Tremulous Forum
General => Feedback => Topic started by: Conzul on July 10, 2010, 03:07:47 pm
-
Well? Do you think tremstats should rank by kill-death-ratio (KDR, sometimes called "Player efficiency), or by score? Discuss. Concerns 1.2
<---voted Score.
-
I vote credits earned * 1.5 - credits fed. Would need some sensible conversion between evo's and creds, although could just split it so it shows aliens and humans separately.
-
score, because it includes destroyed buildings.You can be a hard feeder while you are effective at nading/sawing buildings.
-
Tremstats essentially is sorted by score per death.
If you look, the efficiency ratio is actually the sum of your kill:death ratio and your destruction:death ratio.
Try crunching score per death for all of the top players. I didn't find anyone who was out of order for score/deaths.
I didn't look very hard though.
-
Tremstats essentially is sorted by score per death.
If you look, the efficiency ratio is actually the sum of your kill:death ratio and your destruction:death ratio.
Try crunching score per death for all of the top players. I didn't find anyone who was out of order for score/deaths.
I didn't look very hard though.
I'm referring to Tremstats Too. Go there now and check out the "top" three players. None of them have half the score I do, and yet I'm just over 50th place (for example).
score, because it includes destroyed buildings.You can be a hard feeder while you are effective at nading/sawing buildings.
YES!
-
It should be something else, such as score/deaths or Dave's gain/feed. Score itself is worse than KDR because the players who play the most get the highest score. People like equilibrium, cadynum, dracone, and meisseli illustrate this well, competing with warlock in score/deaths, but not nearly as close to him when ordered just by score.
The current system doesn't reward attacking bases nearly as much as score/deaths would. Fed (an awesome MD) is a good example of this.
however conzul, you're assertion that you have more score than the top three players is misleading, as they have more than double the score/deaths as you do.
-
Score itself is worse than KDR because the players who play the most get the highest score.
Um. No?
(http://blog.mlive.com/taking_notes/2007/09/Mega%20Millions.jpg)
To win one must play.
What's wrong with that? Players who play less shouldn't get higher scores than those who play more. Where did you get that?
-
People like equilibrium, cadynum, dracone, and meisseli illustrate this well, competing with warlock in score/deaths, but not nearly as close to him when ordered just by score.
such as 40+ positions below him. Equilibrium has ~48 score/deaths while warlock has ~32. Equilibrium is in position 48 for score, while warlock is in position 1.
-
As of right now, in Tremstats Too, there is no indication of score/deaths.
-
Has score, has deaths. You do the math.
-
Oh crap. I think I just got what you meant.
-
Score itself is worse than KDR because the players who play the most get the highest score.
Um. No?
What's wrong with that? Players who play less shouldn't get higher scores than those who play more. Where did you get that?
Presumably because we want tremstats to show information that is actually useable/interesting. If it just showed players who played the most... who cares it's just the highest grinders, and doesn't represent skill at all. If it were score/life (what seems the best) (with maybe a minimum total score to get rid of one timers chances) than you can see who actually is a good player. That is useful information.
-
Maybe score/time played. Or grab the data from the site, put them into exel or something and come up with some acceptable "skill" formula.
Btw. do you guys really give a shit about who's the best on tremstats?
-
Not really, but if its going to be changed, it may as well be improved.
-
Btw. do you guys really give a shit about who's the best on tremstats?
Mkay so some guy ingame starts BSing about how great he is. I'd want to look that up. Also, it's handy to know what you're up against. So Yes.
-
Mkay so some guy ingame starts BSing about how great he is. I'd want to look that up. Also, it's handy to know what you're up against. So Yes.
Whether or not you know FrankTheN00BSpanker has a 10 efficiency doesn't change the fact that he has a 10 efficiency. The only think that's useful to know about your enemy is tactics and you can only get minimal tactical information from Tremstats.
-
Mkay so some guy ingame starts BSing about how great he is. I'd want to look that up. Also, it's handy to know what you're up against. So Yes.
