67 replies to this topic

[Jylpah]

New blog post: Vehicle Class imbalance

 

Vehicle Class (Im)Balance

 

I my view, the game became more boring after WG added Vehicle Class Matching rules into the MatchMaker (update 5.1). The exact vehicle class matching rules have been adjusted since, but basically the Vehicle Class Matching limits the number of Tank Destroyers and Light Tanks per team and makes sure there is only +/-1 tank difference between the teams across the vehicle classes. This results mirrored teams causing monotonic battles. In my view, most of the vehicle class matching rules should be abolished, but I know well how I am in the minority here.

 

 

But the MM is not the main story of this post, but a long intro to the issue at hand - the imbalance of the vehicle classes in the game. I believe WG tried to solve fundamentally a balancing issue through the Vehicle Class Matching -rules. And while there is a logic in their thinking, they could have fixed the root cause instead and do better job balancing the tanks. When looking at the best performing tanks chart above, you can notice how light tanks are completely missing from the list and how the mediums are severely under presented. Heavies and heavily armored TDs rule.

 

Continue reading at BlitzAnalysiz.com

 

[_Crusader6_]

Jylpah, on 13 March 2021 - 03:48 PM, said:

New blog post: Vehicle Class imbalance

 

Vehicle Class (Im)Balance

 

I my view, the game became more boring after WG added Vehicle Class Matching rules into the MatchMaker (update 5.1). The exact vehicle class matching rules have been adjusted since, but basically the Vehicle Class Matching limits the number of Tank Destroyers and Light Tanks per team and makes sure there is only +/-1 tank difference between the teams across the vehicle classes. This results mirrored teams causing monotonic battles. In my view, most of the vehicle class matching rules should be abolished, but I know well how I am in the minority here.

 

 

But the MM is not the main story of this post, but a long intro to the issue at hand - the imbalance of the vehicle classes in the game. I believe WG tried to solve fundamentally a balancing issue through the Vehicle Class Matching -rules. And while there is a logic in their thinking, they could have fixed the root cause instead and do better job balancing the tanks. When looking at the best performing tanks chart above, you can notice how light tanks are completely missing from the list and how the mediums are severely under presented. Heavies and heavily armored TDs rule.

 

Continue reading at BlitzAnalysiz.com

 

110% agree.  
   They hide a lot of imbalance issues with vehicle matching. 
 I (like you) would prefer they scrap the type matching aspect of MM and let the chips fall where they may.  
   But to do that requires significant rebalancing of vehicles in the classes. 
 

[Jylpah]

_Crusader6_, on 14 March 2021 - 12:15 AM, said:

110% agree.  
   They hide a lot of imbalance issues with vehicle matching. 
 I (like you) would prefer they scrap the type matching aspect of MM and let the chips fall where they may.  
   But to do that requires significant rebalancing of vehicles in the classes.

 

For me it seems the vehicle class matching was a way to deal with few overly popular tanks like The Death Star. And instead of nerfing the heavies the took s shortcut. But the constant cry for "bad MM" (i.e. when one loses) has likely impacted too. There are already calls for 1:1 tank class matching i.e. make sure the both sides have the same tier, same type tanks. I just wonder what would come next? completely mirrored teams?

[tanker070]

makes sense to me. Playing medium has gotten touch tougher with heavies ruling the field cuz (advance-consumables, Hp/Armor buffs) and there is always a medium on the enemy team ready to counter on flanks.

 

Playing leo1 vs all slow tanks used to be so much fun.. now Well its beyond rare.

 

mm should not look at is8 and e75 the same way.. they are not same  

[__V_O_P__]

Jylap - your website and posts are always amazing. thank you for these.

 

I'm not sure about MM in regard to tank classes. Personally, I don't want MM to determine the outcome of the battle. I see C6's proposal as putting too much in the lap of the gods, especially with the tank classes as imbalanced as the are at the moment. I've mentioned before on the forums... even now when the game throws up Green with one med bottom tier vs Red with two top tier meds it's almost always an auto loss. Because of the lower HP pools and higher DPMs, the medium flank gets resolved quickest, which frequently drives the result of a battle. In other cases - even with the class balancing in place there some really dumb actions... several times I've been on a team with two failtoons and then two normal toons on Red, which is mind-boggling. But the idea of it throwing up matches like 5 tier X heavies and 2 tier 9 TDs vs 5 tier X TDs and 2 tier 9 heavies...  

 

MM's sole aim seems to be to get 14 people into battle as quickly as possible. The less they fiddle with that principle the better, but I would still rather it wasn't completely random w.r.t. the tanks present.

