Jump to content
You need to play a total of 20 battles to post in this section.
bubbleboy264

WG straight up admitting they are balancing legendary upgrades with spreadsheet data.

132 comments in this topic

Recommended Posts

455
[WOLF4]
Members
455 posts
7,127 battles

Recent devblog: https://blog.worldofwarships.com/blog/30

You need to see this for yourself, they are just right out saying that they aren't balancing them based on testing/player feedback, just with win rates/other useless spreadsheet data. No wonder game balance is such a mess right now, when this is WG's attitude. 

Just look at this: 

  • UU's popularity should not be significantly above 65% - if it is, that's a sign that the UU is becoming a no-brainer instead of being an alternative. 
  • UU's popularity should not be significantly lower than 40% - if it is, that's a sign that the UU doesn't offer an interesting enough alternative to the existing upgrades.
  • The UU should not make the relative WR worse (in this case it becomes a downgrade, not a sidegrade).
  • The target WR limit is +2%, as was explained above.
  • Each UU case should be reviewed individually before the suggestion of a final version

This sounds good in theory, but it practice it just leads to incredibly stupid balance decisions based on data that can easily be skewed and unrepresentative (for example, a legendary mod can have a higher win rate than it normally would due to only better/more experienced players using them). They don't take into account how strong or weak the upgrades actually are through player testing/feedback, they just look at the sekrit documents and say "look tovarish! Des Moines legendary is too popular! Therefore it must be OP! Bring in the nerf hammer comrade, da this is good game balance taktik." 

This spreadsheet nonsense leads to stupid nerfs like the Henri speed nerf all the time, and yet they still refuse to nerf the [edited]Kremlin. WG balance department, please stop this.

Edit: Would like to clear something up. I am not saying that statistics/spreadsheet type stuff should not be used at all when deciding whether to nerf/buff something. Stats are very important and useful, but they should not be the main/only metric used. Play testing/feedback should play a big role as well. I just think the framing they use in the devblog is really dumb for example.

Edited by bubbleboy264
  • Cool 6
  • Funny 1
  • Boring 7

Share this post


Link to post
Share on other sites
2,109
[KSC]
Clan Supertest Coordinator
5,295 posts
8,554 battles

I don't know why you're acting so surprised, we've known that they use spreadsheets to balance the game for ages...

  • Cool 4
  • Boring 2

Share this post


Link to post
Share on other sites
455
[WOLF4]
Members
455 posts
7,127 battles
Just now, yashma said:

I don't know why you're acting so surprised, we've known that they use spreadsheets to balance the game for ages...

Not surprised, but to see how openly they try to justify their ridiculous logic with this is just mind boggling. Do they actually play their game? 

  • Cool 3

Share this post


Link to post
Share on other sites
2,109
[KSC]
Clan Supertest Coordinator
5,295 posts
8,554 battles
Just now, bubbleboy264 said:

Not surprised, but to see how openly they try to justify their ridiculous logic with this is just mind boggling. Do they actually play their game? 

But again....this is nothing new.  Whenever the community backlash gets too great they often breakout the spreadsheets to explain their reasoning.

Share this post


Link to post
Share on other sites
298
[SS238]
Members
398 posts
36,614 battles
4 minutes ago, bubbleboy264 said:

Not surprised, but to see how openly they try to justify their ridiculous logic with this is just mind boggling. Do they actually play their game? 

As stated before in many posts, playing the game is not a requirement for a WG employee

Share this post


Link to post
Share on other sites
455
[WOLF4]
Members
455 posts
7,127 battles
1 minute ago, yashma said:

But again....this is nothing new.  Whenever the community backlash gets too great they often breakout the spreadsheets to explain their reasoning.

Of course. I mean just look at this: 

  1. (WR difference between UU equipped ships and those without it) minus (WR difference between players using UU and not using UU but owning it) gives us our actual UU WR difference
  2. Because of the nuances above, as well as the facts that player account WR has more impact on battle performance than the WR on a particular ship, we consider the results of up to +2% acceptable. For example:
    1. X DES MOINES: ships WR difference (6,7%) - Players WR difference (3,9%) = +2,8% difference. Above the acceptable 2%, which means the upgrade is too strong.
    2. X SHIMAKAZE: ships WR difference (-0,1%) - Players WR difference (-0.1%) = 0% difference. Acceptable.
    3. X GEARING: ships WR difference (1,5%) - Players WR difference (0,3%) = +1,2% difference. Fully acceptable.

