Jump to content
Sign in to follow this  
You need to play a total of 50 battles to post in this section.
Hawg

Musings on Class / Ship Balance

47 comments in this topic

Recommended Posts

[HAERT]
Players
414 posts
5,983 battles

What does it mean for a ship or class to be "balanced"?   Presumably that means the combination of its attributes (speed, stealth, ability to damage, ability to tank, manuverability, etc.) are such that it has a "fair chance" to be competitive against other ships / classes within its tier.   

 

If this definition is true, and if WG has managed to achieve balance, then one would expect to see similar win rates among the various classes within the same tier.   Looking at Tier X - for the most part this holds true, until you come to CV's which have a much lower winrate:

 

Tier X Win Rates (Best and Worst)

 

DD's

Khaba 52.06%

Simmy 48.91

 

CA's

Moskva 50.09%

Minotaur 48.85%

 

BB's 

Conqueror 51.30%

Yamato 48.94%

 

CV's

Hakuryu 46.98%

Midway 45.17%

 

Of course the results probably vary by tier, but just looking at this sample, one might think that CV's should be buffed to bring them into 48-51% winrate range.

 

However, if instead of winrate you look at the kill death ratios you get a different picture:

 

Tier X - Kill / Death Ratios (Best and Worst)

 

DD's

Khaba 1.68

Grozovoi 1.10

 

CA's

Zao 1.73

Moskva 1.39

 

BB's

Yamato 1.45

Conqueror 1.89

 

CV's

Hakuryu 6.37

Midway 5.95

 

The numbers above are for all players... to me it would make it seem that CV's are way out of balance when you consider how many times they die compared to how many times they kill another player.    But if you only look at the top 5% of players for those ships the numbers seem even more astounding (Hakuryu 29.43, Midway 27.51) until you consider that the very nature of CV's is not to put themselves in danger, so they in theory will survive many games they loose.   

 

But still, those top 5% Tier X CV players only expereince death one time for every ~28 times they kill another player.   No wonder CV's get such hate.  

 

However I am going off topic, this is not intended to be an anti-CV post (because as shown above, they arguably need help in the win rate catagory).

 

Why am I sharing all of this?   Because it makes me wonder, in terms of game design, how should WG be defining and measuring "balance"?    

 

There are some things that suck that might give a clue?   

 

In a DD, not being able to get un-spotted from a pesky CV -  or the fact that every time someone farts on you you lose an engine or steering module.

 

In a CA, pounding away at a BB that's showing broadside and doing some damage, but not much, only to show a split second of broadside and lose 80% of your hp from one salvo from the same BB.

 

In a BB, not being able to return fire against that unseen Minotaur raining shells down on you, stealth torps from the backfield from an unspotted DD, or from Torpedo Bombers.

 

In a CV, having that Des Moines melt an entire flight group just because you wandered a hair too close to it. 

 

The point being that there are situations each class faces that suck, that feel "unfair" or "unbalanced".   In theory, if all those sucks balance out, then in fact everything is balanced.   Is that the irony?   Do they really balance out?   What do you think?   How should balance be defined and measured?

Share this post


Link to post
Share on other sites
[CATS]
Players
14,753 posts
10,885 battles

Why would you even consider Kill/ Death ratio? It is one of the most useless stats out there.

Kills/Battle is much more interesting.

  • Cool 1

Share this post


Link to post
Share on other sites
[HAERT]
Players
414 posts
5,983 battles
4 minutes ago, ColonelPete said:

Why would you even consider Kill/ Death ratio? It is one of the most useless stats out there.

Kills/Battle is much more interesting.

I decided to look at Kill / Death ratio because I was trying to think about the psychology of playing the game.    Measuring balance is in many ways "in the eye of the beholder",  and the measure of how often you kill someone vs how often you get killed can lead to perceptions on how fair the game is.

 

Kill to death is also used as a measure of success in a lot of first person shooters for that same reason.   

 

I'm not saying its the best way to measure this game, just one indicator of how strong a ship feels when you play it.   If I play X ship and die three times for every time I kill the enemy, I am probably going to hate that ship, even if it has a better win rating compared to Y ship where I kill three people for every time I die.

Share this post


Link to post
Share on other sites
[CATS]
Players
14,753 posts
10,885 battles

In many first person shooters you die multiple times per round. There such a ratio is interesting.

