype that does not include personal scores. When you can't think of one, why not try and figure out why.

Oh yes, and i'm sure sierra knew exactly how ctf would be play 4 years after the game was released.

That's a pretty bad argument.

ok.. name a successful game/gametype that does not include personal scores. When you can't think of one, why not try and figure out why.
Natural Selection.