|
|
| Author |
Thread Statistics | Show CCP posts - 8 post(s) |
John Demonsbane
Unorganized Ninja Infantry Tactics
4354
   |
Posted - 2014.10.31 04:52:00 -
[1] - Quote
CCP Rattati wrote:Leadfoot10 wrote:
Rhetorical Question: How can one hope to balance weapons if the one in charge of balancing them has a never used the gun?
Numbers are certainly important, but there's nothing like looking down the sight and firing it.
That statement is a logical fallacy, sorry. How does an engineer design an airplane if he has never flown one? How does a scientist study gravity if he has never seen it? How does the physicist describe quantum physics without touching atoms? I can make more of these... It's usually called the Scientific Method, http://en.wikipedia.org/wiki/Scientific_method
Am a scientist, can confirm.
(The godfather of tactical logistics)
|
John Demonsbane
Unorganized Ninja Infantry Tactics
4366
|
Posted - 2014.10.31 18:25:00 -
[2] - Quote
Jaysyn Larrisen wrote:
First, good on you and Leadfoot for the mature and positive exchange. Nice to see that on the forums.
Your post above actually helps me understand much more about your vision for balancing (keeping variants balanced vs balancing across whole weapon lines) which makes quite a bit more sense. I would point out that with FOTM and "op" stuff you get to a point in usage that actually creates the cascade effect that you see with RR usage & kills due to player perception. This is similar to my point about reading these trends and how it isn't dissimilar from observing trend analysis in the stock market or other statistically based activites measureing human dynamics and perception.
Basically, when Player X starts playing Dust (or has already been playing for a while) and he sees / hears that a lot of folks like the RR and maybe get's killed by one 5 or 6 times in a match he very naturally starts thinking "I think I need to skill into that RR". At that point, he puts some SP down on it and buys a bunch of SB-39s.
Bare in mind that Player X's K/D may well only vary slightly up or down with the RR versus his previous weapon of choice but he's now further contributing to the usage / kill stats that you accurately track and account for. He's contributing to the percpetion effect that other players see and feel...that they need to skill into the RR or whatever weapon we are talking about to remain competitive but perhaps not actually becoming that much more competitive.
Bottom line: you can possibly (not all the time) find yourself in a situaiton where kills and usage rates are noticably higher and it may not inidcate a vast increase in combat effectiveness for the individual player.
Please note, this isn't about the RR per se, it's just my observation of how player dynamics work and that perhaps it needs to be considered.
You're 100% correct that usage =/= OP, and that the fact that people flock to a weapon alone doesn't mean it's OP, but it should certainly raise the eyebrows of the person who's job it is to track such things.
The example you are giving, to me, is actually what Rattati is getting to in point #1: "even though players migrate over to RR, there is no dilution of efficiency as might be expected"
Take the Scrambler. It's not easy to use, but wrecks in the right hands, like Rattati said. It was around far longer than the RR but you never really saw the same migration to it. Because it's not easy to use.... it's very abusable with modded controllers or macros, but otherwise it's mechanics limited it. So, if every blueberry in the world started using it, you'd better believe the K/S would drop.
Didn't happen with the RR. All comers, from all the scrubs who needed the eHP of a sentinel to stay alive to every brainless noob in my local, they all started using it. And still did well.
50% of the people playing this game are not really good, are they? I doubt you'd answer yes to that.
That's the crux of the whole thing. The overall skill level here is not high enough that it makes any sense that if everyone started using a weapon because it killed them all the time, the overall K/S of the weapon wouldn't drop. It would have to strictly based on regression to the mean, if nothing else!
(The godfather of tactical logistics)
|
John Demonsbane
Unorganized Ninja Infantry Tactics
4377
|
Posted - 2014.11.01 02:35:00 -
[3] - Quote
Jaysyn Larrisen wrote:John Demonsbane wrote:You're 100% correct that usage =/= OP, and that the fact that people flock to a weapon alone doesn't mean it's OP, but it should certainly raise the eyebrows of the person who's job it is to track such things. The example you are giving, to me, is actually what Rattati is getting to in point #1: "even though players migrate over to RR, there is no dilution of efficiency as might be expected" Take the Scrambler. It's not easy to use, but wrecks in the right hands, like Rattati said. It was around far longer than the RR but you never really saw the same migration to it. Because it's not easy to use.... it's very abusable with modded controllers or macros, but otherwise it's mechanics limited it. So, if every blueberry in the world started using it, you'd better believe the K/S would drop. Didn't happen with the RR. All comers, from all the scrubs who needed the eHP of a sentinel to stay alive to every brainless noob in my local, they all started using it. And still did well. 50% of the people playing this game are not really good, are they? I doubt you'd answer yes to that. That's the crux of the whole thing. The overall skill level here is not high enough that it makes any sense that if everyone started using a weapon because it killed them all the time, the overall K/S of the weapon wouldn't drop. It would have to strictly based on regression to the mean, if nothing else! John, I see your point, however there is no way that we can really know what the relative combat effectiveness of players are before and after a weapon migration point. I agree that at least 50% of the players are below average...obvious statistical fact. My point is that if you had an average KDR of 1 before the switch from PR to RR and then have the same KDR it's not about the weapon being OP. OP weapons should create a noticeable and significant uptick in player combat statistics at the individual player level. There is a bit of just finding a weapon style that fits your play style and that is in effect here as well. I would also point out that many on the forums tout being able to "rek face" with just about any weapon they lay hands on and I tried this or that and did fine. Some of this is true and some not...but individual player skill is ultimately the greatest leveler and the most difficult thing to account for in stats in a game like this.
