Jump to content


So, a small sample size about game balance:

KV-2 Marks of Excelllence Artillery Top tier Bottom tier Statistics Luck Sample Size

  • Please log in to reply
35 replies to this topic

Beorn_of_the_NorthernSea #1 Posted May 04 2018 - 19:47

    Major

  • -Players-
  • 25573 battles
  • 2,663
  • [-SCA-] -SCA-
  • Member since:
    08-07-2014

I have been screwing around with my KV-2 (my favorite and most played tank), and a while ago I thought I would try to get some Marks of Excellence on the barrel.

 

Well, first one was easy enough.  So, I started working on the second one...  And about 2 weeks ago or so I started tracking certain things for myself...  The usual; damage done, spotting damage, wins or losses or draws and such.

 

BUT, I also started tracking tier of the match (so, 6, 7, or 8) and the number of artillery in the match {0 [LOL], 1,2, or 3}, and I must say that the results are a bit... interesting.

 

So, far in the past 38 games (I know, small sample size), 19 wins, 19 losses, 0 draws.  m'kay, about right for me.

I am actually LOSING ground towards the 2nd MoE.  Sometimes I made bone-headed moves.  Sometimes I was arty focused [more on that below].  Sometimes?  Bad luck.

 

Number of times top tier in 38 games: 3 (THREE)

Number of times bottom tier in 38 games:21 (TWENTY-ONE) [INCLUDING THE LAST FIVES GAMES IN A ROW]

The remaining games were tier 7 matches.

 

Number of times there was NO ARTY in 38 games:2 {TWO}

Number of times there was ONE ARTY in 38 games:13 {THIRTEEN}

Number of times there were TWO ARTY in 38 games:8 {EIGHT}

Number of times there were THREE ARTY in 38 games:15 {FIFTEEN} including one streak of 8 consecutive matches

 

Y'all know I am not an ARTY hater...  I am not fond of the mechanic, and the better I get, the more I get focused...  BUT, FFS man, does any of the above seem to follow the mathematical Law of Averages to you?

 

I know, I know, extremely small sample size.  But, I am going to CONTINUE to log this until I get 3 marks (IF EVER) so, I even have an excel spreadsheet for this and everything...  And, as long as the game exists, I am going to keep records.

 

Just seems...  a bit off.

 

Any hoo, y'all have a great weekend, and good luck on the battlefield to ALL of my fellow denizens of the digital armored cavalry!

 

~Beorn (who apparently lined his helmet with tin foil overnight)

 


Edited by Beornotns, May 04 2018 - 19:48.


Kamahl1234 #2 Posted May 04 2018 - 19:54

    Major

  • Players
  • 18393 battles
  • 10,078
  • Member since:
    04-06-2012

What about the mathematical law of averages applies to how many arty are present? That count depends on how many are in queue at relevant tiers. Especially as with having a hard limit, you can't dump arty into any battle and make more arty-free battles. As an example, if you limited arty to 2 or even 1 per side, you'll nearly or never see an arty-free battle. Ever. It's just popular enough to be fairly present. That said it's also tier dependent, so you'd probably be more likely to see more arty at some tiers than others, as not many flesh out full lines in arty.

 

 

You may have a point if that was just for tiers, but I can basically assure you that you'll see the expected percentages over enough battles for 3/5/7 to have the expected distribution. 



Treacherous_Neighbour #3 Posted May 04 2018 - 20:09

    Captain

  • Players
  • 26809 battles
  • 1,277
  • Member since:
    01-04-2013

MM is random, 3,5,7 is not.

 

You are noticing the effect of a deliberate move to bias MM towards making you (and everyone else) bottom tier as much as possible.

 

Then you have to fire premium ammo more....

 

Which makes a premium account more neccessary

 

Which equates to more $$$ for WG. The whole business model is diabolic.

