Jump to content
You need to play a total of 10 battles to post in this section.
Rina_Pon

Rina_Pon's Ship OP Ranking

22 comments in this topic

Recommended Posts

Member
325 posts
4,688 battles

Those of you who follow my occasional posts know I like playing with numbers. Today I found wowstats have published a full list of ships with their stats, so went and tried to come up with a way to rank them in terms of how "OP" they are.

"OP" means overpowered, i.e. a ship will score more kills, or do more damage, relative to the other ships she typically sees in battle. The trick is to work out how to balance the weighting between damage dealt and kills, and how to normalize against tier.

Of course the relative power of a ship depends on the captain as well as the ship, so of course firstly "reward ships" and secondly regular premium ships are going to score higher as far fewer of them ever set sail in the hands of anything less than good captains. That's fine, you can still compare "like-for-like", or get an idea of how much better premiums tend to do above regular ones.

I settled on the formula sqrt(dmg/tier)*frags^2. It first I didn't have the square root, but that favored BBs too much I thought. Overall the ordering doesn't change too much with those kinds of tweaks though.

I then scaled the OP rating on a relative, logarithmic scale with the Karlsruhe marking zero point and KamikazaR as a 1.0 - the default reference OP ship in most people's minds.

Ta da!

5a66bdf686430_WoWSOPratingplot.thumb.png.a5c473892acc52495d722f99093e21f7.png

To me, the surprising thing was how well it worked. The ship OP rank is pretty much independent of tier, and doesn't seem to favor any particular ship class. The ships you expect to be at the top are, and the ships you expect to be at the bottom are.

Yes, it's officially proven: the Karlsruhe is really that bad.

The most interesting thing though is the sudden uptick in relative OP over 0.9. The top OP ships are truly monsters.

The full list is here,

 

P.S. Old versions of ships like Kiev and Fubuki are removed from the list, and newer ships with less than 100,000 battles reported are also not included since there maybe insufficient data for good statistics.

P.P.S. I think the list is generally "fair". The only ships that are not given their OP credit are fighter CVs since I couldn't figure out how to factor in plane kills into the formula.

 

 

 

Rina_Pon Ship OP Rank List 180123.txt

Edited by Rina_Pon
postscript added

Share this post


Link to post
Share on other sites
Super Tester
7,585 posts
8,005 battles

Oww Kaay

Thing thing is, certain ships work for you and certain don't. While the server stats may show what is more *likely* to happen, but that's not always the case. For example you have mentioned Karls, I know for a fact that Karl is OP in T4 because I can do 60k-70k consistently in Karls and there is no cruiser that can brawl against her. 

Interesting, nonetheless.

Share this post


Link to post
Share on other sites
[LNA]
Member
1,671 posts
10,499 battles

Hakuryu and Taiho isnt that stronk anymore , the new king is Midway and Essex who outclass both thanks to the buffs they got.

3x4 vs 2x6 , equally dead 

2 v 2 , Amurica win

Dive bomber ? IJN is so puny compare to Amurica muscles

Share this post


Link to post
Share on other sites
Member
157 posts
4,084 battles
2 minutes ago, legionary2099 said:

Hakuryu and Taiho isnt that stronk anymore , the new king is Midway and Essex who outclass both thanks to the buffs they got.

3x4 vs 2x6 , equally dead 

2 v 2 , Amurica win

Dive bomber ? IJN is so puny compare to Amurica muscles

That and the recent changes to DFAA for USN DD's make playing IJN CV painful

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles
14 minutes ago, icy_phoenix said:

Thing thing is, certain ships work for you and certain don't.

By the way can anyone download the full list I posted? It's a plain text file but I get an error.

Anyhoo - sure, it only tells you how well on average the ship tends to do, not how well you will do it in. That's two very different things.

I don't think Bogue is a bad CV, it's just the vehicle of choice for any potato who decides he wants to try his hands at CV life. Ditto most low tier IJN boats, which again have a relatively steep learning curve and the low OP score reflects that. 

So yes, a high score generally means "ship is always played by good player" and a low score means "ship is widely available but hard to play", so bear that in mind.

