rocketbrainsurgeon

WN8: the good, the bad, and the padding

242 posts in this topic

Interesting read, didn't know about the limitation on spots.

If we could only use the spot damage...

 

Also, kills are slightly overrated IMO.

Share this post


Link to post
Share on other sites

Awesome article, and I have to agree: I am not a padder myself, but since I got the T-54 and obj-140, my ratings have rocketted upwards by ~800 recent (even reached 4300 WN8 recent when playing only those in a night-was wondering why). I also agree about the light tank thingy: I'm driving the 12t cause I want to train a crew for the 13 90 without playing a tier 8 scout with a shit crew: 2500dmg+5kills+1k spot? 17k WN8 (one battle).

 

And I also agree that we NEED spotting damage: Another 12t game with 4.8k spots, 1.5k dmg and no kills was "only" 5k WN8 (still a lot, I know-but 4k spot dmg being worth MUCH less than 1k dmg+5kills?....)

 

I really hope the API improves, and the same goes for expected, although I understand it is hard: If a tank was OP but got nerfed (M48, VK2801, etc.), there is no way to know when the battles a player has were done AFAIK.

Share this post


Link to post
Share on other sites

Fair assessment IMO.

 

Every metric used will have issues.  Nothing will ever be perfect.

Share this post


Link to post
Share on other sites

M46 Patton is absurdly easy to pad in, but Im not playing it for the padding I play it because I love American Mediums. Overall though, probably my favorite article. Maybe a *slight* fix to individual game WN8 values would be to add a Win/Loss modifier, so a 5k WN8 loss would lose.. Oh for example 5%, taking it down to 4750. Still a great game, but not as good as a 5k WN8 win by just a smidge. A win would introduce no WN8 modifier. No idea if this is even possible

 

Super Pershing, however? Yeah... Im padding in that tank but I also need an American Medium trainer

Share this post


Link to post
Share on other sites

As you said, it's fine for the 99%.  A perfect system is impossible.  If we know the numbers involved, some of us will always find a way to min/max.  If we don't know the numbers involved, the lack of transparency will make people doubt the system.  That 1% is always going to exist, and if we're worried enough about it, we know exactly how to spot them.  Additionally, anyone capable of padding most of those tanks is equally capable of exceeding ~2500.

Share this post


Link to post
Share on other sites

There have been so many discussions about this recently, the problems with padding we or dpg for wn8. Maybe Praetor should just have the wn9 thread already. Maybe it's time to officially discuss the new metric? Otherwise very well written article.

I'm particularly glad to see more threads like this instead of: Why are Asians stupid, or Yurocum vs Americum.

Share this post


Link to post
Share on other sites

 

 

  1. 2 sets of expected values: historical WN8 vs recent WN8.  This takes care of the OP/UP phenomenon.  The recent WN8 is a full list, while the historical version is just where some tanks need patching.

 

The problem is that WG doesn't track WHEN you achieved the stats you did on a tank, just that you did. If we decided to separate out every tank whose performance has been affected by buffs/nerfs, then we would be left with a very short list indeed.

 

Here's a brief list of tanks who have been readjusted over the life of WoT:

1. Russian heavies

2. Russian mediums

3. American mediums

4. American heavies

5. German heavies

6. German mediums

7. French lights

8. French mediums

9. French heavies

10. All artillery

11. Russian lights

12. German lights

.

I'm sure I missed some...

Share this post


Link to post
Share on other sites

It be easier to say what hasn't changed. British tanks except the tort and cent, Chinese tanks, and jap tanks.

Share this post


Link to post
Share on other sites

It be easier to say what hasn't changed. British tanks except the tort and cent, Chinese tanks, and jap tanks.

 

Except the Chinese tanks got their HEAT pen nerfed across-the-board. I guess that just leaves Brits and Japanese.

Share this post


Link to post
Share on other sites

As you said, it's fine for the 99%.  A perfect system is impossible.  If we know the numbers involved, some of us will always find a way to min/max.  If we don't know the numbers involved, the lack of transparency will make people doubt the system.  That 1% is always going to exist, and if we're worried enough about it, we know exactly how to spot them.  Additionally, anyone capable of padding most of those tanks is equally capable of exceeding ~2500.

 

Right, the real problems only exist with people that are already very good.  It's easy for me to recommend taking the ELC out to collect a big rating, but the average player will go WAT.JPG and then drive off a cliff with it.  There are still some issues such as with the M4 Sherman: the expected values are so high that it's a stat-murderer for anyone attempting it post-nerf.  Or that some of the best tanks in game are also rating padders.  -_-

 

I probably re-wrote the article 5 or 6 times, starting with a very harsh stance and gradually coming to accept it as I got a better understanding.  But as I understand it now, it seems best to turn it into a comparative damage metric as the other pieces of the formula are marginalized.

Share this post


Link to post
Share on other sites

I do like the idea you mentioned, of a fluctuating "recent" measure.  Help keep things straight when tanks get buffs/nerfs.  I wonder if that's too much effort for too little reward, though.