 



Backfire_ #4 Posted May 04 2018 - 20:12

    Major

  • Players
  • 28974 battles
  • 2,205
  • [X-OUT] X-OUT
  • Member since:
    03-08-2013
I like that you're taking a factual look at data.  Much more interesting than the average ermg MM rigged kinda post that is way more common.  Keep logging it :)  I'd be interested to see what it looks like at a few hundred battles.

Kamahl1234 #5 Posted May 04 2018 - 20:36

    Major

  • Players
  • 18393 battles
  • 10,078
  • Member since:
    04-06-2012

View PostTreacherous_Neighbour, on May 04 2018 - 19:09, said:

MM is random, 3,5,7 is not.

 

You are noticing the effect of a deliberate move to bias MM towards making you (and everyone else) bottom tier as much as possible.

 

Then you have to fire premium ammo more....

 

Which makes a premium account more neccessary

 

Which equates to more $$$ for WG. The whole business model is diabolic.

 

 

Something I don't get with this thought process is, wasn't the old MM where you often found more top tiers than currently, something which would have you fire premium more than 3/5/7?

 

Like if I only see 3 top tiers, that's at most 3 I "have" to shoot premium at, but the old MM often had 5+, sometimes as many as 15 enemy +2 tier tanks. IMO that MM wanted you to fire premium much more often. Limiting top tiers to at most 20% of the enemy team would greatly reduce premium ammo "needs" because now you have a set limit, and not players preparing for the worst possible. Heck I carry less premium now in battle than I've ever done before. 



Treacherous_Neighbour #6 Posted May 04 2018 - 20:45

    Captain

  • Players
  • 26809 battles
  • 1,277
  • Member since:
    01-04-2013

View PostKamahl1234, on May 04 2018 - 20:36, said:

 

Something I don't get with this thought process is, wasn't the old MM where you often found more top tiers than currently, something which would have you fire premium more than 3/5/7?

 

Like if I only see 3 top tiers, that's at most 3 I "have" to shoot premium at, but the old MM often had 5+, sometimes as many as 15 enemy +2 tier tanks. IMO that MM wanted you to fire premium much more often. Limiting top tiers to at most 20% of the enemy team would greatly reduce premium ammo "needs" because now you have a set limit, and not players preparing for the worst possible. Heck I carry less premium now in battle than I've ever done before. 

 

delusional....

GeorgePreddy #7 Posted May 04 2018 - 20:45

    Major

  • Players
  • 14345 battles
  • 10,390
  • Member since:
    04-11-2013

View PostTreacherous_Neighbour, on May 04 2018 - 16:09, said:

MM is random, 3,5,7 is not.

 

You are noticing the effect of a deliberate move to bias MM towards making you (and everyone else) bottom tier as much as possible.

 

 

To make players bottom tier "as much as possible" would require a 1/1/13 MM.

Treacherous_Neighbour #8 Posted May 04 2018 - 20:46

    Captain

  • Players
  • 26809 battles
  • 1,277
  • Member since:
    01-04-2013

View PostGeorgePreddy, on May 04 2018 - 20:45, said:

 

To make players bottom tier "as much as possible" would require a 1/1/13 MM.

 

delusional.....

Drone157 #9 Posted May 04 2018 - 20:46

    Staff sergeant

  • Players
  • 21356 battles
  • 317
  • [GBOTS] GBOTS
  • Member since:
    01-19-2012

First thing’s first, I also love the KV-2, and have over 1.2k games in it. I’ve got 2 marks and let me tell you, don’t try to get the 3rd one with the derp gun, go with the 107.

 

As far as the results of the data go:

Your low win rate is a significant factor in not having high MOE progress.

I’d love if you could include how many games were 3/5/7 5/10 and all 15.

 

From my rough calculations, I believe we should see a split as follows over 1,000 games. 26% Top tier, 25% Mid Tier, and 49% Bottom tier. (I consider all same tier as mid tier.)

 

I’ll probably start logging my tier “8” games in order to test these numbers.

 

But because of tier 10 there will probably be a 10-15% skew towards bottom tier.

 

good luck with your progress.