What? The Gearing and Shima upgrades are objectively trash, but since the data is "acceptable" they don't get buffed, and the DM upgrade is OP because it is slightly higher than the "acceptable" win rate difference, then though the Kremlin is still in the game... 

Have they considered that ship win rate can easily be skewed/manipulated based on the player population that is using said ships? It's why very rare ships often have super high win rates even though they actually aren't that good, because only the best players have them. Have they considered that this might happen with Ultimate upgrades too? Of course not, because the spreadsheet rules all...

  • Cool 2

Share this post


Link to post
Share on other sites
455
[WOLF4]
Members
455 posts
7,127 battles
4 minutes ago, mushmouthmorton said:

As stated before in many posts, playing the game is not a requirement for a WG employee

And they wonder why we criticize them for being so out of touch...

Share this post


Link to post
Share on other sites
616
[PELTS]
Beta Testers
2,145 posts
30,115 battles
2 minutes ago, bubbleboy264 said:

Of course. I mean just look at this: 

  1. (WR difference between UU equipped ships and those without it) minus (WR difference between players using UU and not using UU but owning it) gives us our actual UU WR difference
  2. Because of the nuances above, as well as the facts that player account WR has more impact on battle performance than the WR on a particular ship, we consider the results of up to +2% acceptable. For example:
    1. X DES MOINES: ships WR difference (6,7%) - Players WR difference (3,9%) = +2,8% difference. Above the acceptable 2%, which means the upgrade is too strong.
    2. X SHIMAKAZE: ships WR difference (-0,1%) - Players WR difference (-0.1%) = 0% difference. Acceptable.
    3. X GEARING: ships WR difference (1,5%) - Players WR difference (0,3%) = +1,2% difference. Fully acceptable.

What? The Gearing and Shima upgrades are objectively trash, but since the data is "acceptable" they don't get buffed, and the DM upgrade is OP because it is slightly higher than the "acceptable" win rate difference, then though the Kremlin is still in the game... 

Have they considered that ship win rate can easily be skewed/manipulated based on the player population that is using said ships? It's why very rare ships often have super high win rates even though they actually aren't that good, because only the best players have them. Have they considered that this might happen with Ultimate upgrades too? Of course not, because the spreadsheet rules all...

- the Gearing UU is not trash...it is well worth it imo. 

- and the Shimmy UU is getting a buff!!!!

  • Cool 3
  • Funny 1

Share this post


Link to post
Share on other sites
1,793
[SOFOP]
Members
2,486 posts
15,380 battles

Just want to add:

These values are not strict and serve as indicators - the decision to more precisely check the performance of UU is based on them. Besides these values we also consider feedback and expert review. 

That, and they are also using popularity as a metric.  People won't use a bad UU, but it will be fairly obvious when one is overperforming because it is going to be hugely popular.  I don't really see a huge problem using WR and Popularity as at least a guide post to say "Hey, you need to look at this".

 

Edited by Old_Baldy_One
  • Cool 2
  • Boring 1

Share this post


Link to post
Share on other sites
1,179
[PISD]
Members
1,921 posts
6,332 battles
3 minutes ago, bubbleboy264 said:

Of course. I mean just look at this: 

  1. (WR difference between UU equipped ships and those without it) minus (WR difference between players using UU and not using UU but owning it) gives us our actual UU WR difference
  2. Because of the nuances above, as well as the facts that player account WR has more impact on battle performance than the WR on a particular ship, we consider the results of up to +2% acceptable. For example:
    1. X DES MOINES: ships WR difference (6,7%) - Players WR difference (3,9%) = +2,8% difference. Above the acceptable 2%, which means the upgrade is too strong.
    2. X SHIMAKAZE: ships WR difference (-0,1%) - Players WR difference (-0.1%) = 0% difference. Acceptable.
    3. X GEARING: ships WR difference (1,5%) - Players WR difference (0,3%) = +1,2% difference. Fully acceptable.

