Jump to content


Let's talk about tanks - AMX 30 B UPDATE!


  • Please log in to reply
13 replies to this topic

RaiBOT01-x #1 Posted 04 May 2020 - 06:35 PM

    Publications (Cat) Manager

  • Administrator
  • 3325 battles
  • 15,089
  • [WGW]
  • Member since:
    06-28-2013

Hey everyone!

 

My name is Jeff Gregg, and I am the Lead Designer on World of Tanks Valor.  Welcome back to "Lets talk about tanks! AMX 30 B edition".

As a reminder, the goal was to highlight specific tanks, one at a time, and really dig down into what makes them tick, how they are being used, and how approaches may be taken to improve them if necessary.

 

Lets see how our changes to the AMX 30 B.

 

What we did

As a reminder: the following was done to this tank:

  • Buff the penetration and shell speed on this gun to be more than now but not as much as it was before we nerfed it before. This will also impact the Tier 9 AMX 30
  • 105mm mle F1 gun
    • 105mm OFL 105 F1 shell
      • Speed: 1100 → 1525
    • 105mm OCC 105 F1 shell
      • Speed:  800 → 1000
    • Penetration: 248/300/53 → 255/310/53

Results

Lets take a look at what occurred.

All data from below is from Random Battle from April 22nd (first full day of changes) till April 30th

NOTE: We ran the loaded dice event this same week. (Damage RNG is +/-10% and Pen RNG is 0%)

 

Here is a breakdown of what we saw:

Stat % or # Desc
Battles 32,000

Usage of this tank went up considerably. Seems like players wanted to check out what the changes were about. This is good as it gave us a decent chunk of data to analyze (tho not enough to be conclusive, see more on that below)

  • ~ 3,555 a day (up from ~1,700 a day before)

AMX-UPDATE-1.jpg

Win Rate 51.59%

The win rate is up from a previous average of 49.9%. Expected, but to be honest a bit higher than we desired and expected.

Popularity Graph N/A

We don't have enough info re: popularity yet to make a conclusion. See the graphs below. It SEEMS the vehicle may be more popular than it was before with the better players, but its not possible to say that emphatically (note saw-tooth graph).

Previous

AMX-UPDATE-1.png

Current

AMX-UPDATE-2.jpg

 

Skill Graph N/A

There is a minute yet interesting dip in performance for lower skill players. But higher skill players are playing as we expect. This MIGHT mean the tank is balanced fine, but until we get more data on popularity, we can't say that for 100% sure. We need to keep collecting data.

Previous

AMX-UPDATE-2.png

 

Current

AMX-UPDATE-3.jpg

Accuracy 76.30%

The gun's accuracy is up from a previous average of 72%

AMX-UPDATE-4.jpg

Pen 52.97%

The gun's penetration is up from a previous average of 48.2%

AMX-UPDATE-5.jpg

Damage per Hit 249.3

Damage per hit is up from a previous average of ~230 (which makes sense as it is penetrating more).

AMX-UPDATE-6.jpg

Damage per Penning Hit 360.9

No real change (360 before) which makes sense as we buffed the pen, not the damage output on a successful pen.

AMX-UPDATE-7.jpg

Damage Per Battle 1,914

A nice bump from the previous average of ~1,600

AMX-UPDATE-8.jpg

Survival 27.4%

All this extra penetration as risen the survival from a previous average of 24.6%

AMX-UPDATE-9.jpg

Lifespan 4min 37s

Up marginally from 4min 32s

AMX-UPDATE-10.jpg

Action Items and Next Steps:

For now we are going to keep watching and make no further changes. If the win-loss stays high but the popularity evens out, we may have an issue. If the win-loss stays high but the higher skill players clearly are playing more, we may NOT have an issue. Either way we need to collect more data. Once we have enough to make additional calls, we will let you know.

 

Thanks and see you on the battlefield!

 

-CHUJI


Edited by RaiBOT01, 08 May 2020 - 02:33 PM.

