Jump to content
You need to play a total of 5 battles to post in this section.
Baskerville77

Stats Analysis and Averages

13 comments in this topic

Recommended Posts

151
[STURM]
Members
302 posts
2,921 battles

 

So, lately I've been keeping a closer eye on my stats, and really trying to improve my win rate. Just doing more basic things, like looking at the mini-map and not rushing in to die needlessly, and...

RecentWinrate.jpg.b0027c9cbacf5c59a102d011f0e5e530.jpg

I think it might be paying off.

WinrateBig.thumb.jpg.d855acde35f11d8f097ff81c33f7ecb0.jpg

Just a tad.

That's not even what I'm most excited about. Take a look at this:

WinrateGeneral.jpg.862690969800f1c75079bec96dca0035.jpg

Look! Look!

Winratecruiser.jpg.5b849ef61051f4b76fb28f10e408ff23.jpg

It's over 50%! I can finally say that I am just slightly, almost insignificantly better than average in cruisers!

index.jpg.5bb13af0e8678ca6f79d2ebaaf7fb809.jpg

But, as I was looking at all these pretty numbers, I realized I don't actually have much to compare them to, and thus lots of the information loses a lot of meaning to me.

For example, I have a 34.89% and 8.44% hit ratio for main battery and torpedoes respectively, but I don't know if those are high or low compared to other players. I suspect that the main battery ratio is high, and the torpedo ratio is low, but I'm just guessing. And it would be even better to compare numbers like that in specific ships to the averages in those ships.

So, then, I have two requests for everybody:

First, what are the best sites to find and compare information like that? I mostly use WoWS Stats & Numbers, but would another site be better for viewing these, or even complimentary?

Second, could some of you with more experience than me take a look at my stats and tell me what kind of information you can gather and how? Cause I'm always looking to shore up weaknesses and enhance strengths, and I cannot do that without knowing what they are.

And be savage! I'm feeling pretty good after looking at my recent win rates, so my ego could use some deflating before I do something stupid, like YOLO rush in a Neptune.

Thank you all!

 

EDIT: Gosh darn it, I should probably put some links to my stats here if I'm asking people to look at them, right?

  https://na.wows-numbers.com/player/1021440414,Baskerville77/

https://worldofwarships.com/en/community/accounts/1021440414-Baskerville77/!/pvp/overview/

 

Edited by Baskerville77

Share this post


Link to post
Share on other sites
472
[QNA]
[QNA]
Members
1,838 posts
5,881 battles

Look at stats for fun but do not take them too serious. I look forward to how you play in battle than I do at stats. There are plenty of stat shamers that will hate on you because of your stats... because you know... I suck with Ryujo. The main website I use is https://na.warships.today/

image.png

  • Bad 1

Share this post


Link to post
Share on other sites
896
[SYN]
[SYN]
Members
2,254 posts
8,649 battles

All the stats can be gamed in some way or another, but I think the nitty gritty ones like main battery and torpedo hit rates are especially meaningless. Not only will they vary hugely between ships, but they will vary hugely depending on how you play a specific ship.... there's too many confounding variables to get any meaningful information out of those numbers. 

Share this post


Link to post
Share on other sites
30
[B2P]
[B2P]
Beta Testers
136 posts
7,207 battles
15 minutes ago, pikohan said:

All the stats can be gamed in some way or another, but I think the nitty gritty ones like main battery and torpedo hit rates are especially meaningless. Not only will they vary hugely between ships, but they will vary hugely depending on how you play a specific ship.... there's too many confounding variables to get any meaningful information out of those numbers. 

Me personally, I think the hit ratios should only be used to compare your own average hit chance based on the overall average recorded for that ship on those sites. That way you can analyze if there's improvement needed in regards to your aiming.

The main battery overall ratio will be higher if you mostly play cruisers or destroyers, while battleships will be lower due to their high dispersion rates. So that's something to consider. 

But to measure a form of skill compared to others regarding hit radio, there's no point. I've seen players hold their fire just because they want to preserve their hit radio, it's pointless to worry about it.

