Jump to content


Thoughts on Matchmaking and Win Rate

Win Rate Matchmaking random universe

  • Please log in to reply
84 replies to this topic

worldoftanksblitzcanada #41 Posted 28 July 2015 - 07:15 AM

    Lance-corporal

  • Players
  • 34310 battles
  • 17
  • Member since:
    01-25-2015
I love the words very wise and calm mate excuse my grammar

Beastnumber2 #42 Posted 28 July 2015 - 09:47 AM

    First Sergeant

  • Players
  • 10168 battles
  • 1,113
  • Member since:
    06-28-2014

Give this man CWG! :) 

 

 

Between this post and your others, like the Ace Mastery thread and such, your a really great contributor man. Keep it up! :)



glass2707 #43 Posted 28 July 2015 - 02:02 PM

    First Sergeant

  • Players
  • 30004 battles
  • 3,070
  • [GRIM]
  • Member since:
    09-27-2014

View PostCrimsonMG, on 27 July 2015 - 06:45 PM, said:

Let's hope this ends the unfair MM threads, I remember Skittles posting a similar thread.

 

I can only laugh at the notion of this actually being happening.  The people who complain about MM aren't using reason for the most part.  In threads when I try to introduce reason to their claims they usually attack me, claim I work for WG, etc...  

Savvybuilder #44 Posted 28 July 2015 - 04:12 PM

    Lance-corporal

  • Players
  • 17742 battles
  • 28
  • [MOM]
  • Member since:
    04-27-2015

Thank you for the very thoughtful post based on fact and not a rant based on conspiracy theory.  MM is like flipping a coin for every battle and the results are a zero sum game.  For every potato there is a Unicum.  Even the average potato will win 40% of the time and the rest of us fight for the 10% left on the table by carrying harder.  Potato bashing and in-game raging will not change this equation (foul language is on the rise however) AND potatoes have the strong belief that it is their tank that is producing the poor results rather than the tanker.  The potato must spend copious amounts of money to get a better tank since they will not accumulate and earn the necessary credit to get one for free.  Say, perhaps, you owned Wargaming.....who supplies the money to pay your salary?  Why would you cater to the freeloading Unicums who do not pay a red cent to your salary?  You let the elite players play for free to get the potatoes to pay more money.  If the potatoes (interchangeable with Wargaming ATM machine) don't win 40% then they will quit and stop spending money.  Hug a potato.....they let us play for free.

 

 



Mere_Anarchy #45 Posted 28 July 2015 - 04:25 PM

    First Sergeant

  • Players
  • 35681 battles
  • 1,118
  • Member since:
    07-07-2014

View PostMistake_, on 27 July 2015 - 07:59 PM, said:

 

 

Thanks!

 

View PostXavier5A, on 27 July 2015 - 08:24 PM, said:

I haven't seriously platoon in the Matilda in quite some time! I'm totally up for this! Before the update comes out (which I'm guessing will be August 4th).

I mainly haven't been playing it because of my recent found love for my A-32. SPEED! VRRRM VRRRM! BOOM!

 

Sounds good.  At times its a bit slow for me, but some days it's just what the Dr. ordered.

 

View PostPhodilus, on 27 July 2015 - 08:27 PM, said:

I'll just save the link to this topic for future use....

 

This is an absolutely amazing post. Bravo and well done sir

 

Thanks for the kind words.

 

View Postitsnotpersonal, on 27 July 2015 - 08:38 PM, said:

Well done! Kudos. :honoring:

 

TY!

 

View Postzippojinx, on 27 July 2015 - 09:59 PM, said:

The only time I win is when I get lumped in with people who know how to play. So basically everyone else is to blame for my win rate

 

Lol.

 

View PostNevirSayDie, on 27 July 2015 - 10:38 PM, said:

Great post, of course! The only thing I'd say is that the official statement, while informative, isn't really a good resource. I can't really recommend anyone with questions to read it, because it is technically accurate, but makes it appear that Blitz actually has a system for matchmaking. All MM really does is put 14 random players from across 2-3 tiers together (and make sure each team has 1 platoon if there are 2 toons in those 14 random players). 

 

I agree that the best way to enjoy the game is to accept the fact that the matchmaker isn't out to get anyone, that it's completely random, and that most teams will not be even and large swings in short-term WR are guaranteed. From a standpoint of pure principle, the matchmaker bothers me because it's so poorly designed, but the gameplay is still fun enough to make up for its design flaws. 