In WoWs it matters what you accomplish in an average match, no matter if you sink or not.

Share this post


Link to post
Share on other sites
[MASLO]
Beta Tester
663 posts
1,238 battles

CVs win ratio is kinda... different thing. First of, there are always same number of them per team, so overall their winrate should be slightly under 50% (odd draw or two). Even if there is more of one than another, their number will go towards 50% because of more mirror matches that result in 50% win ratio. So... their win ratio stats are close to useless, at least regarding their balance with other classes, as its decent to show balance between different CVs though.

At least looking at their win ratio, they seem fairly balanced between themselves.

 

P.S. tho its kinda interesting that their win rate on warships today is above 50%... how exactly, i dont know, unless there are games where only one team has CV, it should be max 50% between them, but both above 50% in last 2 weeks? Lol

Luckily WOWs numbers stats make more sense.

Share this post


Link to post
Share on other sites
[DAVY]
Players
1,542 posts
9,023 battles

i think you also have to consider the newer players to the game that want to grind their first T10's and the choices they make, ooh the yamato has big guns, the shima has lots of torps and fail when they get them...

  • Cool 1

Share this post


Link to post
Share on other sites
[HAERT]
Players
414 posts
5,983 battles
3 minutes ago, ColonelPete said:

In many first person shooters you die multiple times per round. There such a ratio is interesting.

In WoWs it matters what you accomplish in an average match, no matter if you sink or not.

PUBG -- you only die once per match.  K/D ratio is arguably the most important metric behind win ratio.

 

But really that's beside the point.   I wasn't arguing that its "the" ratio to use in measuring balance, just one of several to think about.

Share this post


Link to post
Share on other sites
Weekend Tester
769 posts
3,524 battles
34 minutes ago, Hawg said:

CV's

Hakuryu 46.98%

Midway 45.17%

 

Where do you get these numbers from? The draw epidemic has been gone for a long time now, the numbers should have averaged much closer to 50% since then surely.

Share this post


Link to post
Share on other sites
[MASLO]
Beta Tester
663 posts
1,238 battles
2 minutes ago, rvfharrier said:

 

Where do you get these numbers from? The draw epidemic has been gone for a long time now, the numbers should have averaged much closer to 50% since then surely.

 

I guess its from https://wows-numbers.com/ships/

I checked there and exactly the same as he posted. Just select carriers and T10

  • Cool 1

Share this post


Link to post
Share on other sites
[HAERT]
Players
414 posts
5,983 battles
3 minutes ago, genai said:

CVs win ratio is kinda... different thing. First of, there are always same number of them per team, so overall their winrate should be slightly under 50% (odd draw or two). Even if there is more of one than another, their number will go towards 50% because of more mirror matches that result in 50% win ratio. So... their win ratio stats are close to useless.

 

Yeah CV's are a bit difficult to measure... only have 2 at tier X doesnt help.   Tier 8 with three CV's (excluding Graff due to low number of battles) shows the best/worst as:

 

Shokaku 50.75%, 4.15 k/d. (36.60 k/d for top 5%!!)

Lexington 46.64%, 2.77 k/d (11.18 k/d for top 5%)

Share this post


Link to post
Share on other sites
[UNICS]
Beta Tester
4,481 posts
7,994 battles

I want to start by questioning your numbers, and I'm quite certain you can imagine how ridiculous the idea of both the Haku and Midway having substantially below 50% WR is.

 

Now if we leave your numbers aside and talk about balance as a concept.

You really have a few points to cover.

  • Team balance: The ships facing each other needs to be somewhat comparable given equal skill. Either through a tight MM algorithm or through maintaining strict ship v ship balance.
  • Ship balance: Ships need to adhere to some sort of zero sum power score balance. Basically if a ship is good at something it needs to bad at something else of equal value.
  • Reward balance: Ships need to be able to have equal access to rewards. Doesn't mean that all ships need to have the same max and min reward potential, there is room for more or less skill dependent ships.

Team Balance

Pretty straight forward in theory but limited by how much WG wants to tax the matchmaker, or how much they want to normalize ships so that the exact team composition matters less.

 

Ship balance

What most people think about when talking about new ships and how much they will own or get owned versus other certain ships.