<a data-cke-saved-href='http://www.pixeljoint.com/files/icons/walkingcat_by_thetaupe.gif' href='http://www.pixeljoint.com/files/icons/walkingcat_by_thetaupe.gif' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...by_thetaupe.gif</a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...hetaupe.gif</a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...upe.gif</a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...gif</a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'>http://www.pixeljoin.../a></a></a></a></a>

Player's passion for our games is the fuel that keeps my tank engine running every day.

"Never let your sense of morals get in the way of doing what's right" - Isaac Asimov

Follow me on Twitter: @RAIBot01 

Customer Support 


BrogueOne-p #2 Posted 11 May 2020 - 01:15 PM

    Major

  • Supertest - PS4
  • 24821 battles
  • 11,202
  • [ELDRZ]
  • Member since:
    01-07-2016
I would like to highlight your final statement because I think everyone reading this article should understand it:

Action Items and Next Steps:
For now we are going to keep watching and make no further changes. If the win-loss stays high but the popularity evens out, we may have an issue. If the win-loss stays high but the higher skill players clearly are playing more, we may NOT have an issue. Either way we need to collect more data. Once we have enough to make additional calls, we will let you know.

This means that the change was good, but it might be too good, but we’re gonna give it some more time to make sure.  

As a personal note/observation.... I think it should be left alone regardless.  The AMX 30b will always fall short in popularity because it has a far more popular relative in the BatChat.  This leaves the 30b to be really great tank that’s going to be rewarding for the players who choose it over or instead-of the BatChat.

Koneko Da Best-x #3 Posted 11 May 2020 - 01:35 PM

    First lieutenant

  • Players
  • 31157 battles
  • 692
  • [KMD]
  • Member since:
    02-03-2014
I'm actually baffled that you guys could even say that these buffs are enough let alone suggest they might have been too much. I think at this point it might be time to drop the data standpoint and just straight up compare the AMX 30B stats to other tier 10 mediums that alone should show that it isn't really a great tank by any means other than mobility everything else about it is terrible. :sceptic:

natedaishmaster-x #4 Posted 11 May 2020 - 02:29 PM

    First lieutenant

  • Players
  • 28835 battles
  • 731
  • [IMTLS]
  • Member since:
    07-29-2014

It is a pretty sharp increase in battles by the 55-65% win rate players and a slight decrease in battles from the 43-47% bracket, almost to the point that the peak of the data (as hard as it is to read with the sawtooth) ends up right around the 51.6% mark that you see as the average win rate. This skews all of the numbers up since better players tend to hit more shots and aim better to pen more shots. The change definitely should not be reverted, instead you should compare pens and dpg across tier X mediums, itemized by win rate bucket (dpg for players between 52-57% win rate). Assuming a fairly normal distribution, this will give you the data to accurately see the effect of the change. The current data only shows that the skill of players in the 30b has increased, and that skews all of the data to the point that none of it is really usable until either the curve shifts back to peaking at 50% or you compare to different mediums using win rate buckets. I personally think that it needs a bit more of a buff (the fact that the 30b win rate compared to the player win rate is worse across most skill levels indicates that imo), but leaving the tank alone is ok for now, as it is at least usable in its current state

 

View PostKoneko Da Best, on 11 May 2020 - 05:35 AM, said:

I'm actually baffled that you guys could even say that these buffs are enough let alone suggest they might have been too much. I think at this point it might be time to drop the data standpoint and just straight up compare the AMX 30B stats to other tier 10 mediums that alone should show that it isn't really a great tank by any means other than mobility everything else about it is terrible. :sceptic:

 

They can use different statistical analysis to do that. Compare certain win rate buckets across all tier X mediums and use those numbers to do analysis. Their current method renders most of the data points as useless due to an increase in player skill. All it shows is that better players pen more shots and do more damage. Data analysis is still the way to go, but they need to analyze the correct data to see the real results

Edited by natedaishmaster, 11 May 2020 - 02:30 PM.