GeorgePreddy #10 Posted May 04 2018 - 20:51

    Major

  • Players
  • 14345 battles
  • 10,390
  • Member since:
    04-11-2013

View PostTreacherous_Neighbour, on May 04 2018 - 16:45, said:

 

delusional....

Ad Hominem Fallacy.

 

Disregarding facts and resorting to ad hominem attacks is a childish strategy.



GeorgePreddy #11 Posted May 04 2018 - 20:54

    Major

  • Players
  • 14345 battles
  • 10,390
  • Member since:
    04-11-2013

View PostTreacherous_Neighbour, on May 04 2018 - 16:46, said:

View PostGeorgePreddy, on May 04 2018 - 20:45, said:

 

To make players bottom tier "as much as possible" would require a 1/1/13 MM.
delusional.....

 

Actually, this is a fact, nothing to do with anyone being delusional. A 1/1/13 would make the maximum number of players bottom tier.

Ad hominem attacks do nothing good for you.



Treacherous_Neighbour #12 Posted May 04 2018 - 21:08

    Captain

  • Players
  • 26809 battles
  • 1,277
  • Member since:
    01-04-2013

View PostGeorgePreddy, on May 04 2018 - 20:51, said:

Ad Hominem Fallacy.

 

Disregarding facts and resorting to ad hominem attacks is a childish strategy.

 

Delusional and touchy....



Pipinghot #13 Posted May 04 2018 - 22:56

    Major

  • Players
  • 25960 battles
  • 8,924
  • [IOC] IOC
  • Member since:
    11-20-2011

View PostBeornotns, on May 04 2018 - 13:47, said:

I have been screwing around with my KV-2 (my favorite and most played tank), and a while ago I thought I would try to get some Marks of Excellence on the barrel.

Hi Beorn, hope all is well, and wishing you good fortune on your quest. You know me, so you know this is going to have a lot of words, but hopefully this will help you in pursuit of your goal.

View PostBeornotns, on May 04 2018 - 13:47, said:

So, far in the past 38 games (I know, small sample size), 19 wins, 19 losses, 0 draws.  m'kay, about right for me.

I am actually LOSING ground towards the 2nd MoE.  Sometimes I made bone-headed moves.  Sometimes I was arty focused [more on that below].  Sometimes?  Bad luck.

Glad to see you're trying to take a data driven approach, it will be interesting to see what things look like when you get to about 75 or 100 battles.

 

On a side note, I'm hoping you know that it's going to be very difficult for an average player like ourselves to 3-mark anything. I'm hoping that your primary reason for doing this is a mini-quest to improve your KV-2 playing, and that you won't be overly disappointed if you can't make it to 3 marks. To be fair, you're better in the KV-2 than most of your tanks, but even dark green stats usually are not going to get someone a third mark, your game play will have to be at least a blue-stats level if you want to get that third mark. I'm not trying to be a bummer, just hoping that even if you don't get that third mark you're mentally in the right place to enjoy the process and hopefully get something out of it.

View PostBeornotns, on May 04 2018 - 13:47, said:

Number of times top tier in 38 games: 3 (THREE)

Number of times bottom tier in 38 games:21 (TWENTY-ONE) [INCLUDING THE LAST FIVES GAMES IN A ROW]

The remaining games were tier 7 matches.

This brings to mind a couple of questions.

 

1) Were you playing only the KV-2 during these 38 battles or did you play other tanks in between your KV-2 battles?

Depending on whether you like to use noobmeter or wotlabs as an information source, they're indicating that you've played either 337 battles or 161 battles in the last 7 days, and that strongly indicates that you've been playing other tanks too. If that's true, if you've been playing other tanks in between your KV-2 battles, then the data you've presented above means absolutely nothing. Sorry to be a bummer (again) but that's just reality, you can't selectively look at only your KV-2 battles when all of those other tanks affect how the 3/5/7 system treats you.

