MBT K/D ratios

[Stanis]

I'm struggling with this



We have three tanks with a KD ratio.
We know they died on average 1 time each to get that D statistics.

Problem is we don't have have 3 Kills.

The MBT stats Magrider: 1.71 Vanguard: 0.87 Prowler: 0.63

1.71 + 0.87 + 0.63 = 3.21 kills

1.17:1
0.87:1
0.63:1
---
3.21:3

so .. where did the extra 0.21 tank to get killed come from?

[Assist]

Originally Posted by Stanis

I'm struggling with this

https://twitter.com/Staniskal/status/297299016759255041


We have three tanks with a KD ratio.
We know they died on average 1 time each to get that D statistics.

Problem is we don't have have 3 Kills.

The MBT stats Magrider: 1.71 Vanguard: 0.87 Prowler: 0.63

1.71 + 0.87 + 0.63 = 3.21 kills

1.17:1
0.87:1
0.63:1
---
3.21:3

so .. where did the extra 0.21 tank to get killed come from?

I'd love for them to clarify these stats because it is now everyones argument for nerfing when in fact those stats are completely meaningless without the parameters of it.

For all we know there was 20,000 Magriders, 1000 Vanguard's, and 1000 Prowlers in that sample. If they try to tell us that there's 10,000 strictly 1v1 MBT fights in a day in PS2 they're full of shit and if they try to say otherwise then why are they ignoring population in k/d numbers?

Using statistics like this to balance out a game is really dumb when you're not going to explain them. Posting them on a forum like reddit where the users believe every word they read on the internet is even worse.
If they want to use statistics like that to justify their changes, which is exactly what they tried to do by posting those, then they need to post their parameters for them. Otherwise they're just pushing propaganda. It's just like the US Government telling everyone how the unemployment numbers are dipping every time tax season rolls around, idiots buy into every year not even thinking that they're all temp workers until tax season is over.

[Meecrob]

I agree that these statistics seem a little suggestive. Are they the average stats overall per tank right now? Or just some sampleset taken by someone?

The post also seems to suggest that every unit in game should (on average) have a 1 k/d, which is just not true. Im pretty sure the average k/d of liberators is way above 1 and with fighters the same, but this is due to their role on the battlefield.

Tho if it's meant to be an argument that the magrider is more powerfull than the rest, than quit trying, it's allready a known fact that mag's are. They were designed this way and their allready working on balancing it better. (altho i think mag's will allways be the best, and this is not necessarily a bad thing).

[Figment]

Originally Posted by Stanis

I'm struggling with this

https://twitter.com/Staniskal/status/297299016759255041


We have three tanks with a KD ratio.
We know they died on average 1 time each to get that D statistics.

Problem is we don't have have 3 Kills.

The MBT stats Magrider: 1.71 Vanguard: 0.87 Prowler: 0.63

1.71 + 0.87 + 0.63 = 3.21 kills

1.17:1
0.87:1
0.63:1
---
3.21:3

so .. where did the extra 0.21 tank to get killed come from?

Can you even assume it's pure tank on tank combat K/D or that one empire has more advantageous air (more maneuvrable) or infantry AV (more reliable) that softens up the enemy tanks?

We would require a complete numerical breakdown. :/

If they survive longer, why?

• What if an empire pulls more consistently because there's more players with a tank fetish? (should be assumed pretty evenly distributed)
• What if they move in larger groups and use numerical advantages? (Presense of zergfits - should be assumed pretty evenly distributed)
• What if they are more organised? (Outfit and leader quality - should be assumed pretty evenly distributed)
• What if their home terrain and invasion routes are advantageous to vehicles or advantageous to killing enemy vehicles? (High ground over Indar's canyons and seabed, Eisa usualy in VS hands)
• What if one unit has a (re-)positioning advantage? (Mag hovering and afterburner)
• What if one unit has a dodging advantage? (Mag strafing and afterburner)
• What if one unit has more platform stability and therefore an accuracy advantage? (Mag hovering)
• What if one unit has dogfighting maneuvring advantages? (No dedicated driver for rotating turret units: bumping into objects, Magrider driver drives where he aims, while strafing)
• What if one unit is more effective at using armour increases? (Mag frontal armour usualy faces enemy, particularly at range)
• Are they using the same kind of weaponry (HEAT/HE/AP)?
• What gunner weaponry is used and how effective is it?
• Are some using more gunners, or more solo-units?
• What is the hit-ratio in direct combat between the units?

>.>

It's anyone's guess right now, even if we have some clear indicators that the Magrider does have significant advantages in several fields. Particularly hovering, Tank Destroyer setup under non-dedicated driver advantage and strafing.