3 MOE: Hellcat, Lycan, Sp1C, T20, M41 Bulldog, Motherland, STA 1, T95E2, Pershing, AMX 50 100, AMX 13 90, T54ltwt, T49, RU251, M46 Patton, Type 61, Leopard PTA, T54, E50, Batchat AP, Skoda T50, Centurion 7/1, Sharpshooter, T10, Conquerer, AMX 13 105, Sheridan, T-100LT, RHM PZW, M60, Chisel, M48 Patton, Leopard 1, E50M, Obj 140, Batchat 25t, TVP 50/51, AMX 50B, Valour, IS7, Super Conqueror, The Machine, Chieftain, FV215B, T110E4


RaiBOT01-x #5 Posted 11 May 2020 - 02:43 PM

    Publications (Cat) Manager

  • Administrator
  • 3325 battles
  • 15,089
  • [WGW]
  • Member since:
    06-28-2013
Bump

<a data-cke-saved-href='http://www.pixeljoint.com/files/icons/walkingcat_by_thetaupe.gif' href='http://www.pixeljoint.com/files/icons/walkingcat_by_thetaupe.gif' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...by_thetaupe.gif</a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...hetaupe.gif</a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...upe.gif</a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin...gif</a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'><a href='http://www.pixeljoin.../a></a></a></a></a>' class='bbc_url' title='External link' rel='external'>http://www.pixeljoin.../a></a></a></a></a>

Player's passion for our games is the fuel that keeps my tank engine running every day.

"Never let your sense of morals get in the way of doing what's right" - Isaac Asimov

Follow me on Twitter: @RAIBot01 

Customer Support 


LacqueredBacon-x #6 Posted 11 May 2020 - 03:16 PM

    Major

  • Players
  • 42006 battles
  • 6,693
  • Member since:
    02-02-2015

I don't think you can draw any conclusions from this data.  First off, the changes themselves are so minimal that the majority of the player base won't even feel a difference.  Second, comparing the results to the previous results can't tell the whole story.  I read these and think, Ok a tank with a slight buff is slightly better.

 

But does that mean it is balanced?  How is this comparing to other tier 10 mediums?

And quit with the win rate.  Win rate is still a horrible metric in a game with this many variables and with skill the most important variable that affects win rate.


 



 


natedaishmaster-x #7 Posted 11 May 2020 - 03:33 PM

    First lieutenant

  • Players
  • 28835 battles
  • 731
  • [IMTLS]
  • Member since:
    07-29-2014

View PostLacqueredBacon, on 11 May 2020 - 07:16 AM, said:

I don't think you can draw any conclusions from this data.  First off, the changes themselves are so minimal that the majority of the player base won't even feel a difference.  Second, comparing the results to the previous results can't tell the whole story.  I read these and think, Ok a tank with a slight buff is slightly better.

 

But does that mean it is balanced?  How is this comparing to other tier 10 mediums?

And quit with the win rate.  Win rate is still a horrible metric in a game with this many variables and with skill the most important variable that affects win rate.


 

 

Most of the changes are due to better players playing the tank. Better players = more wins = more damage = better accuracy. Win rate is valuable if analyzed correctly. The win rate vs overall win rate metric is a good indicator if something is over or underperforming, and overall win rate is usable if the tank is played at an equal distribution across all player skills, which is not the case here. They identified that this is the case and made the decision not to revert the change, but the tank (probably) still is performing below average. However, the data analysis has not been done in a way to detect that

3 MOE: Hellcat, Lycan, Sp1C, T20, M41 Bulldog, Motherland, STA 1, T95E2, Pershing, AMX 50 100, AMX 13 90, T54ltwt, T49, RU251, M46 Patton, Type 61, Leopard PTA, T54, E50, Batchat AP, Skoda T50, Centurion 7/1, Sharpshooter, T10, Conquerer, AMX 13 105, Sheridan, T-100LT, RHM PZW, M60, Chisel, M48 Patton, Leopard 1, E50M, Obj 140, Batchat 25t, TVP 50/51, AMX 50B, Valour, IS7, Super Conqueror, The Machine, Chieftain, FV215B, T110E4