 

So if you've played any other tanks in between your KV-2 battles then you either need to present the data for all of your battles during this time, or you need to throw out what you've collected and start over using the right methods until you're done with your exercise. I'm sure that sounds like a bummer, but it's the only way to get honest results. There are many kinds of data analysis that can be valid by extracting a subset of the data, but this is not one of them. If you want to get an accurate picture of how the 3/5/7 MM is working for you then you have to include every battle that you play, this is not something that can be examined by extracting a subset of the data, or looking at it on a per-tank basis, because that amounts to cherry picking which skews the results.

 

2) When you say, "The remaining games were tier 7 matches" does this mean that in the remaining 14 battles you were middle tier? I think that's what you mean, but I don't want to make an assumption.

 

3) During this time, were you in any battles that had only 1 or 2 tiers in them, rather than the full three tier spread of a 3/5/7 battle? The major flaw with the 3/5/7 system is that any battles that are all same-tier, like 15/15 Tier VI tanks, count as "top tier". This means that even a Tier 1 tank can be "top tier" if it gets put into a 15/15 battle with all Tier 1 tanks. And of course every battle in a Tier X tank counts as top tier, which is something that a lot of people forget or don't know. Personally, I think this is crappy decision making by by WG, I think 15/15 battles should be excluded from any and all 3/5/7 calculations, but that's the reality we have to deal with for now.

View PostBeornotns, on May 04 2018 - 13:47, said:

Number of times there was NO ARTY in 38 games:2 {TWO}

Number of times there was ONE ARTY in 38 games:13 {THIRTEEN}

Number of times there were TWO ARTY in 38 games:8 {EIGHT}

Number of times there were THREE ARTY in 38 games:15 {FIFTEEN} including one streak of 8 consecutive matches

 

Y'all know I am not an ARTY hater...  I am not fond of the mechanic, and the better I get, the more I get focused...  BUT, FFS man, does any of the above seem to follow the mathematical Law of Averages to you?

The law of averages doesn't apply here. The MM has to make battles based on the tanks that are in the queue, if you're playing at a tier where there is more arty then you're going to see more arty. If you're playing during period when there are more arty players in the queue then you're going to see more arty. The MM can't ever guarantee you any specific number as the average number of arty per battle. There is no law of averages for this issue. The only expectation you can roll with is that you'll see a maximum of 3 of them, and any battle with less than 3 you should count yourself lucky and make the most of it. Hoping for fewer arty is not a strategy for 3-marking a tank, the game doesn't change just because you're pursuing a specific goal.

 

As a side note, I'm betting that you've always gotten focused more in your KV-2 because it's a big, slow obvious target that's easy to recognize even for n00b players. Arty players love to go after the KV-2, even when it's not the biggest threat on the battlefield. But in all of those other battles you weren't striving towards a goal based on significantly improving your performance, so all of those arty in previous battles didn't annoy you more in the KV-2 than they did in any other tank. This is almost guaranteed a mental issue rather than a game issue.

View PostBeornotns, on May 04 2018 - 13:47, said:

I know, I know, extremely small sample size.  But, I am going to CONTINUE to log this until I get 3 marks (IF EVER) so, I even have an excel spreadsheet for this and everything...  And, as long as the game exists, I am going to keep records.

Just a follow up on the sample size point. The KV-2 is by it's nature more prone to extreme swings of the pendulum. When you combine the factors of being large, slow and with a long reload time, this is a tank that has more battles on both ends of the spectrum. In the KV-2, it's easier to have one battle with 6 kills and a bunch of damage followed by a battle in which you get killed by arty before even firing a shot than it is with most tanks (especially at Tier VI). It's like a baseball player who always swings for the fences, they get more home runs but they also strike out more. Even blue and purple players, who are more consistent in their KV-2 than you or I might be, have more strikeouts in their KV-2 than they do in other tanks.