Lol, yes certainly you can't trust how much awesomer everyone on the forums is. Or base any kind of weapon balancing on forum buzz. I used to follow regynum around on the forums and post:
"Results not typical. Your results may vary, Regynum is not approved by the FDA to diagnose or treat any disease."
And you could certainly be right. Maybe not one single persons KDR changed when switching to the RR. And maybe MIna is right and there were not that many RR kills under 40m.
Personally I doubt that seriously, more so Mina's contention than yours. Thing is, though, even if CCP didn't track it, you would honestly have to assume that, given the objective-based nature of most of the game modes, there was a lot of CQC involved.
(If you didn't you would run the risk of being anti-science in that you were using your opinion as a basis of interpreting data instead of the other way around,)
(The godfather of tactical logistics)
|
John Demonsbane
Unorganized Ninja Infantry Tactics
4384
|
Posted - 2014.11.01 19:51:00 -
[4] - Quote
Leadfoot10 wrote:
The Gallente have their AR which is superior at close to mid range. Same with the minmatar CR. The Amarr has the ScR, which is obviously takes a bit more skill, but has no niche as it works well at CQC and range. What do the Caldari have? The worst CQC weapon already just got 4x harder to control, and the delay increased by 60%.
Well, the ScR's niche is as a skill-shot weapon but I digress.
Leadfoot10 wrote:
And we have no idea how well or not the RR truly worked in CQC in comparison to other guns. All we have is anecdotal evidence and a lot of differing viewpoitns. A few of us have been calling for K/S statistics to include some range information so we can see what's really going on. I've not seen anything. Have you?
Excellent point, but... what number would we look for? How much of a dropoff should we say is acceptable? K/S down to 1? 1.5? 30% of kills coming in CQC? 20?
What you, I, or Rattati think is appropriate efficacy is more important than the numbers themselves (unless they are obvious, like no dropoff at all or over 1/2 the kills coming inside 40m).
Data are only as useful as the cut-points that are applied to it.
Leadfoot10 wrote:Fox Gaden wrote:The RR was OP in the sense that it was the best weapon in the least skilled hands. It was too easy for low to mid level players to get exceptional results with it. It was too easy to use. Again, simply wrong. It was the most popular weapon, not the best. If it was the best, it would have had a clear K/S advantage. The data shows it does not...despite having a range advantage that would lead towards higher K/S situation -- killing at range.
That point is simply not as clear-cut as you think it is. Again, if a large number of players, which statistically would have to be lower skilled on the basis of sheer numbers alone, start to use a weapon, it by all logic should have had a K/S drop as the migration occurs. (Total kills, as you noted before, would obviously go up and be meaningless in terms of evaluating weapon performance)
The contrary is much more true: If it was used by the majority of the players and had any K/S advantage, no matter how small, it would be de facto OP.
What might be interesting is to see similar data for the old TAR (which nobody would argue was not OP), did it's K/D change as people migrated to it?
Leadfoot10 wrote:And the most telling question of all: If the RR was so superior, why was it only on one of your fits? I'll wait for your answer to that one.
That's obviously a lolargument. There's any number of reasons someone would not use the RR.
- For one, Fox runs Sentinel a lot afaik, and no self-respecting vet would catch themselves dead running RR sentinel on the regular.
- Myself, I have zero fits with the RR. The charge up time annoys me, and also as an LR user I consider the RR to be the spawn of satan himself and refuse to use it. More importantly, the Amarr logi suits I use are terribly gimped on PG and so the CR is a perfect fit because it basically uses up none.
- Some people wouldn't for RP reasons or to take advantage of the bonus of a suit they like for other reasons.
I can't use any more quotes, but your point about PC is totally valid. However, like you said, we do need to balance for both and we get into the same thing about how much of as disadvantage in PC would balance out an advantage in pubs?
(The godfather of tactical logistics)
|
|
|
| |
|