Sign in to follow this  
Followers 0
Liberty75

Is something wrong with the new Match Maker ?

34 posts in this topic

I've been busy at work a lot lately, so haven't gotten too many battles in, but, in the mid-tiers at least, I'm winding up top tier enough.  I get the logic you're offering, and it makes sense.  Perhaps I've been lucky.  

Perhaps T8 winds up bottom more now, and likely the sheltered T8 premiums do, because of higher numbers playing T9 and T10.  

Share this post


Link to post
Share on other sites
12 hours ago, DHP said:

But you can only be top tier only so much with 3-5-7 format or else someone else is going to be bottom tier way more. 

 

 

So far it seems that tier 6 is bottom tier most of the time. I have continued on my quest and I am playing another 50 games in the VK3001D to verify the first 50. I am 28 games in and 13 of them have been bottom tier and 2 have been top tier. The disparity is still great. And this makes sense, as DHP says, with the breakdown 3-5-7. I'm guessing tiers 1, 2, 9 and 10 are the sweet spots when playing now. Playing in tiers 3-8 you are going to be on the bottom or middle a large majority of the time. I am not sure if the population at tier 8 and 10 throws this assessment off. There just might be more tier 8 and 10 dumps than other tiers.

Share this post


Link to post
Share on other sites

I've been getting a lot of top tiers in Type 62 (t7 light)

Share this post


Link to post
Share on other sites

Ive been getting a fair bit of top tier in 5, 8 and 9. Tier 6-7 not so much but occansional +1s. Single tiers I get too. 

To my suprise I only see 3 artys in around half the games with a few without it at all. Small sample thou around 40 games. 

Share this post


Link to post
Share on other sites

I see shitton of tier 7s with pref tier 8 (kv-5). As for other tanks, i typically see +/- 0 or +/- 1 matches; +/-2 are rare.

Share this post


Link to post
Share on other sites

Final Analysis

I played another 50 games in a row with the VK3001D over a few days and the trend seemed less pronounced, although similar, but with a greater share of the games going to middle tier matches. I played in the afternoon and morning more, so that may have a little to do with it. Playing at a different time from the first 50 games should also give a better all around view of tiering frequencies when the results are combined (see below).

Here are the results of the second set of 50 games (win rate is in there for reference again): *

Top = 3 (2-1 66%)
Partial Top = 0 (NA)
Middle = 17 (10-6-1 59%)
Only Tier Six = 10 (7-3 70%)
Partial Bottom = 0 (NA)
Bottom = 20 (10-10 50%)

My win rate was the exact opposite and now I am overall 50% with this tank. As I stated in an earlier post, I don't mind top tier battles and they are EASIER to influence under this new system, but it is not ideal. The higher your position in tier, the more likely you are to influence a match, as these 100 games seem to support, along with the logic of having better firepower, armor, and hit points when you are high tier.

Here are the results of all 100 battles combined together:

Top = 8 (6-2 75%)
Partial Top = 0 (NA)
Middle = 22 (13-8-1 59%)
Only Tier Six = 23 (12-11 52%)
Partial Bottom = 1 (1-0 100%)
Bottom = 46 (18-28 39%)

 

From 100 games, here is the frequency per tier level for a tier 6 tank:

Top = 8%
Partial Top = 0 (NA)
Middle = 22%
Only Tier Six = 23%
Partial Bottom = 1%
Bottom = 46%

 

100 games is not enough to get a perfect picture. The data, while insufficient, is enough to show a general trend. Players at tier 6 will more likely be bottom tier, which will place more emphasis on the random nature of your "teammates" than on your skill to determine your win rate. In other words, the control you have over your win rate destiny will decrease (even if only slightly) when playing at this tier under 9.18. I don't have any data to compare the old Match Maker (MM) frequency to this new version. Anecdotally, I would testify that I was a top tier tank more than once in every ten games at tier 6. If fact, I recall understanding that I had a decent chance at being top tier every time I hit the battle button (I am guessing that it was about 25-35% of the time). If we combine top tier with only tier six matches we will get 31% and this would give us a similar result to the old MM. The downside of all tier six matches is that you are again at a decrease in win rate destiny (depending on the caliber of your teammates) compared to being top tier with the old MM. This decrease is most likely slight, yet still a factor.

If this helps anyone, I'm glad. I saw a trend when I first started playing the VK3001D and it made me curious: was I getting bad luck or is there a new trend in the MM? As I discovered, there is a new trend with the MM. This makes sense with the new breakdown of teams into a set 3-5-7 format and the amount of tier 8 tanks being played on the servers.

I'm still neutral about this new MM until I see how it works over a longer period of time. I am concerned that there will be "sweet-spots" at certain tiers that will be exploited to the detriment of other tiers and this may create an unhealthy environment overall in the game as people try to grind though those disadvantaged tiers and give up on the game and hurt the population. I would conjecture that tiers 5 and 6 seem the most likely to be affected. Then again, this concern may be totally unfounded. In any case, good luck out there and thank you for your patience if you reached this far in the post!

-------------------------------------------------

* Here are the first 50 games again for quick reference:

Top = 5 (4-1 80%)
Partial Top = 0 (NA)
Middle = 5 (3-2 60%)
Only Tier Six = 13 (5-8 38%)
Partial Bottom = 1 (1-0 100%)
Bottom = 26 (8-18 31%)

Share this post


Link to post
Share on other sites

Really good data collection and analysis @Liberty75. I've not collected any detailed data myself but post patch I've seen a rise in win rate. I attribute this to what @GehakteMolen noted above, the absence of matches where I'm simply fodder for 6-14 +1/2 tier tanks. I see 3 8s in a tier 6 and I know I can still work over the 7s and 6s and those 8s simply can't be EVERYWHERE like they used to be.

Share this post


Link to post
Share on other sites
On 27.4.2017 at 9:31 PM, CompeAnansi said:

They only additional change I want is to try and also balance heavies at least within a range of +/- 2. Having 7 mediums on one team and 7 heavies on the other is stupid. There were clearly enough tanks for this match to have a relatively even split. I don't think it would make matches take any longer to fire, it would just result in a better balance of tanks between teams.

that is quite intentional. WG said not balancing the number of mediums and heavies is good because it adds variety, lol

 

Quote

Q: Why doesn’t Matchmaker balance all the classes for each team?

A: The matchmaker currently tries to balance 3 classes: LT, TD and SPG in both teams, and allow the matchmaker to avoid balancing HT and MT specifically.  This is done to avoid making matchmaking timing slow down, and also to allow of more variety in battles.

https://thearmoredpatrol.com/2017/04/22/wot-0-9-18-developer-qa-42017/

Share this post


Link to post
Share on other sites
3 hours ago, CraBeatOff said:

Really good data collection and analysis @Liberty75. I've not collected any detailed data myself but post patch I've seen a rise in win rate. I attribute this to what @GehakteMolen noted above, the absence of matches where I'm simply fodder for 6-14 +1/2 tier tanks. I see 3 8s in a tier 6 and I know I can still work over the 7s and 6s and those 8s simply can't be EVERYWHERE like they used to be.

Thanks. And I agree. I would rather face a maximum of three high tier tanks than 6 or more. In that last run, I had a nice carry on Prokhorovka in a tier 8 match which probably wouldn't have happened if I was swamped with tier 8s. Thanks again for the feedback.

Share this post


Link to post
Share on other sites
Sign in to follow this  
Followers 0

  • Recently Browsing   0 members

    No registered users viewing this page.