Higby on Magriders - Page 6

Rothnang · 2013-02-19, 06:26 AM

I'm not saying the graph isn't a piece of evidence, just that it isn't enough to call the entire debate.

There are a ton of metrics we aren't seeing, like how the tanks stack up against each other when it comes to killing other things than tanks for example, or how often they get killed by infantry or aircraft.

There is a very real possibility that all the data supports exactly what Higby is saying, and he's spot on in the claim that the changes are the total package when it comes to balance. If we're going to use metrics as definitive proof of it one way or another though I'd like to see all the data.

I still believe the Magrider has a low skill ceiling at this point, and that some tanks got unintended buffs, like Prowlers being a pretty good AA unit now because of absurdly fast projectiles.

Assist · 2013-02-19, 07:31 AM

K/D graphs of specific classes/vehicles/etc are irrelevant because they don't take population size into account.

Look at the screenshot provided by one player, where a new player clan on Waterson was running a conga-line of Magriders. Do you think the K/D of those tanks vs other tanks is going to be even ? No, not even close. Firepower of numbers supercedes firepower of individuals.

K/D in tank vs tank is a HORRIBLE WAY TO BALANCE. I've stated this for so long it's depressing to see people argue for it still. You don't balance a game based on these metrics, just like the game shouldn't be balanced on the NC AI MAX compared to the VS/TR AI MAX. SO WHY IS IT OK TO IT IN THIS SITUATION BUT NOT IN OTHERS? Because people are bad at the game and don't know how to play, so we must cater the game towards them. That's the only answer I can come up with.
One side spawns 50 Lightnings a day, another faction spawns 500 Lightnings a day. Whose going to have a higher K/D at the end of the week? Imagine you're defending a tech plant, you spawn a Lightning vs. the 50 Lightnings sitting outside, how many is your solo lightning going to kill? ZERO. Imagine you're taking Indar Excavation with your 10 buddies and a single tank comes from Hvar/Allatum, is he going to have a high effective K/D? What if 100 tanks come from Hvar/Allatum, how is your K/D looking? Population matters immensely in K/D numbers why does no one understand this.
These metrics for K/D are a complete joke and the dev's putting them out there as the reason for balance is pathetic. Some of you don't stop and even think. Run the numbers for a certed Prowler vs. certed Vanguard battle and explain to me how it's possible for the Prowler to win in that 1v1 situation? I'll give you a big hint, it's NOT. Yet that graph states otherwise.

Population size is incredibly important in K/D of vehicles, anyone who believes otherwise is ignorant of the way the game plays.

Rothnang · 2013-02-19, 07:38 AM

Also a good point. This whole graph thing has been pretty disappointing to me, because it just seems like they just busted out something that looks right to people who don't think about it too much to kill the discussion.

Fewer people are using Magriders now than before, that's for sure. The number of prowlers and Vanguards doesn't seem to have gone up significantly as far as I can tell though.

At any rate, we need more meaningful metrics to settle this argument once and for all.

Twido · 2013-02-19, 07:52 AM

The data is really interesting and should not be dismissed because it doesn't go into detail. It shows the big picture and what has happened overall, most of the details being argued arn't even faction specific and therefore not so relevant.

I don't think they are finished with balancing, but I don't think anyone can argue that the balance isn't closer now than it was a month ago.

Rothnang · 2013-02-19, 08:22 AM

I disagree that the data shows the big picture, to see the big picture you have to have a rough understanding of all the relevant information.

Like I said before, what if it turned out that the Magrider was getting half as many infantry kills as the other tanks before the patch. Don't you think that kind of information would be pretty essential to having a "big picture"?

I'm not saying that that's the case, but that's the kind of information the graph simply doesn't give you.

The graph is simply misleading. It shows a bunch of bars that make it look as though before the patch the Magrider was twice as good as the other tanks and now they are all even. If you really delve into what the graph means and what it doesn't give any information on though it becomes pretty clear that there is more to it than just what we see.

Twido · 2013-02-19, 08:39 AM

You are so close to understanding the statistics! You are right about the graph not showing how many infantry the tanks can kill, and I expect that you are right that the prowler can kill more than the magrider (although I have no data on this and could be wrong).
Now instead of looking at what the graph doesn't show, try to look at what it does show remembering that it is a broad overview.