Really difficult to pin down prior to large scale usage on the live server, and even then it's going to be very subjective depending on what you want from the ship. Do you think every DD should be good at capping? Should every ship be viable in ranked? Should every ship have the same general impact on the result of a battle?

 

Reward balance

This is probably what is important to most players who don't spend time arguing over balance. Does the ship bring in the rewards you expect at the end of the day.

The advantage to this way of balancing is that it can be done centrally and does not require balancing individual ships as much.

In theory you reward everything you want players to do in the game and then you balance the ships to bring in the same average rewards.

 

In the end it's going to come down to a combination of these three approaches to balancing, simply because the game should cater to a wide audience: the players who wants a fair team battle and doesn't care that much about individual balance, those who wants to have equal impact on the outcome of the battle given that they can utilize their ship's strengths, and finally the players who just wants to play the game and be greeted by decent rewards after the battle.

Most players will be some combination of all three.

Share this post


Link to post
Share on other sites
Weekend Tester
769 posts
3,524 battles

There's no way those can be right though, can they? Due to the way CV MM is mirrored then the average WR for T10 CVs as a whole should be 50% before the effect of any draws. Draws simply haven't been common enough for long enough to cause the combined numbers to be so low.

 

To make it even more confusing warships today gives something completely different with the Hakuryu having a WR of 53.28% and the Midway 47.76%.

 

 

Share this post


Link to post
Share on other sites
[MASLO]
Beta Tester
663 posts
1,238 battles
5 minutes ago, Nechrom said:

I want to start by questioning your numbers, and I'm quite certain you can imagine how ridiculous the idea of both the Haku and Midway having substantially below 50% WR is.

 

 

Why exactly? It makes perfect sense. There still CAN be draws, altough its rare. So if two CVs are decently balanced between themselves, it HAS to be under 50% for both. its logic and common sense.

Even if you check other tiers, you see that avg winrate of all CVs on that tier is under 50%. Sure some have above 50%, but if one has 51% for example, other has <49%, just because 50% avg between them is nearly impossible (0 draws in huge number of games). Only tiers where one has above 50% are very imbalanced tiers where one CV is much stronger than other.

Calling it ridiculous is kinda... ridiculous...

Share this post


Link to post
Share on other sites
Players
2,248 posts
14,326 battles
47 minutes ago, Hawg said:

...if WG has managed to achieve balance...

WG does put much effort into ship balance... it just isn't an area of high concern for them.

Share this post


Link to post
Share on other sites
[UNICS]
Beta Tester
4,481 posts
7,994 battles
1 minute ago, genai said:

Why exactly? It makes perfect sense. There still CAN be draws, altough its rare. So if two CVs are decently balanced between themselves, it HAS to be under 50% for both. its logic and common sense.

Even if you check other tiers, you see that avg winrate of all CVs on that tier is under 50%. Sure some have above 50%, but if one has 51% for example, other has <49%, just because 50% avg between them is nearly impossible (0 draws in huge number of games).

Calling it ridiculous is kinda... ridiculous...

Draws are less than 1% of battles, way less.

So claiming that 7,85% of tier 10 battles with CVs are draws is indeed ridiculous.

  • Cool 1

Share this post


Link to post
Share on other sites
[MASLO]
Beta Tester
663 posts
1,238 battles
15 minutes ago, Nechrom said:

Draws are less than 1% of battles, way less.

So claiming that 7,85% of tier 10 battles with CVs are draws is indeed ridiculous.

And anyone who claimed that would indeed be ridiculous. But 3.925% of battles is much less ridiculous. Remember that 1 draw with CVs counts as 2 battles with no win, 1 for each!

 

Share this post


Link to post
Share on other sites
Players
2,248 posts
14,326 battles

Player Average for Ships [ at 2018/02/10 ]
2 months 
    
                     win       draw       lose
Hakuryu      44.04    0.01      55.95


Midway       53.04     0.03       46.93    

 

-------------------------------------------------------------

 

[Facts with Fem] Midway Changes 7.1 - World of Warships [Beginner]

 

The need for a re-balance is discussed by Femennenly

Share this post


Link to post
Share on other sites
[UNICS]
Beta Tester
4,481 posts
7,994 battles
11 minutes ago, genai said:

And anyone who claimed that would indeed be ridiculous. But 3.925% of battles is much less ridiculous. Remember that 1 draw with CVs counts as 2 battles with no win, 1 for each! A total of 9785 battles out of 301068

Stop this. It's not even an argument.

