Jump to content
You need to play a total of 5 battles to post in this section.
Lert

Lert's tier 5 experiment

260 comments in this topic

Recommended Posts

646
[HINON]
Members
2,680 posts
19,792 battles

Lately I find the MM has been pretty favorable around tier 6, a lot better than perhaps a month or 2 ago. Looking forward to the experimental results.

Share this post


Link to post
Share on other sites
3,294
Members
20,825 posts
5,614 battles

 

That's fine. My own experiment is likely to take a few weeks too.

 

Makes sense. You probably have other things to accomplish with your playtime, whereas I've reached the "play just to play" stage. :)

Share this post


Link to post
Share on other sites
24,406
[HINON]
Supertester
20,092 posts
13,558 battles

since Lert is in the Netherlands, I'm assuming his matches would be mid-late afternoon here. So that should cover non-prime time, other than morning.

 

Late morning / mid afternoon for the most part.

 

You probably have other things to accomplish with your playtime, whereas I've reached the "play just to play" stage.

 

Between this experiment, my own play / grind, divisioning, supertest duties etc I play more than enough floaty boats as is.

Share this post


Link to post
Share on other sites
1,182
[SCTFB]
[SCTFB]
Beta Testers
4,021 posts
16,002 battles

I like this idea. But I disagree on one point I think 100 is a great sample size and please note weather or not you are having more or less fun in the matches 

Share this post


Link to post
Share on other sites
3,294
Members
20,825 posts
5,614 battles

 

Between this experiment, my own play / grind, divisioning, supertest duties etc I play more than enough floaty boats as is.

 

Heh, with my limited playtime, Tier 7-8 is about the highest I can actually shoot for without it becoming a frustratingly long grind. (was the same for tanks)

 

I only WISH I could play more boatz. :) (although, limited play means I'll probably never get bored or need a break lol)

Share this post


Link to post
Share on other sites
3,294
Members
20,825 posts
5,614 battles

I like this idea. But I disagree on one point I think 100 is a great sample size and please note weather or not you are having more or less fun in the matches 

 

Plus, with more of us doing it, sample size will be even larger.

 

As far as fun, I know that tiering makes little difference for me in KBerg, Konig, Kongo, and Nicholas. Kirov will be more tiering dependent, and Emerald will be less than fun no matter what lol.

Share this post


Link to post
Share on other sites
325
[WSU]
Members
1,494 posts
10,207 battles

 

I'm not TELLING Lert to do a thing. I was only asking if when you see a couple T7 in battle if that truly makes it a tier 7 battle, which under this data collection scheme will obviously result in most battles being T7. I could do that at tier 8 and get the same result currently considering most times there is at least a single 10 in there.

 

I'm going to predict this one right now without all the battles and headaches. Tier 5 sees tier 7 most of the time under the data collection terms laid out for this experiment. 

 

The issue is not about the average tier that T5 ships play against.  The isssue is how often are they top tier (against T4) bottom tier against T6, and bottom tier against T6 & T7.  The difference, which I am sure you are aware of but for some reason are ignoring, is that in T7-T10 you always get the opportunity to be top tier and then fight against ships 2 tiers lower...but that cannot happen in T5 & T6.

Share this post


Link to post
Share on other sites
7,753
[NMKJT]
Beta Testers
21,984 posts
3,895 battles

I think I'll get in on this too. I have a good selection of tier 5s to pick from.

 

So I'll throw together a quick Excel doc and note what ship winds up in what tier spread, and share the results after 100 matches.

 

Tier 5 used to be my favorite place, so anything that can help my beloved Kongo be less of a drag is something I can get behind.

Edited by Destroyer_Kiyoshimo

Share this post


Link to post
Share on other sites
20
[OGRES]
Members
96 posts
3,528 battles

Here Lert. Maybe this info will help to add numbers to your experiment.