Or take it as a badge of honor that you do well in "low OP" ships. Knowing that most people suck at Furutaka makes doing well in her all the more sweet.

btw and this is just a rando thing I discovered in the numbers: if you look at all the battles reported for all the ships, the win rate is 52%, not 50%, because, essentially, only winners use stat trackers. So all the data is weighted slightly towards the more skilled player base.

 

Edited by Rina_Pon

Share this post


Link to post
Share on other sites
Member
455 posts
8,038 battles

I think the key fault in your analysis is that you're taking server average to be the defining measure of a ship's ability. 

Some ships are simply designed to be harder to play for the average player but more rewarding than their counterparts if played right.

For instance, I have always heard people complain about the Pensacola being really bad even after the concealment buff, but when I played her she was pretty powerful. Her glass cannon like characteristic means that the average player gets deleted fast and lowers the server avg.

Instead, I'd say you should consider top5% or 10% player base stats as that would more accurately represent the ship's full potential

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles

I don't think that's so much a fault so much as the working definition. To me "OP" means "ship typically does better than average", not "how fast can flamu get a kraken in it".

Again, OP does not mean you will do well in one particular ship. The statistics applies to ensembles, not to individuals. Of course it is important to keep in mind that "ship typically does better than average " need not equate to "ship is better than average". It don't think it would ever be possible to normalize the data for some hypothetically constant captain skill.

All that said theirs some interesting conclusions to be drawn from the exercise. First let me post the results since I don't think WG is letting me upload txt files.

I'll split it, T1-5 first:

Tier, Ship Name, OP Score

V, Kamikaze R, 1
V, Fujin, 1
V, Kamikaze, 1
V, Gremyashchy, 0.9
V, Zuiho, 0.8
V, Murmansk, 0.8
V, Texas, 0.8
V, ARP Kongo, 0.7
V, Konig, 0.7
V, Iron Duke, 0.7
V, Minekaze, 0.7
V, Marblehead, 0.7
V, ARP Haruna, 0.6
V, ARP Hiei, 0.6
V, ARP Kirishima, 0.6
V, Emile Bertin, 0.5
V, Kongo, 0.5
V, Konigsberg, 0.4
V, Omaha, 0.4
V, New York, 0.4
V, Kirov, 0.3
V, Gnevny, 0.3
V, Nicholas, 0.3
V, T-22, 0.2
V, Furutaka, 0.2
V, Mutsuki, 0.1
V, Podvoisky, 0.1
V, Emerald, 0.1
V, Bogue, 0

IV, Imperator Nikolai I, 1.2
IV, Orion, 1.1
IV, Arkansas Beta, 1
IV, Kaiser, 0.8
IV, Ishizuchi, 0.8
IV, Hosho, 0.8
IV, Isokaze, 0.7
IV, Yubari, 0.7
IV, Duguay-Trouin, 0.6
IV, V-170, 0.6
IV, Clemson, 0.6
IV, Wyoming, 0.5
IV, Kuma, 0.5
IV, Svietlana, 0.4
IV, Danae, 0.4
IV, Izyaslav, 0.3
IV, Langley, 0.3
IV, Phoenix, 0.3
IV, Myogi, 0.2
IV, Karlsruhe, 0

III, Konig Albert, 1.2
III, Nassau, 0.9
III, Bellerophon, 0.8
III, Aurora, 0.8
III, Bogatyr, 0.7
III, Campbeltown, 0.7
III, St. Louis, 0.6
III, G-101, 0.6
III, Friant, 0.6
III, Caledon, 0.4
III, Wakatake, 0.4
III, Kawachi, 0.4
III, Tenryu, 0.3
III, Derzki, 0.3
III, South Carolina, 0.3
III, Wickes, 0.3
III, Kolberg, 0.2

II, V-25, 0.8
II, Tachibana, 0.7
II, Umikaze, 0.7
II, Emden, 0.7
II, Albany, 0.6
II, Dresden, 0.6
II, Jurien de la Graviere, 0.5
II, Mikasa, 0.5
II, Novik, 0.4
II, Weymouth, 0.3
II, Storozhevoi, 0.3
II, Chikuma, 0.3
II, Sampson, 0.2
II, Chester, -0.1

I, Hashidate, 0.7
I, Orlan, 0.7
I, Black Swan, 0.7
I, Hermelin, 0.6
I, Erie, 0.5
I, Bougainville, 0.3
 

 

 

 

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles

T5-10 next.

