Jump to content


Downtier Abbey

Downtier Abbey

  • Please log in to reply
23 replies to this topic

Captain_Underpants53 #1 Posted 14 December 2018 - 06:14 PM

    Second Lieutenant

  • Member
  • 15333 battles
  • 1,531
  • [H_PUN] H_PUN
  • Member since:
    04-17-2017

I asked this question 3 days ago in "Questions to Developers" but evidently they didn't see it because I never got a reply.

 

What determines, in direct plain and simple English please, whether or not you will be the downtiered plane in a battle?  If it were pure chance it should be happening around 50% of the time there are mixed tier battles.  But I had over 12 battles in a row the other day where I was in the lower tier.  This is so far outside the probability curve for simple chance as to be vanishingly small.

 

blindfold assures us that karma has no place in the MM but I was on a big winning streak at the time.  If not karma, what?


MSgt, USAF, (ret)

BB3_Oregon_Steel #2 Posted 14 December 2018 - 06:32 PM

    Master Sergeant

  • Member
  • 571 battles
  • 78
  • [H_PUN] H_PUN
  • Member since:
    10-26-2015
I think there was a pretty detailed article about exactly how the matchmaker works on the main WOWP website which should be right here https://worldofwarpl...matchmaker_209/ .  I think you can find your answer there. 

SenatorTH #3 Posted 14 December 2018 - 09:53 PM

    Master Sergeant

  • Member
  • 2583 battles
  • 80
  • Member since:
    07-04-2018

View PostBB3_Oregon_Steel, on 14 December 2018 - 06:32 PM, said:

I think there was a pretty detailed article about exactly how the matchmaker works on the main WOWP website which should be right here https://worldofwarpl...matchmaker_209/ .  I think you can find your answer there. 

 

That article has about the same value as The Fairy Tales by brothers Grimm or H.C.Andersen.

 

Everything was more or less fine with MM before 2.0.7 iirc, battles contained same tier mostly, waiting time was the same as now or even shorter. Suddenly from nowhere that new concept of MM appeared with mixed tiers, and since then all the complains started - sadly no one is listening.

 

Let's take standard situation with 4 humans each side.

 

WAS: I am t4, have 3 teammates t4, we fight against other 4 in t4 planes, all ground is t4. Very honest situation, isn't it?

 

IS: two possibilities.

1) I am t4, one more t4, two teammates t3. Other team contains 2x t4, 2x t3, ground and bots are t4 level. I don't care much about t3, as I fight vs t4. Situation for me is about the same as was earlier - fighting against equal opponents.

2) everything the same, but I am one of those t3's. Situation sucks completely - I fight vs higher tier, bots are higher tier, ground targets are higher tier, even the altitude for high-AAs is different. In fact it looks like predetermined conditions for loss. Remember, we're not considering any skills, just pure matchmaking.

Now if we add to this that +2 tier allowance, the situation with matchmaking comes anecdotic at all, like one t3 vs flight of two t4.

 

Conclusions you can make by yourself.



Captain_Underpants53 #4 Posted 14 December 2018 - 11:57 PM

    Second Lieutenant

  • Member
  • 15333 battles
  • 1,531
  • [H_PUN] H_PUN
  • Member since:
    04-17-2017

View PostBB3_Oregon_Steel, on 14 December 2018 - 01:32 PM, said:

I think there was a pretty detailed article about exactly how the matchmaker works on the main WOWP website which should be right here https://worldofwarpl...matchmaker_209/ .  I think you can find your answer there. 

 

I don't see anything in the article that answers my direct, simple question.  Just a lot of verbiage that, frankly, gives me a headache.
MSgt, USAF, (ret)

BB3_Oregon_Steel #5 Posted 15 December 2018 - 01:58 AM

    Master Sergeant

  • Member
  • 571 battles
  • 78
  • [H_PUN] H_PUN
  • Member since:
    10-26-2015

Fair enough.

 

It was pretty complex and in evaluating that complexity (analysis is kind of my life) it occurs to me that there is no simple answer to your question.  A lot seems to be dependent on which player planes are in the que, what type of planes they are, what tiers are there, how long the planes have been in the que amongst other things, all of which impact how the MM works. 

 

This is a FAR more complex system than that used is WOT or WOWS.  There, you're just looking at types of ships and tiers and you have a plentiful supply of human players to work with.  

 

My impression here is that the MM is so complex because it was designed primarily to reduce que times and try to balance out bots while attempting to be as fair as reasonable in the process.  I suppose, if they took historical averages over a representative period of time, someone might be able to come up with an estimate of how things, on average, have worked.  However it also occurs to me that as the variables shift and as new players enter the game in larger numbers, those historical averages would be inconsistent with present conditions which would make those averages, and any estimate based on them, to one extent of another, invalid.