Hence why the Magrider in PS1 had the least armour of the three tanks, despite of having the weakest weaponry of the three: damage over time vs endurance should be equal between the three MBTs. What happens with the Mag in the current situation is that the Damage over Time is equal, but the endurance much larger as they are missed more often due to the strafing TD platform.


It's clear from Higby's comments however that the Prowler's HE is so attractive to killing infantry, more HE is used than AP. Could this influence the statistics in tank combat? In theory it may very well do so, since farming is more appealing to the point the healthy HE-amongst-tanks ratio is broken. In essence, a Prowler HE nerf might actually make the AP more attractive and therefore their tanks more successful in tank combat as more would use AP rounds.

[psijaka]

Originally Posted by Stanis

I'm struggling with this

https://twitter.com/Staniskal/status/297299016759255041


We have three tanks with a KD ratio.
We know they died on average 1 time each to get that D statistics.

Problem is we don't have have 3 Kills.

The MBT stats Magrider: 1.71 Vanguard: 0.87 Prowler: 0.63

1.71 + 0.87 + 0.63 = 3.21 kills

1.17:1
0.87:1
0.63:1
---
3.21:3

so .. where did the extra 0.21 tank to get killed come from?

Originally Posted by Figment

Can you even assume it's pure tank on tank combat K/D or that one empire has more advantageous air (more maneuvrable) or infantry AV (more reliable) that softens up the enemy tanks?

We would require a complete numerical breakdown. :/

I'm really struggling with what these K/D figures mean. I don't see that they can be pure MBT V MBT combat as such a thing doesn't exist in reality.

Are they just the ratio of any kills obtained whilst in a tank, versus deaths whilst in the tank? In which case they seem unbelievably low for the Prowler and Vanguard.

Edit - or more likely, MBT kills per MBT death.

Whatever the figures are, they do show the superiority of the Magrider. Wish we had more info though.

Originally Posted by Figment

It's clear from Higby's comments however that the Prowler's HE is so attractive to killing infantry, more HE is used than AP. Could this influence the statistics in tank combat? In theory it may very well do so, since farming is more appealing to the point the healthy HE-amongst-tanks ratio is broken. In essence, a Prowler HE nerf might actually make the AP more attractive and therefore their tanks more successful in tank combat as more would use AP rounds.

Good point.

I'm pretty sure that more Magriders are pulled overall than Vanguards and Prowlers, which makes sense as people know that they are the superior tank, and the easiest to use.

[Sunrock]

Originally Posted by Assist

I'd love for them to clarify these stats because it is now everyones argument for nerfing when in fact those stats are completely meaningless without the parameters of it.

For all we know there was 20,000 Magriders, 1000 Vanguard's, and 1000 Prowlers in that sample. If they try to tell us that there's 10,000 strictly 1v1 MBT fights in a day in PS2 they're full of shit and if they try to say otherwise then why are they ignoring population in k/d numbers?

Using statistics like this to balance out a game is really dumb when you're not going to explain them. Posting them on a forum like reddit where the users believe every word they read on the internet is even worse.
If they want to use statistics like that to justify their changes, which is exactly what they tried to do by posting those, then they need to post their parameters for them. Otherwise they're just pushing propaganda. It's just like the US Government telling everyone how the unemployment numbers are dipping every time tax season rolls around, idiots buy into every year not even thinking that they're all temp workers until tax season is over.

I could not have sad it better my self.

[Mietz]

Originally Posted by Assist

I'd love for them to clarify these stats because it is now everyones argument for nerfing when in fact those stats are completely meaningless without the parameters of it.

For all we know there was 20,000 Magriders, 1000 Vanguard's, and 1000 Prowlers in that sample. If they try to tell us that there's 10,000 strictly 1v1 MBT fights in a day in PS2 they're full of shit and if they try to say otherwise then why are they ignoring population in k/d numbers?

Using statistics like this to balance out a game is really dumb when you're not going to explain them. Posting them on a forum like reddit where the users believe every word they read on the internet is even worse.
If they want to use statistics like that to justify their changes, which is exactly what they tried to do by posting those, then they need to post their parameters for them. Otherwise they're just pushing propaganda. It's just like the US Government telling everyone how the unemployment numbers are dipping every time tax season rolls around, idiots buy into every year not even thinking that they're all temp workers until tax season is over.

It's the same shit they did with the infantry kills vs vehicle kills statistics a few weeks back.
The graph was just thrown out without any explanation on how the stats were obtained and all the people that wanted to support their "game is fine for infantry" agenda latched onto it.

I find releasing stats like these without explaining methodology is just a dick move to stir up shit in the community as someone can use those stats to justify any position they want.

Transparency is good, but if you are going to do it half-assed then just don't do it at all. This kind of behavior creates more problems than alleviating player concerns.