Share this post


Link to post
Share on other sites
1,550
[OPGS]
Beta Testers
3,420 posts
5,852 battles
19 minutes ago, pikohan said:

All the stats can be gamed in some way or another, but I think the nitty gritty ones like main battery and torpedo hit rates are especially meaningless. Not only will they vary hugely between ships, but they will vary hugely depending on how you play a specific ship.... there's too many confounding variables to get any meaningful information out of those numbers. 

I would disagree with you about main battery hit rate. I do not think I have seen a high-level player like yourself with a main battery hit rate of below 30% on a battleship. However, when I see a 42% win rate player with a 600 WTR and look at the main battery hit rate it will universally be at 20-23%. If you can't aim and hit the target consistently than you are not going to be a very good player.

  • Cool 1

Share this post


Link to post
Share on other sites
744
[HYDRO]
Members
1,600 posts
3,753 battles

I could argue about how important stats are etc etc but I would't be telling the whole truth. Most have loopholes to farm them anyway. You can division and farm WR and Damage, the two most important factors in WTR rating for example. That said, a marked improvement in them can be a good indicator.

The sign of improvement you should look out for however, are getting that bit extra influence in each battle. A strategically good smoke, spotting DDs that are killed by your team, or even missing with torps but having the red enemy show broadside to someone, these are small factors that if seen separately seem insignificant, but if combined can give you a good idea on how you are improving.

  • Cool 2

Share this post


Link to post
Share on other sites
2,057
[ARGSY]
Members
6,452 posts
4,302 battles
43 minutes ago, pikohan said:

I think the nitty gritty ones like main battery and torpedo hit rates are especially meaningless.

Especially for those who like to dump whole salvoes of torps for area denial or forcing out of smoke/cap, not necessarily expecting to hit anything at all. 15 torpedoes for no hits (Shimakaze) is going to make a dent in that statistic, especially if you do it more than once per game. And let's face it - after the first four or five hits on an enemy warship, she goes kablooie and I'm not sure the rest are even counted, even if they ALL hit home on the sinking wreck.

Edited by Ensign_Cthulhu

Share this post


Link to post
Share on other sites
3,937
[-K-]
Supertester
3,099 posts
6,832 battles

na.wowsnumbers.com is the best place for you to compare yourself.  If you look at your individual ships, on the right hand side is a button called "details."  Click it.  It opens a window where you can compare yourself to the best 50%, 25%, 10% and 5% of players in that ship.  It's a great place for you to measure and plan your progress.

For example, your Scharnhorst is your most played ship.  Here are some screens of how you compare to the top 50% of performers in the Scharn:

 

vnvvXpM.png

 

What this means is you are better then the potatos, but not quite as good as the top 50%.  The criteria are 20 PVP battles in the ship for the stats to be considered.

  • Cool 4

Share this post


Link to post
Share on other sites
4,866
[WOLF3]
[WOLF3]
Members
16,217 posts
14,630 battles
1 hour ago, pikohan said:

All the stats can be gamed in some way or another, but I think the nitty gritty ones like main battery and torpedo hit rates are especially meaningless. Not only will they vary hugely between ships, but they will vary hugely depending on how you play a specific ship.... there's too many confounding variables to get any meaningful information out of those numbers. 

You can game things like WR% by cheesing the system... You know, the guys that pad their stats with Kamikaze / Kamikaze-R / Fujin? :Smile_teethhappy:  But if you see diversity in the player's experience, lots of non-low tier experience, then it's good.  I am actually more impressed by someone that has Average Tier 8 or higher and still maintains an excellent record.  People tend to drop in performance with High Tier, especially Tier X ships because it's so brutal and unforgiving there.

 

My Average Tier is 7.2 because long ago I liked playing A LOT of Tier VI, now it's more Tier VII-VIII.  My Division Buddy, is 8.7 and he literally lives in High Tier, with Tier X being his Baby, yet he still maintains an excellent performance.  When something like Clan Battles and Ranked VIII / X came about, he was freaking ecstatic.

 

I remember when I was newer, eventually getting to the point that Tier VIII and below I was good to go.  But with Tier IX, things got way harder, and getting a Tier X wasn't any easier, either.  Matter of fact, my first 2 Tier X ships were the old Pre-Buffs Montana and Yamato back when she had the old Mega Healz.  I was getting rekt with them, even with Yamato, and it took me ages to figure out how to get better with them.