RabidSasquatch0-x #8 Posted 11 May 2020 - 07:17 PM

    Captain

  • Beta Tester
  • 22655 battles
  • 1,425
  • [VYPER]
  • Member since:
    11-09-2013

Even with these buffs, this has the undisputed worst shell hard stats at tier 10 (worst pen, worst velocity for the archetype).  This is not an inherent issue, after all someone has to be the worst, right?  The problem is, nothing else about the tank is remarkable.  The dpm is mediocre at best, the gun handling, while not bad, decidedly average compared to other tier 10 medium tanks, and while this tank did receive an armor buff that was substantial in numeric value (30mm to over 200 on the mantlet!), in reality, this has done next to nothing to improve the damage blocked by the vehicle (not a given stat here, but I feel confident in saying this, maybe % wise its much higher, because anything over the near 0 from before is huge here, but not in-game).  Mobility, camo, and view range are all just ok.  Before, this tank was almost directly inferior to the m46 patton, 1 tier below it (worse gun stats, armor, view range, camo, maneuverability, all it had was slightly better pen, slightly better top speed, and more HP).  I can't imagine the justification for this other than nobody bothered to look at other tanks when deciding to balance this in the first place (which still seems to be an issue here).  So ignoring every chart and result given here, my question is this; why would I play the 30b over say a leopard one?  The short answer is there isn't any (atleast no good ones, asthetic?, nationality?, Masochism?).

 

One of the key reasons I can assume this issue exists is only looking at backend result data.  While win rate and average damage are very important to consider, the blatant disregard for comparing one tank hard stats to another baffles me.  Nobody who has half a brain playing this game will ever choose to grind the 30b over the leopard one for the purpose of playing well* (and if you ask me most people shouldn't even bother with the leopard one, but that's an argument for a different day).  *I emphasize this because there are a lot of reasons to play other than winning, such as checking out changes or for laughs or for XP towards a merc contract, you get the idea.  The fact is, you can easily look at this gun, say "it is the worst of any tier 10 105 non-autoloading medium, and as such should have the worst stats of them (pen, hit%, whatever)", and look at your results data to conclude how misleading it can be.  The 30b is not a 51% winrate tank by virtue of its stats.  It is worse than the leopard one, anyone playing a 30b could take a leopard one out, do exactly the same thing, and have better results.  Every stat that exist shows this to be true.  The only reason this would ever not be true is not due to the tanks stats, but rather player's abilities (it's newly changed, or only a small percentage of players use the tank, either of which cause the tank to have a player choice than the leopard one, which is old and hasn't been put in the spotlight).

 

This is unfortunately a jarbled mess and not super clear (I've tried rewriting but haven't had much success, and not having stats infront of me don't help; sidebar question, any chance we can get full stat rundowns on the website here? it's kinda frustrating I can't look at those PC hidden stats we were given on console anywhere but on the console itself.  Seems wasted to supply data but limit the ways you can access it).  My takeaways are

  • Focus on direct tank to tank comparisons, a lot of the tanks being addressed have competitors that are directly superior or inferior to them.  Using other tanks in game would streamline the balancing process drastically, and arbitrary values wouldn't be necessary if you started with the already established basis in game rather than trying to come up with values on backend analysis.  Looking at one tank and disregarding the rest of the game is a disservice and harmful to player experience.  Apologies for the meme, but play your own game, it would help alot with some of the decisions (keep it simple stupid)
  • Realize how drastically backend stats vary seemingly random.  There's no point in spending manhours trying to define human response, if you could anticipate how players would play, then gaming wouldn't be a thing.  A tank could have a 90% winrate one day just because.  Now there are some factors we know will cause fluctuations (forum posts about a tank, adjustments made to it, on track events, merc contracts that inclue it, etc).  It's impossible to accurately gauge a vehicle's performance when any of these are happening, and while, yes, this means longer turnaround times for backend data, you don't have to use only results to make adjustments, see point one.


