Jump to content


BlitzAnalysiz.com - Final Stats for update 7.5 added

blitzanalysiz.com new site stats

  • Please log in to reply
155 replies to this topic

Just_merpug #141 Posted 12 December 2020 - 08:31 AM

    First Sergeant

  • Player
  • 58206 battles
  • 10,261
  • [AFK]
  • Member since:
    03-05-2013

View Posterwin10001, on 12 December 2020 - 12:52 AM, said:

Great work on the stats.

 

What I found interesting are the most played tanks by players in tier 6 and tier 7 are KV-2 and SU-152 respectively. Basically, tanks for camping/sniping. 

 

Weirder still is that AT8 is the top tier 6 tank by relative winrate

:child:

If you check the Tank Zeitgeist data on Blitzstars, you'll see that the most played tanks generally are ones with big guns. Says a lot about the playerbase. 


Denoobing member of the Active Frontline Klan - and proud of it. Be nice to me..https://www.blitzstars.com/player/eu/Chairman_merpug


erwin10001 #142 Posted 12 December 2020 - 08:38 AM

    Sergeant

  • Player
  • 2171 battles
  • 366
  • [BT_SV]
  • Member since:
    11-16-2019

I was a little bit surprised by the Chimera being the no.1 tank, but then I noticed there were far fewer players suggesting that those that buy the tank actually know something about gun depression. 

The reload time is so painful so it is more of a platoon tank rather than a solo tank. I took it out for a spin and got 6 wins in a row... something I haven't achieved with Hellcat so far. :D



gallabru #143 Posted 12 December 2020 - 09:47 AM

    First Sergeant

  • Player
  • 18214 battles
  • 1,038
  • [LAGER]
  • Member since:
    09-01-2016
Nice as always.
Just two quick remarks : 1/ I find some charts would gain not being displayed full scale, like Hit Rate and another one about WR. They look like straight horizontal lines (unless this is the impression you want to give, ofc). 2/ collector tanks appear as premium. Some high tier premium are collectors, but some are old techtree tanks (DW2 for instance). I have no suggestions on that point.

3 comments.
1/ I have the feeling that the tanks with the highest delta wr and/or highest wr are also played by above average players. They are generally premium tanks too. Is there some stat management behind it in addition to p2w commercial strategy ?
2/ the most played tanks are, by far, techtree tanks.
3/ most played tier is tier 8. It would be interesting to have the proportion of premium tanks played per tier. This could be displayed as a two color bar in your histogram. An additionnal information would be the fraction of existing premium tanks to techtree per tier. Maybe too much to put on a single chart.

My team stats in real time  : http://forum.wotblit...ats-repository/

Carlsberg clan family is recruiting. Contact us.


jylpah #144 Posted 15 December 2020 - 07:46 PM

    First Sergeant

  • Player
  • 27036 battles
  • 5,676
  • [AFK]
  • Member since:
    08-05-2014

View Postgallabru, on 12 December 2020 - 11:47 AM, said:

Nice as always.
Just two quick remarks : 1/ I find some charts would gain not being displayed full scale, like Hit Rate and another one about WR. They look like straight horizontal lines (unless this is the impression you want to give, ofc). 2/ collector tanks appear as premium. Some high tier premium are collectors, but some are old techtree tanks (DW2 for instance). I have no suggestions on that point.

3 comments.
1/ I have the feeling that the tanks with the highest delta wr and/or highest wr are also played by above average players. They are generally premium tanks too. Is there some stat management behind it in addition to p2w commercial strategy ?
2/ the most played tanks are, by far, techtree tanks.
3/ most played tier is tier 8. It would be interesting to have the proportion of premium tanks played per tier. This could be displayed as a two color bar in your histogram. An additionnal information would be the fraction of existing premium tanks to techtree per tier. Maybe too much to put on a single chart.

 

While the focused scale would show differences bigger, I prefer to show them from zero so minor differences do not get exaggerated. Hit rate is the only chart that has minor differences between the tanks. I don't think the chart were any better if it would show the differences larger when the data is between 82-80%.