Assist · 2013-02-19, 09:01 AM

Originally Posted by Twido

You are so close to understanding the statistics! You are right about the graph not showing how many infantry the tanks can kill, and I expect that you are right that the prowler can kill more than the magrider (although I have no data on this and could be wrong).
Now instead of looking at what the graph doesn't show, try to look at what it does show remembering that it is a broad overview.

What does it show? That pre-patch Magriders had a higher effective MBT K/D in battles that had MBT's that hit each other and at least one of them died.

It doesn't show that Magriders had a higher K/D in 1v1 MBT fights. It doesn't show the number of spawned MBT's. It doesn't show the number of killed MBT's. It doesn't show how or what killed the MBT's. It doesn't show anything that reflects maneuverability. It doesn't show which MBT did more damage in those fights. It doesn't show how many MBT's did damage in those K/D ratios. It doesn't show if any other vehicle, unit, friendly, or rock was involved in those K/Ds. It doesn't show many deaths to infantry. It doesn't show how many deaths to air. It doesn't show how many MBT's are spawned pre/post patch. It doesn't show a ridiculous number of things, and all it truly shows us is that numbers like K/D are the worst thing in the world to balance an evolving game off of because they don't tell even a tenth of the story.

Would you like weapons to be balanced by K/D? Should we compare infantry units by K/D? Do we need to show that players who get a .5 K/D need a buff compared to players with a 3.5 K/D? When does it stop and why does it stop? Why does it only take effect in this situation and not others? If you want to balance a game by K/D then it needs to have specific parameters to balance off of. CoD, BF, CS can all be balanced by K/D to an extent, because they have specific number of players in every fight. There's nothing consistent in every fight in Planetside 2. That's why we play it, because it's always different. So why are we balancing the game as if everything is a controlled fight?

What this really boils down to is there's a huge difference between statistical data and the interpretation of that data. Using the data as the interpretation is not how statistics are meant to be used.

Rothnang · 2013-02-19, 09:02 AM

That's exactly the problem though, it isn't a broad overview. All it shows is that in tank vs. tank battles the Magrider did better on average.

It doesn't show why the Magrider did better though, and it doesn't show if it did better in every single area. We need other data to make any kind of statement about what this particular graph means other than just "Magriders killed more tanks than other tanks and now they don't."

What's most troubling here is simply the fact that I can't possibly imagine that the power of Vanguards and Prowlers hasn't gone up drastically against all other units as well, and unless the Magrider was outkilling them by 2 to 1 for every other unit as well that should skew a lot of other statistics significantly in favor of Prowler/Vanguard.

psijaka · 2013-02-19, 09:02 AM

Originally Posted by Rothnang

And your argument that the average on tank to tank K/D proves that the tanks are perfectly balanced at all levels of play is also pure speculation.

Like I said, if they can show us metrics that show the actual distribution of kills achieved (and not just tank kills) within the tanks and it turns out that top end Prowler and Vanguards aren't doing any better than top end Magriders since the patch I'll concede that I'm wrong about all of this.

Where did I say that tanks are perfectly blanced at all levels of play?

I would love to see more detailed data released, but this then poses another problem: how do we determine who is a "High End" player and who is "Mr Average" or "Newb". I suspect that it would be difficult to capture this data; normal game k/d is meaningless as they might just be spawn farmers.

My own experience of driving tanks is limited; it's just not my thing; but I tried them all in Beta and the Magrider was the easiest to handle for a Newb; especially appreciated the strafing ability. How this translates into "High End" performance, I'm not qualified to say. The only people who are really qualified to pass judgement on this without bias are those that have mastered all of the MB tanks and do not have any affiliation to a particular empire.

I do have a lot of experience of taking on tanks playing as infantry though (NC AV MAX), and I can say without doubt that, pre-balance, the Magrider was harder to deal with than the Prowler. Their strafing ability in particular makes their movement unpredictable; tricky to hit, and enables them to keep their small armoured front profile facing towards me at all times. And their hill climbing ability pre balance enabled them to catch me out time and time again when I thought that I had the advantage on difficult ground; no such worries about the Prowler; with them I definately had the upper hand on the steep gullies to the NE of Crossroads, for instance.

Kerrec · 2013-02-19, 09:07 AM

The reason that people are not on the same page in this thread is many DO NOT understand statistics.

Rothnang seems to be convinced that when the word "average" is thrown about, that means that the data eliminates noobs and highly skilled players from the data, and what remains (average) is used in the statistics. This is THE flaw in all of his arguments.