 

Tier, Ship Name, OP Score

X, Hakuryu, 1.3
X, Midway, 1.1
X, Minotaur, 0.9
X, Conqueror, 0.8
X, Groser Kurfurst, 0.8
X, Montana, 0.8
X, Khabarovsk, 0.8
X, Des Moines, 0.8
X, Zao, 0.8
X, Yamato, 0.8
X, Henri IV, 0.7
X, Hindenburg, 0.7
X, Moskva, 0.7
X, Z-52, 0.6
X, Gearing, 0.6
X, Shimakaze, 0.5

IX, Taiho, 1.2
IX, Essex, 1
IX, Missouri, 0.7
IX, Neptune, 0.6
IX, Z-46, 0.6
IX, Saint-Louis, 0.5
IX, Fletcher, 0.5
IX, Udaloi, 0.5
IX, Iowa, 0.5
IX, Friedrich der Grose, 0.5
IX, Dmitri Donskoi, 0.4
IX, Izumo, 0.4
IX, Roon, 0.4
IX, Ibuki, 0.4
IX, Yugumo, 0.4
IX, Kagero, 0.3
IX, Baltimore, 0.2
IX, Tashkent, 0.2

VIII, Shokaku, 0.9
VIII, Alabama, 0.7
VIII, Amagi, 0.7
VIII, Lexington, 0.7
VIII, Bismarck, 0.7
VIII, Monarch, 0.6
VIII, Tirpitz, 0.6
VIII, Mikhail Kutuzov, 0.5
VIII, North Carolina, 0.5
VIII, Charles Martel, 0.5
VIII, Edinburgh, 0.5
VIII, Akizuki, 0.4
VIII, Atago, 0.4
VIII, Mogami, 0.4
VIII, Benson, 0.4
VIII, Chapayev, 0.4
VIII, Z-23, 0.4
VIII, Kiev, 0.3
VIII, ARP Takao, 0.3
VIII, Kagero, 0.3
VIII, Loyang, 0.2
VIII, Tashkent, 0.2
VIII, HSF Harekaze, 0.2
VIII, Admiral Hipper, 0.1
VIII, Ognevoi, 0.1
VIII, Prinz Eugen, 0.1
VIII, New Orleans, 0.1

VII, Kaga, 1.3
VII, Belfast, 0.9
VII, Scharnhorst, 0.9
VII, Saipan, 0.9
VII, King George?V, 0.8
VII, Hiryu, 0.8
VII, Fiji, 0.8
VII, Hood, 0.7
VII, Gneisenau, 0.6
VII, Shiratsuyu, 0.6
VII, Nagato, 0.6
VII, Leningrad, 0.5
VII, Shchors, 0.5
VII, Blyskawica, 0.5
VII, Colorado, 0.5
VII, Ranger, 0.5
VII, Leberecht Maass, 0.5
VII, Algerie, 0.5
VII, ARP Myoko, 0.4
VII, Myoko, 0.4
VII, Atlanta, 0.4
VII, Indianapolis, 0.3
VII, Akatsuki, 0.3
VII, Sims, 0.3
VII, Minsk, 0.3
VII, ARP Nachi, 0.2
VII, ARP Ashigara, 0.2
VII, ARP Haguro, 0.2
VII, Pensacola, 0.2
VII, Yorck, 0.2
VII, Mahan, 0.2