Pontiac Pat-x #9 Posted 11 May 2020 - 07:19 PM

    Major

  • Players
  • 27978 battles
  • 7,753
  • [47R]
  • Member since:
    08-11-2014

The skill graph says it all.  The overall win rate is up, but the skill graph is actually very slightly down within the displayed data range (45-55%).  This verifies that the overall win rate increase is from the increased play of players with win rates over 55%.  The data is right there, but you have to know how the data interacts across the various graphs.

 

Jeff, you should be able to control for the number of games played per skill bucket to extrapolate a modified overall win rate assuming equal games played per skill bucket.  That would remove the player skill popularity skew effect.


If you want to be one of the 47 Ronin check out our recruitment post and then send a message to Prof935 (xbox) or dimka935 (ps4).

Estimated Marks of Excellence thresholds - https://wotclans.com.br/tanks/moe


JDOGx34 #10 Posted 15 May 2020 - 03:38 AM

    Corporal

  • Players
  • 22290 battles
  • 22
  • Member since:
    01-30-2018

rabit Sasquatch said it right

 

only looking at stats is absolutely retarded, there I said it

 

I have the amx 30b,  leopard one andn the stb,  they all suck,  everyone knows it who actually plays this game a lot

 

every tank in the game should have a reason to play it and these three tanks have no real reason other than you run out of tanks if you've done every other grind

 

we're gaming would make more money if every single tank had a reason to play it and we all know they only care about money so we need to get the point across to them that every tank needs to have a fun reason to play!

 

and stats are never going to say that!   

 

 

 

 



MrWiggles24 #11 Posted 15 May 2020 - 08:59 PM

    Sergeant

  • Players
  • 14323 battles
  • 186
  • Member since:
    11-26-2015

View PostJDOGx34, on 14 May 2020 - 08:38 PM, said:

rabit Sasquatch said it right

 

only looking at stats is absolutely retarded, there I said it

 

I have the amx 30b,  leopard one andn the stb,  they all suck,  everyone knows it who actually plays this game a lot

 

every tank in the game should have a reason to play it and these three tanks have no real reason other than you run out of tanks if you've done every other grind

 

we're gaming would make more money if every single tank had a reason to play it and we all know they only care about money so we need to get the point across to them that every tank needs to have a fun reason to play!

 

and stats are never going to say that!   

 

 

 

 

It's not the tanks.



Poppyluv77-p #12 Posted 08 June 2020 - 07:59 PM

    Captain

  • Players
  • 17036 battles
  • 1,344
  • Member since:
    01-05-2017

View PostRaiBOT01, on 04 May 2020 - 06:35 PM, said:

Hey everyone!

 

My name is Jeff Gregg, and I am the Lead Designer on World of Tanks Valor.  Welcome back to "Lets talk about tanks! AMX 30 B edition".

As a reminder, the goal was to highlight specific tanks, one at a time, and really dig down into what makes them tick, how they are being used, and how approaches may be taken to improve them if necessary.

 

Lets see how our changes to the AMX 30 B.

 

What we did

As a reminder: the following was done to this tank:

  • Buff the penetration and shell speed on this gun to be more than now but not as much as it was before we nerfed it before. This will also impact the Tier 9 AMX 30
  • 105mm mle F1 gun
    • 105mm OFL 105 F1 shell
      • Speed: 1100 → 1525
    • 105mm OCC 105 F1 shell
      • Speed:  800 → 1000
    • Penetration: 248/300/53 → 255/310/53

Results

Lets take a look at what occurred.

All data from below is from Random Battle from April 22nd (first full day of changes) till April 30th

NOTE: We ran the loaded dice event this same week. (Damage RNG is +/-10% and Pen RNG is 0%)

 

Here is a breakdown of what we saw:

Stat % or # Desc
Battles 32,000

Usage of this tank went up considerably. Seems like players wanted to check out what the changes were about. This is good as it gave us a decent chunk of data to analyze (tho not enough to be conclusive, see more on that below)

  • ~ 3,555 a day (up from ~1,700 a day before)

AMX-UPDATE-1.jpg

Win Rate 51.59%

The win rate is up from a previous average of 49.9%. Expected, but to be honest a bit higher than we desired and expected.

