Jump to content
LittleWhiteMouse

How to Control your Win Rate

505 comments in this topic

Recommended Posts

1,043
[TOG]
Members
4,627 posts
29,487 battles
17 hours ago, Spud_butt said:

how exactly are you supposed to accomplish making sure your team is never down by more than one ship...?  well obviously, by not being sunk, and sink or chase off enemy vessels (damage) that may sink you or your teammates in the localized engagement you're in. i

By yourself no. Unless you're someone who is supremely confident and works at W/R 's of 60% minimum. You do the best you can. And sometimes is not enough.

17 hours ago, Spud_butt said:

I have zero clue. I'm not saying nor blaming it on potato teams, because they can't all be potato teams. can they??

I've been on matches where sometimes the reds were spuds (woohoo!) and the spuds (except me of course)  were on my side (grumble). W

 

17 hours ago, Spud_butt said:

also, in more than a few battles a few players made herculean efforts to guide and direct the team. these efforts were responded to very negatively, verbally profusely and profanely, and we lost them slaughter level owie.

Worse are players who deliberately go Leroy Jenkins.

 

17 hours ago, Spud_butt said:

also, in more than a few battles a few players made herculean efforts to guide and direct the team. these efforts were responded to very negatively, verbally profusely and profanely, and we lost them slaughter level owie.

Yes, because those folks do not care whether they win or lose. FPS mentality.

Share this post


Link to post
Share on other sites
3,483
[META_]
Members
2,317 posts
8,119 battles
On ‎6‎/‎14‎/‎2018 at 8:01 PM, Bill_Halsey said:

What LHM said about being a ROFL  stomp 1/6 of them time does happen. 

I would happily start playing regularly, and even buying stuff from WG again if I could get anywhere close to a ROFL stomp defeat 1/6 of the time.

Friday last week played 11 battles, and won 2. of the 9 defeats, 7 were all green boats sunk, the other 2 stopped when we went under 0 points, averaging 700 pts down. I was not trying to lose, nor make my team lose.

that's 7/11 slaughter losses. it followed Thursday, where I played 4 or 5 with one victory, and 3 were slaughter losses.

hell, at this point, I would even be delighted experiencing ROFL stomp defeat 1/3 of the time. that would be a monster improvement over seeing more than half the battles played soul crushing defeats.

the enemy has taken the lead. the enemy has taken the lead. the enemy is about to win. the enemy is about to win. the enemy is about to win. 

Share this post


Link to post
Share on other sites
1,043
[TOG]
Members
4,627 posts
29,487 battles

I am pleased to announce I  have achieved 50% W/R solo from a 43.6% when I started.

 

  Overall 7/26, 11:01 PM
to 7/28, 1:06 PM
7/26, 12:08 AM
to 7/26, 11:01 PM
7/25, 6:17 AM
to 7/26, 12:08 AM
7/24, 8:29 AM
to 7/25, 6:17 AM
7/22, 10:08 PM
to 7/24, 8:29 AM
7/22, 4:53 PM
to 7/22, 10:08 PM
7/21, 11:47 AM
to 7/22, 4:53 PM
 
Battles 9 866 25 7 9 7 3 4 25
Win rate
Next level: Good (+2%)
50%
Average
56%
Great
+ 0.02% 57.14%
Great
+ 0.01% 44.44%
Bad
- 0.01% 57.14%
Great
+ 0.01% 100%
Super Unicum
+ 0.02% 50%
Average
- 48%
Below Average
- 0.01%

 

 

  • Cool 1

Share this post


Link to post
Share on other sites
703
[NSEW]
Members
2,615 posts
12,116 battles
On 28/07/2018 at 7:49 PM, Bill_Halsey said:

I am pleased to announce I  have achieved 50% W/R solo from a 43.6% when I started.

Well done, I too am slowly revisiting the WR after not caring for some time.

 

Share this post


Link to post
Share on other sites
613
Members
696 posts
8,275 battles

What is a realistic "good" solo win rate for random battles?

Right now I have a 50.x% solo wr for t10 in my yamato over 100+ battles, and I feel like I can't do much more to influence the outcome.

I have a 66% solo wr in my t4 IJN premium bb over ~60 battles, yet I have no idea how I am able to do that.