VI, Arizona, 0.9
VI, Warspite, 0.7
VI, Dunkerque, 0.7
VI, Fuso, 0.7
VI, Mutsu, 0.7
VI, Shinonome, 0.7
VI, Ryujo, 0.6
VI, Queen Elizabeth, 0.6
VI, Molotov, 0.6
VI, Bayern, 0.6
VI, Perth, 0.6
VI, New Mexico, 0.6
VI, Leander, 0.6
VI, Anshan, 0.5
VI, Admiral Graf Spee, 0.5
VI, Budyonny, 0.5
VI, Cleveland, 0.5
VI, Duca d'Aosta, 0.4
VI, Ernst Gaede, 0.3
VI, Hatsuharu, 0.3
VI, Aoba, 0.3
VI, La Galissonniere, 0.3
VI, Farragut, 0.2
VI, Nurnberg, 0.2
VI, Ognevoi, 0.2
VI, Independence, 0.2
VI, Fubuki, 0.1
VI, Gnevny, 0.1

Share this post


Link to post
Share on other sites
Super Tester
7,585 posts
8,005 battles
1 hour ago, Adm_Kunkka said:

Instead, I'd say you should consider top5% or 10% player base stats as that would more accurately represent the ship's full potential

^^ This.

Or even look for the top 50% players (i.e. cutting off below average players). Because the only thing they do is drag the average down, they do not contribute to the stat in positive way. They show what will happen if you "Always" make the wrong play with a ship.

Wait, that site does not track these information. 

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles

No, sadly this cannot be separated. So if you are looking for some absolute sense of which ship is more powerful than that ship, you must leave disappointed.

The captain cannot be taken out of the equation.

Edited by Rina_Pon

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles

So, as I was saying, general conclusions:

RN BB often the top OP BB in the tier but nor exclusively so. I believe the numbers support the accusation that the ships are OP since the captain skills should not be different from other nations. Usually BBs (and carriers) sit at the top of the tier, but there are odd exceptions, see T9 for example. This reflects the simple fact that BBs are strong as a class, as are CVs (I'd argue they should be). The BBs people consider to be good generally score highly.

The advantageous MM of T7 actually boosts the scores of that Tier as a whole relative to T8 or T5. Same deal with T10.

Belfast, Scharnhorst --- most of the premium ships considered to be powerful show up at the top of the lists.

Cruisers show a very wide range of scores. Just within T7 the difference between Yorck and Myoko and Fiji and Belfast covers almost the entire list. Why does Schorrs score so much better than Pensacola for example? I thought both were quite fragile.

Very few Japanese ships other than BBs score highly.

DDs are kind of hit or miss. The worse their guns the worse they tend to place.

 

Edited by Rina_Pon

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles
6 minutes ago, Gesterbein said:

>hakuryu OP
since when ?
 

Since 109k average dmg, 1.64 kills/game. 

 

or did I miss the /s tag?

Edited by Rina_Pon

Share this post


Link to post
Share on other sites
Member
2,225 posts
13,751 battles
On 1/23/2018 at 4:23 PM, Rina_Pon said:

By the way can anyone download the full list I posted? It's a plain text file but I get an error.

....

 

More WG shenanigans. Downloading is broken atm. But you can download with right click "Save file as..."

edit: no apparently you can't. I thought I had confirmed it but I guess I hadn't.

Edited by keskparane

Share this post


Link to post
Share on other sites
Member
274 posts
23,699 battles
9 hours ago, Rina_Pon said:

I don't think that's so much a fault so much as the working definition. To me "OP" means "ship typically does better than average", not "how fast can flamu get a kraken in it".

Again, OP does not mean you will do well in one particular ship. The statistics applies to ensembles, not to individuals. Of course it is important to keep in mind that "ship typically does better than average " need not equate to "ship is better than average". It don't think it would ever be possible to normalize the data for some hypothetically constant captain skill.

All that said theirs some interesting conclusions to be drawn from the exercise. First let me post the results since I don't think WG is letting me upload txt files.