Whether or not you know FrankTheN00BSpanker has a 10 efficiency doesn't change the fact that he has a 10 efficiency. The only think that's useful to know about your enemy is tactics and you can only get minimal tactical information from Tremstats.
I'm not sure what you're arguing here. You don't find the information useful. I get that. Many people (myself included) find the information useful and interesting to see. Should we take down tremstats?
-
Mkay so some guy ingame starts BSing about how great he is. I'd want to look that up. Also, it's handy to know what you're up against. So Yes.
Whether or not you know FrankTheN00BSpanker has a 10 efficiency doesn't change the fact that he has a 10 efficiency. The only think that's useful to know about your enemy is tactics and you can only get minimal tactical information from Tremstats.
I'm not sure what you're arguing here. You don't find the information useful. I get that. Many people (myself included) find the information useful and interesting to see.
^ This.
I will treat a player very differently ingame depending on if I know if he is a total nOOb, or a veteran. It irks me when a 20-50 hour player will PM me because he thinks my barricades are stupid. Because he doesn't know that I'm a 300+ hour player (or other stats on me), he doesn't know that I have the sense to take them down at s2, so he PM's me with BS advice. Tremstats is a very useful tool, and if it indeed ranks by score-per-death NOW, then I'm fine with it. I thought It ranked by KDR because only those with the highest KDR are "top players". I may die once every 8 seconds, but I may take out a helluva lot of human infrastructure in that time. I may be playing motorman (a lead alien who draws fire and dies, but the alien group as a whole PWNs the human group). There are things unrelated to KDR that an individual can do for their team.
-
Mkay so some guy ingame starts BSing about how great he is. I'd want to look that up. Also, it's handy to know what you're up against. So Yes.
Whether or not you know FrankTheN00BSpanker has a 10 efficiency doesn't change the fact that he has a 10 efficiency. The only think that's useful to know about your enemy is tactics and you can only get minimal tactical information from Tremstats.
I'm not sure what you're arguing here. You don't find the information useful. I get that. Many people (myself included) find the information useful and interesting to see.
^ This.
I will treat a player very differently ingame depending on if I know if he is a total nOOb, or a veteran. It irks me when a 20-50 hour player will PM me because he thinks my barricades are stupid. Because he doesn't know that I'm a 300+ hour player (or other stats on me), he doesn't know that I have the sense to take them down at s2, so he PM's me with BS advice. Tremstats is a very useful tool, and if it indeed ranks by score-per-death NOW, then I'm fine with it. I thought It ranked by KDR because only those with the highest KDR are "top players". I may die once every 8 seconds, but I may take out a helluva lot of human infrastructure in that time. I may be playing motorman (a lead alien who draws fire and dies, but the alien group as a whole PWNs the human group). There are things unrelated to KDR that an individual can do for their team.
Moose: I'm not saying the information isn't interesting; I'm saying it's not useful. It's good for determining how oneself does in certain situations but even then it's not all that great. It's pretty to look at and it's certainly rather entertaining but it's got no use tactically, which is what our buddy Conzul claims.
Fancy segue.
Conzul: You're going to be pissed off X_AMOUNT if Y_PLAYER tells you to fuck off with the barricades. Knowing if Y_PLAYER has Z_SKILL doesn't affect that. Knowledge of your opponent's time played, kill/death, favorite map, and win/loss is completely and utterly useless. You're going to play hard either way and you're going to learn about your opponent's tactics dynamically and on the fly as Tremstats is incapable of giving you tactical information. It spits very, very basic numbers at you which do not help you make tactical decisions pre, mid, or post match.
-
Conzul: You're going to be pissed off X_AMOUNT if Y_PLAYER tells you to fuck off with the barricades. Knowing if Y_PLAYER has Z_SKILL doesn't affect that. Knowledge of your opponent's time played, kill/death, favorite map, and win/loss is completely and utterly useless. You're going to play hard either way and you're going to learn about your opponent's tactics dynamically and on the fly as Tremstats is incapable of giving you tactical information. It spits very, very basic numbers at you which do not help you make tactical decisions pre, mid, or post match.
