29 replies to this topic

[_GraceToo]

javajoejava, on 13 October 2016 - 06:46 PM, said:

If you want to learn what STATISTICALLY matters the most (based on fact, not gut), check out this thread:

 

http://forum.wotblit...te-to-win-rate/

 

 

Cool thread.  This one and the one you linked to.

I think if I were running a clan (which I'm not so this is total armchair quarterbacking) from what I've seen and read I'd focus far less on WR and far more on survival % and damage ratio.  

 

I'd love to actually check the correlation between WR and some other factors.  Does anyone understand the API the folks use to get data for blitzstars etc.?

[javajoejava]

Rage_And_Love, on 13 October 2016 - 01:18 PM, said:

 

I'd love to actually check the correlation between WR and some other factors.  Does anyone understand the API the folks use to get data for blitzstars etc.?

 

Thanks.  You don't need the API to check the correlation ratios or run a regression analysis.  I just pasted the data from WOTBStars and did it in Excel.

 

It intuitively makes sense too... Remember you can do a ton of damage and still lose, but by definition (supremacy mode/draws/capping aside) you can't lose if you survive.  Even campers that wait until the very end to engage, still end up dying.

 

 

Edited by javajoejava, 13 October 2016 - 08:27 PM.

[TydiriumX]

Above^^^

  Even campers that wait until the very end to engage, still end up dying.

yeah but its because they cant play.

 you should not be losing because you finally got taken out after doing 50% more damage than the rest of your team just to have a camper who has been alive from the beginning finally 'try' to engage. This is a team game and people need to help others out to win. Not go afk or camp till they are the last ones.

 

btw, I try my best to win while putting the hurt on the reds when I can. So 'winning' is my option while doing severe damage to the reds

Edited by TydiriumX, 13 October 2016 - 08:22 PM.

[BobboEvans]

javajoejava, on 13 October 2016 - 07:54 PM, said:

 

Thanks.  You don't need the API to check the correlation ratios or run a regression analysis.  I just pasted the data from WOTBStars and did it in Excel.

 

It intuitively makes sense too... Remember you can do a ton of damage and still lose, but by definition (supremacy mode aside) you can't lose if you survive.  Even campers that wait until the very end to engage, still end up dying.

 

 

Hmmm... What the heck happened in my T1 Heavy??  I did the grind way before supremacy.

 

WR:  56.69%

Survival:  63.78%

Games Played:  127

No... I wasn't just camping in back.  My damage ratio was 1.8 with 3.6 kills per death.

 

Correlation = Broken. 

 

 

[Gitgudm9]

I hate getting carried and/or barely doing any damage. However, I hate 4K damage losses at tier 7 more...

[javajoejava]

BobboEvans, on 13 October 2016 - 02:10 PM, said:

Hmmm... What the heck happened in my T1 Heavy??  I did the grind way before supremacy.

 

WR:  56.69%

Survival:  63.78%

Games Played:  127

No... I wasn't just camping in back.  My damage ratio was 1.8 with 3.6 kills per death.

 

Correlation = Broken. 

 

 

 

No.  That just means you lost a portion of games due to draws or capping.

 

Here are your actual stats for all your tanks with 100+ battles - nothing even comes close to Survival %.....

 

	Survival	Dmg Ratio	Kills/Btl	Kills/Death	Hit Rate	Avg Def	Avg XP	Avg Dmg	Avg Spot
Correlation %:	72%	55%	59%	69%	-12%	9%	3%	-21%	30%

 

 

 

Edited by javajoejava, 13 October 2016 - 08:21 PM.

[Player_5773117413]

In this era of truly terrible teams, players wanting to win will usually have to "do it themselves". They'll have to carry the burden of others to victory. This usually requires one player dealing a lot of damage to make up for the rest. Therefore, both kinda go hand in hand, though neither will guarantee the other.

[Posit1ve_]

javajoejava, on 13 October 2016 - 12:46 PM, said:

If you want to learn what STATISTICALLY matters the most (based on fact, not gut), check out this thread:

 

http://forum.wotblit...te-to-win-rate/

 

I personally believe aiming for survival will get you nowhere. And the correlation does not imply causation, so even though higher WR is correlated with a higher survival, you can't prove what causes what.

 

There's no way to prove higher survival rate causes higher WR or vice versa. Which brings us back to the debate on which matters more, damage, winning, or (if you really care) survival

[Leadf00t]

panzermk6, on 14 October 2016 - 04:30 AM, said:

I personally believe aiming for survival will get you nowhere.

 

I'm not sure I agree with the "nowhere" part, but I do think that too much focusing on survival can breed poor behaviors.  Specifically, playing too passively.   I think my best rounds are those when I am balancing on the knife edge between being aggressive and not taking too much damage (particularly early).  If I focus too much on surviving, I tend to play too passively only to be the last man standing against 4 or more enemies -- which typically doesn't end well.

Edited by Leadf00t, 13 October 2016 - 08:41 PM.

[BobboEvans]

javajoejava, on 13 October 2016 - 08:16 PM, said:

 

No.  That just means you lost a portion of games due to draws or capping.

 

Here are your actual stats for all your tanks with 100+ battles - nothing even comes close to Survival %.....

 

	Survival	Dmg Ratio	Kills/Btl	Kills/Death	Hit Rate	Avg Def	Avg XP	Avg Dmg	Avg Spot
Correlation %:	72%	55%	59%	69%	-12%	9%	3%	-21%	30%

 

 

 

A portion!  That's 9 games lost where the other team capped AT A MINIMUM assuming I survived every win (which I'm sure I didn't).  That stat just jumped out at me the other day... That's a lot of cap losses.

 

Getting back on point... 99.9% of players can't be relied upon to cover or carry you.  So if you want wins... You better do damage.  Damage is the Means and Wins is the Ends and the Ends justify the Means.  Damage is primary focus.  And live while you're at it.