It is statistically improbable (heh) that a large sample size will produce a situation where one team has all the highly skilled players, and the other two teams do not. That is just nonsense.

Also, situations where one team has a 20:1 ratio in a head on battle can happen for one team on one server at one time. However over time, the same will happen for another team, on some other server, at some other time. Over a large sample size, all the situations end up being represented. It is wrong (to be correct, it is not probable) in Statistics to assume otherwise.

The graphs (the data) itself is now suspect due to the IMPROBABILITY that two graphs for two very different data sets end up identical. However, if the assumption is made that the data is correct, then there is no "situation" that is not represented within the data.

psijaka · 2013-02-19, 09:10 AM

I would very much like to see data on the overall K/D of the various MBTs, including all kills and deaths whilst driving the tank; not just Tank v Tank.

bpostal · 2013-02-19, 09:14 AM

Originally Posted by BlaxicanX

It's hard to hate the guy when you read his posts. He seems so reasonable.

Can't comment much on the Magrider. From an NC/TR perspective, I've never had problems fighting Mags that I haven't had fighting the other tanks. So.

I know! I can't think of a lower level of hell than having to play as VS. On the day off no less.
Also, idk about you but I'm loving the new AP lockdown mode

Rothnang · 2013-02-19, 09:17 AM

Originally Posted by psijaka

I would love to see more detailed data released, but this then poses another problem: how do we determine who is a "High End" player and who is "Mr Average" or "Newb". I suspect that it would be difficult to capture this data; normal game k/d is meaningless as they might just be spawn farmers.

Well, the data he has is the result of averaging out all the tank kills by Magriders, right?

So let's say we take all the individual datapoints in that average, and plot them out into a curve.

Let's say we have 100 Magrider samples.
39 of them scored 0 tank kills,
38 of them scored 1 tank kill,
14 of them scored 2 tank kills,
7 of them scored 3 tank kills,
2 of them scored 4 tank kills,

Then we have 100 Vanguard samples.
48 of them scored 0 tank kills,
20 of them scored 1 tank kill,
15 of them scored 2 tank kills,
9 of them scored 3 tank kills,
5 of them scored 4 tank kills,
3 of them scored 5 tank kills

In this hypothetical data set what you would see is that much fewer people do OK in a Vanguard than in a Magrider, but there are more people who do really well in a Vanguard than in a Magrider.

This kind of graph would tell us something about the skill distribution in the vehicle, and it would show that certain vehicles are much harder to learn, but stronger when you do master them.

The average simply doesn't tell the whole story, distribution does matter because it shows if everyone is clustered closely around the average, or if its a mass of data points on one side that are balanced by some huge outliers on the other... there are all sorts of things that only the distribution can tell you.

What if a tank with an average K/D of 1 had 99 out of 100 people getting 0 kills, and then one guy who got 100 kills. Don't you think that would be noteworthy, despite the fact that it doesn't show up on the average?

Twido · 2013-02-19, 09:23 AM

Originally Posted by Rothnang

That's exactly the problem though, it isn't a broad overview. All it shows is that in tank vs. tank battles the Magrider did better on average.

It doesn't show why the Magrider did better though, and it doesn't show if it did better in every single area. We need other data to make any kind of statement about what this particular graph means other than just "Magriders killed more tanks than other tanks and now they don't."

Thats exactly it! All it shows is that in tank vs tank the Magrider did significantly better on average. That is important isn't it?

It is not so important to find why the magrider was better and then nerf that aspect into the ground; we don't want the tanks to become carbon copies of each other.

From the statistics shown, I would say the prowler could lose about 5% damage from it's HEAT and AP gun as minor tweak since it seems just a bit better than the others right now.

To reply to Assist, you must be getting tired of telling us the same thing. To be fair you do have a point, the statistics do not go into fine detail. Most of us understand that but the validity of the data still stands regardless. Most of the variations you talk about arn't even specific to one faction.

I do wonder if when they do the Max review they release a graph showing the VS Max has a worse K/D ratio than the TR and NC, will you argue that the VS Max should not be buffed because the K/D is not detailed enough?

Rothnang · 2013-02-19, 09:28 AM

Of course its important, it's just not the "Haha gotcha, the magrider was super OP and now everything is great" piece of evidence people make it out to be.