This list dates from November 04, 2016 until January 07, 2017

 

  • Average: average tier of enemies
  • Count: number of games in that tier
  • Top: fraction of games where you're top tier
  • Top-1: fraction of games where you're 1 tier lower than top
  • Top-2: fraction of games where you're 2 tiers lower than top

 

 

NiS8FLw.jpg

Share this post


Link to post
Share on other sites
521
[WOLFD]
[WOLFD]
Beta Testers
5,057 posts
1,501 battles

Put my replays through the tool linked up thread, got this:

 

Iv0Pv79.png

Share this post


Link to post
Share on other sites
6,802
[SYN]
Alpha Tester, Beta Testers
30,523 posts
4,836 battles

 

Late morning / mid afternoon for the most part.

 

 

Between this experiment, my own play / grind, divisioning, supertest duties etc I play more than enough floaty boats as is.

 

Hey you black cat, I'm doing a tier 6 version. I'll keep the results on the first page so we can compare notes.

Share this post


Link to post
Share on other sites
7,753
[NMKJT]
Beta Testers
21,984 posts
3,895 battles

Here Lert. Maybe this info will help to add numbers to your experiment.

This list dates from November 04, 2016 until January 07, 2017

 

  • Average: average tier of enemies
  • Count: number of games in that tier
  • Top: fraction of games where you're top tier
  • Top-1: fraction of games where you're 1 tier lower than top
  • Top-2: fraction of games where you're 2 tiers lower than top

 

 

NiS8FLw.jpg

 

Wow. Once I had a moment to really examine this, the numbers are quite telling.

 

It goes from 45/55 (close enough, IMO, to half-and-half) in 0.5.13 to 29% top tier, 57% bottom tier in 0.5.14. IIRC that's the patch that added tier 4 protected MM, too.

Edited by Destroyer_Kiyoshimo

Share this post


Link to post
Share on other sites
24,406
[HINON]
Supertester
20,092 posts
13,558 battles

First batch:

 

1) 6.2k pop, tier 7 battle, victory

2) 6.5k pop, tier 5 battle, victory

3) 6.8k pop, tier 7 battle, defeat

4) 7.3k pop, tier 5 battle, defeat

5) 7.6k pop, tier 7 battle, victory

6) 8k pop, tier 7 battle, victory

7) 9.2k pop, tier 5 battle, defeat

8) 9.6k pop, tier 7 battle, defeat

9) 10.6k pop, tier 6 battle, victory

10) 11.3k pop, tier 7 battle, victory

 

Tier 5 battle: 3 / 10

Tier 6 battle: 1 / 10

Tier 7 battle: 6 / 10

Share this post


Link to post
Share on other sites
26,178
[WG-CC]
WoWS Community Contributors
9,684 posts
8,169 battles

 

Wow. Once I had a moment to really examine this, the numbers are quite telling.

 

It goes from 45/55 (close enough, IMO, to half-and-half) in 0.5.13 to 29% top tier, 57% bottom tier in 0.5.14. IIRC that's the patch that added tier 4 protected MM, too.

 

Keep in mind those are very small sample sizes.  For each of his examples, one game equals a shift of...

  • 0.5.13 - 9.1%
  • 0.5.14 - 4.8%
  • 0.5.15 - 3.2%
  • 0.5.16 - 7.7%

 

Add them all together and you get 76 games which breaks down to the following:

 

  • Tier 5:  29 games  (38%)
  • Tier 6:  9 games (12%)
  • Tier 7:  38 games (50%)

 

The Matchmaking changes came into place in 0.5.9 on July 26th, 2016.

Share this post


Link to post
Share on other sites
5
[O7]
[O7]
Members
34 posts
6,266 battles

 

Yup. That's why I'm collecting the data that I'm interested in.

 

Lert, I would suggest you use 2 data sets - since you're using solely Premium ships, you should perform another sample with non-premium tier 5s in case that is a confounding factor.  There is no way to tell whether it would be or not, but best to eliminate it with a sufficient sample size at the outset.

