Suboptimal testers?

[koliber_1984]

Ship's performance should be tested by skilled and not-that-much-skilled players IMO, it's fine for me.

[ColonelPete]

On 7/29/2021 at 1:50 PM, MannequinSkywalker said:

That's how it should be anyway. 1) You can always buff a pre-nerfed ship but if you already released a (i.e) a broken prem ship like Thunderer, Smolensk, etc. then what r you gonna do? Well, what you do is removing them which is another wrong step. (2) You let below average Joes or bad Joes to test ship chars, analytics, armors, shells, penet. dmg. this and that, etc. while they can't even sail properly or have no idea about the game mechanics. Another wrong step.  (3) If you'd let only or mostly unicum/sunicum players test the ships you would see the true potential of those ships in the right hands and can make adjustments which you call nerfs accordingly. You saying "but below avg players  wouldn't be able to play these ships as good as better players, what would they do then?" Well, (a) They have to learn better (b) For years we have seen what happens when you test ships with bad players and all these Thunderers, CV reworks, etc. come out and all these points may have been leading to a bad game design. That's why we see so many land slide wins/losses because a bad joes can take advantage of poorly tested ships from the map end and spam he (just as an example). I believe, instead of WG "nerfing" its ship testing capabilities with bad, below avg and avg players, they need to put more steps in the game to teach game mechs., game play and such so players can improve themselves.

Do you not think your answer was a bit too late?

[Capra76]

On 7/29/2021 at 1:15 PM, Verblonde said:

This. If you only have unicums doing the testing, you end up with almost zero meaningful data regarding how the ship will perform in game 'in real life' for the simple reason that most players are (probably very obviously) not unicums.

 

A purely hypothetical example of why a range of testers is needed and what might happen if WG only looked at unicum testing results:

 

Let's imagine that WG want to launch a new T10 premium aircraft carrier, they want to make as much money as possible off it so they make it exceptionally noob-friendly, AA has no effect whatsoever, the planes are immortal, the strikes are fully automated, there's no skill whatsoever and you are guaranteed to average ~ 80k damage and about 50% WR whoever you are.

 

Now let's think about how the testers are going to respond to this:

• the unicum group are going to hate it, they're used to averaging 150k and 60% WR, now they're only managing 80k and 50%;
• the potato group will love it, they're used to averaging 40k and 40%, now they're also at 80k and 50%.

Based on the unicum group alone the conclusion would be that it's a terrible ship and needs massive buffs, but if that happened the final ship in game would be absurdly OP, because most players in that ship would now be performing at the level of the very best in any ship and well above their normal level.

 

Thus it's important for WG to know both what the upper limit of this ship's potential is, but also what the median performance level is and to try and balance around both.

[admiraldelorin]

For unbiased testing only AI can be trusted. Period. A heavely invested player can not be trusted to have no bias. Even if a player would not have a bias, a player could have a cold or the weather could be bad.  Testing by players rather is a PR measure, testing the waters how a market would accept a product. I'm pretty sure any decision on a product will in the end be revenue oriented (long term economic success, continuity of the game ) weather a tester is representative will be known by determining the correlation between the tester's feedback and continuity of the game.  If they don't listen to you then you may suspect that that correlation is low. In that context a testers success as a player is totally irrelevant.

[Ze_Reckless]

7 hours ago, Capra76 said:

Let's imagine that WG want to launch a new T10 premium aircraft carrier, they want to make as much money as possible off it so they make it exceptionally noob-friendly, AA has no effect whatsoever, the planes are immortal, the strikes are fully automated, there's no skill whatsoever and you are guaranteed to average ~ 80k damage and about 50% WR whoever you are.

Unrealistic scenario, this isn't how it works. It would only make sense if there was no player input at all a.k.a. no game play. Just click battle and the rest is on autopilot, you watch it like a movie.

 

Unicorns don't necessarily need to aim better to reach their results. They have better target and tool selection and positioning because they can predict the flow of the game. The ship capabilities don't matter for this.

[lameoll]

7 hours ago, Capra76 said:

the planes are immortal, the strikes are fully automated, there's no skill whatsoever and you are guaranteed to average ~ 80k damage and about 50% WR whoever you are.

 

the unicum players will still do better  LUL  there is also other skills as the ships itself.

 

The unicums will  hate it because they actualy see its an easy to play ship and get bored with it.

the potatos ofc will love it as they dont see how dumb the ship is and only see they are actualy doing good for a change.

[Fat_Maniac]

On 7/28/2021 at 3:41 PM, lovelacebeer said:

 

they just ignore what the testers say.

^THIS^

 

Add in the growing trend of CC's telling us WG ignores the feedback they give.

 

I'm sure WG will claim they do listen. But just taking onboard the bits you like or want to hear, is NOT listening.

[Capra76]

2 hours ago, Ze_Reckless said:

Unrealistic scenario,

 

Of course, that's the point, take the most extreme scenario and see what it tells us, then if we wish we can apply it to more realistic scenarios and see if the lessons still hold.

 

 

2 hours ago, Ze_Reckless said:

It would only make sense if there was no player input at all a.k.a. no game play. Just click battle and the rest is on autopilot, you watch it like a movie.

 

Unicorns don't necessarily need to aim better to reach their results. They have better target and tool selection and positioning because they can predict the flow of the game. The ship capabilities don't matter for this.

 

In this case I tried to remove as much player input as possible to bring the potatoes and unicums as close as possible, here you might see a 70k to 90k or 60k to 100k damage split, the exact figures are irrelevant, what's important is that the ship brings unicums and potatoes far closer to each other than normal.

 

The key point is that it's possible to design a ship that would be simaltaneously horribly UP for unicum players but at the same time absurdly OP for the vast majority, which is why testing needs to include both player groups.

[Ze_Reckless]

2 hours ago, Capra76 said:

The key point is that it's possible to design a ship that would be simaltaneously horribly UP for unicum players but at the same time absurdly OP for the vast majority, which is why testing needs to include both player groups.

Such a design has never existed and I can't imagine one that would work like that.

 

The problem in your design idea is that unicorns are the most adaptive players, they can just play like a potato if that's what it takes to gain an OP ship. They can push every design to the max.

 

You can design the tool but you can't tell the unicorn how to use and abuse it.

 

The only reason to include below average players into testing is to test the usability of the design for those players. You don't learn much from them about ship balance.