Jump to content


Thank you WG for resisting MM by PR, win rate, or WIN-whatever


  • This topic is locked This topic is locked
55 replies to this topic

awildseaking #41 Posted Jan 03 2018 - 01:02

    First lieutenant

  • -Players-
  • 9702 battles
  • 566
  • [NEET] NEET
  • Member since:
    08-05-2015

Removing 357 is a good idea. Top tiers are too important now. MM should address the reality of the playerbase: the bulk of games are T8-T10 and there are very few new players. Templates force +/-2 games more often than +/-1 or 0, which has made T1-T7 bottom tier the overwhelming majority of the time. And I don't mean like 50.01% of the time, I mean like 90% of the time. The other problem, template or not, is that MM tries to force 15v15 games. This is why T1-T7 was bottom tier the majority of the time long before 357. 357 only worsened the problem, but removing 357 won't fix it.

 

MM needs to allow for smaller games. For instance, if you were a new player and tried T1 today, you would play nearly 100% of your games +1. You have to wait 3-5 minutes for a match and the result is terrible every time. This is just bad matchmaking. MM needs to give preference to +/-0 over 15v15, ensuring that T1 never sees T2. Most importantly, you won't wait as long. Game populations decrease with time. If there aren't enough players to support 15v15 at T1, stop trying to force 15v15 and make smaller matches. The same logic should apply to every tier.

 

SBMM, 357, and random are all false choices. Trolls want you to believe there is no third option. There are many factors that WG could consider in MM but does not. For instance, tanks could be weighted by their equipment, modules and crew. Nobody should consider a stock top tier with unskilled crews and no equipment equal to a decked out top tier with 4 skills and rare equipment.



Striker_70 #42 Posted Jan 03 2018 - 01:02

    Major

  • Players
  • 19409 battles
  • 6,241
  • Member since:
    04-02-2013

View PostFeargrim27, on Jan 01 2018 - 18:17, said:

Hey bud, listen the discussion of skill/stat based MM has been talked about for years now.

 

The problem with WoT and a skilled based MM is what would happen to the skilled player base? They would leave the game, why? Insane long wait times for a match.

 

There would be no difference in wait times.  All they'd have to do is rearrange the teams of 30 they had already selected in a more balanced manner.



NeatoMan #43 Posted Jan 03 2018 - 01:04

    Major

  • Players
  • 25385 battles
  • 16,716
  • Member since:
    06-28-2011

View PostStiffWind, on Jan 02 2018 - 18:40, said:

The rigging is triggered by an upward or downward trend in stats, just as it's stated in the patent WoT has on the methodology.  If you don't have a change in stats that triggers the rigging, you're not going to see it.  As such, any "data" must be accompanied by a historical trend of WN8 and Win Rate in order to be considered valid.  I have yet to see that posted by anybody who says the game isn't rigged.

Then it must be an increase over the short term in order to trigger rigging, because the more games you include in the "trigger" scope, the less likely it is to activate the trigger because win rates/stats generally smooth out over the long term.   Plus, if it is as pervasive and obvious as the riggers claim, it would have to be a short term phenomenon in order to appear and disappear from session to session.

 

So which is it?  is it happening on a daily/session basis, or is it long term, and therefore not happening as often as people on the forums claim?  Both can be tested.  Which do you suggest I try first, and how many games to include in the "trigger" scope?  10 games?  20 games?  50 games? 100 games? 500 games?  1000 games?

 

Or do you throw out nebulous ideas that you don't actually have to back up with anything concrete?



spud_tuber #44 Posted Jan 03 2018 - 01:10

    Major

  • Players
  • 47221 battles
  • 5,363
  • Member since:
    08-26-2013

View PostNeatoMan, on Jan 02 2018 - 18:04, said:

*snip*

Or do you throw out nebulous ideas that you don't actually have to back up with anything concrete?

This is my bet.  Notice how he ignored my point that he could test for himself, and easily make the data available to others, by running vbaddict's adu.



diego999 #45 Posted Jan 03 2018 - 01:21

    Major

  • Beta Testers
  • 31195 battles
  • 4,695
  • [-4K-] -4K-
  • Member since:
    11-22-2010

View PostStriker_70, on Jan 02 2018 - 21:02, said:

 

There would be no difference in wait times.  All they'd have to do is rearrange the teams of 30 they had already selected in a more balanced manner.

 

Take this unicum in a Tiger 2 and give me the tomato in a Maus. Oh wait, now I have 1 more tier X and you're complaining again.

Kelly_Sereda #46 Posted Jan 03 2018 - 01:26

    Staff sergeant

  • -Players-
  • 10666 battles
  • 397
  • Member since:
    12-28-2015

View PostFeargrim27, on Jan 02 2018 - 01:34, said:

 

I think what is meant is do nothing, get nothing. Spotting damage would count, etc.