Share this post


Link to post
Share on other sites
24,406
[HINON]
Supertester
20,092 posts
13,558 battles

 you should perform another sample with non-premium tier 5s

 

1) Other people are doing that

2) I'm not going to rebuy tier 5 tech tree ships and retrain captains just for this experiment

3) You're entirely welcome to do the same 100 match run with tier 5 tech tree ships yourself

  • Cool 2

Share this post


Link to post
Share on other sites
Supertester
3,005 posts
10,086 battles

 

Lert, I would suggest you use 2 data sets - since you're using solely Premium ships, you should perform another sample with non-premium tier 5s in case that is a confounding factor.  There is no way to tell whether it would be or not, but best to eliminate it with a sufficient sample size at the outset.

 

Im running solely non premium ships for this, for that exact reason.

Share this post


Link to post
Share on other sites
7,753
[NMKJT]
Beta Testers
21,984 posts
3,895 battles

 

Keep in mind those are very small sample sizes.  For each of his examples, one game equals a shift of...

  • 0.5.13 - 9.1%
  • 0.5.14 - 4.8%
  • 0.5.15 - 3.2%
  • 0.5.16 - 7.7%

 

Add them all together and you get 76 games which breaks down to the following:

 

  • Tier 5:  29 games  (38%)
  • Tier 6:  9 games (12%)
  • Tier 7:  38 games (50%)

 

The Matchmaking changes came into place in 0.5.9 on July 26th, 2016.

 

Yes I'm aware of the sample sizes. It does establish a trend though.

 

And wow it's been longer than I thought.

Share this post


Link to post
Share on other sites
Members
605 posts
7,291 battles

Bought a Kamikaze and would be willing to track this as well.  Will do this only in solo games and see what happens.  My guess is uptiered 70%, top tier 30%.  Will track tier and server population and update data as I go along.

 

1) T7, 8.7k, W

2) T7, 5.1k, L

3) T6, 5.7k, L

4) T6, 5.8k, W

5) T5, 6.1k, L

6) T6, 6.8k, W

7) T7, 6.9k, W

8) T7, 7.3k, W

9) T7, 8.5k, W

10) T5, 10k, W

11) T5, 10.3k, L

12) T7, 8.4k, W

13) T5, 10.8k, W

14) T5, 11.4k, L

15) T7, 11.5k, W

16) T5, 3.7k, W

17) T5, 3.7k, W

18) T7, 4k, W

19) T7, 4k, L

20) T6, 4.1k, W

21) T5, 4.2k, W

22) T5, 5.6k, L

23) T7, 7k, W

24) T6, 6.3k, W

25) T7, 6.7k, L

26) T7, 7.2k, W

27) T5, 7.5k, L

28) T5, 7.6k, L

29) T5, 8.1k, W

30) T6, 8.4k, L

31) T7, 17k, W

32) T7, 17.2k, W

33) T6, 3.6k, L

34) T6, 3.7k, L

35) T6, 3.9k, L

Total

T5 - 12/100

T6 - 9/100

T7 - 14/100

 

 

Edited by Aeries1

Share this post


Link to post
Share on other sites
26,178
[WG-CC]
WoWS Community Contributors
9,684 posts
8,169 battles

Here's the running tally for everything submitted up til now.  Contributions from Lert, LittleWhiteMouse, Carl and Feragus

 

  • Tier 5 Game:  73  (38.6%)
  • Tier 6 Games:  26 (13.8%)
  • Tier 7 Games:   90 (47.6%)
  • TOTAL:  189 games.
  • Cool 1

Share this post


Link to post
Share on other sites
Members
1,351 posts
4,319 battles

My data from earlier this year:

 

Played from August 30 to September 5 at various times and I kept count of the match making for the German BBs I was playing:

 

Konig, tier 5 BB, 22 games

 

Top tier (game of 5s and 4s) - 7 games / 32% - these games only happened at EST prime time