I didn't say it helps tactically. I know that. I helps strategically. There are so many examples of this that should be obvious, I'm not even going to post them.
-
Conzul: You're going to be pissed off X_AMOUNT if Y_PLAYER tells you to fuck off with the barricades. Knowing if Y_PLAYER has Z_SKILL doesn't affect that. Knowledge of your opponent's time played, kill/death, favorite map, and win/loss is completely and utterly useless. You're going to play hard either way and you're going to learn about your opponent's tactics dynamically and on the fly as Tremstats is incapable of giving you tactical information. It spits very, very basic numbers at you which do not help you make tactical decisions pre, mid, or post match.
I didn't say it helps tactically. I know that. I helps strategically. There are so many examples of this that should be obvious, I'm not even going to post them.
I call semantics.
Tactic =/= strategy?
Definitions of tactic on the Web:
* tactic: a plan for attaining a particular goal
Definitions of strategy on the Web:
* Strategy refers to a plan of action designed to achieve a particular goal
NO! Tactic = strategy!
I will say again (in more detail) because you've yet to refute me and have possibly forgotten:
Knowing your opponent's time played will not reduce his skill or make you play harder.
Knowing your opponent's K/D ratio will not make him kill less or die more or make you play harder.
Knowing your opponent's best weapon will not make you stronger against him when he uses it or make you play harder.
Knowing your opponent's greatest weakness will not make you any better with that weapon or make you play harder.
Knowing your opponent's most built structures will not better prepare you for his base or make you play harder.
Knowing your opponent's most destroyed structures will not make you better prepare your base or make you play harder.
Knowing your opponent's favorite stage will not make you any stronger when he reaches it or make you play harder.
Knowing your opponent's build order, build style, positioning of buildings and self, favorite routes, coordination with teammates, etc., will help you prepare for him. This information is not supplied by Tremstats.
You should be playing to the best of your ability all the time. If you have the patience to search for every opponent on Tremstats but not to play hard every game there is something very very wrong with you. Everything you'll learn about your opponents to give you a strategical (or tactical - same thing) advantage you'll learn as you fight them. This is because there's no strategy to fighting a higher K/D player than a lower. You dodge when you see him, kill when you can, and eat the base the rest of the time.
There's no strategy to fighting a player that likes las compared to pulse. You dodge. Yes, you dodge differently, but you dodge. You take the shots you need to to kill him and you dodge the ones you don't. And there's even less difference between playing against one pulser compared to another. You're going to come across every possible weapon in every match and two pulse rifle users are the same. One will be statistically worse but knowing which one will not help you kill him. Knowing how he moves and when he shoots will and you won't learn this from Tremstats.
There's no strategy to fighting someone who's weak against mara - you simply always use whatever class you're best at. No one's weak against a weak mara unless they're weak in general in which case you use your best class. ETC. ETC. ETC.
-
I like tremstats because I can figure out everyone's aliases.
-
Conzul: You're going to be pissed off X_AMOUNT if Y_PLAYER tells you to fuck off with the barricades. Knowing if Y_PLAYER has Z_SKILL doesn't affect that. Knowledge of your opponent's time played, kill/death, favorite map, and win/loss is completely and utterly useless. You're going to play hard either way and you're going to learn about your opponent's tactics dynamically and on the fly as Tremstats is incapable of giving you tactical information. It spits very, very basic numbers at you which do not help you make tactical decisions pre, mid, or post match.
I didn't say it helps tactically. I know that. I helps strategically. There are so many examples of this that should be obvious, I'm not even going to post them.
I call semantics.
Tactic =/= strategy?
Definitions of tactic on the Web:
* tactic: a plan for attaining a particular goal
Definitions of strategy on the Web:
* Strategy refers to a plan of action designed to achieve a particular goal
NO! Tactic = strategy!
Not sure I agree with Conzul's points or anything, but strategies and tactics are most assuredly different things.
-
Not sure I agree with Conzul's points or anything, but strategies and tactics are most assuredly different things.
Pics or it didn't happen.
In other words explain.
And, dammit mooseberry, you overtook me for the highest poster.
