Essex, the (Ess)ence of suck

[urk_the_red]

There was a thread the other day that was pointing out the ships with the lowest win rates. American CVs were consistently at the bottom of the list. This presented an interesting statistical problem for a very boring workday (measuring rheology is absolutely mind numbing). Essex only faces two possible match ups, Essex vs. Essex or Essex vs. Taiho. So I set out to find not Essex's aggregate WR, but Essex's WR against Taiho. 

 

Based on 2 week data from Warships Today, Essex has a WR 47% with 7324 battles and Taiho has a WR of 55% with 4965 battles. Now for all matches TvT or EvE the WR is 50% so any deviance of the WR from 50% is the result of Essex vs Taiho battles only. I'll spare you the math, it was worse than I anticipated (been a while since I played with multivariable systems of equations or probability). But based on an assumption that 60% of the time a Taiho lines up for battle (the relative percentage of Battles fought by Taihos vs battles fought by Essexes) it faces an Essex, the average WR for Essex vs. Taiho is roughly 40%. So only 2/5 battles fought by an Essex against a Taiho is successful, for the Essex. 

 

Disclaimer: I have neither CV, I just thought it was an interesting stat.

[saagri]

There was a thread the other day that was pointing out the ships with the lowest win rates. American CVs were consistently at the bottom of the list. This presented an interesting statistical problem for a very boring workday (measuring rheology is absolutely mind numbing). Essex only faces two possible match ups, Essex vs. Essex or Essex vs. Taiho. So I set out to find not Essex's aggregate WR, but Essex's WR against Taiho. 

 

Based on 2 week data from Warships Today, Essex has a WR 47% with 7324 battles and Taiho has a WR of 55% with 4965 battles. Now for all matches TvT or EvE the WR is 50% so any deviance of the WR from 50% is the result of Essex vs Taiho battles only. I'll spare you the math, it was worse than I anticipated (been a while since I played with multivariable systems of equations or probability). But based on an assumption that 60% of the time a Taiho lines up for battle (the relative percentage of Battles fought by Taihos vs battles fought by Essexes) it faces an Essex, the average WR for Essex vs. Taiho is roughly 40%. So only 2/5 battles fought by an Essex against a Taiho is successful, for the Essex. 

 

Disclaimer: I have neither CV, I just thought it was an interesting stat.

 

Sounds about right. 

[Wo_9]

Meh i personally have a 55% winrate in Essex, but i am still pretty tilted how the stats are so low compared to its peers.  

 

overnerfed unjustly by biased hate. 

[urk_the_red]

Considering that it's only one out of 12 ships, that's a horrifyingly low WR. You could probably run 12 vs. 11 matches (without CVs) and still wind up with a narrower W/L than Essex vs Taiho.

[urk_the_red]

Meh i personally have a 55% winrate in Essex, but i am still pretty tilted how the stats are so low compared to its peers.  

 

overnerfed unjustly by biased hate. 

Congrats on finding a way to make her work!

[Submarine_Wahoo]

Meh i personally have a 55% winrate in Essex, but i am still pretty tilted how the stats are so low compared to its peers.  

 

overnerfed unjustly by biased hate. 

 

I'm too put off by how strong Taiho is to move up from Lexington to Essex. I'll stick with outplaying Shokaku for now.

[RogueFlameHaze]

Upgraded from lexy but dont really feel like touching essex based on what the cv balance sounds like at top tier

[Edgecase]

 

Meh i personally have a 55% winrate in Essex, but i am still pretty tilted how the stats are so low compared to its peers. 

 

Congrats on finding a way to make her work!

 

Not being a jerk here, just giving an obligatory statistical caution: global and personal WR stats for carriers are not directly comparable, because a bunch of those wins were against other Essexes. They would have canceled out of the global stat, but not Wo_9's personal win rate for the ship.

 

That said, I agree that maintaining a positive WR in USN CVs just seems like a moral victory, no matter what.

[urk_the_red]

 

 

Not being a jerk here, just giving an obligatory statistical caution: global and personal WR stats for carriers are not directly comparable, because a bunch of those wins were against other Essexes. They would have canceled out of the global stat, but not Wo_9's personal win rate for the ship.

 

That said, I agree that maintaining a positive WR in USN CVs just seems like a moral victory, no matter what.

Fair point. It would be interesting to see how well the best Essex players do vs. Taiho.