Lastly when I play dd (only t5 so far), I am 100% team objective focused (scouting first, killing dd 2nd, capping 3rd) and yet my solo wr is somewhere in the ~30%.

I would love to know the secret sauce(s) to get 60% wr as a solo player.

Share this post


Link to post
Share on other sites
Members
1,380 posts
4,100 battles
Quote

What is a realistic "good" solo win rate for random battles?

If 60% is the best  you can get and while rare is quite achievable, and 40% the worst (this is not totally accurate - most players are around 48-49%); then (approximately) 52%WR you are in top 40%, 54%WR you are in top 30%, 56%WR you are in top 20%, 58% WR top 10%....

Quote

Right now I have a 50.x% solo wr for t10 in my yamato over 100+ battles, and I feel like I can't do much more to influence the outcome.

I have a 66% solo wr in my t4 IJN premium bb over ~60 battles, yet I have no idea how I am able to do that.

That's simply because T10 is far more punishing than T4. Your mistakes are punished by more accurate, long range weapons and focus fire.

You simply are "average"at Yamato and, accordingly, games WILL feel like a coin flip. You aren't influencing the game one way or another.

Quote

Lastly when I play dd (only t5 so far), I am 100% team objective focused (scouting first, killing dd 2nd, capping 3rd) and yet my solo wr is somewhere in the ~30%.

I find a strong correlation between survival and win rate in destroyers. If you die in the first 5min, you derped. If you last 5-10 min - not good, if you last beyond 10 min, you were probably useful to your team. Each time I die in a DD, I consider this.  Destroyers become more useful the longer into the game you last.  T4 destroyers are universally strong - stay at T4 and practice.  T5-6 tend to get bad MM. The next best tier is T7.

Finally: Sample size. When considering stats, you need lots of games for stats to be reliable. I play 3 games in my new DD and win only 1 with a 33% win rate....   ...that's not reliable as I could have got bad teams etc.  Don't stress.  If I played 1000 games with 33% win rate it would mean I was a bad player. Losing an extra 167 games probably involves me rather than random luck.

Edited by evilleMonkeigh

Share this post


Link to post
Share on other sites
565
[BS]
Members
2,008 posts
11,839 battles

Here's an important tip, you have to know WHEN to shoot when you're looking at a target. Just because you can, doesn't mean you should. The guy might turn a little bit more if you wait a few seconds, and his ship will end up perfectly broadside when your shells arrive...Things like that help your winrate because it has "stopping power." Quickly maiming a ship influences said players behavior afterwards. Some people say screw it, i'm yoloing now... Some people hide for a little while. Either way, that's helping you win.

Edited by HorrorRoach

Share this post


Link to post
Share on other sites
1,043
[TOG]
Members
4,627 posts
29,487 battles
3 hours ago, mixmkz said:

What is a realistic "good" solo win rate for random battles?

Right now I have a 50.x% solo wr for t10 in my yamato over 100+ battles, and I feel like I can't do much more to influence the outcome.

I have a 66% solo wr in my t4 IJN premium bb over ~60 battles, yet I have no idea how I am able to do that.

Lastly when I play dd (only t5 so far), I am 100% team objective focused (scouting first, killing dd 2nd, capping 3rd) and yet my solo wr is somewhere in the ~30%.

I would love to know the secret sauce(s) to get 60% wr as a solo player.

Good? - WoWs Stats and numbers defines it as 52% or greater.

The secret sauce on how to get 60%? Pick a ship and be stupidly good in it.

Share this post


Link to post
Share on other sites
613
Members
696 posts
8,275 battles

Those are all really great suggestions, so thanks so much for sharing them with me.  I agree with the shot timing and survival rate comments a lot, as I catch myself making those mistakes often.

The reason why I ask about solo win rate is I wonder if the statistic itself is deceiving. I can't help but wonder if the 52% wr is a factor of the population over indexing higher wr in the lower tiers plus that the median lies in the 40s while there are a few truly great players with high wr that drags the mean artificially higher than where the bulk of the population is distributed.  I do pretty well in t4 (both the clemson and the IJN prem BB), but while blapping newbies faster than they can blap me back will be good for stats, it doesn't teach me to be a better player.

Edited by mixmkz

Share this post


Link to post
Share on other sites
Members
95 posts
3,169 battles

I tend to be the baby seal bravely sailing forward to get killed. I am getting better and my damage and survivablity rates are going up.