 

Thanks for the compliments.  I appreciate your additional thoughts on MM.

 

View Post_Diesel_41_, on 27 July 2015 - 10:42 PM, said:

Thanks brother! I needed this. Been pretty hard on myself lately thinking I needed to carry harder but, it is what it is. Thanks Van.

 

Glad it helped Diesel!  I won't pretend that I still don't get frustrated and angry with myself, but I am learning to be more accepting.

 

View Postfredlf, on 27 July 2015 - 11:06 PM, said:

Nicely done, Van. Wise words that apply to much more than a tank video game. Kudos, man.

 

Thanks fredlf - high praise.

 

View Postworldoftanksblitzcanada, on 28 July 2015 - 01:13 AM, said:

Awesome post nice picture CC Sly and well my WR has risen q bit but by not much

 

Thanks!


 

 


Mere_Anarchy #46 Posted 28 July 2015 - 04:28 PM

    First Sergeant

  • Players
  • 35681 battles
  • 1,118
  • Member since:
    07-07-2014

View Postworldoftanksblitzcanada, on 28 July 2015 - 01:15 AM, said:

I love the words very wise and calm mate excuse my grammar

 

No excuse or apology needed.  Thanks for the compliment!

 

View PostBeastnumber2, on 28 July 2015 - 03:47 AM, said:

Give this man CWG! :) 

 

 

Between this post and your others, like the Ace Mastery thread and such, your a really great contributor man. Keep it up! :)

 

Thanks - I try in the few ways I know how.  Mostly its just random stuff that I think about when I should be sleeping.  Every once in a while a feel motivated to actually type something out and post it, if nothing else b/c it helps me to think through it fully.  Glad that others find them useful as well!

 

View Postglass2707, on 28 July 2015 - 08:02 AM, said:

I can only laugh at the notion of this actually being happening.  The people who complain about MM aren't using reason for the most part.  In threads when I try to introduce reason to their claims they usually attack me, claim I work for WG, etc...  

 

Lol - can't say I disagree with you.

 

View PostSavvybuilder, on 28 July 2015 - 10:12 AM, said:

Thank you for the very thoughtful post based on fact and not a rant based on conspiracy theory.  

 

Thank you - your positive feedback means a lot.


 

 


1carburo #47 Posted 28 July 2015 - 05:22 PM

    Sergeant

  • Players
  • 54858 battles
  • 419
  • Member since:
    06-30-2014

Excellent post as usual vanpeenen. It's a shame that no amount data will convince the tin foil hat brigade.

 



Eddie_Toms #48 Posted 28 July 2015 - 05:29 PM

    Sergeant

  • Players
  • 53184 battles
  • 310
  • [T0ES]
  • Member since:
    03-01-2015

Over the long term MM appears to be pseudo-random as far as evenly matched lineups go. 

 

But players would be less frustrated if the standard deviation of the MM algorithm was lower. This would reduce the wide fluctuations in balanced vs un-balanced matches. Long loosing streaks (and winning streaks) would be less common. Since it's only natural to remember long loosing streaks... the level of frustration for even moderately skilled players would decrease.

 

The problem is the standard devitiation in MM. I don't know how come WG thinks they will make more profit with wide swings in unbalanced MM. Maybe people shot more pramo to try and compensate for lopsided teams? Maybe rage selling due to unbearable frustration during a long loosing streak generates more revenue?

 

One thing for sure, customers don't enjoy long loosing streaks where unbalanced MM neutralizes their hard earned skills. I simply stop playing when a bad streak starts. Sometimes I switch to Premium tanks with preferential MM. Or, I switch to low tiers where anything can happen. But most of the time I just quit. WG makes less money when less people play.


 

 


vvk2 #49 Posted 28 July 2015 - 05:47 PM

    First Sergeant

  • Players
  • 71320 battles
  • 2,128
  • [SPUDX]
  • Member since:
    11-29-2014

What an exceptional post. Well researched, well written. Quite rare qualities these days!

Thank you!

 

 



NevirSayDie #50 Posted 28 July 2015 - 08:36 PM

    First Sergeant

  • Players
  • 10005 battles
  • 1,411
  • [TAN60]
  • Member since:
    03-19-2012

View PostEddie_Toms, on 28 July 2015 - 05:29 PM, said:

Over the long term MM appears to be pseudo-random as far as evenly matched lineups go. 