 

Relative WR charts (as all the other charts) have an issue that the data is not stratified. Therefore the player base distribution impacts on the numbers. I have a to-do item about stratified sampling.

 

Therfore, most of the well performing tanks in terms or relative are actually better in the hands of the average players - since there are so many more average players than exceptionally good players. Well-armored tanks do well in Relative WR at the moment. I do not think stratification will have major impact, but it will remove the impact of player base distribution. In addition to stratification, I need to think how to add stats for different player skill-categories without multiplying the page count with 5. It takes 6 hours to build the site already now.

 

The feature request #3 is easy to make. Approved :) I could for example add premium tank battles with different color into this chart.


BlitzAnalysiz.com  In-depth analysis of WoT Blitz game statistics

Jylpah @EU  Jylpah @Asia  Blitz-tools  


llo1 #145 Posted 17 December 2020 - 08:35 AM

    First Sergeant

  • Player
  • 32144 battles
  • 7,685
  • [1RTR]
  • Member since:
    06-27-2014

Jylpah...as previously posted when you first issued the latest stats, I really do think that there is a “skew” around the Emil 1, and since the (slight) nerf, this is even more pronounced. I really think that as people become more “used” to these tanks in the game then players are more easily overcoming them. 
I feel that they have always been vulnerable with such a long reload between clips, and have taken advantage of this myself, but since I have taken mine out again, I am really struggling to actually pen the reds, the increased dispersion seems to have ensured that 39% of shots do not land where aimed, and to cap it all, the pen seems to have been reduced (although not announced), but could simply be the dispersion!

Do you have any updated results beyond your last grab of WG data?



_RagingRabbit_ #146 Posted 17 December 2020 - 09:37 AM

    First Sergeant

  • Player
  • 18156 battles
  • 1,140
  • [0XIDE]
  • Member since:
    08-20-2016

Wow, statistics are obviously abracadabra for me :D 

 

Nice work though Jylpah! 



jylpah #147 Posted 17 December 2020 - 11:27 AM

    First Sergeant

  • Player
  • 27036 battles
  • 5,676
  • [AFK]
  • Member since:
    08-05-2014
@llo1, I gonna fetch preliminary stats next week. Due to the way the scripts have been built, there needs to be enough time between the snapshots for the data to make any sense. Otherwise it is just noise.

BlitzAnalysiz.com  In-depth analysis of WoT Blitz game statistics

Jylpah @EU  Jylpah @Asia  Blitz-tools  


eliasdepelias #148 Posted 17 December 2020 - 01:36 PM

    First Sergeant

  • Player
  • 43582 battles
  • 1,287
  • [EMRS]
  • Member since:
    12-21-2014

The Emil1 has gotten a serious nerf. It feels like a vulnerable little autoloader now, not a tank that could boss every hull down position. Looking at the stats, it deserved one but maybe they overcooked it. Like Ilo siad, once players get used to a tank, stats should go down a bit.

It is also interesting that premiums with the same OP stats get left untouced, but it takes them only 1 update to correct the emil. And that is a shame because this game could use some of the tech tree tanks being stronger / a little bit OP.

The Emil2 feels almost as good as before. 



jylpah #149 Posted 17 December 2020 - 01:42 PM

    First Sergeant

  • Player
  • 27036 battles
  • 5,676
  • [AFK]
  • Member since:
    08-05-2014
Yeah, the ambivalence with OP premiums and the urge to nerf almost-as-good tech-tree tanks makes it just very obvious how WG sees the game as pay2win where they are all fine with money giving a sizeable advantage over others. It has not always been like that.

Edited by jylpah, 17 December 2020 - 01:47 PM.

BlitzAnalysiz.com  In-depth analysis of WoT Blitz game statistics

Jylpah @EU  Jylpah @Asia  Blitz-tools  


jylpah #150 Posted 27 December 2020 - 01:03 AM

    First Sergeant

  • Player
  • 27036 battles
  • 5,676
  • [AFK]
  • Member since:
    08-05-2014

Update 7.5 preliminary stats added

 

Preliminary stats for update 7.5 have been added.