Popularity Graph N/A

We don't have enough info re: popularity yet to make a conclusion. See the graphs below. It SEEMS the vehicle may be more popular than it was before with the better players, but its not possible to say that emphatically (note saw-tooth graph).

Previous

AMX-UPDATE-1.png

Current

AMX-UPDATE-2.jpg

 

Skill Graph N/A

There is a minute yet interesting dip in performance for lower skill players. But higher skill players are playing as we expect. This MIGHT mean the tank is balanced fine, but until we get more data on popularity, we can't say that for 100% sure. We need to keep collecting data.

Previous

AMX-UPDATE-2.png

 

Current

AMX-UPDATE-3.jpg

Accuracy 76.30%

The gun's accuracy is up from a previous average of 72%

AMX-UPDATE-4.jpg

Pen 52.97%

The gun's penetration is up from a previous average of 48.2%

AMX-UPDATE-5.jpg

Damage per Hit 249.3

Damage per hit is up from a previous average of ~230 (which makes sense as it is penetrating more).

AMX-UPDATE-6.jpg

Damage per Penning Hit 360.9

No real change (360 before) which makes sense as we buffed the pen, not the damage output on a successful pen.

AMX-UPDATE-7.jpg

Damage Per Battle 1,914

A nice bump from the previous average of ~1,600

AMX-UPDATE-8.jpg

Survival 27.4%

All this extra penetration as risen the survival from a previous average of 24.6%

AMX-UPDATE-9.jpg

Lifespan 4min 37s

Up marginally from 4min 32s

AMX-UPDATE-10.jpg

Action Items and Next Steps:

For now we are going to keep watching and make no further changes. If the win-loss stays high but the popularity evens out, we may have an issue. If the win-loss stays high but the higher skill players clearly are playing more, we may NOT have an issue. Either way we need to collect more data. Once we have enough to make additional calls, we will let you know.

 

Thanks and see you on the battlefield!

 

-CHUJI

 

View PostLacqueredBacon, on 11 May 2020 - 03:16 PM, said:

I don't think you can draw any conclusions from this data.  First off, the changes themselves are so minimal that the majority of the player base won't even feel a difference.  Second, comparing the results to the previous results can't tell the whole story.  I read these and think, Ok a tank with a slight buff is slightly better.

 

But does that mean it is balanced?  How is this comparing to other tier 10 mediums?

And quit with the win rate.  Win rate is still a horrible metric in a game with this many variables and with skill the most important variable that affects win rate.


 

 

It’s unbelievable they still haven’t completely reversed the unnecessary idiotic 4.6 nerfs. Gregg’s simplistic obsession with win rate continues 

BrogueOne-p #13 Posted 08 June 2020 - 09:34 PM

    Major

  • Supertest - PS4
  • 24821 battles
  • 11,202
  • [ELDRZ]
  • Member since:
    01-07-2016

View PostPoppyluv77, on 08 June 2020 - 07:59 PM, said:

 

 

It’s unbelievable they still haven’t completely reversed the unnecessary idiotic 4.6 nerfs. Gregg’s simplistic obsession with win rate continues 

You have a simplistic obsession with the idea that Jeff Gregg is a monolith; he’s not. He oversees much of the technical work, but decisions are made by committee.  



Poppyluv77-p #14 Posted 08 June 2020 - 11:54 PM

    Captain

  • Players
  • 17036 battles
  • 1,344
  • Member since:
    01-05-2017

View PostBrogueOne, on 08 June 2020 - 09:34 PM, said:

You have a simplistic obsession with the idea that Jeff Gregg is a monolith; he’s not. He oversees much of the technical work, but decisions are made by committee.  

 

He’s the lead developer so should be accountable in that sense




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users