Mid tier (impossible on tier 5 and tier 4 cannot see tier 6) - 0 games

Bottom tier (game of 5s, 6s and 7s) - 15 games / 68%

 

Bayern, tier 6 BB, 36 games

 

Top tier (game of 6s and 5s) - 13 games / 36% 

Mid tier (game of 5s, 6s and 7s) - 11 games / 31%

Bottom tier (game of 6s, 7s and 8s) - 12 games / 33%

 

Scharnhorst, tier 7 BB, 4 games

 

Top tier (game of 5s, 6s and 7s) - 2 games / 50%

Mid tier (game of 6s, 7s and 8s) - 1 games / 25%

Bottom tier (game of 7s, 8s and 9s) - 1 games / 25%

 

Played 5 games in a Kongoon Oct 3

  1. T6 game, bottom tier, lost, 74k damage
  2. T7 game, bottom tier, lost, 76k damage
  3. T7 game, bottom tier, lost, 87k damage
  4. T7 game, bottom tier, lost, 31k damage
  5. T4 game, top tier, win, 102k damage

 

More Kongo games

  1. T5 game, top tier
  2. T7 game, bottom tier
  3. T5 game, top tier
  4. T5 game, top tier
  5. T6 game, bottom tier

 

Total for T5 games:

  • Bottom 21
  • Top 11

Share this post


Link to post
Share on other sites
Members
1,351 posts
4,319 battles

Data for multiple tiers if anyone cares over the December NA holidays
 

 

 

Fri Tier Time Delta  
Mogami 10 18:52 0 CA/CL
Mogami 8 19:11 1 CA/CL
Chapayev 9 19:27 2 CA/CL
Benson 10 19:40 2 DD
Chapayev 10 19:57 2 CA/CL
Sat        
Mogami 9 17:20 1 CA/CL
Mogami 8 18:01 0 CA/CL
Chapayev 8 18:19 0 CA/CL
Kutuzov 9 18:30 1 CA/CL
Chapayev 8 19:16 0 CA/CL
Chapayev 9 19:39 1 CA/CL
Chapayev 10 19:58 2 CA/CL
Chapayev 10 20:13 2 CA/CL
Tripitz 8 21:03 0 BB
Tripitz 10 21:17 2 BB
Tripitz 10 21:34 2 BB
Tripitz 9 21:46 1 BB
Kongo 7 22:16 2 BB
Sun        
Tripitz 10 7:27 2 BB
Tripitz 8 7:51 0 BB
Tripitz 10 8:11 2 BB
Bismarck 10 8:36 2 BB
Izumo 9 9:00 1 BB
Izumo 10 18:17 1 BB
Izumo 10 18:29 1 BB
Izumo 10 18:48 1 BB
Mon        
Izumo 10 17:06 1 BB
Izumo 9 15:26 0 BB
Izumo 9 17:40 0 BB
Izumo 10 17:59 1 BB
Bismarck 8 18:17 0 BB
Bismarck 8 16:33 0 BB
Bismarck 10 18:50 2 BB
Bismarck 9 19:11 1 BB
Bismarck 8 19:27 0 BB
Bismarck 9 19:52 1 BB
Bismarck 10 20:08 2 BB
Tues        
Izumo 10 15:50 1 BB
Izumo 9 16:05 0 BB
Kutuzov 8 16:24 0 CA/CL
Chapayev 10 17:38 2 CA/CL
Chapayev 9 17:58 1 CA/CL
Chapayev 8   0 CA/CL
Izumo 9 18:18 0 BB
Izumo 10 18:34 1 BB
Izumo 9 18:51 0 BB
Bismarck 10 19:09 2 BB
Bismarck 8 19:18 0 BB

 

 

Edited by BossmanSlim

Share this post


Link to post
Share on other sites
Members
1,351 posts
4,319 battles

Data continued

 

 

 

 

  Tier Time Delta

 