Edited by HazeGrayUnderway

Share this post


Link to post
Share on other sites
408
[XBRTC]
Members
1,424 posts
7,482 battles

If you want to have some real fun with your stats, and if you want to get some truly valuable information from them, you need to download and install MXStat, and then spend several hours playing with Excel.

Here's a screenshot from the clan member analysis worksheet that I built, though it hasn't been updated since clan battles ended and it only covers a very specific period of time.

image.thumb.png.372f64456feb4af8be11e9423a21b81a.png

 

These are my own numbers for random battles, drawn from the excel export from MXStat, though I did have to fold / spindle / mutilate a bit to get the information and present it this way.

After each session of clan battles, I'd get all my clanmates to run MXStat and send me their exported data file, then add their battle data to mine.

Using this spreadsheet I have pulldown boxes for selecting random / coop / ranked / clan / scenario, and I can compare various metrics per player on a ship-by-ship basis, breaking down the information across wins, losses, or against all the games, and see how each person does in that ship compared to the clan average in that type of battle. I haven't gotten around to putting in fires and flooding yet, or hit percentages, torpedo use, etc., but you can definitely still draw some good information here when it comes to performance and how someone's play style impacts the game in terms of winning or losing.

Some of the labels are not exactly self-explanatory unless you built the spreadsheet yourself, so let's look at the Kamikaze R line for explanations and an example of how I analyze the data. We use these metrics (and a few others from a different spreadsheet) in trying to put together the optimal lineup in terms of ships and players out of the available folks on any given night for clan battles.

Matchmaking: This shows the player average and the clan average for how badly screwed you are in matchmaking, by showing how far that ship is below the average battle tier that it sees. During the period of time covered by this data, I was seeing an average of 6.3 for the battle level faced in my Kamikaze, while the overall clan average was 6.25 for everyone that owned Kami-R. (5 + 1.3 = 6.3)

Battles Fought: I fought 50 battles in that time period using Kami-R. My clan as a whole fought 95 (including my 50). Therefore, I fought 53% of the clan's total battles in Kami R.

Battles Won: My clan as a whole won 52 of the 95 total battles, including my 29 wins out of 51. I fought 53% of the clans battles in that ship, but earned 56% of the wins. That indicates that I'm doing slightly better in that ship than is the average for my clan.

Win Rate: 58% during that time period average for me, 55% for other clan members.

Damage (all battles): When all battles (wins and losses) are considered, I'm at 42,042 average damage during that time period, while my clan is at 42,082 overall. This is so close as to be statistically insignificant. However, when considered against other factors it can indicate that there's a difference in play style. The same damage when compared against a 3% difference in win rate may indicate that I'm more aggressive in capping, or there may be other factors at play that can be quantified from other parts of the spreadsheet.

Damage (wins): 46,837 for me vs 44,490 clan average. This shows only damage in battles that were wins. I have a higher average than my clanmates' average, in a win. This indicates that I'm probably more aggressive, most of the time, and that it pays off.

Damage (losses): 34,316 (me) vs 38,527 (clan average). I'm significantly lower in damage, here. This could indicate more passive play, taken by itself, but since we've also got the average damage for battles won to compare against, it's more likely to indicate that I was too aggressive, and I'm getting shut down far more often in losses than I am in wins.

Kills: My average across all battles is 1.24, which is lower than the clan average of 1.31. On the face of it, this seems like it might be an issue. Let's dig a little deeper, though. In wins, I'm dead on the clan average of 1.62 kills per battle. In losses, though... I'm lagging. .71 to .93. This lag in losses makes sense from the difference in damage during a loss. But, since the difference in damage during wins is pretty significant and I'm quite a bit higher... why am I dead on the average of 1.62? The higher damage indicates more aggressive play, but the dead-on average for kills indicates that I'm also backing off where appropriate, and not pushing to the point of suicide just to finish a damaged target.