 

This fine for normal play, because average performance is all that matters. But for the marking system recent battles are weighted more heavily than older battles, which means that even a few bad battles can noticeably drag your ranking down. I'm having trouble finding it at the moment, but there is a good threat over at wotlabs that discusses in detail the math behind the marking system, and the formula is pretty non-intuitive. If you're not interested in digging up that thread that's cool, in that case just be aware of this - it's not good enough to improve your average game play if you want to get a third mark, you have to be trending upwards at just the right time to cross over to that next mark. There are a lot of purple players frustrated with their inability to 3-mark tanks at various times, because all it takes is one or two bad battles at the wrong time and then they have to play quite a few battles to get past that bad patch.

 

So even though you've won 50% of your 38 battles that doesn't mean much for marking tanks, because for marking tanks only your own personal performance matters. By your own admission, "I am actually LOSING ground towards the 2nd MoE.  Sometimes I made bone-headed moves.  Sometimes I was arty focused", all of which makes it sound like those 19 wins aren't telling you anything meaningful, the real story is what those "bone-headed moves" tell you about your personal performance in service to your goal of getting more marks. Overall performance doesn't help if that performance is up-and-down, averages aren't as good for getting marks as people think they are, your personal performance has to trend upwards at the right time in order to get to that next mark.

 

 



Urabouttudie #14 Posted May 04 2018 - 23:04

    Captain

  • Players
  • 21257 battles
  • 1,830
  • Member since:
    11-11-2013

View PostGeorgePreddy, on May 04 2018 - 11:51, said:

Ad Hominem Fallacy.

 

Disregarding facts and resorting to ad hominem attacks is a childish strategy.

 

...somebody sure has a solid understanding of the tactics and methods of online shillery...

 

Troll Vernacular: A+

Subtly: C-

Subterfuge: F

 

Summer school for you George...

 



Urabouttudie #15 Posted May 04 2018 - 23:06

    Captain

  • Players
  • 21257 battles
  • 1,830
  • Member since:
    11-11-2013

View PostPipinghot, on May 04 2018 - 13:56, said:

The law of averages doesn't apply here. 

 

Oh, now OP you've really kicked the can! 

 

You said 'spreadsheet"!!

 

It's Pipinghot...Neatoman's sock puppet shill account...

 

the veneer is thinning...

 

and the conversation just ended...



Urabouttudie #16 Posted May 04 2018 - 23:11

    Captain

  • Players
  • 21257 battles
  • 1,830
  • Member since:
    11-11-2013

View PostTreacherous_Neighbour, on May 04 2018 - 11:09, said:

MM is random, 3,5,7 is not.

 

You are noticing the effect of a deliberate move to bias MM towards making you (and everyone else) bottom tier as much as possible.

 

Then you have to fire premium ammo more....

 

Which makes a premium account more neccessary

 

Which equates to more $$$ for WG. The whole business model is diabolic.

 

 

...and the trolls ask endlessly..."but why would they do it?" when it's pretty simple to figure...

 

...any good detective will tell you "always follow the money"

 



Guido1212 #17 Posted May 04 2018 - 23:32

    Community Contributor

  • Players
  • 77358 battles
  • 8,082
  • [CARTL] CARTL
  • Member since:
    06-11-2011

View PostTreacherous_Neighbour, on May 04 2018 - 19:09, said:

MM is random, 3,5,7 is not.

 

You are noticing the effect of a deliberate move to bias MM towards making you (and everyone else) bottom tier as much as possible.

 

Then you have to fire premium ammo more....

 

Which makes a premium account more neccessary

 

Which equates to more $$$ for WG. The whole business model is diabolic.

 

 

lmbo, MM doesn't have to make you bottom tier more, 3/5/7 does that naturally unless you play a lot of 9/10

8_Hussars #18 Posted May 04 2018 - 23:45

    Staff sergeant

  • Players
  • 53542 battles
  • 358
  • [SHIRE] SHIRE
  • Member since:
    09-17-2013

The continued belief that the "law of large numbers" applies to whatever small sample sub-set of the data a player chooses is astounding!

The law of large numbers is a mathematical theorem; the law of small numbers is an observation about human psychology...