 

In other words, it's entirely possible there may be no answer to your question since there are too many variables and the data set is not stable but is instead fluctuating beyond historical norms.  

 

I don't know, maybe someone at WG has taken the time to try and extract that type of data from the system, maybe applying some sort of linear regression to the data might allow them to predict where things are now and where they are likely headed.  I'm just saying, maybe they can't give you a simple answer because one doesn't exist. 

 

If that's the case, maybe it would be better if they just told us so, but I can also see how presenting that data would lead to even more challenges with their players than they already have.  I'm pretty sure, for example, Ham Hock Jones would have a field day with an answer like that.   

 

 


Edited by BB3_Oregon_Steel, 15 December 2018 - 01:59 AM.


Bobby_Tables #6 Posted 15 December 2018 - 02:37 AM

    Second Lieutenant

  • Member
  • 4642 battles
  • 1,378
  • [-DOW-] -DOW-
  • Member since:
    06-16-2014

View PostBB3_Oregon_Steel, on 14 December 2018 - 07:58 PM, said:

 I'm pretty sure, for example, Ham Hock Jones would have a field day with an answer like that.   

 

Say... whatever happened to HHJ.  He was doing daily posts and now is gone.  Hope he is OK. 

 

Maybe he is just on a winning streak.  Anyone seen him in game?



Captain_Underpants53 #7 Posted 15 December 2018 - 04:19 AM

    Second Lieutenant

  • Member
  • 15333 battles
  • 1,531
  • [H_PUN] H_PUN
  • Member since:
    04-17-2017
I had a battle against him last night.  He did a foolish thing.  I know, right?
MSgt, USAF, (ret)

SenatorTH #8 Posted 15 December 2018 - 06:42 AM

    Master Sergeant

  • Member
  • 2583 battles
  • 80
  • Member since:
    07-04-2018

View PostBB3_Oregon_Steel, on 15 December 2018 - 01:58 AM, said:

This is a FAR more complex system than that used is WOT or WOWS. 

 

Bad joke.

It doesn't work as intended, but even if it would - it's algorithm is very primitive, according to the blog.



Captain_Underpants53 #9 Posted 15 December 2018 - 09:00 AM

    Second Lieutenant

  • Member
  • 15333 battles
  • 1,531
  • [H_PUN] H_PUN
  • Member since:
    04-17-2017

I was hoping for an answer from the devs to my question.  But no.  I was expecting a tap dance but just crickets is not acceptable.

 

Well, keep on keepin' on.  Maybe they will get tired of pretending not to see it on 'Questions to the Developers'.

 

:sceptic:


MSgt, USAF, (ret)

EZP_MRKARP #10 Posted 15 December 2018 - 12:31 PM

    Senior Master Sergeant

  • Closed Beta Tester
  • 5021 battles
  • 270
  • Member since:
    01-26-2012
cap the answer is they dont care

Greg_Pattinson #11 Posted 15 December 2018 - 02:14 PM

    Senior Master Sergeant

  • Member
  • 10774 battles
  • 182
  • [GW__S] GW__S
  • Member since:
    06-02-2018

1- 12 is NOT "far outside the probability curve".  If you use a simple hi low count (+1, -1) in a 50/50 system (like blackjack) seeing a +12 or -12 is actually pretty common.  

 

2-  this is NOT a 50/50 system because its based on players in que and those players may be highly concentrated in certain tiers.  For example if most players are concentrated at tier 5 and you are playing tier 4 with very few players at t3 than there is a much higher probability of being down tiered.  You will be matched with T5 much more frequently than T3.   

 

3- application- since there is a jump in equipment slots and consumables in level 2 and than again at level 3 than t5 and t8 have a significant advantage.  Therefore players tend (not always) to concentrate at those tiers.  If you play at t4 or T7 you will than have a further disadvantage as a result of the MM.  By playing t6 you have a higher probability of actually being up tiered because of the high concentration of players at t5 and the low concentration at t7 with players avoiding that tier so they don't have to face t8s. 

 

4- a word of caution- I would like to stress I said TEND this does not mean always.  If a lot of people randomly decide to grind their t7 planes all at the same time than there is a high probability that T6 will be down tiered instead.  Also keep in mind how probability works.  The outcome (up tiered or down tiered) of the previous battle has no effect on the probability of the next.  Even if you are playing t6 and have say a 70% chance of being up tiered due to player concentrations at that time than 12 battles in a row being down tiered would still be well with in probability.  Also player concentrations are constantly changing as people enter and leave the game or choose what plane they want to fly.  You will not have the same probability for all 12 battles played in a row



Captain_Underpants53 #12 Posted 15 December 2018 - 04:07 PM

    Second Lieutenant

  • Member
  • 15333 battles
  • 1,531
  • [H_PUN] H_PUN
  • Member since:
    04-17-2017

So you are saying that if I was to flip a coin 12 times, it would be reasonable to expect 12 heads in a row?

