I have proven that it is appropriate to replace ± 1.0BR matchmaker with ± 0.7BR or smaller matchmaker in Air RB

Game Discussion Realistic Battle
,
1
  1. Please allow me to introduce the settings of my program first. I assumed some aircraft and assigned different BRs to them (the BRs I gave are the ideal BR decompression schemes I think, and only include some aircraft I am familiar with. If you don’t like it, you can ignore it and don’t need to comment on it. I just mentioned it in passing)
    aircraft_list = [
    (15.3, [“EF2000_DE”, “F2000A_IT”, “Typhoon FGR4_UK”, “Rafale_FR”, “F15E_US”, “F15I_IR”]),
    (15.0, [“F15C_US”, “F15JM_JP”, “J11B_CN”, “J10A_CN”]),
    (14.7, [“Su27SM_RU”, “JAS39C_SW”, “JAS39C_UK”, “JAS39C_IT”, “F16C_US”, “Mirage2000-5F_FR”]),
    (14.3, [“Mig29SMT_RU”, “Su34_RU”, “J11A_CN”, “JF17_CN”]),
    (14.0, [“F4FICE_DE”, “Kfir-C10_IR”, “F14B_US”, “F14A-IRAF_US”, “JA37DI_SW”, “Av8B+_US”]),
    (13.7, [“J8F_CN”, “TornadoF3L_UK”, “F15A_US”, “F15J_US”, “Su27_RU”, “Su33_RU”, “SeaHarrier FA2_UK”]),
    (13.3, [“Yak141_RU”, “J11_CN”, “F14A_US”, “Mig29G_DE”, “Mirage4000_FR”]),
    (13.0, [“F16ADF_US”, “Mig29_DE”,“Mig29_RU”, “F16AJ_JP”]),
    (12.7, [“F4EJKai_JP”, “Mig21Bison_UK”, “F20A_US”, “F16A_US”])
    ]
  2. I set up a generator. Its initial setting is: every 0.01 seconds, a random player is generated to enter the matching queue. Every time 32 people who can fight are matched, a battle will start. Print who participated in this battle, what is the BR range, and then output how long this match took.
    There are 2 Teams in a battle. Each Team must match 16 people to start a battle between the two Teams. The BR range between the two Teams must be the same. My initial setting is ±1.0 (max_min_BR = 1.0). I try to ensure that the distribution of aircraft owned by the two teams is as consistent as possible in BR.
  3. I did not complete the country allocation part because the current top tier is a mess in terms of country allocation.
    4.Experiment Results

========== ±1.0 ==========
Max match time: 3.75 s
Min match time: 0.62 s
Total Time Spent for 10 Battles: 21.00 s
Average Max Wait Time for All Battles: 0.93 s
Average Min Wait Time for All Battles: 0.00 s
Remaining aircraft in queue: 37

========== ±0.7 ==========
Max match time: 3.77 s
Min match time: 0.91 s
Total Time Spent for 10 Battles: 22.39 s
Average Max Wait Time for All Battles: 1.24 s
Average Min Wait Time for All Battles: 0.00 s
Remaining aircraft in queue: 39

========== ±0.4==========
Max match time: 3.85 s
Min match time: 1.31 s
Total Time Spent for 10 Battles: 25.52 s
Average Max Wait Time for All Battles: 1.62 s
Average Min Wait Time for All Battles: 0.00 s
Remaining aircraft in queue: 37

========== ±0==========
Max match time: 5.19 s
Min match time: 2.31 s
Total Time Spent for 10 Battles: 35.49 s
Average Max Wait Time for All Battles: 3.24 s
Average Min Wait Time for All Battles: 0.00 s
Remaining aircraft in queue: 135

Note:
1.The time spent may not be exactly the same each time, but it will not differ too much.
2.The max/min match time refers to the maximum time it takes for a group of games to be generated, starting from the first player entering the matching queue until the last player enters the matching queue, and then the game starts.
3.I also tested the player waiting time, namely Max/Min Wait Time. This indicator represents who has the longest/shortest waiting time among all the players in this lobby when it is established, and how long they waited. The Average Max/Min Wait Time is the average value of 10 battles.
4.I have discovered a small bug in my code, but it is not a big deal. I have fixed it and it does not affect the final conclusion.

From the experimental results, we can see that only when ±BR==0, the time spent increases significantly, and the time changes of ±0.7/0.6 ​​or ±0.4/0.3 on the matching time is negligible.