[ckupf]

Honestly surprised it's that high. When your team has a Lexington or an Essex and you face off against a Shokaku or a Taiho you are at a huge disadvantage. At T8 and T9 the CVs are not unbalanced they are mismatched. You will lose the air battle barring a huge difference in player quality. It's that simple. Winning takes an extraordinary effort if two CVs are even quality and quickly becomes impossible if the IJN CV player is better (which is common as better players are drawn towards more powerful lines).

[Wo_9]

Fair point. It would be interesting to see how well the best Essex players do vs. Taiho.

 

It depends on how skilled the player is.  if its a unicum i have to work my [edited]just to keep up.  if its a super-unicum CV player...well he will have a major advantage over me because of the IJN loadouts.  also if enemy team is using teamwork with an IJN CV on there team, its pretty damn one sided. 

[Ujenele]

OP - I love you

 

Will you do the same this for the Midway?

please

[urk_the_red]

Honestly surprised it's that high. When your team has a Lexington or an Essex and you face off against a Shokaku or a Taiho you are at a huge disadvantage. At T8 and T9 the CVs are not unbalanced they are mismatched. You will lose the air battle barring a huge difference in player quality. It's that simple. Winning takes an extraordinary effort if two CVs are even quality and quickly becomes impossible if the IJN CV player is better (which is common as better players are drawn towards more powerful lines).

It may very well be worse than what I quoted. The assumption that Taiho's face Essex's in 60% of their matches is just an assumption. I'll have to recheck my spreadsheet but if Taiho faces nothing but Essex's it has a 55% WR against Essex by definition, on the other end, the more TVT battles there are, the lower Essex's WR vs. Taiho becomes. All this is stated in terms of Taiho's battles because there were fewer of them, so Taiho is the limiting factor. Theoretically, Taiho's WR could be as high as 100% vs. Essex. But that's a theoretical limit, not a practical assumption. The numbers based off of 60% were the best assumption.

[vak_]

I do okay in Essex (118k average / 72% WR damage after 29 battles). But it is a bit of an underdog vs Taiho, no denying then.

 

Then again, what's new? Arguably USN tech tree CVs are worse than IJN ones, with the sole exception of Langley.

 

PS What's curious, on the RU server Essex is doing a lot better than on NA (though it's still behind Taiho there as well)

[urk_the_red]

OP - I love you

 

Will you do the same this for the Midway?

please

Hahaha thanks, now that I have the spreadsheet, it ought to be fairly easy. Maybe I'll put some of the other CVs up over the next few days. 

[ckupf]

 

It depends on how skilled the player is.  if its a unicum i have to work my [edited]just to keep up.  if its a super-unicum CV player...well he will have a major advantage over me because of the IJN loadouts.  also if enemy team is using teamwork with an IJN CV on there team, its pretty damn one sided. 

 

Fair point. It would be interesting to see how well the best Essex players do vs. Taiho.

 

FWIW there are 14 qualifying (100 battles or more) Essex players with winrates above 70% and 0 with 80%. There are 50 qualifying Taiho players with 70% winrates and 8 with at least 80%.

 

You can't make the same assumptions with this group so it's harder to extrapolate.

 

Then again, what's new? Arguably USN tech tree CVs are worse than IJN ones, with the sole exception of Langley.

 

 

It's not arguable and the Langley is not an exception.

[Herr_Reitz]

There was a thread the other day that was pointing out the ships with the lowest win rates. American CVs were consistently at the bottom of the list. This presented an interesting statistical problem for a very boring workday (measuring rheology is absolutely mind numbing). Essex only faces two possible match ups, Essex vs. Essex or Essex vs. Taiho. So I set out to find not Essex's aggregate WR, but Essex's WR against Taiho. 

 

Based on 2 week data from Warships Today, Essex has a WR 47% with 7324 battles and Taiho has a WR of 55% with 4965 battles. Now for all matches TvT or EvE the WR is 50% so any deviance of the WR from 50% is the result of Essex vs Taiho battles only. I'll spare you the math, it was worse than I anticipated (been a while since I played with multivariable systems of equations or probability). But based on an assumption that 60% of the time a Taiho lines up for battle (the relative percentage of Battles fought by Taihos vs battles fought by Essexes) it faces an Essex, the average WR for Essex vs. Taiho is roughly 40%. So only 2/5 battles fought by an Essex against a Taiho is successful, for the Essex. 