I'll split it, T1-5 first:

Tier, Ship Name, OP Score

V, Kamikaze R, 1
V, Fujin, 1
V, Kamikaze, 1
V, Gremyashchy, 0.9
V, Zuiho, 0.8
V, Murmansk, 0.8
V, Texas, 0.8
V, ARP Kongo, 0.7
V, Konig, 0.7
V, Iron Duke, 0.7
V, Minekaze, 0.7
V, Marblehead, 0.7
V, ARP Haruna, 0.6
V, ARP Hiei, 0.6
V, ARP Kirishima, 0.6
V, Emile Bertin, 0.5
V, Kongo, 0.5
V, Konigsberg, 0.4
V, Omaha, 0.4
V, New York, 0.4
V, Kirov, 0.3
V, Gnevny, 0.3
V, Nicholas, 0.3
V, T-22, 0.2
V, Furutaka, 0.2
V, Mutsuki, 0.1
V, Podvoisky, 0.1
V, Emerald, 0.1
V, Bogue, 0


 

 

 

 

And there's the evidence that the model you're working with is flawed. 

Bogue is the single most powerful ship at T5. It has been at every point of the game, before they removed manual drop, after, after US CV alteration, its never budged from its position at the top. 

The distortion is data is probably due to the fact that the 110 loadlout is what made bogue brokenly powerful  but almost all users resorted to the 201 loadout.

 

This is probably true for a lot of your other findings as well. 

I applaud the effort and this is a very interesting discussion but we must be wary of the gross oversimplification that will result from the approach we're using.

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles

It seems a lot of the objections raised here is some variant on "you're holding it wrong". i.e. a given ship is better than the score indicates because a good player can score much better than the average.

Which, when you consider what "average" means, must be true of all ships in all cases. To explore that any further (whether some ships have a higher spread of results than others) requires more detailed breakdown of the statistics than the available data allows, so it's a bit of a dead end.

On the other hand, there's a counter-argument that goes like this:

What makes a ship "balanced" vs. "OP" so far as the game (or WG) is concerned is not whether very good players do well, but how a ship performs on average. In that sense, "OP" means one or both of "easy to do tons of damage and score lots of kills in"  and "exclusively available to the highest skilled players". If you remove premiums (or compare within premiums exclusively) so that the availability is broadly comparable, what's left is a measure of ship difficulty level factoring in predominant MM. To my way of thinking at least, if a ship is exceptionally easy to do well in, it is OP.

Even if you disagree, and that's fine, just look at the OP rating as a measure of ship difficulty. The "challenge factor" if you will.

Share this post


Link to post
Share on other sites
231
[151ST]
Member
1,017 posts
4,530 battles

My argument against this is simple;

The amount of damage/frags doesn't have any influence on getting caps, denying caps and power projection.

A good example are Lo Yang and Z-52. Both are arguably best in tier at taking and holding caps, which often leads to victories and better positioning, however both rank very low.

IMO all you've done is a 'which ships hurt stuff more' so it's no surprise that BBs often come out higher, they have bigger damage and better staying power.

And there's no metric for WR, which is a huge factor.

Just my 2-bob.

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles
4 hours ago, S4pp3R said:

IMO all you've done is a 'which ships hurt stuff more' so it's no surprise that BBs often come out higher, they have bigger damage and better staying power.

Well, the score is literally defined as "which ship hurts stuff more" so, yes - I agree with you! :fish_happy:

And as you noticed, hard hitting BBs and some CVs tend to run on the high side as a result. In addition to not counting "paid work" like capping and spotting, there is also the "unpaid jobs" like protecting other ships and other defensive tasks. Boats which excel at support roles at the expense of offense are going to score low, what if any mitigating strengths the ship might bring to a fight remaining TBD. Or to put it another way, if a ship scores exceptionally low you'd want to look carefully to see what features it might have in return for giving up that offense. 

Share this post


Link to post
Share on other sites
231
[151ST]
Member
1,017 posts
4,530 battles
53 minutes ago, Rina_Pon said:

Well, the score is literally defined as "which ship hurts stuff more" so, yes - I agree with you! :fish_happy:

And as you noticed, hard hitting BBs and some CVs tend to run on the high side as a result. In addition to not counting "paid work" like capping and spotting, there is also the "unpaid jobs" like protecting other ships and other defensive tasks. Boats which excel at support roles at the expense of offense are going to score low, what if any mitigating strengths the ship might bring to a fight remaining TBD. Or to put it another way, if a ship scores exceptionally low you'd want to look carefully to see what features it might have in return for giving up that offense. 