[Mordelicius]

Originally Posted by Stanis

I'm struggling with this

https://twitter.com/Staniskal/status/297299016759255041


We have three tanks with a KD ratio.
We know they died on average 1 time each to get that D statistics.

Problem is we don't have have 3 Kills.

The MBT stats Magrider: 1.71 Vanguard: 0.87 Prowler: 0.63

1.71 + 0.87 + 0.63 = 3.21 kills

1.17:1
0.87:1
0.63:1
---
3.21:3

so .. where did the extra 0.21 tank to get killed come from?

Perhaps:

1.71 Magriders
0.87 Vanguard
0.36 Prowler
----
2.94 ~ 3

Higby said this data is from a random day with 10k tanks involved in Tank vs Tank battles. They probably counted tank kills + tank deaths and tallied them up. The missing fragment from tank deaths on suicide that's not counted?

If it's really 0.36 then maybe that's why they gave HEAT +25% bonus as a drastic measure.

[Beerbeer]

Higby pulled this crap before and it honestly makes me wonder if they have any idea how to do actual, unbiased statistical and quantitative analysis.

Just because one person on staff has a college statistics course under their belt doesn't mean they know what heck he or she is doing and if anything, the use of improper statistics will almost always lead to bad decisions.

[psijaka]

I don't think that we can argue that these stats do anything other than point to the fact that the Magrider does have a statistical advantage over the other tanks, but I really would like a bit more clarity - why even bother publishing the figures if there is a question mark hanging over them?

[Beerbeer]

They should completely stop releasing these--what I hope are--partial data sets.

It's almost like they want to justify changes to us.

First, unless they release everything, these partial releases mean absolutely nothing. Second, it makes them look insecure about their decisions, IMO. Last, I hope to god they know how to do good, unbiased statistical analysis and are not basing their decisions on simple data shoved into an excel spreadsheet that any sixth grader can do these days. Hire a statistician or pure mathematics major.

Let him or her crunch the numbers and make sound decisions. And never, ever release or even infer about these numbers to us, because it's a double-edge sword.

[Shamrock]

Many Magriders have an advantage because the vast majority of Mag's use Saron an AP dedicated secondary while most prowlers spec HE/HEAT to farm infantry, its a total no-brainer that an AP spec tank will wipe the floor with a HE/HEAT tank, AS IT SHOULD. The TR players made a concious CHOICE to spec their tanks with HE/HEAT, they like farming infantry.

The better metric would have been to compare AP prowlers Vanny's and Mags. I know that will be hard as I cant remember the last time (if ever) I got a kill on an AP prowler they are so dam rare as TR are too busy farming infantry kills with their tank's spamming the hell out of towers and outpost windows.

By this logic we should nerf AP lightnings as well; I often pull one and go one on one with 1/2 Prowlers confident that 90% of the time my prowler opponent is kitted out to kill infantry, and so my lightning's additional maneuverability /speed allows me to get a back shot on them while they sit there waiting for their slow ass reload on HE/HEAT. But wait we are going to buff these guys with 25% more HEAT damage so they can farm sundies and outposts even faster, insane.

[james]

Is this tank v tank or not, becuase if you have less than a 1 with a tank you are awful

[Hamma]

Here are Higby's exact words:

Quick thoughts behind these changes:

Looking purely at performance of tank vs tank the balance is far from ideal. The Prowler makes up for it's vs tank shortcomings quite a bit with infantry farming, but on a tank vs. tank basis it is very weak. We don't really want people to feel like they HAVE to roll vs infantry loadouts with tanks to be useful, but the current state of the prowler makes that seem like the only real useful role for it. These changes (along with the previous HE nerfs) are hopefully going to encourage a lot more tank vs tank combat from each empire.

Picking a random day from last month, the effective K:D of MBTs vs. other MBTs across about 10k tank vs tank fights was:

Magrider: 1.71 Vanguard: 0.87 Prowler: 0.63

There is a lot more that goes into the "usefulness" of tanks, and non-tank based counters that keep this MBT vs. MBT balance from being as severe to the overall game balance as it is to the tank game, but I don't think that anyone is surprised to see the Magrider with such an undisputed advantage in MBT vs MBT. The overall kills per vehicle type are much better balanced, due largely to the Prowler's higher capacity for farming infantry, but again, we want to make sure tanks are tuned around combat with other tanks, not combat with infantry, and large changes were and are warranted to bring them into line.
Edit: We're going to hold off on changing anything with the Magrider performance items (Hover height changes) until we see how the rest of these changes play out. Thanks for the feedback about that.

Emphasis mine.

[Bloodlet]

Edit: nvm I was wrong. This is supposed to be purely tank v tank from a single day with a 10k sample size.