 

But players would be less frustrated if the standard deviation of the MM algorithm was lower. This would reduce the wide fluctuations in balanced vs un-balanced matches. Long loosing streaks (and winning streaks) would be less common. Since it's only natural to remember long loosing streaks... the level of frustration for even moderately skilled players would decrease.

 

I'm not sure if Blitz matchmaking can be called an algorithm. It basically just puts 14 random players together from across 2-3 tiers. Algorithms usually imply multiple steps to a goal, and Blitz MM basically just has one step: gather 14 random players from similar tiers. 

 

 

I agree that the matchmaker is bad, but the OP's point is still valid. It's not out to get anyone. It may be crude and poorly-designed, but that's all. 

 



Dianite_2014 #51 Posted 29 July 2015 - 12:04 AM

    Junior Sergeant

  • Players
  • 31398 battles
  • 110
  • [FSX]
  • Member since:
    09-20-2014
Mm is bad and unfair.

SieurPersil #52 Posted 29 July 2015 - 01:04 AM

    Knight of the Round Potato

  • Players
  • 25048 battles
  • 1,164
  • [SPUDX]
  • Member since:
    12-21-2014

 

   Thank you sir for a well thought and explained post. If you would write a book, I would read it.


Volunteer Tater Ambassador - Potentate Relationship Manager - Inter Clan Camping Liaison Officer. Knight of the round potato. Information regarding the hamac or our adirondack chairs available. Quote of the moment : "Will focus fire on WriterDude for beer"


Mere_Anarchy #53 Posted 29 July 2015 - 11:20 AM

    First Sergeant

  • Players
  • 35681 battles
  • 1,118
  • Member since:
    07-07-2014

View Post1carburo, on 28 July 2015 - 11:22 AM, said:

Excellent post as usual vanpeenen. It's a shame that no amount data will convince the tin foil hat brigade.

 

Thanks for the compliment - much appreciated.  I'd like to think that you're mistaken about the second part, but the realistic side of my brain thinks you're probably right.

 

View PostEddie_Toms, on 28 July 2015 - 11:29 AM, said:

Over the long term MM appears to be pseudo-random as far as evenly matched lineups go. 

 

But players would be less frustrated if the standard deviation of the MM algorithm was lower. This would reduce the wide fluctuations in balanced vs un-balanced matches. Long loosing streaks (and winning streaks) would be less common. Since it's only natural to remember long loosing streaks... the level of frustration for even moderately skilled players would decrease.

 

The problem is the standard devitiation in MM. I don't know how come WG thinks they will make more profit with wide swings in unbalanced MM. Maybe people shot more pramo to try and compensate for lopsided teams? Maybe rage selling due to unbearable frustration during a long loosing streak generates more revenue?

 

One thing for sure, customers don't enjoy long loosing streaks where unbalanced MM neutralizes their hard earned skills. I simply stop playing when a bad streak starts. Sometimes I switch to Premium tanks with preferential MM. Or, I switch to low tiers where anything can happen. But most of the time I just quit. WG makes less money when less people play.

 

Fair points - but one thing to keep in mind is that if WG were to have a more 'balanced' MM system that would mean that they would take into account player win rate when putting together teams.  For some players to win, others have to lose.  The upshot of that would be that everyone's win rate would be more similar.  Or to put it a different way, it would actually be more difficult to have a win rate above 50%.  I'm not sure that would be more fun in the long term.

 

Don't get me wrong, I also hate the horrible loss streaks, but I'm not sure I would trade them for the awesome win streaks.  The other night my platoon partner ripped off 9 or 10 straight wins - both nail biting matches and blow outs.  It was a blast.  It sounds to me like you are doing what you need to deal with the frustration of loss streaks. 

 

View Postvvk2, on 28 July 2015 - 11:47 AM, said:

What an exceptional post. Well researched, well written. Quite rare qualities these days!

Thank you

 

Thank you, much appreciated.  And thanks as well for the Aces you've submitted!

 

View PostDianite_2014, on 28 July 2015 - 06:04 PM, said:

Mm is bad and unfair.

 

To each his own, but I would disagree with the unfair.  Unfair implies that it's biased, and I have yet to see empirical proof that this is the case.  But, I'm always open to consider evidence counter to this.  All the same - you are definitely entitled to your opinion.

 

View PostSieurPersil, on 28 July 2015 - 07:04 PM, said:

Thank you sir for a well thought and explained post. If you would write a book, I would read it.