Share this post


Link to post
Share on other sites

"Damage gets the largest number because that is what the creators want to matter the most."

 

  • Pls cut out crap like that - it was never about what we wanted but what the numbers told us. Damage has the strongest correlation to wins - simple as that - even for light tanks (t5 and above). It has nothing to do with what Praetor, Crab or others wanted to matter. I have above 3000 WN8 in my Chaffee, 60% WR, 1K damage, 1.2 kills/game, 3 spots/game all solo - pls go and look up the stats of 50% WR chaffee drivers ... I m certain I know what you ll find there. Would be nice if such examples would also be shown in articels like yours.

 

"Get rid of the spotting/defense/win factors of the equation since they don’t really matter anyway."

 

  •  why should that be done, if it would decrease the accuracy of WN? And that would happen ... read the thread, the numbers are there for everyone to check.

 

WN8 isnt perfect ...

 

  • Everyone knows that myself included

 

 

Overall though good article :thumbup:

Share this post


Link to post
Share on other sites

If damage leads to wins, why are you purple in WN8 and green & blue in winrate?

 

I'm not picking on you specifically, look at my stats.  Super-uni WN8... but it didn't lead to higher winrate for me.  During the experiment, my winrate consistently went down as I was reacting less to flanks/defense/spotting and just focusing on that next shot of damage.  Rexxie has a 4.2K WN8, and 1% higher recent winrate than I do despite dealing a truck-ton more damage.  Etc etc etc.

 

Higher damage obviously does correlate strongly with wins, but only to a point.

Share this post


Link to post
Share on other sites

 

Higher damage obviously does correlate strongly with wins, but only to a point.

it seems more like as long as you bring unicum platoon-mates your win rate sky rockets, it also depends on the time of day, and if it's a X5 weekend.

Share this post


Link to post
Share on other sites

I have a pivot point of about 2k WN8 and 60% winrate in the WZ-132 and have been playing around with raising and lowering my damage output. I notice that playing for damage significantly affects my solopub winrate in that tank. When I start trying to get more damage, my WN8 will start trending into the 3k range and my winrate will dip into the 50s. When I focus on being at the front and only taking opportunistic shots, my WN8 dips into the lower 1k range but my winrate jumps up to the upper 60s.

 

It really sucks because I want to see my WN8 trending up but playing the WZ-132 to win pushes it down.

Share this post


Link to post
Share on other sites

.... why are you purple in WN8 and green & blue in winrate?

 

...

 

 

Simple - because I mostly screw around nowdays.

 

Everyone camping in the woods Komarin or Murovanka? Fine, but without me - I ll take bath or suicide scout in my arti and try my luck with the next bunch of pubbies. E100s hill climbing Himmelsdorf (yes they do on Eu-Server, together with IS4, Maus and Tortoise) -> back to garage. Lemming train? Without me! Some dumb pubbie blocks me, rams me whatever - I ll finish it and switch to an other account if I have to for the next 1h - 24 hours. I dont give a fuck - just play for my fun and thats it. Let the dumb unwashed masses carry their own sorry asses - I wont anymore.

Share this post


Link to post
Share on other sites

So we agree: simply dealing damage doesn't lead to wins.  It's a lot of other things besides "more damage", like where or why that damage is being dealt.

Share this post


Link to post
Share on other sites

Simple - because I mostly screw around nowdays.

 

Everyone camping in the woods Komarin or Murovanka? Fine, but without me - I ll take bath or suicide scout in my arti and try my luck with the next bunch of pubbies. E100s hill climbing Himmelsdorf (yes they do on Eu-Server, together with IS4, Maus and Tortoise) -> back to garage. Lemming train? Without me! Some dumb pubbie blocks me, rams me whatever - I ll finish it and switch to an other account if I have to for the next 1h - 24 hours. I dont give a fuck - just play for my fun and thats it. Let the dumb unwashed masses carry their own sorry asses - I wont anymore.

 

That's not an explanation for high WN low WR. That's more of an explanation for low WR low WN.

Share this post


Link to post
Share on other sites

So we agree: simply dealing damage doesn't lead to wins.  It's a lot of other things besides "more damage", like where or why that damage is being dealt.

 

I agree. The difficulty is capturing that and separating it out from platoon effect.

Share this post


Link to post
Share on other sites

Pre 8.11 for a couple of months every time I got erlenburg I would just yolo into the enemy while maintaining a2500 wn8 rating and 60% wr recent. Only time I wouldn't was when I platooned and even then sometimes I would convince my plat to do the same.

Share this post


Link to post
Share on other sites

That's not an explanation for high WN low WR. That's more of an explanation for low WR low WN.

My damage output is still good enough for 2500WN and if I would play serious, I d get around 60% WR solo random, because thats what I do - playing solo random mostly (last couple of months around 98%). Throw in arti solo random, where the WR is lower anyway ... . I only have 26 platoon awards.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.