What? The Gearing and Shima upgrades are objectively trash, but since the data is "acceptable" they don't get buffed, and the DM upgrade is OP because it is slightly higher than the "acceptable" win rate difference, then though the Kremlin is still in the game... 

Have they considered that ship win rate can easily be skewed/manipulated based on the player population that is using said ships? It's why very rare ships often have super high win rates even though they actually aren't that good, because only the best players have them. Have they considered that this might happen with Ultimate upgrades too? Of course not, because the spreadsheet rules all...

Gearing upgrade is not trash and highly competitive in competitive or in division.

Shima UU will receive a boost. That boost is not justified by the WR difference but by the small number of people using it.

Players feeling are a thing, but raw data do not lies. And WG took the time to show us how they decided it, which is great.

  • Cool 10

Share this post


Link to post
Share on other sites
Members
348 posts
5,767 battles
4 minutes ago, bubbleboy264 said:

And they wonder why we criticize them for being so out of touch...

How do you expect them to balance things? By feels?

  • Cool 7
  • Boring 1

Share this post


Link to post
Share on other sites
1,793
[SOFOP]
Members
2,486 posts
15,380 battles

I don't agree with everything WG does based and the sheets, but I don't really see the issue here.  Seems to be a perfectly acceptable starting point to review and address.  The adjustments seem somewhat reasonable, though I don't have experience will all of them.

Throwing Kremlin references in here is just weird, since Kremlin has no UU.

 

  • Cool 2

Share this post


Link to post
Share on other sites
1,511
[WOLFC]
Members
2,779 posts
11,304 battles

I trust objective data much more than a collection of diverse player’s options and (mis)conceptions. It appears WG does too. And, as noted above, WG simply uses these numbers as indicators of what they should look at.

The people making the decisions at WG would be idiots to do it any other way.

  • Cool 3

Share this post


Link to post
Share on other sites
12,832
[WOLF3]
Members
30,951 posts
26,039 battles

RIP Des Moines, Yamato, Zao Legendaries.

Even Montana's Legendary was questionable before the proposed nerf.

Edited by HazeGrayUnderway

Share this post


Link to post
Share on other sites
16,739
[ARGSY]
Members
24,468 posts
18,417 battles
34 minutes ago, bubbleboy264 said:

You need to see this for yourself, they are just right out saying that they aren't balancing them based on testing/player feedback, just with win rates/other useless spreadsheet data.

Would you rather that they chose completely subjective criteria and justified it SOLELY on a "Because that's how it will be" basis?

5 minutes ago, Grimm262 said:

How do you expect them to balance things? By feels?

 No, I'm guessing he expects them to listen exclusively to him and to like-minded players. Or he expects them to make omnipotent, Godlike decisions that will please everybody all the time.

Neither of those things is ever going to happen. I wish my Minotaur UU fitted into slot six, dropped the gun reload to a quarter-second, tripled the turret rotation speed, added 500m/s to the muzzle velocity and 10km to the range, and threw in 1% fire chance on top (which is all it would need). That would make me very happy indeed. Not gonna happen, and I know it. 

  • Cool 3

Share this post


Link to post
Share on other sites
Members
2,441 posts
14,072 battles
1 minute ago, Ensign_Cthulhu said:

Would you rather that they chose completely subjective criteria and justified it SOLELY on a "Because that's how it will be" basis?

No.. They should actually PLAY the game and use those experiences, along with other employees feedback along with USER feedback... but they don't.

Imbalance is what WG does on purpose as is evidenced by the Venezia and T10 CV

 

  • Cool 2
  • Boring 2

Share this post


Link to post
Share on other sites
1,793
[SOFOP]
Members
2,486 posts
15,380 battles
8 minutes ago, Ensign_Cthulhu said:

Neither of those things is ever going to happen. I wish my Minotaur UU fitted into slot six, dropped the gun reload to a quarter-second, tripled the turret rotation speed, added 500m/s to the muzzle velocity and 10km to the range, and threw in 1% fire chance on top (which is all it would need). That would make me very happy indeed. Not gonna happen, and I know it. 