 

Now if you really want to balance the Heavy meta... I would say (and I have said)... change the Line of Sight algorithms. Make them relative the direction that a tank's turret is looking in - 100% view range 90o front... say 75% the next 45% and then the rear 180o gets 50%. Why? because it favors tanks flanking. Once they are behind the enemy they get all the view range advantages and it makes it possible to break contact. The downside is that LOS calculations are computationally expensive. Plus it's likely incompatible with the LOS model WOTB uses now - which just uses modifiers for terrain and actions. But still...

[Jylpah]

__V_O_P__, on 20 March 2021 - 02:59 AM, said:

Jylap - your website and posts are always amazing. thank you for these.

 

I'm not sure about MM in regard to tank classes. Personally, I don't want MM to determine the outcome of the battle. I see C6's proposal as putting too much in the lap of the gods, especially with the tank classes as imbalanced as the are at the moment. I've mentioned before on the forums... even now when the game throws up Green with one med bottom tier vs Red with two top tier meds it's almost always an auto loss....

 

 

Now if you really want to balance the Heavy meta... I would say (and I have said)... change the Line of Sight algorithms. Make them relative the direction that a tank's turret is looking in - 100% view range 90o front... say 75% the next 45% and then the rear 180o gets 50%. Why? because it favors tanks flanking.

 

Thanks!

 

We had a quite debate about the mediums vs. heavies on EU forums. Some players say the mediums (in good hands) matter the most, but I cannot find statistical evidence for that. Even the very best players win more with heavies than with mediums. I think the really good players often prefer more mobile tanks since playing heavies rinse-repeat is ... rather boring.

 

I find your idea of spotting being related to the direction of the turret interesting. Some folks proposed shorter 6th sense wake-up time and camo reset time for light tanks. WG could experiment these. That is the only way to really know how would they impact on the gameplay.

[Jylpah]

Update 7.7 final stats added

 

Final stats for update 7.7 have been added.

• In total, there were 236M battles by 481k players recorded in the database during the update 7.7.
• Familiar & well-armored tanks took the top spots in tank performance rankings: 1) Object 252U, 2) Annihilator and 3) T54E1. There was only one Tech-Tree tank in the TOP 5 (Maus) and two in TOP 10 (AT 8).

 

 

• Tier X performance ranking was the same armor-spam: 1) Maus, IS-4, E 100 and so on. All the TOP 10 tanks have a strong armor profile. What is interesting is that the order is pretty much the same for better players (55%+ WR at tier X).

 

 

 

Czechoslovakian tanks

 

So how did the Czechoslovakian tanks do? Short answer is: “pretty bad”, but in reality it is still too early to say. The Relative WR of the Czechoslovakian tanks will likely to improve in the next update as players learn to play the tanks better and a larger share of the player population has finished the stock-grind. Blitzanalysiz[] tries to eliminate the impact of the stock-grind by requiring players to have a tier-specific number of battles in a tank before their stats can be included into the calculations, but that seems not to fully eliminate the stock-grind.

 

Tier X TVP T 50/51

 

The tier X TVP T 50/51 performed pretty average being among the lowest 1/3rd. It may have an unique playstyle, but it seem not to perform that well.

 

 

If you constantly see TVP T 50/51s to perform superb on the battlefield, that is simply because it has been played by very-very good players on average at this point. The rest of the players will grind it soon and the true (non-)performance of the tier X TVP will become more apparent.

 

 

Tier IX Škoda T 50

 

The tier IX Škoda T 50 performed bit better, but it still just average.

 

 

Tier VIII TVP VTU Koncept

 

The tier VIII TVP VTU Koncept was a big disappointment to play. Based on stats it was the second worst performing tier VIII medium and the worst performing tier VIII medium for players with 55%+ WR at tier VIII.

 

 

Tier VII Konštrukta T-34/100

 

I personally felt the tier VII Konštrukta T-34/100 was pretty decent, but based on stats it was actually one of the worst performing tier VII mediums.

 

 

Tier VI Škoda T 25

 

The tier VI Škoda T 25 fits the pattern and was one of the worst performing tier VI medium tanks.

 

 

I still cannot believe the Czechoslovakian tanks were so bad. I dislike the tanks, but I believe their stats will fair bit better (i.e. average) in update 7.8. Let’s see.

[_Crusader6_]

I’ve actually done remarkably well in the VTU - so much that I can’t fathom why as it’s a pretty trash tank.  
  I’ve fallen a bit under 70% at 178 games, down to 67.38% and it’s a bit below my 30-60-90 day average for solo tier 8 meds - for both WR and DMG but not considerably 

But compared to what I expected, it’s doing better for me. 

   It really has a gun and that is it for redeeming qualities.