 

Thank you!  As far as the book, how about I send you my dissertation?  I guarantee that would cure you of any desire to read anything I write ever again.  :P


 

 


Lephturn #54 Posted 29 July 2015 - 03:06 PM

    First Sergeant

  • Club Wargaming
  • 22174 battles
  • 3,647
  • [III-S]
  • Member since:
    07-28-2014

View PostNevirSayDie, on 27 July 2015 - 11:38 PM, said:

Great post, of course! The only thing I'd say is that the official statement, while informative, isn't really a good resource. I can't really recommend anyone with questions to read it, because it is technically accurate, but makes it appear that Blitz actually has a system for matchmaking. All MM really does is put 14 random players from across 2-3 tiers together (and make sure each team has 1 platoon if there are 2 toons in those 14 random players).

 

Great post Vanpeenen!

 

Bold added by me above - that is an excellent point, and helps explain why my platoon rate doesn't seem to correlate with my win rate to the same extent as my STAR1 penalty would imply. The benefit of having one other solid player on my team is often cancelled out by the other team also having a platoon. I see more platoons than I did a few months ago - and that's a good thing, but I suspect the correlation with WR is going down as a consequence. That's just my guess at this point, Face4star will have some actual data. :)

 


Lephturn Triarii Signature Media Producer

Host of the World of Tanks Blitz podcast TANK BOOM!

 

Search for "tank boom" on iTunes, Stitcher, SoundCloud or your podcast player of choice!

Android user? Get the Tank Boom! App: https://play.google.com/store/apps/details?id=com.tank.boom


glass2707 #55 Posted 29 July 2015 - 03:51 PM

    First Sergeant

  • Players
  • 30004 battles
  • 3,070
  • [GRIM]
  • Member since:
    09-27-2014

View PostLephturn, on 29 July 2015 - 03:06 PM, said:

 

Great post Vanpeenen!

 

Bold added by me above - that is an excellent point, and helps explain why my platoon rate doesn't seem to correlate with my win rate to the same extent as my STAR1 penalty would imply. The benefit of having one other solid player on my team is often cancelled out by the other team also having a platoon. I see more platoons than I did a few months ago - and that's a good thing, but I suspect the correlation with WR is going down as a consequence. That's just my guess at this point, Face4star will have some actual data. :)

 

 

Yeah though for every good platoon I see many that are bad.  You know those guys who ask random people because of the tank they are driving?  Those cant be that beneficial to them or their team when they randomly get that platoon instead of fullestnuke or that person who shan't be named or they will close the thread.

 



Eddie_Toms #56 Posted 29 July 2015 - 05:00 PM

    Sergeant

  • Players
  • 53184 battles
  • 310
  • [T0ES]
  • Member since:
    03-01-2015

If MM's computations are so simple and unbiased, then how come the empirical standard deviation is so high (long streaks are common)?

 

A simple system should result in even distributions of Ws and Ls... especially on weekends and during special events when the player pool is very large. However anecdotal  evidence seems to indicate these periods are when streaks are more common. This is inconsistent with a simple system.

 

Tier I and II battles should have less streaks as player skill levels are widely distributed. Of course seal clubbers can make a difference.


 

 


glass2707 #57 Posted 29 July 2015 - 06:44 PM

    First Sergeant

  • Players
  • 30004 battles
  • 3,070
  • [GRIM]
  • Member since:
    09-27-2014

View PostEddie_Toms, on 29 July 2015 - 05:00 PM, said:

If MM's computations are so simple and unbiased, then how come the empirical standard deviation is so high (long streaks are common)?

 

A simple system should result in even distributions of Ws and Ls... especially on weekends and during special events when the player pool is very large. However anecdotal  evidence seems to indicate these periods are when streaks are more common. This is inconsistent with a simple system.

 

Tier I and II battles should have less streaks as player skill levels are widely distributed. Of course seal clubbers can make a difference.

 

anecdotal evidence is just that anecdotal.  People don't make posts or really even take note of things when it happens as they expect it to.  Streaks happen if you flip a coin enough times.  Though you would expect that in something like this the person's play can make streaks more frequent.  By that I mean if someone is on a losing streak maybe they press more, are over aggressive, or even the opposite of that overly protective.  As a result they aren't playing as well.

 

 



__Crusader6__ #58 Posted 29 July 2015 - 07:30 PM

    Annoying Mod

  • Players
  • 59543 battles
  • 11,566
  • [-C6-]
  • Member since:
    12-08-2014

 

I'm copying some posts I made on the Triarii site.