 

I had this happen 12 times in a row playing, in order, Tiers III, IV, V, VI, VII, and VIII.  Rinsed and repeated.

 

I still say it is outside the probability curve.


Edited by Captain_Underpants53, 15 December 2018 - 04:11 PM.

MSgt, USAF, (ret)

Greg_Pattinson #13 Posted 15 December 2018 - 05:27 PM

    Senior Master Sergeant

  • Member
  • 10774 battles
  • 182
  • [GW__S] GW__S
  • Member since:
    06-02-2018

as a matter of fact yes.  one of my favorite sayings I learned in university statistics was "any statistical probability will become a statistical certainty given enough trials"  the coin flip was actually one of the things that was used as an example quite frequently.  12 trials is a low number.  so if there are only 12 trials the probability is low that they will all be the same.  However In 12 trials it is also a low probability that it will be exactly 50% either.  Try it and see if you come up with 6 heads and 6 tails. I would bet money that you don't.   I actually have an advantage because the picture on both sides is different giving a different weight.  But in a string of trials say 14,445 for example the probability of getting 12 in a row is highly likely.  It will also tend to come close if not exactly 50% (if it was perfectly balanced).  

 

But the main point is that it is not perfectly balanced.  It is based on player concentration which is not even close to balanced.  I will use the example you gave to demonstrate but it is purely hypothetical.  you play a game at t3.  there are not a lot of new people so there are more people who are doing daily missions at t4 to get achievements than at t2 so there is a high probability you were down tiered.  t4 was again a high probability you were down tiered for the reasons stated before.  people got sick of spitfire spam so moved up to t6 so high probability again at t5.  t6 you just got unlucky.  t7 was high probability again of being down tiered for previous stated reason.  and t8 there happened to be a couple of flights at t9 giving you the short end of the stick.  so really you only got unlucky once out of 6 trials or 2 out of 12.  



Captain_Underpants53 #14 Posted 15 December 2018 - 05:37 PM

    Second Lieutenant

  • Member
  • 15333 battles
  • 1,531
  • [H_PUN] H_PUN
  • Member since:
    04-17-2017

:medal:

 

Now that makes sense the way you explain it.  Thanks.  I must confess my last statistics class was over 35 years ago. 


MSgt, USAF, (ret)

BB3_Oregon_Steel #15 Posted 15 December 2018 - 07:34 PM

    Master Sergeant

  • Member
  • 571 battles
  • 78
  • [H_PUN] H_PUN
  • Member since:
    10-26-2015

View PostGreg_Pattinson, on 15 December 2018 - 09:27 AM, said:

as a matter of fact yes.  one of my favorite sayings I learned in university statistics was "any statistical probability will become a statistical certainty given enough trials"  the coin flip was actually one of the things that was used as an example quite frequently.  12 trials is a low number.  so if there are only 12 trials the probability is low that they will all be the same.  However In 12 trials it is also a low probability that it will be exactly 50% either.  Try it and see if you come up with 6 heads and 6 tails. I would bet money that you don't.   I actually have an advantage because the picture on both sides is different giving a different weight.  But in a string of trials say 14,445 for example the probability of getting 12 in a row is highly likely.  It will also tend to come close if not exactly 50% (if it was perfectly balanced).  

 

But the main point is that it is not perfectly balanced.  It is based on player concentration which is not even close to balanced.  I will use the example you gave to demonstrate but it is purely hypothetical.  you play a game at t3.  there are not a lot of new people so there are more people who are doing daily missions at t4 to get achievements than at t2 so there is a high probability you were down tiered.  t4 was again a high probability you were down tiered for the reasons stated before.  people got sick of spitfire spam so moved up to t6 so high probability again at t5.  t6 you just got unlucky.  t7 was high probability again of being down tiered for previous stated reason.  and t8 there happened to be a couple of flights at t9 giving you the short end of the stick.  so really you only got unlucky once out of 6 trials or 2 out of 12.  

 

Yep, agreed.  Kind of what I was trying to say but failing at.  

 

Thanks!! :medal:



SenatorTH #16 Posted 15 December 2018 - 07:41 PM

    Master Sergeant

  • Member
  • 2583 battles
  • 80
  • Member since:
    07-04-2018

View PostGreg_Pattinson, on 15 December 2018 - 05:27 PM, said:

But the main point is that it is not perfectly balanced.  It is based on player concentration which is not even close to balanced.