 

  • In total, there were 151M battles by 481k players recorded in the database during the update 7.5 so far. Please note, these are note all the WoT Blitz players, just those tracked by BlitzAnalysiz[].

 

Top 10 tanks by Average WR

 

  • Maus has taken the title of the best performing (most OP) tank in the game just ahead of T28 Concept and Object 252U. The nerfed Emil I to 15th, but still doing fine. It is nice to see the so-miserable AT line to feature at the strongest performing tanks charts, so my tip is to grind that line, before it gets power-creeped again.

 

Best performing tanks, update 7.5


BlitzAnalysiz.com  In-depth analysis of WoT Blitz game statistics

Jylpah @EU  Jylpah @Asia  Blitz-tools  


erwin10001 #151 Posted 27 December 2020 - 08:50 AM

    Sergeant

  • Player
  • 2171 battles
  • 366
  • [BT_SV]
  • Member since:
    11-16-2019

The Tank with the Highest Damage and Highest Average Kills and no.5 in Average WR in Tier 6 is simultaneously in the Bottom 20 Relative WR list!  :bajan:

 

This sort of confirms Hellcat as a skill intensive tank. Sadly, WG doesn't like skill-intensive tanks, so I guess I'll just enjoy the Hellcat till it gets nerfed!


Edited by erwin10001, 27 December 2020 - 08:50 AM.


jylpah #152 Posted 28 December 2020 - 12:29 PM

    First Sergeant

  • Player
  • 27036 battles
  • 5,676
  • [AFK]
  • Member since:
    08-05-2014

The playerbase still impacts on the Relative WR. I have ideas how to eliminate that impact through stratified sampling. The idea is to sample the player base for each tank based on the average player base of the tier. This will eliminate the differences in the tanks' playerbase that is very pronounced for new tech-tree tanks or very expensive OP premium tanks.

 

But which is more understandable performance measure: Average WR OR Relative WR?

 

When stratified, the following holds:

 

(stratified) Relative WR = Player base average WR at the tier - (stratified) Average WR


BlitzAnalysiz.com  In-depth analysis of WoT Blitz game statistics

Jylpah @EU  Jylpah @Asia  Blitz-tools  


gallabru #153 Posted 28 December 2020 - 01:49 PM

    First Sergeant

  • Player
  • 18214 battles
  • 1,038
  • [LAGER]
  • Member since:
    09-01-2016
For some one just looking at charts, the average wr is straightforward.
But is it a good performance measure ? 

Edited by gallabru, 28 December 2020 - 01:49 PM.

My team stats in real time  : http://forum.wotblit...ats-repository/

Carlsberg clan family is recruiting. Contact us.


llo1 #154 Posted 28 December 2020 - 05:23 PM

    First Sergeant

  • Player
  • 32144 battles
  • 7,685
  • [1RTR]
  • Member since:
    06-27-2014
Jylpah, uploaded 30 gravity fails for you today!

jylpah #155 Posted 28 December 2020 - 07:45 PM

    First Sergeant

  • Player
  • 27036 battles
  • 5,676
  • [AFK]
  • Member since:
    08-05-2014

View Postgallabru, on 28 December 2020 - 03:49 PM, said:

For some one just looking at charts, the average wr is straightforward.
But is it a good performance measure ? 

 

If it is stratified (=scaled for the average played base of the tier, not the tank), it should be. Or what do you think?


BlitzAnalysiz.com  In-depth analysis of WoT Blitz game statistics

Jylpah @EU  Jylpah @Asia  Blitz-tools  


net4force #156 Posted 28 December 2020 - 11:44 PM

    Junior Sergeant

  • Player
  • 60622 battles
  • 214
  • [ADULT]
  • Member since:
    11-17-2014

The two graphs serve different purposes, both are interesting. If it isn't too much additional work for you, just keep them both.

 

BTW: very nice work!!






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users