You are off by a factor of 10^-2

Share this post


Link to post
Share on other sites
[ADRIA]
Players
4,268 posts
6,795 battles
1 hour ago, Hawg said:

CV's

Hakuryu 46.98%

Midway 45.17%

Yeah, not sure what your sources are (edit: found them down lower lol) but those numbers are clearly off especially with how the CVs are forced into 100% mirror matches

 

18 minutes ago, genai said:

There still CAN be draws, altough its rare. So if two CVs are decently balanced between themselves, it HAS to be under 50% for both.

Yeah, bcuz they can happen doesn't mean they happen every game. Lets say - 1 draw in 1000 battles, with just one CV for simplicity. So result would be 499 wins, 499 losses and 2 draws - 49.9% WR. And 1 in a thousand is probably a massive overestimation for how often they would be happening, thy are far rarer than that.

 

Also - I wonder where did you get that "3.925%" or whatever you had? Some statistics? I bet that statistics includes "all time" meaning it would still count the time when draws were common from the beginning of the game.

 

And yes, 3.925% still is pretty ridiculous.

 

20 minutes ago, rvfharrier said:

warships today

At least on my end (so I'd assume it's the same for everyone) they haven't updated any data since about Christmas. Still usable, just as more and more new patches are rolling out it's quickly getting outdated :Smile_teethhappy:

Share this post


Link to post
Share on other sites
[MASLO]
Beta Tester
663 posts
1,238 battles
15 minutes ago, Nechrom said:

Stop this. It's not even an argument.

You are off by a factor of 10^-2

My bad in that quoted post, i halved the wrong number, but same thing.

 

should just halve the number of battles, but halving the % accomplishes the same.

Example: 10 matches, 4 wins by each, 2 draws

10 40%

10 40%

if we were to add both up, its 20% of 20 battles, but its actually 10% of 20 or 20% of 10.

 

Also - I wonder where did you get that "3.925%" or whatever you had? Some statistics? I bet that statistics includes "all time" meaning it would still count the time when draws were common from the beginning of the game.

 

And yes, 3.925% still is pretty ridiculous.

 

 

Like i said above, i halved the % instead averaging the CV battles when it comes to calculating number of total draws, my bad. It was quick maths, not good one! It does end with same number of draws though. But yes, 7.85% of all games with CVs ended in draws according to wows-numbers... with unknown time period and only those that didnt hide stats.

 

But yes, sadly on wows numbers you cannot select time period. So who knows since when are those stats? The other guy posted last 2 months.

 

My point is just that its logical to be under 50%. By how much... thats different story.

Share this post


Link to post
Share on other sites
Players
579 posts
5,014 battles

Some top CV players hidding theire stats maybe?

We are few players , so even like 20 guys hidding there stats with above 50% winrate screw the picture.

 

(Bloggis has his stats closed afaik e.g.)

 

Btw. i ran into a 24% WR, 50k avg.  >100 games midway yesterday - i feel really sorry for that team....

Share this post


Link to post
Share on other sites
[UNICS]
Beta Tester
4,481 posts
7,994 battles
1 minute ago, genai said:

My bad in that quoted post, i halved the wrong number, but same thing.

 

should just halve the number of battles, but halving the % accomplishes the same.

Example: 10 matches, 4 wins by each, 2 draws

10 40%

10 40%

if we were to add both up, its 20% of 20 battles, but its actually 10% of 20 or 20% of 10.

Just, just...

Look at @Culiacan_Mexico's post for the actual server data:

 

Share this post


Link to post
Share on other sites
[NWP]
[NWP]
Players
8,132 posts
11,682 battles

I'm laughing at the suggestion both T10 CV have potato WR. As surely someone has to actually be winning?

 

Although I see some utter garbage top tier CVs. Ones that have clearly not played much CV before getting up there.

Share this post


Link to post
Share on other sites
[HAERT]
Players
414 posts
5,983 battles
59 minutes ago, Salentine said:

i think you also have to consider the newer players to the game that want to grind their first T10's and the choices they make, ooh the yamato has big guns, the shima has lots of torps and fail when they get them...

 

This was me :)

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×