Don't get me wrong, I applaud the effort involved and I enjoyed the read.

While we all know some ships are favoured by better players, perhaps a metric to include win rate would balance the results more?

Share this post


Link to post
Share on other sites
Member
325 posts
4,688 battles
1 hour ago, S4pp3R said:

perhaps a metric to include win rate would balance the results more?

There is a strong correlation between the OP rating and the WR. Yes, there is quite a bit of scatter, but ships which score low OP are indeed less likely to win.

As you predicted though, the outliers tend to be destroyers: Sims, Loyang, (Indianapolis), Kiev, and Hatsuharu all have exceptionally high win rates compared to their low OP rating.

So I'd need to work out how to bring WR into the rating in such a way to boost those ships up without over-emphasizing it. It's kinda arbitrary, but what the heck, we've come this far...

These are the revised ranking, according to an OP+ formula which factors in WR to a modest extent . Most ships changed less than 20 places in the overall ranking of 200+ ships as a result of the change.

 

Tier Ship Name OP+ Score
X Hakuryu 1.9
X Khabarovsk 1.2
X Midway 1.1
X Conqueror 1.0
X Des Moines 1.0
X Minotaur 0.9
X Z-52 0.9
X Großer Kurfürst 0.9
X Yamato 0.8
X Zao 0.8
X Montana 0.8
X Moskva 0.8
X Henri IV 0.7
X Hindenburg 0.7
X Gearing 0.6
X Shimakaze 0.4
IX Taiho 1.7
IX Missouri 1.3
IX Z-46 1.0
IX Udaloi 0.8
IX Essex 0.8
IX Fletcher 0.8
IX Saint-Louis 0.7
IX Neptune 0.7
IX Dmitri Donskoi 0.5
IX Friedrich der Große 0.5
IX Roon 0.4
IX Iowa 0.4
IX Yūgumo 0.3
IX Ibuki 0.3
IX Izumo 0.2
IX Baltimore 0.2
IX Kagero 0.1
IX Tashkent 0.0
VIII Shokaku 1.3
VIII Alabama 1.0
VIII Mikhail Kutuzov 0.8
VIII Amagi 0.8
VIII Monarch 0.8
VIII Bismarck 0.8
VIII Charles Martel 0.7
VIII Akizuki 0.7
VIII Tirpitz 0.6
VIII Z-23 0.6
VIII Benson 0.6
VIII Edinburgh 0.5
VIII Atago 0.5
VIII North Carolina 0.5
VIII Loyang 0.4
VIII Lexington 0.4
VIII Chapayev 0.4
VIII Mogami 0.3
VIII Kiev 0.3
VIII ARP Takao 0.3
VIII HSF Harekaze 0.3
VIII Kagero 0.2
VIII Tashkent 0.2
VIII Prinz Eugen 0.0
VIII Admiral Hipper 0.0
VIII Ognevoi -0.1
VIII New Orleans -0.2
VII Kaga 2.4
VII Saipan 1.7
VII Belfast 1.6
VII Scharnhorst 1.2
VII King George V 1.2
VII Fiji 1.0
VII Hood 1.0
VII Shiratsuyu 1.0
VII Hiryu 1.0
VII Leningrad 0.9
VII Błyskawica 0.9
VII Leberecht Maass 0.7
VII Gneisenau 0.7
VII Shchors 0.5
VII ARP Myoko 0.5
VII Nagato 0.5
VII Algérie 0.5
VII Atlanta 0.4
VII Sims 0.4
VII Myoko 0.4
VII Indianapolis 0.3
VII Colorado 0.3
VII Minsk 0.2
VII Ranger 0.1
VII Akatsuki 0.1
VII ARP Nachi 0.1
VII ARP Ashigara 0.1
VII Mahan 0.0
VII Yorck 0.0
VII ARP Haguro 0.0
VII Pensacola -0.1
VI Arizona 1.5
VI Shinonome 1.2
VI Anshan 1.1
VI Perth 1.0
VI Dunkerque 1.0
VI Warspite 1.0
VI Admiral Graf Spee 0.9
VI Molotov 0.9
VI Mutsu 0.8
VI Fuso 0.8
VI Ryujo 0.8
VI Leander 0.7
VI Queen Elizabeth 0.6
VI Budyonny 0.6
VI Duca d'Aosta 0.6