 

In general, people grossly underestimate the variability in small samples.  No one has good intuition about probability. Statisticians do better than the general public, not because their intuition is much better, but because they know not to trust their intuition; they know they need to rely on calculations....



Urabouttudie #19 Posted May 05 2018 - 00:12

    Captain

  • Players
  • 21257 battles
  • 1,830
  • Member since:
    11-11-2013

View Post8_Hussars, on May 04 2018 - 14:45, said:

The continued belief that the "law of large numbers" applies to whatever small sample sub-set of the data a player chooses is astounding!

The law of large numbers is a mathematical theorem; the law of small numbers is an observation about human psychology...

 

In general, people grossly underestimate the variability in small samples.  No one has good intuition about probability. Statisticians do better than the general public, not because their intuition is much better, but because they know not to trust their intuition; they know they need to rely on calculations....

 

A whole lot of supposedly highly educated and intellectually superior people in lofty positions of authority at NIST tried to convince me that 3 buildings in one day defied the laws of physics and my own intuition and collapsed at free-fall speed into their own footprints...

 

...as a result of office fires...the numbers said so...analysis and statistics compiled and proclaimed infallible by those powers that be...contrary to all the laws of nature known to man.

 

fast forward...

 

People were lining up at the news desks to take turns proclaiming HRC the next POTUS....every poll, statistic and talking head literally proclaimed a preordained victory...

 

The numbers said so...based upon well-established data...analysis and statistics compiled and proclaimed infallible by those powers that be...

 

It was guaranteed...right?

 

Way back...JFK and a magic bullet...again, intelligent men with lots of letters after their names proclaimed a single bullet traveled the proposed trajectory...

 

contrary to my own simple understand of these laws of physics...but the numbers said so...right there in black and white.

 

Who you gonna believe?

 

Here is what I know with a certainly, regarding "the numbers"...

 

People lie. People are self-serving. People are deceitful and deceptive and they will use whatever is at their disposable to convince you...to show you how it is really your own lying eyes which are deceiving you. Whatever you happen believe in and hold truthful...be it numbers or otherwise...will be the tool used to manipulate those beliefs.

 

This is what I know about the numbers....and people.

 

Trust. Your. Instincts.

 

...and if some POS ever tells you not to believe your own intuition and instincts...run, don't walk away, or prepare to fight.

 



Flarvin #20 Posted May 05 2018 - 00:17

    Major

  • Players
  • 52289 battles
  • 13,944
  • Member since:
    03-29-2013

View PostUrabouttudie, on May 04 2018 - 18:12, said:

 

A whole lot of supposedly highly educated and intellectually superior people in lofty positions of authority at NIST tried to convince me that 3 buildings in one day defied the laws of physics and my own intuition and collapsed at free-fall speed into their own footprints...

 

...as a result of office fires...the numbers said so...analysis and statistics compiled and proclaimed infallible by those powers that be...contrary to all the laws of nature known to man.

 

fast forward...

 

People were lining up at the news desks to take turns proclaiming HRC the next POTUS....every poll, statistic and talking head literally proclaimed a preordained victory...

 

The numbers said so...based upon well-established data...analysis and statistics compiled and proclaimed infallible by those powers that be...

 

It was guaranteed...right?

 

Way back...JFK and a magic bullet...again, intelligent men with lots of letters after their names proclaimed a single bullet traveled the proposed trajectory...

 

contrary to my own simple understand of these laws of physics...but the numbers said so...right there in black and white.

 

Who you gonna believe?

 

Here is what I know with a certainly, regarding "the numbers"...

 

People lie. People are self-serving. People are deceitful and deceptive and they will use whatever is at their disposable to convince you...to show you how it is really your own lying eyes which are deceiving you. Whatever you happen believe in and hold truthful...be it numbers or otherwise...will be the tool used to manipulate those beliefs.

 

This is what I know about the numbers....and people.

 

Trust. Your. Instincts.

 

...and if some POS ever tells you not to believe your own intuition and instincts...run, don't walk away, or prepare to fight.

 

 

If I don’t understand it, it’s not real. lol






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users