-
Not sure I agree with Conzul's points or anything, but strategies and tactics are most assuredly different things.
Pics or it didn't happen.
In other words explain.
And, dammit mooseberry, you overtook me for the highest poster.
This website (http://ezinearticles.com/?The-Difference-Between-Strategy-and-Tactics&id=1558273) seems to give a good explanation.
Yesh, I hacked to get top poster. >.>
[/hijack]
-
This may be a bit offtopic, but what if Tremstats only calculated your place in the "Top Players" list based on the data (KDR, Score/death, or whatever) from the last 30 days or another reasonable length of time. This would make the rankings more dynamic and also would allow newer players to shed their initial games where they have over 50 deaths and only a few kills. This also has the added benefit of moving inactive players off the rankings and let the active players take up their spots.
I'm not sure how useful this would be, but, what if Efficiency was measured by score/time. That way, number of deaths don't matter in stats, only how ever much you score. Also, if you build, your stats do not suffer as much.
-
Strategy is quite different from tactics. Sipmlified, here's how I understand it:
A strategy is PLAN that you make before the outset of the conflict. It's a plan that you will try to stick to, because if everything happens according to your strategy then you expect it to go well.
A tactic is a somewhat reusable course of action that has no specific context. You use a tactic within the scope of your strategy. A tactic is an action or set of actions that you perform during a conflict when a means to fullfiling your strategy becomes more likey to succeed than you had thought before. Use moose's link, Sun Tzu is badass (I've read his book, UberNerd).
I said I wouldn't, but you need help. Oh and, yes, I'll browse Tremstats to look up any player who I'm seeing with any regularity. I usually don't remember much more than their total hours played, and human play/alien play ratio.
*1* So it's a small game and I'm on aliens. 3 humans join. One is "><Warlock", one is "-sg-aequitas", and one is "arnold". Based on what I don't know about arnold (from tremstats), I assume that he is new, and I formulate the strategy that when I rush the human base, I will dretch HIM. It will be easy to know which he is because new players are allways looking in circles, or walking to odd places in the map.
Another example:
*2* I notice that KiLL3R (fictional) joins. I'm thinking of going alien builder, but I suddenly remember that on his kill-graph, there is a large green spike right near the beginning. This lets me know that KiLL3R likes to rush, and that I shouldn't rebuild the Overmind and pray that he trips on acid. So I make the tactical decision to evolve and dretch him.
Sheesh.
-
voting for player score
-
Situation 1 is incredibly stupid because you will learn about the skill level of your enemies the moment you either kill them or get killed by them.
Situation 2 is really complicated. Why don't you go alien builder and coordinate a proper move? Your base is rather inadequate at defending itself, anyway. And what the hell are you doing moving the OM at S1 beginning of the match? That's what eggs are for. If he rushes then you'll evolve. If he doesn't you'll build. These decisions are made on the fly.
-
I must agree with plague bringer here. Admittedly it can be useful to keep track of what important players such as warlock are doing (defending base or attacking yours), but the real applications of tremstats are completely outside of in-game situations.
Its makes tremulous more fun if you try to avoid stacking one side too heavily. If you know who is good and who isn't (which tremstats isn't too shabby at, provided you correct for players that have been scumming for position), you can place yourself in such a way that you have a better chance of having a good game.
Clan recruitment. Who is this guy you don't know anything about? Look at first seen, last seen, activity,
Pride. Sometimes it just feels good to have a metric of how you're doing, its why commercial games have an "achievements" section. Of course, this can be abused, but I don't think its that hard to pick out a poser.
Is this so important that tremstats needs to be changed before 1.2 (or ever, for that matter)? No. Actually, I don't really think this is that important at all, just trivial to implement. But I don't see why making such metrics more accurate is a bad thing. So unless someone things 1.2 could absolutely not be released without this change, I do ask that we discuss what we are here for, not whether this matter is important, but "what would be the best metric we could have for this?".
-
Oh I'm not worried about having it fixed in time for 1.2.
Situation 1 is incredibly stupid because you will learn about the skill level of your enemies the moment you either kill them or get killed by them.