 

I also want to talk briefly on the topic of "get punished" for a bad game. I cannot recall a game in which I did no damage and no spotting. It is incredibly easy to do SOMETHING. Again, WG has to add some educating point here, if someone got a 0 exp battle, I think a pop-up should appear (for new players) and advise them of why.

 

Solid idea.  Being an effective player in WoT is not obvious at first to most players and sometimes never to some players.  Being bottom tier in an underpowered tank still leaves a number of ways to creatively make yourself useful to the team and score XP.  Educating players on how to contribute and score points with lessons along the way would help get players more quickly to a useful state.  I wouldn't stop at new player though.  Everyone is new to a certain degree the first time they hit a new tier, with new vehicles/maps/tactics.  Could have optional advanced tutorials that are suggested if a player underperforms consistently in certain areas to focus on unique skill sets.

StiffWind #47 Posted Jan 03 2018 - 03:43

    Major

  • -Players-
  • 13456 battles
  • 2,054
  • Member since:
    03-15-2017

View Postspud_tuber, on Jan 03 2018 - 01:10, said:

This is my bet.  Notice how he ignored my point that he could test for himself, and easily make the data available to others, by running vbaddict's adu.

 

Since when does logging off and going to eat dinner constitute "ignoring your post"?  You guys are pretty quick to judge, which often results in misjudgments aplenty.

 

A portion of the evidence being referred to can actually be seen by all in the charts provided by Wotlabs.  I can't answer to exactly the time-table they use in WoT that triggers rigging, but I agree that it would have to be relatively short.  And although I would like to see the WN8 as well, as a possible indicator of other things, the Win Rate is specifically mentioned in the patent as the triggering stat.  Unfortunately, the graphs don't go back as far as I'd like them to, and I haven't figured out if I can change that or not....but it should be clear to anyone who can read a chart as to when the "rigging" hit me this last time around.

http://wotlabs.net/na/player/StiffWind

Hint:  When the rigging hits, you quite often see an increase in your WN8 at the same time, as there are now a lot more targets for you to hit than there were before.  Poor players and/or bots are not very good at doing damage and taking out enemy tanks.  Others have noted this as well, so it makes a good secondary indicator.

My advice to you, with the way you "bet", is don't go to Vegas.


 

 



NeatoMan #48 Posted Jan 03 2018 - 04:01

    Major

  • Players
  • 25385 battles
  • 16,716
  • Member since:
    06-28-2011

View PostStiffWind, on Jan 02 2018 - 21:43, said:

....but it should be clear to anyone who can read a chart as to when the "rigging" hit me this last time around.

http://wotlabs.net/na/player/StiffWind

from those charts it looks like your win rate started improving at ~10600 games and lasted over a span of ~180 games before the rigging hit.  That means you were running ~57 to 58% win rate during that time you improved when the rigging hit.   So now we got numbers to go by.

 

Good.  I'll check my data, shadora's data, and budha's data to see if any similar rate of improvement over 50 to 200 games triggered worse teams than usual over the following 50 or 100 games.



spud_tuber #49 Posted Jan 03 2018 - 04:13

    Major

  • Players
  • 47221 battles
  • 5,363
  • Member since:
    08-26-2013

View PostStiffWind, on Jan 02 2018 - 20:43, said:

 

Since when does logging off and going to eat dinner constitute "ignoring your post"?  You guys are pretty quick to judge, which often results in misjudgments aplenty.

 

 

*snip*

 

My advice to you, with the way you "bet", is don't go to Vegas.


 

 

Ok, let's clarify a few things here.  First off, my point, not my post.  Secondly, my point was in a post you quoted, so you had to have seen it. 

 

As for the rest of your post above, I will respond to it separately later when I have more time.



Striker_70 #50 Posted Jan 04 2018 - 03:53

    Major

  • Players
  • 19409 battles
  • 6,241
  • Member since:
    04-02-2013

View Postdiego999, on Jan 02 2018 - 18:21, said:

 

Take this unicum in a Tiger 2 and give me the tomato in a Maus. Oh wait, now I have 1 more tier X and you're complaining again.

 

Why do something dumb like that?  Say 1 team has 3 blues as their 3 top tier tanks and the other has 3 reds.  Just trade a red for a blue and the match will be a bit more balanced in general.



Trauglodyte #51 Posted Jan 04 2018 - 04:20

    Captain

  • -Players-
  • 12926 battles
  • 1,870
  • Member since:
    06-04-2016

View PostRevoc_cn, on Jan 02 2018 - 00:47, said:

Had a wonderful week in WoT with red tomatoes pushing or shooting teammates or doing anything but helping the team to win. And I think this is the worst time I have experienced since I started to play WoT in 2012.