 

I only had time for 5 games this am (silly work ;)
For ease of typing I am identifying Green/Red players over 50% - followed by battle result, and brief description of outlying factors) and my results
Game 1 5/2 for a 7-2 victory (C6 OBJ704 w/4 kills survived)
Game 2 6/1 for a 7-1 victory, Red IS-7 AFK, he was a 42% WR w/ 3k battles (C6 IS-7 w/ 2 kills Survived)
Game 3 5/2 for a 7-4 victory (C6 ST-1 2 kills survived)
Game 4 4/2 for a 6-7 loss, complete green fiasco on BlackGoldville, should have been win but team split three ways - enemy JadgTiger at 44% WR did most damage and kills - got me when it was 3-4 for them, and I should have had him (C6 IS-8 1 kill, destroyed)
Game 5 6/2 for a 7-6 victory, enemy had 1 68% WR player - who interestingly did not do most of the damage but was good at pinning folks down. (C6 IS-8 2 kills survived)

WR (Writedude) - one reason I hate campers - I'm not sure if our ISU, and JgPz E100 on game 4 where AFK, or just stupid, but they and a T-54 died with zero damage .

 

Update:
Game 6: 1/1 for a 7-2 victory (C6 T-54, zero kills and destroyed - but top damage and still more than the next three combined on my team) - stunning (I kid) performance from Red Leo PTA 0 damage, drove around looking at stuff at Castilla - 1313 battles,36% WR, 34% in PTA with 331 battles in it - proving you can loose your way to Tier X as he is attempting to do so. Pers rating of 256, 0 mastery badges, and 447 avg damage. .35 Destruction ratio, and .43 Damage Ratio.


Game 7: 2/5 for a 1-7 loss at Dead Rail (C6 JgPz E100 3x my team total damage, and only 2.9k sigh, ineffective E100 , E75, T-28 Proto, JP II and IS-7) - zero communication.

Game 8: 3/1 for a 7-0 win at Port Bay (C6 JgPZ E100 2kills 3.8k damage - survived) Nek3D (60.45%WR) in E50M called for flank (he did 3.4k) and JimmJames00 (55.32WR) did 3.8k too no real damage by others. All team followed instructions to hills/flank and supported (and we had two kills by other team mates at end).

Game 9: 3/6 for a 6-7 loss at Castilla (C6 E100 destroyed early 1.5k d) Darius in Green JgPz E100 (61.2% WR) did 3k damage he called Mill (E-75 and Leo PTA [55%WR]did not follow and drove to cap). All Red went to Mill area Red lead by platoon of DudeTanksALot in his M48A1 (5.8k dam and 4 kills) and RCC74 in his T110E5 (2.8k dam and 1 kill) supported by mobthool [56.62 WR] in a ST-1 w/ 2.6k Damage and 2 kills. Interestingly enough our E5 did 3.1k dam, and 4 kills is a 45% WR(646 battles in E5 at 44%)

 

 

So I tossed stuff into excel:

 

 

Green over 50%WR

Red over 50% WR

W/L

R/G +50 Ratio

5

2

w

2.5

6

1

w

6

5

2

w

2.5

4

2

l

2

6

2

w

3

1

1

w

1

2

5

l

0.4

3

1

w

3

3

6

l

0.5

35

22

6W/3L

2.322222

3.888889

2.444444

0.666667

 

 

Deducting me from all 9 games - I ended up with 26 players who filled 54 (6 players x 9 games) spots - or a 48% chance of each player being over 50%, other than me, with an average per game of 2.8 players plus myself for those games.

 

It's not yet a valid sample size - but I am going to track over the next 100 games to see what happens and someone with more than second year stats (from a long time ago) can make some sense of.

 

 

BUT it's clear that have lemons on your team is not good...

 


 
Tank Hoarder: 383 tanks in Garage:  375/385 aced (Mk1 Heavy and T49A repo),    wallet warrior.  Loyal Original M60 owner
 
 
Sorry I haven’t aced the Smasher yet - So I’m still ruining tier 6-8 

 


Mere_Anarchy #59 Posted 29 July 2015 - 09:01 PM

    First Sergeant

  • Players
  • 35681 battles
  • 1,118
  • Member since:
    07-07-2014

View PostEddie_Toms, on 29 July 2015 - 11:00 AM, said:

If MM's computations are so simple and unbiased, then how come the empirical standard deviation is so high (long streaks are common)?

 

A simple system should result in even distributions of Ws and Ls... especially on weekends and during special events when the player pool is very large. However anecdotal  evidence seems to indicate these periods are when streaks are more common. This is inconsistent with a simple system.