5.Possible Questions
Q1: ±0.7 requires a redesign of the BR system, and gaijin have to analyze all vehicles in the game, decide where they’ll go.
A1: Please imagine an extreme case where each aircraft is fighting against an aircraft with the same BR as itself. Is this fair and does it require a redesign of the BR system? Smaller number gets fairer game.

Q2: Over 90% of vehicles in War Thunder are balanced, 0.7 is more work than decompression.
A2: The time cost of 0.7 can be ignored, and as for what it will bring to the game, we cannot be completely sure, but at least it will give players a chance to try!

Q3: Where’s your code?
A3: machmaker - Google Documenten

30 Likes
2

That’s very cool, really professional research.

Get ready to have your whole post disregarded by players who FEEL that it will not work

11 Likes
3

(15.3, [“EF2000_DE”, “F2000A_IT”, “Typhoon FGR4_UK”, “Rafale_FR”, “F15E_US”, “F15I_IR”]),
(15.0, [“F15C_US”, “F15JM_JP”, “Su30SM_RU”, “J11B_CN”, “J10A_CN”]),

Hahaha. 10-12 AMRAAM carriers sitting .3 BR below? Thrust vectoring jet sitting .3 BR below top?
Cmon now

1 Like
4

I didn’t think there even needs to be concrete proof a ±0.7 matchmaker would improve the quality of matches, but good work.

3 Likes
6

If you don’t like my BR decompression scheme, you can ignore it . I just mentioned it in passing :) And you can pay more attention to my main contribution.

1 Like
7

I’m letting you know that 8-12 AMRAAM carriers, at this current point in the game, should 100% be at the highest BR rating.

8

It was proposed on this forum before, and a lot of people were doing mental gymnastics to prove it would ruin the game

1 Like
9

Haha, this reminds me of some guy XD

1 Like
10

Okay, I accept

1 Like
11

there are no planes with 12 amraams in wt

12

Thanks. Point still stands. 10 AMRAAM should 100% be the highest BR.

13

No, compressing the matchmaker both requires more effort as you need to do ALL the work of decompression only to compress the BRs and matchmaker.

Worse idea.

Decompression is both easier and objectively superior.

Not only that but ALL current 14.0s along with F-15C/JM/BazM and J-11B should be the same BR.

There is no notable difference between them that warrants BR differences.

14

I have thought much the same. Rather than endlessly trying to decompress which to me, sounds an aweful lot like this:

Spoiler

butter-scraped

Reducing the BR range by 0.3 would have basically the same affect without all the issues.

2 Likes
15

Did I miss the part in the starting post that 0-4 players are top BR in each team? And so on…

Also, is the term “max/min match time” defined clearly anywhere? What exactly is this, how is it computed?

1 Like
16

I even guessed that you would oppose me, because you opposed another post about matchers in the past.

  1. I am not against decompression, I even gave my own decompression scheme.
  2. I want to ask, if all vehicles only fight with vehicles with the same BR as themselves, do you think such a battle is fair and better than the current one?
  3. If you insist that changing the matcher requires more effort, can you explain it with actual examples? Or write a program like me?
  4. Even if changing the matcher does require more effort, that is gaijin’s business. gaijin has made a lot of money, so what if he does more?
1 Like
17

This is just a term I defined to show the maximum and minimum time it takes to complete a battle, in a total of ten battles.

If a test includes 3 battles, starting from a certain moment, they take 0.5 seconds, 4 seconds, and 2 seconds to match, then max match time==4s, min match time == 0.5s

18

A decompressed match maker is superior to a compressed one.

19

Allow me to complain here about the 12.7-13.0 planes, they are too strong against lower BR planes, too weak against higher BR planes (especially those with fox3), I have skipped them/given up playing them.

20

But what can be done about that though?

Yes, some decompression can occur, but with the amount of aircraft we have, you could at most increase the BR to about 14.3 or 14.7. Would help sure, but not fix the problem entirely. But that combined with a reduction of MM size to +/-0.7 on top of a 0.3-0.7 BR increase and minor decompress would do quite a bit.

But…

Aircraft like the Gripen A, F-16, F-15A, Su-27, etc should 100% be at the same BR as aircraft like the Sea Harrier FA2 or F-14A.

and more than a few non-fox-3 aircraft at 13.0 would probably move up as well, because they are too strong in a downtier. (fighting F-15A or Su-27s in something like a Phantom sucks)

21

0.7 MM is definitely one of the best changes gaijin could do. Decompression up to 16.0 would be better, but gaijin completely refuses to do that.

The one downside would be that some vehicles would need to be rebalanced, but the work needed to do that is much less than what decompression needs.

1 Like