WG just can not face the facts that all the fancy MM system they imaged is a SH*T. In a 357 MM if the three top-tier are red tomatoes then the game is over, and magically most of the time you have all 3 [edited]or all 3 purple re-rolls and the enemies have the opposite. So in the new MM, the bottom tier facing the situation that if your top tiers are [edited], there is no space for you to do anything while in the old MM time, bottom tiers can at least do something because having 10 red tomatoes top tiers is impossible at that time.(it is possible now). And in the new 357 MM, bottom tiers get less exp because they have no chance to earn them from higher tiers and top tiers are getting less cause there are fewer hit points available. red tomatoes are crying about re-rolls(I can't get why people always crying about re-roll, a re-roller get a higher penetration or RoF or what? If a re-roller is a true noob and you are not why you care about whether he rolls)

And after all the years, WG cannot face the fact that win rate is the only parameter to judge a player and at least they should give players chance to compete with players of the same level. The rank battle is the hugest joke in the history of rank systems that in the second season I have seen the second rank is full of 47% win rate players. red tomatoes are angry about everyone because they are always crushed by the enemies and they can never face the fact that win rate lower than 50% can only be the result of their lack of skills, experience, and ability to learn. "re-rollers" are seeking for data rather than a win when finding out the top-tiers are overwhelmed by the opponents. What makes MM by win rate so hateful to WG?

I love this game but I'm sure WG is trying to cheap out in making the game better and they are trying the get the last coin from it. I do not know how long this game can last 

 

Happy new years guys. it is just a game but I hope it can better than now

 

The irony is that I have had a much worse plethora of tools tonight than I had all of last week.  Perhaps [edited]ters are just [edited]ters?

HiBan #52 Posted Jan 05 2018 - 00:24

    Staff sergeant

  • -Players-
  • 18996 battles
  • 486
  • Member since:
    01-08-2015

Changing the 3-5-7 template wont really change anything, since the frequency of being top/mid/lowest tier that we currently get doesn't correlate with that template anyway.

 

I've been tracking my games for the last week, and i got the following:

 

61% lowest tier

30% mid tier

9% top tier

 

while according to the 3-5-7 template it should be:

 

46% lowest tier

34% mid tier

20% top tier


Edited by HiBan, Jan 05 2018 - 00:25.


spud_tuber #53 Posted Jan 05 2018 - 03:09

    Major

  • Players
  • 47221 battles
  • 5,363
  • Member since:
    08-26-2013

View PostHiBan, on Jan 04 2018 - 17:24, said:

Changing the 3-5-7 template wont really change anything, since the frequency of being top/mid/lowest tier that we currently get doesn't correlate with that template anyway.

 

I've been tracking my games for the last week, and i got the following:

 

61% lowest tier

30% mid tier

9% top tier

 

while according to the 3-5-7 template it should be:

 

46% lowest tier

34% mid tier

20% top tier

What about 5/10/0 and 15/0/0 matches?  How do they figure into either number above?



HiBan #54 Posted Jan 05 2018 - 04:02

    Staff sergeant

  • -Players-
  • 18996 battles
  • 486
  • Member since:
    01-08-2015

View Postspud_tuber, on Jan 05 2018 - 03:09, said:

What about 5/10/0 and 15/0/0 matches?  How do they figure into either number above?

 

As for the frequency we get 5-10-0 and 15-0-0 games, we just don't know what influences them, so i just applied the main 3-5-7 template, and simplified everything into "top tier", "mid tier" and "lowest tier" based on the following criteria:

 

Lets say you play a tier 6:

in  tiers 4-5-6 and 5-6 games you are "Top Tier", because you only play against lower tiers and you dont meet higher tiers.

in tiers 6 and 5-6-7 games you are "Mid tier", because you're neither Top tier nor Lowest Tier.

in tiers 6-7 and 6-7-8 games you're "Lowest Tier", because you play against higher tiers and you don't meet lower tiers.

 

But if i just take the 3-5-7 games into account and discard the 5-10-0 and 15-0-0 games, then i get the following numbers:

68% lowest tier

28% mid tier

4% top tier

 

while according to the 3-5-7 template it should be:

46% lowest tier

34% mid tier

20% top tier


Edited by HiBan, Jan 05 2018 - 04:09.


spud_tuber #55 Posted Jan 05 2018 - 04:29

    Major

  • Players
  • 47221 battles
  • 5,363
  • Member since:
    08-26-2013
Thank you for breaking it out. Interesting, but not entirely unexpected, results.

I really need to find out if wotnumbers is working atm so I can record some data the easy way, and break this stuff down by tier played as well as full percentages for every match format.

Dirty_Camel #56 Posted Jan 05 2018 - 07:02

    Captain

  • Players
  • 50873 battles
  • 1,905
  • [F-3] F-3
  • Member since:
    11-29-2013

View Postdiego999, on Jan 03 2018 - 01:21, said:

 

Take this unicum in a Tiger 2 and give me the tomato in a Maus. Oh wait, now I have 1 more tier X and you're complaining again.

I'll gladly take the unicum in a tiger, as the tomato in a Maus has yet to figure out how to pen tanks ......so he is useless. Fact, proven in most T10 matches.






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users