Survival (Full Battle): This shows the percentages of battles where I survived the entire time. This is a stat which is interpreted from the raw data by taking the exported MXStat document's "Time in Battle" and dividing it by "Battle Duration" to get a percentage. Any percentage greater than 98 is considered to be a full battle survived. (There is a little bit of a slop factor how the game handles time. The battle duration shown is usually about 5-10s longer than the time in battle even for a game that you survived, so there's a 2% correction factor allowed in my calculations). I survive 36% of Kami-R battles on average, while my clan survives 20% on average. This means I'm better at taking advantage of my stealth and not getting spotted. In wins, I'm at 45% to the clan average of 25%,  which tells me that my assessment above from kills is accurate: I back off when I need to, rather than just always boring straight in for the kill. In losses, we see 24% survival (me) and 14% (clan average). When viewed against the average damage in losses and average kills in losses... I may have to reassess the too aggressive idea from Damage. It's unclear whether I'm being too timid or too aggressive in losses, from the numbers, at this point. There's still some more factors to look at, though, which may give a bit of clarity.

Survival (% of battle): This shows the average percentage of a battle that I survive. On average, I'll survive 77% of the battle in a win, and 68% in a loss. (78/69 for my clan's average in wins and losses). This slightly lower length of survival for me in a win does indicate that I'm being more aggressive. I'm taking bigger risks, and though they pay off (higher than average win rate, higher than average damage, higher than average full battle survival), they are still risks and they can sometimes come out badly. The shorter average duration of survival in a loss... well, it still really doesn't clear up much.

 

I don't have all the metrics modeled in yet that I intend to show, which could be useful in shedding more light on why my performance in a loss is so far below the clan average, but eventually it will all come in.

(in case anyone is wondering, I do know why. It's a mix of things, but mostly taking risks and being aggressive against red teams that actually work together to shut me down. the things that can work out just fine against enemies who are barely above potato level will be suicidal against a BB who has a cruiser and a couple destroyers who are actively working to protect him and trying to win the game rather than just farming damage and lighting fires. I don't run into these guys that often, but when I do it's going to result in a game where I have <5,000 damage and die almost instantly, and that skews the rest of the averages substantially.)

 

 

 

 

 

 

  • Cool 1

Share this post


Link to post
Share on other sites
151
[STURM]
Members
302 posts
2,921 battles
44 minutes ago, Pope_Shizzle said:

na.wowsnumbers.com is the best place for you to compare yourself.  If you look at your individual ships, on the right hand side is a button called "details."  Click it.  It opens a window where you can compare yourself to the best 50%, 25%, 10% and 5% of players in that ship.  It's a great place for you to measure and plan your progress.

For example, your Scharnhorst is your most played ship.  Here are some screens of how you compare to the top 50% of performers in the Scharn:

 

vnvvXpM.png

 

What this means is you are better then the potatos, but not quite as good as the top 50%.  The criteria are 20 PVP battles in the ship for the stats to be considered.

Thank you so much for showing me this! This is exactly the kind of data that I was looking for!

I can draw so many more conclusions from this kind of stuff. For example, my win rate in the Missouri is about 2% better than the average for the top half of players, but my damage and kills are below average for all players in the Missouri. I think that probably means that I do well when I don't over extend and get deleted instantly trying to use radar, but if I do overextend I accomplish squat and bring down my average damages and kills. Plus even then I might get a victory just because my team drags my burning corpse across the finish line.

Actually, it's kind of sobering seeing that I don't measure up to the top half on many of my favorite ships, but that's just more incentive to improve. 

Share this post


Link to post
Share on other sites
491
[NWNG]
[NWNG]
Members
2,039 posts
3,580 battles

Excellent work. Using those stats in an attempt to improve oneself, for one's own sake, rather than trying to beat another person, is honestly, in my opinion, the best use of stats overall.

Share this post


Link to post
Share on other sites
3,937
[-K-]
Supertester
3,099 posts
6,832 battles
20 hours ago, Baskerville77 said:

Actually, it's kind of sobering seeing that I don't measure up to the top half on many of my favorite ships, but that's just more incentive to improve. 

 

Just remember that sample size is important.  The site only tracks people in the top X % who have played 20+ battles.  So all the people with 5, 10, 15 battles whose stats are factored into "averages" aren't tracked.  Also, at 20 battles, a few really good or really bad games can have a disproportionately large impact on overall numbers.  

 

That you are using the metrics to better yourself is to be commended.  Too many people use stats as a way to demean others.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×