 

The question remains, why do we need balancer which only brings us imbalance? Well, it's kinda rhetorical, as whatever you'd answer devs don't care.

Reitousair #17 Posted 15 December 2018 - 08:09 PM

    Command Chief Master Sergeant

  • Member
  • 7499 battles
  • 550
  • [HVAR] HVAR
  • Member since:
    08-10-2013

View PostGreg_Pattinson, on 15 December 2018 - 10:27 AM, said:

 

But the main point is that it is not perfectly balanced.  It is based on player concentration which is not even close to balanced.  I will use the example you gave to demonstrate but it is purely hypothetical.  you play a game at t3.  there are not a lot of new people so there are more people who are doing daily missions at t4 to get achievements than at t2 so there is a high probability you were down tiered.  t4 was again a high probability you were down tiered for the reasons stated before.  people got sick of spitfire spam so moved up to t6 so high probability again at t5.  t6 you just got unlucky.  t7 was high probability again of being down tiered for previous stated reason.  and t8 there happened to be a couple of flights at t9 giving you the short end of the stick.  so really you only got unlucky once out of 6 trials or 2 out of 12.  

 

You sir, deserve a +1

 

Though I've been finding the mid-tiers a lot more populated as of late. Flying tier 6 is about 40/60 whether or not you get up/downtiered from when I was screwing around in tier 6 yesterday (some of it depends on the time of day too as different people play at different hours,) while tier 7 is almost always getting uptiered I can't stop using WT terminology, fite me and fighting tier 8's.

 

I've generally found better chances of being top tier when I play during the evening (I live on the best coast of the United States,) but the chances to be top tier seem to increase dramatically during the day if you fly tier 6 or 9 due to how many people reside in tiers 5 and 8 (though bear in mind almost everybody who plays at tier 10 will have fully kitted out planes, possibly even specialized so watch out when flying 9's.)

 

View PostSenatorTH, on 15 December 2018 - 12:41 PM, said:

 

The question remains, why do we need balancer which only brings us imbalance? Well, it's kinda rhetorical, as whatever you'd answer devs don't care.

 

We need just a little bit more balance in the matchmaking to make matches closer and more satisfying to play.

 

...I say that but then I realize somebody thought Cold Peak - Gem of the Mountain was a good idea :sceptic:


Edited by Reitousair, 15 December 2018 - 08:09 PM.

I do fly Chinese aircraft a fair amount so... *ahem*

CHINA NUMBAH WAN

 

I even made a custom skin for my IL-10M!

 


Azis_ #18 Posted 15 December 2018 - 10:39 PM

    Second Lieutenant

  • Member
  • 773 battles
  • 1,398
  • [RBL_] RBL_
  • Member since:
    02-20-2013

The only real balancing that should be going on or even really probable, is matching the bots IMO. Honestly would be nice to leave some amount of realism and incentive to play. Why improve pilot or plane if not to gain an advantage. Real battles were rarely "balanced" in the sense that most whine about.

 

I have noticed lately that flying certain class of plane I see some members in certain planes, but if I switch class then I don't, switch back and there they are again....No I am not whining about head hunting or any such, just that when you get a good game and fly against some real people, it would be nice to have another chance in perhaps a different class than they are using...also, I have made many friends who began as my enemy.

 

Fly what ya brung and make the most of it!

 

I previously made a suggestion about a battle mode to build our own battles for clan wars. I am willing to guestimate that not many of the windbags in the forum would care for this mode. I would imagine that would deflate some of the napalming as it could be put up or shut up.

 

Then again I am just some stupid ReBeL who IIRC, always beats the Dark Side  ;)


rblsquad.com                                                                                              Key vehicle - Bristol Beaufighter

trikke #19 Posted 15 December 2018 - 11:03 PM

    First Lieutenant

  • Member
  • 2639 battles
  • 2,218
  • [R-A-W] R-A-W
  • Member since:
    01-26-2012
Plus 1 on the topic title, Cap!
Spittoon says #smarterpilotswinmore

CorvusCorvax #20 Posted 15 December 2018 - 11:13 PM

    First Lieutenant

  • Member
  • 2212 battles
  • 2,136
  • [JG52] JG52
  • Member since:
    01-26-2012

View PostAzis_, on 15 December 2018 - 10:39 PM, said:

 

Then again I am just some stupid ReBeL who IIRC, always beats the Dark Side  ;)

 

Hmmm, I never see you in those battles.  Do you fly under a different username?




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users