Dude...how did you get a look at the Smolensk UU?

5 minutes ago, Elo_J_Fudpucker said:

No.. They should actually PLAY the game and use those experiences, along with other employees feedback along with USER feedback... but they don't.

Imbalance is what WG does on purpose as is evidenced by the Venezia and T10 CV

 

I don't agree with the idea that the developers don't play their own game - this is a common theme across various games as a complaint against anything they don't like, and it is really a pretty absurd idea.  Not saying they don't miss the mark sometimes with changes, but in all honesty, balancing a game with so many factors at play is a pretty big feat that most games struggle with.  Small changes have huge, unforseen impacts.

I'm still looking for an explanation from anyone as to why using solid data from the game is a bad idea when looking for things to adjust.  I mean, I know players spend a lot of time telling people that stats don't matter and its all random luck of the draw....but this seems like a pretty sound practice to me.

Edited by Old_Baldy_One
  • Cool 1
  • Funny 1

Share this post


Link to post
Share on other sites
1,179
[PISD]
Members
1,921 posts
6,332 battles
2 minutes ago, Ensign_Cthulhu said:

 I wish my Minotaur UU fitted into slot six, dropped the gun reload to a quarter-second, tripled the turret rotation speed, added 500m/s to the muzzle velocity and 10km to the range, and threw in 1% fire chance on top (which is all it would need). That would make me very happy indeed. Not gonna happen, and I know it. 

What are you smoking dude.... oh wait, your a UU Mino player! ;-)

Share this post


Link to post
Share on other sites
8,413
[GGWP]
Members
7,582 posts
17,928 battles

So what is the problem with using data to balance?? 

I mean, I find that article suspect when it says up to +2% wr is fine and over 65% is too popular and in line to be adjusted. Then the gk at precisely 2% is too strong and it's too popular at 57% utilization.... Completely contradicting their parameters for adjustment.... 

  • Cool 1

Share this post


Link to post
Share on other sites
206
[CS7]
Members
320 posts
5,609 battles
3 minutes ago, Elo_J_Fudpucker said:

No.. They should actually PLAY the game and use those experiences, along with other employees feedback along with USER feedback... but they don't.

I appreciate they provided an explanation on how they are balancing the unique upgrades and I'm glad they are using statistics to do it. More importantly, I do think they listened to player feedback on the Henri and Des Moines UU.

The way you provide feedback matters.  No one likes ranting or whining.  A+ to wargaming development in how they have reacted to feedback on the Unique Upgrades.

Share this post


Link to post
Share on other sites
1,555
[KSE]
Beta Testers, In AlfaTesters
1,732 posts
19,258 battles
6 minutes ago, Ducky_shot said:

So what is the problem with using data to balance?? 

I mean, I find that article suspect when it says up to +2% wr is fine and over 65% is too popular and in line to be adjusted. Then the gk at precisely 2% is too strong and it's too popular at 57% utilization.... Completely contradicting their parameters for adjustment.... 

Exactly.   Publishing stuff like this doesn't help the reputation of the company as being competent at balancing their own game.

  • Cool 1

Share this post


Link to post
Share on other sites
Members
2,441 posts
14,072 battles
9 minutes ago, Old_Baldy_One said:

don't agree with the idea that the developers don't play their own game - t

..*the* head of balance admitted in a live interview last summer, as a knee jerk reaction to the DryDock outrage,  that he did not play the game to determine balance, but that he worked a spread sheet all day.  This directly counters your assertion.

  • Cool 2

Share this post


Link to post
Share on other sites
1,793
[SOFOP]
Members
2,486 posts
15,380 battles

I think this would all be a lot easier if they just created a UU slot, that was outside the normal slots and was only for the UU.  They could even go as far to make a bunch of generic UUs, instead of these boat specific ones, and let players choose (admittedly that would be even harder to balance I guess).

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×