Share this post


Link to post
Share on other sites
Members
253 posts
8,820 battles

Using this formula, the Sims and Nurnberg I was improving. Mostly played a certain time of day and always stopped at 50/50 w/l or one loss more.Just a short amount of time each session.

I was thinking in a few months time I'd have been close to 50%, which even where I was at I was always near being the average player I've always claimed to be. More support than a player who carries. 

Well yesterday morning, after Comcast crashed twice within an hrs time. I called them up, again, cussed em out for a few minutes, using a lot more salt than ever I put into battle chat, while cancelling my service, 1st of next month.

At one point I told the agent I didn't care if God himself was the reason.

Many years ago I learned only a complete fool keeps doing the same thing over and over expecting different results.

Back to off-line Xbox again. Not owned a console since 2005. Things probably  still buggy.

When that picture was took I played Xbox offline, CivII and Outpost2 on a Win98, and was just learning about Twitter and facebook, sometime around 2004 in San Antonio. 

 

 

Edited by IOWADragon

Share this post


Link to post
Share on other sites
1,043
[TOG]
Members
4,627 posts
29,487 battles

Closing in on being statically insignificant. 0.02 % to go. It's getting harder now. To keep improving I need an average of 55% or better. 

Share this post


Link to post
Share on other sites
85
[BONKS]
Alpha Tester
242 posts
7,854 battles
On 6/14/2018 at 9:28 PM, LittleWhiteMouse said:

Again, it is quite easily provable by looking at win rate distribution.  Here's what I mean.

 

blT2A5Z.png

Here's my quick and dirty MS Paint example of win rate distribution.   Please forgive it's crudeness.

  • In RED we see a standard distribution curve.  It's not symmetrical with a big chunk of the population sub 50% and a gradual long slope towards higher win rates.
  • In ORANGE we see a  manipulated distribution curve.  There are almost no players with win rates higher than 55% or lower than 42.5% and most are hedged between 45% and 53% win rates.  This indicates that the game is hard capping maximum and minimum win rates through manipulation either of teams or through mechanics.
  • In GREEN we see a manipulated distribution curve.  It's too symmetrical to be indicative of a normal player population.

@LittleWhiteMouseHere is a real distribution of player winrates (by ship by player) where they have more than 25 battles in a ship to weed out extremes as a result of luck as opposed to skill (previous analysis has supported this number). The normal distribution curve is overlayed and the chart above is the descriptive stats of the sample.

image.thumb.png.c6ff0eaaf58cd643fafda54bb34969d2.png

Share this post


Link to post
Share on other sites
85
[BONKS]
Alpha Tester
242 posts
7,854 battles

I have also written a recent article explaining winrate as a viable metric for player skill but in it go over briefly two other 'metrics,' survival rate and damage that relate and can help account for winrate. http://warshipnews.blog/2018/10/02/winrate-as-a-player-metric/

  • Cool 1

Share this post


Link to post
Share on other sites
1,043
[TOG]
Members
4,627 posts
29,487 battles
22 hours ago, SnipeySnipes said:

I have also written a recent article explaining winrate as a viable metric for player skill but in it go over briefly two other 'metrics,' survival rate and damage that relate and can help account for winrate. http://warshipnews.blog/2018/10/02/winrate-as-a-player-metric/

Read it and found it informative. I'm finding out from my own experience it's a bit more nuanced than that. Making survival rate as one of the metrics for supporting W/R is a bit misleading. It's saying a player didn't contribute anything, save being cannon fodder, toward winning. Better questions to ask is

1.  When did he die? - Was it 2 mins into the match? 5 min? 15 min? 19 min? 

2. What he do in the mean time?  A DD player can contribute more to winning the game just by spotting, cap control and killing ships rather than just farming damage.  If a DD player had taken 2 solo caps and 3 killed DD's before dying at the 17 min mark, he already had done more to win than the Conqueror  who had farmed 110k damage on BB's and didn't kill a single one.

A better metric to look at is the K/D ratio along side damage rather than survivability to see what contributes  to win rate.  Especially for destroyers. A K/D ratio >1 I see as being effective. 

I've been playing DD's for a while and my average survival rate last night is 42.86%.  My K/D avg ratio is 2.25. W/R was 57.14%. Most of HP score was taken off destroyers. So I say K/D ratio instead of survivability is a better metric.