It's handy to know beforehand. Just because you don't see the value in it, doesn't mean that it doesn't have any. (and I'm not trying to be rude).
Situation 2 is really complicated. Why don't you go alien builder and coordinate a proper move? Your base is rather inadequate at defending itself, anyway. And what the hell are you doing moving the OM at S1 beginning of the match? That's what eggs are for. If he rushes then you'll evolve. If he doesn't you'll build. These decisions are made on the fly.
For instance, in atcs, I can spawn as a granger the moment the game starts (or countdown ends) and replace the Overmind backwards 3 feet. This helps alien winning chances quite well, so I do this whenever I can. (LOL not anymore :P ) . If I know that a rusher is on the foreign team, I won't do this the moment the game starts, I'll wait for a minute or two until I detect the mood. (Human team has three moods: Camp-happy, feed-happy, or creep-happy).
-
Oh I'm not worried about having it fixed in time for 1.2.
Situation 1 is incredibly stupid because you will learn about the skill level of your enemies the moment you either kill them or get killed by them.
It's handy to know beforehand. Just because you don't see the value in it, doesn't mean that it doesn't have any. (and I'm not trying to be rude).
Situation 2 is really complicated. Why don't you go alien builder and coordinate a proper move? Your base is rather inadequate at defending itself, anyway. And what the hell are you doing moving the OM at S1 beginning of the match? That's what eggs are for. If he rushes then you'll evolve. If he doesn't you'll build. These decisions are made on the fly.
For instance, in atcs, I can spawn as a granger the moment the game starts (or countdown ends) and replace the Overmind backwards 3 feet. This helps alien winning chances quite well, so I do this whenever I can. (LOL not anymore :P ) . If I know that a rusher is on the foreign team, I won't do this the moment the game starts, I'll wait for a minute or two until I detect the mood. (Human team has three moods: Camp-happy, feed-happy, or creep-happy).
This is turning into a real debate of strategy.
Player and team mood changes all the time. It's not constant and it differs not only from game-to-game but minute-to-minute.
Okay. So, this player, RUSHERofN00bZ is incredibly good taking down A early game, right? If you're getting rushed right off the bat it's for one of two reasons.
ONE: You're being cheesed. If your opponent is 4-pooling you (lol StarCraft), their base is that much more undefended and they're only doing it to put a lot of early game pressure on you and chance a win. A base rush within the first minute is going to be unsuccessful unless it's a coordinated team effort, in which case you're fucked whether or not you know it's coming. Being better prepared is not going to help you. Acids do very little to help in these situations.
TWO: You're being stomped. The opposing team is a party (lol Modern Warfare 2), and they're more coordinated than you. You can kiss your ass goodbye because you're going to be spanked so hard it's going to fall off.
Now, let's assume for a moment that I'm completely wrong.
This is a hypothetical:
In this hypothetical, there's you, Conzul, and a small (eight to ten person) game on US1. You join five, maybe ten minutes in and there's plenty of time left. You don't feel pressure to join a team right away because the game's progressed so much and so slowly so you take your time and search for everyone on Tremstats. By the time you finish figuring out who's who, the game's coming to a close and you have no time to play. You start the next game thinking you know the dirt on everyone in the match; thinking you've got their strategy - their big plan - figured out.
No.
You know their tactics.
No. Even less.
You know roughly their potential tactics.
Now, had you not checked Tremstats and played the game you would've learned plenty more about your opponents and been more prepared for the next game.
There is a second hypothetical about joining a game during warmup but I don't think I have to tell you why taking the time to get the dirt on your opponents this early in a match (for the purposes of predicting their movements in the first 20 seconds of the game [which is all you could ever do with Tremstats]) is backwards and self-defeating.
-
Mkay for one, I'm no where near that dependent on tremstats. I don't watch to see who's on and mentally bring up their rap sheet. However, some stats shown in tremstats can bring up a red flag for me when I see the player in game. And it's occurring to me now, I might be the only one who does this.
-
It has to be teamkills. No team could hope to win without the best teamkillers around.