 

Tier I and II battles should have less streaks as player skill levels are widely distributed. Of course seal clubbers can make a difference.

 

First, I just want to start by saying that I very much appreciate your input and willingness to contribute to the conversation generated by this thread.  My point is not simply to have everyone agree, but also to stimulate larger thought and conversation on the topic.

 

I want to respond to your thoughts in two ways.

 

The first way is based on my own approach to the topic - from an empirical perspective.  Your primary question seems to be driven by the idea that the oscillations of wins and losses have too large a frequency to be random.  My primary response to that is to note that the swings of WR in my original graph have durations of one or more weeks because that is the time frame in which the data were gathered.  In reality the frequency of these swings is a direct product of the time scale in which they are recorded.  See the following graph which combines my original WR data (which is a 7-day scale of measurement) with one that uses the roughly 100 battle (roughly 2 day) scale provided by wotbstars.com (thank you face4stars!).  Notice how the swings in win rate are both of a greater amplitude and also occur with greater frequency.  Each of the larger oscillations is made up of smaller ones.  I know anecdotally that if I were to measure this at an even smaller (or larger) unit of time I would see the same.  I have some evenings where I will go from having a tough loss stretch in 10-15 battles to winning most of the next 10-15.  Overall my primary point was that if you empirically consider these over a sufficient time period and they do even out as you would expect with a random process.

 

Graph


 

But, in all honesty the second part of my response to you is more theoretical (or really epistemological if you want).  When it comes down to it, neither one of us can actually ‘prove’ that MM is random or not.  Only actually having the matchmaker mechanics available to us can answer that question.  I can only present data that should provide insight into the mechanics of this system.  My tendency to see the world in an empirical way makes me predisposed to interpret this data as representative of the underlying structure, which in this case appears to be random.   I am in no way naive enough to assume that Wargaming would never try and develop an app that employs subtle means of inducing their customers to spend more money.  However, in this particular case I don’t believe that MM is the way in which that occurs (note that I used the word believe).  Again, my approach to understanding this issue, as well as my likelihood of interpreting it as random is inherently colored by the larger way in which I view the world around me.

 

Let’s assume that MM really is non-random and designed to get more money from Blitz players. What can we do about it?  This is really one of the points I was trying to make as well in my original post (and linked to Omega’s earlier comments about it).  No matter what the reality of matchmaking, there is little we can do about it.  Honestly we have two choices (one of which you note) - we can accept it for what it is and tank on, or we can delete the app.  If it truly is an unfair system that antagonizes players to the point that they quit the game, in the long term it will not be a valid business strategy.  But if we choose to continue to play then we are knowingly accepting Wargaming’s conditions for how the game will work and have to simply deal with it.

 

Trust me, I get very upset about my poor performance in this game.  LIke others, I have those moments when I am cursing my teammates, RNG, myself, and want to throw my iPad across the room.  And, I’m not sure what makes me more upset, the fact that I have those loss streaks or that I let myself get so emotionally invested in a stupid-[edited]pretend tank game.  Yet, I keep coming back (over 18k battles) because there is something in it that is too much fun to put down.  This, in my mind, is where WG makes their money.  

 

All right - I’m out of breath and off my soapbox.  Sorry for such a long post!

 

 

 


Mere_Anarchy #60 Posted 29 July 2015 - 09:06 PM

    First Sergeant

  • Players
  • 35681 battles
  • 1,118
  • Member since:
    07-07-2014

View PostLephturn, on 29 July 2015 - 09:06 AM, said:

 

Great post Vanpeenen!

 

Bold added by me above - that is an excellent point, and helps explain why my platoon rate doesn't seem to correlate with my win rate to the same extent as my STAR1 penalty would imply. The benefit of having one other solid player on my team is often cancelled out by the other team also having a platoon. I see more platoons than I did a few months ago - and that's a good thing, but I suspect the correlation with WR is going down as a consequence. That's just my guess at this point, Face4star will have some actual data. :)

 

 

Thanks Lephturn - much appreciated.  Definitely explains why platooning is less effective at times.

 

View Post__Crusader6__, on 29 July 2015 - 01:30 PM, said:

 

I'm copying some posts I made on the Triarii site.

 

BUT it's clear that have lemons on your team is not good...

 

 

No kidding!  Check out the post I did much earlier where I look at this (in a little less detail but with greater sample size).  Keep at it, I'm interested to see what you come up with!


 

 





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users