 

Edited by Bill_Halsey

Share this post


Link to post
Share on other sites
85
[BONKS]
Alpha Tester
242 posts
7,854 battles

@Bill_HalseyFrom the data that I have, Survival rate is black and white, they either survived the battle or they didn't, and using that I can draw conclusions from the analyses. Though it is probably no stretch to say the LONGER you survive, the more you contribute, but I don't have that data. As for K/D, I can't remember if I ran analyses on that or not. I do have that information so I could do that and get back to you. But in the survival rate, as an FYI, the average for my data set (over 3 million points) was about 33%, so pulling 40%+ is pretty good.

  • Cool 1

Share this post


Link to post
Share on other sites
1,043
[TOG]
Members
4,627 posts
29,487 battles
4 minutes ago, SnipeySnipes said:

@Bill_HalseyFrom the data that I have, Survival rate is black and white, they either survived the battle or they didn't, and using that I can draw conclusions from the analyses. Though it is probably no stretch to say the LONGER you survive, the more you contribute, but I don't have that data. As for K/D, I can't remember if I ran analyses on that or not. I do have that information so I could do that and get back to you. But in the survival rate, as an FYI, the average for my data set (over 3 million points) was about 33%, so pulling 40%+ is pretty good.

Yes, the detailed report doesn't say how long you're alive so you can't compile a report. BTW, can you do a report of overall survivability average by ship class? Even better if you can do it by tier and ship class.

Share this post


Link to post
Share on other sites
85
[BONKS]
Alpha Tester
242 posts
7,854 battles

I should be able to run that report. What would you like to see, just averages with standard deviations or correlations to something? As an FYI, the data that I have is older, so it will not include some of the new ships, but the data should be able to carry over

 

Share this post


Link to post
Share on other sites
1,043
[TOG]
Members
4,627 posts
29,487 battles
1 hour ago, SnipeySnipes said:

I should be able to run that report. What would you like to see, just averages with standard deviations or correlations to something? As an FYI, the data that I have is older, so it will not include some of the new ships, but the data should be able to carry over

 

Correlation to W/R would work. But you'll have to tell me if it's valid or not since I'm not a statistician. Survivability avg per ship class and tier would do. LHM said the CA's have a worse survivability rate than dd's.  True or not?

Share this post


Link to post
Share on other sites
85
[BONKS]
Alpha Tester
242 posts
7,854 battles

Seeing as this will likely be a big table or several tables, I'll probably end up writing a whole article up on it. It will be on my site and I'll post the link here when it's up.

Edit: This will most certainly be stat intensive so an article will be a MUST. I'll try to boil it down as best I can for the general population. @n00bot If you want in on this let me know. Should be fun

Share this post


Link to post
Share on other sites
643
Members
1,185 posts
3 hours ago, SnipeySnipes said:

From the data that I have, Survival rate is black and white

There are definitely times when suicide missions help the team, and high-level callers will use that tactic. The ship that dies may be able to delay 3+ ships attacking a flank, or pull a radar cruiser out of position so the team can win a critical cap. Nothing is black-and-white, even survival, but these special cases are “washed out” in aggregate data.

Share this post


Link to post
Share on other sites
85
[BONKS]
Alpha Tester
242 posts
7,854 battles

What I mean is they survived or didn't. I have not data to show why or, in your examples, if their survival posed beneficial to the team

Share this post


Link to post
Share on other sites
643
Members
1,185 posts
1 hour ago, SnipeySnipes said:

Seeing as this will likely be a big table or several tables, I'll probably end up writing a whole article up on it. It will be on my site and I'll post the link here when it's up.

Edit: This will most certainly be stat intensive so an article will be a MUST. I'll try to boil it down as best I can for the general population. @n00bot If you want in on this let me know. Should be fun

Not sure how much time I’d spend but yah, I’d love to take a look at some data.  PM me.

Share this post


Link to post
Share on other sites
643
Members
1,185 posts
14 minutes ago, SnipeySnipes said:

What I mean is they survived or didn't. I have not data to show why or, in your examples, if their survival posed beneficial to the team

Not dissing your analysis just bringing up the point that nothing is “black&white” in this game.  Personally I found my WR went up when I learned to break the rules.

Edited by n00bot

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×