Wed        
Izumo 10 12:13 1 BB
Bismarck 9 12:33 1 BB
Izumo 10 12:50 1 BB
Scharnhorst 8 13:11 1 BB
Izumo 10 13:24 1 BB
Izumo 9 14:15 0 BB
Izumo 9 14:28 0 BB
Izumo 10 14:44 1 BB
Izumo 9 15:02 0 BB
Bismarck 9 16:08 1 BB
Ognevoi 8 16:35 2 DD
Ognevoi 8 16:49 2 DD
Molotov 7 17:05 1 CA/CL
Molotov 7 17:19 1 CA/CL
Konigsberg 6 17:25 1 CA/CL
Konigsberg 5 17:42 0 CA/CL
Bismarck 9 15:52 1 BB
Danae 5 18:47 1 CA/CL
Danae 4 19:02 0 CA/CL
V-25 2 19:15 0 DD
Minekaze 5 19:30 0 DD
Minekaze 7 19:45 2 DD
Akatsuki 8 19:52 1 DD
Thur        
Izumo 9 9:12 0 BB
Bismarck 8 9:26 0 BB
Bismarck 9 9:38 1 BB
Izumo 9 9:56 0 BB
Bismarck 8 10:10 0 BB
Izumo 9 10:29 0 BB
Izumo 10 11:10 1 BB
Bismarck 8 11:19 0 BB
Izumo 10 11:39 1 BB
Bismarck 8 14:02 0 BB
Bismarck 9 14:22 1 BB
Bismarck 8 14:39 0 BB
Bismarck 8 19:13 0 BB
Bismarck 9 19:25 1 BB
Bismarck 8 19:42 0 BB
Izumo 9 20:09 0 BB
Fri        
Izumo 9 8:32 0 BB
Izumo 9 8:45 0 BB
Izumo 9 9:16 0 BB
Izumo 9 9:37 0 BB
Bismarck 8 10:04 0 BB
Izumo 9 10:14 0 BB
Chapayev 10 10:20 2 CA/CL
Chapayev 10 10:49 2 CA/CL
Izumo 9 11:06 0 BB
Izumo 9 11:21 0 BB
Sat        
Izumo 10 10:26 1 BB
Bismarck 9 10:37 1 BB
Izumo 9 10:52 0 BB
Izumo 10 11:24 1 BB
Izumo 9 11:41 0 BB
Izumo 9 12:21 0 BB
Izumo 9 12:45 0 BB
Izumo 10 13:02 1 BB
Chapayev 8 13:17 0 CA/CL
Chapayev 10 13:39 2 CA/CL
Bismarck 8 13:49 0 BB
Izumo 10 16:21 1 BB
Izumo 9 14:33 1 BB
Sun        
Izumo 9 13:19 0 BB
Izumo 10 13:32 1 BB
Izumo 10 13:54 1 BB
Izumo 9 14:09 0 BB
Izumo 10 14:23 1 BB
Izumo 9 14:40 0 BB
Izumo 9 14:53 0 BB
Izumo 10 15:11 1 BB
Mon        
Izumo 9 7:42 0 BB
Izumo 10 7:57 1 BB
Izumo 10 8:17 1 BB
Izumo 9 8:45 0 BB
Izumo 10 11:14 1 BB
Izumo 10 11:33 1 BB
Izumo 9 11:52 0 BB
Izumo 9 12:08 0 BB
Izumo 9 12:25 0 BB
Izumo 10 12:42 1 BB
Tues        
Bismarck 9 18:31 1 BB
Bismarck 10 18:49 2 BB
Bismarck 10 19:11 2 BB
Bismarck 9 19:24 1 BB
Wed        
Bismarck 8 6:03 0 BB
Bismarck 8 6:19 0 BB

 

 

Edited by BossmanSlim

Share this post


Link to post
Share on other sites
1,326
[CUTE]
Members
5,202 posts
3,461 battles

I'll now have a "reason" to play all these Gremmy games this weekend now.

 

 

If only people could stop putting in emergency calls, I could get down to ships.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×