VI New Mexico 0.5
VI Bayern 0.5
VI Ernst Gaede 0.5
VI Cleveland 0.4
VI Hatsuharu 0.3
VI La Galissonnière 0.2
VI Ognevoi 0.2
VI Aoba 0.2
VI Farragut 0.2
VI Nürnberg 0.0
VI Independence 0.0
VI Gnevny -0.1
VI Fubuki -0.2
V Fūjin 1.7
V Kamikaze R 1.7
V Kamikaze 1.7
V Gremyashchy 1.5
V Texas 1.2
V Murmansk 1.2
V Marblehead 1.0
V ARP Kongō 1.0
V Zuiho 1.0
V König 0.8
V Iron Duke 0.8
V ARP Haruna 0.8
V ARP Hiei 0.7
V ARP Kirishima 0.7
V Minekaze 0.7
V Émile Bertin 0.6
V Königsberg 0.4
V Kongo 0.3
V Omaha 0.2
V Kirov 0.2
V New York 0.0
V Nicholas 0.0
V T-22 -0.1
V Podvoisky -0.1
V Furutaka -0.1
V Mutsuki -0.2
V Emerald -0.2
V Bogue -0.3
IV Imperator Nikolai I 2.3
IV Arkansas Beta 1.5
IV Orion 1.3
IV Ishizuchi 1.1
IV Hosho 0.9
IV Yūbari 0.9
IV Kaiser 0.8
IV Duguay-Trouin 0.7
IV Isokaze 0.7
IV V-170 0.7
IV Clemson 0.6
IV Svietlana 0.5
IV Kuma 0.4
IV Danae 0.3
IV Wyoming 0.3
IV Izyaslav 0.2
IV Langley 0.1
IV Phoenix 0.0
IV Myogi -0.2
IV Karlsruhe -0.3
III König Albert 2.0
III Aurora 1.0
III Bogatyr 0.9
III Nassau 0.9
III Campbeltown 0.9
III Bellerophon 0.9
III G-101 0.6
III St. Louis 0.6
III Friant 0.5
III Caledon 0.3
III Wakatake 0.3
III Derzki 0.2
III Kawachi 0.2
III Tenryu 0.1
III Wickes 0.0
III South Carolina 0.0
III Kolberg 0.0
II V-25 1.0
II Umikaze 1.0
II Emden 0.9
II Tachibana 0.8
II Mikasa 0.7
II Albany 0.7
II Dresden 0.6
II Jurien de la Gravière 0.6
II Weymouth 0.2
II Novik 0.2
II Chikuma 0.1
II Storozhevoi 0.1
II Sampson -0.2
II Chester -0.5
I Black Swan 0.7
I Hashidate 0.7
I Orlan 0.7
I Hermelin 0.4
I Erie 0.2
I Bougainville 0.0

 

 

 

WR correlation.png

 

OP plus.png

Edited by Rina_Pon

Share this post


Link to post
Share on other sites
Member
274 posts
23,699 battles
11 hours ago, Rina_Pon said:

It seems a lot of the objections raised here is some variant on "you're holding it wrong". i.e. a given ship is better than the score indicates because a good player can score much better than the average.

Which, when you consider what "average" means, must be true of all ships in all cases. To explore that any further (whether some ships have a higher spread of results than others) requires more detailed breakdown of the statistics than the available data allows, so it's a bit of a dead end.

 

I think this too is an oversimplification. 

The criticism here isn't "good players do really well with it" but rather "there are multiple builds and loadouts for most ships, and certain specific builds/loadouts result in massively better performances than other ships of the same tier while a different build/loadout on the same ship will result in a worse performance than said peers.

I recognize that there are limitations constructed by data available but even within those restrictions, generalizing the model for the whole player base is problematic. Especially when (as pointed out) stratified data for exists (top 50%, top 10%, etc).

Share this post


Link to post
Share on other sites

×