 

Disclaimer: I have neither CV, I just thought it was an interesting stat.

 

Maybe it's been said... but you know the WR is gonna be based - as will ALL stats - on the players who play the ship in the time frame you are discussing. Maybe all the good players gave it up for a bit and moved on to something else? Maybe the good players quit playing her completely? Maybe... you see where I'm going, right? 

 

Stats are meaningless without context. You would need to dig deeper, analyze all the players tracked playing the ship to see what their stats were for all the other ships they played. Then you might be able to suggest something but I'm not exactly sure it would be legitimate. You know? 

[cometguy]

Fair point. It would be interesting to see how well the best Essex players do vs. Taiho.

I think the more interesting comparison is how a player performs in each ship. The mechanics aren't very different between the lines and it eliminates player skill, since it's the same player.

 

For example, while I'm not yet at t9, my win rate in my Ranger is 51.19% and my win rate in my Hiryu is 63.24%

[ckupf]

 

Maybe it's been said... but you know the WR is gonna be based - as will ALL stats - on the players who play the ship in the time frame you are discussing. Maybe all the good players gave it up for a bit and moved on to something else? Maybe the good players quit playing her completely? Maybe... you see where I'm going, right? 

 

Stats are meaningless without context. You would need to dig deeper, analyze all the players tracked playing the ship to see what their stats were for all the other ships they played. Then you might be able to suggest something but I'm not exactly sure it would be legitimate. You know? 

 

WRs may become inflated or deflated, but it is the best determination we get without being able to regress on player quality. It should always point in the right direction at least.

[urk_the_red]

 

Maybe it's been said... but you know the WR is gonna be based - as will ALL stats - on the players who play the ship in the time frame you are discussing. Maybe all the good players gave it up for a bit and moved on to something else? Maybe the good players quit playing her completely? Maybe... you see where I'm going, right? 

 

Stats are meaningless without context. You would need to dig deeper, analyze all the players tracked playing the ship to see what their stats were for all the other ships they played. Then you might be able to suggest something but I'm not exactly sure it would be legitimate. You know? 

There's some truth to that. I considered taking all the stats, but there have been too many huge changes to CV game mechanics for all stats to be valid for the current state of the Essex. Preferably, I would have used the stats for all games since the last change to CVs, but I didn't have that option. There isn't really a right answer if you're looking for a stand in for pure combat power. However, the two week stats are valid for how Essex is CURRENTLY performing.

[TenguBlade]

Meh i personally have a 55% winrate in Essex, but i am still pretty tilted how the stats are so low compared to its peers.  

 

overnerfed unjustly by biased hate. 

Because being cross-dropped by 2 6-plane torpedo squads, with a torpedo spread tight enough to trap DDs even with DF on, and arming time, damage, flood chance, and speed superior to that of the IJN torpedoes, is clearly reasonable.  IJN anvil-and-hammer has a better chance to be foiled or mitigated because AA mechanics favor large plane squads.

 

Not to mention that this was going on when it was common practice for CV players to just martyr their planes for the sake of launching another wave faster than having them fly back and rearm.  I don't like how USN CVs are right now either, but the double torpedo bomber squad was stupid beyond belief.  Those 12 planes alone could sink any ship besides a Yamato or Des Moines.  Both of those would just require you to bring in the dive bombers to finish them off.

[saagri]

Because being cross-dropped by 2 6-plane torpedo squads, with a torpedo spread tight enough to trap DDs even with DF on, and arming time, damage, flood chance, and speed superior to that of the IJN torpedoes, is clearly reasonable.  IJN anvil-and-hammer has a better chance to be foiled or mitigated because AA mechanics favor large plane squads.

 

Not to mention that this was in the heyday of Midway and Essex players just martyring their planes to launch another wave faster.  I don't like how USN CVs are right now either, but the double torpedo bomber squad was stupid beyond belief.  Those 12 planes alone could sink any battleship besides a Yamato.

 

And yet their damage output was still on part with their peers. 

[vak_]

It's not arguable and the Langley is not an exception.

 

Langley is about as good as Hosho. Sure, she's got a win rate that's two percentage lower than Hosho's in the past two weeks, but that isn't a huge difference.

[saagri]

 

Langley is about as good as Hosho. Sure, she's got a win rate that's two percentage lower than Hosho's in the past two weeks, but that isn't a huge difference.

 

Hosho also does way more damage and a slightly worse plane kill rate. 

[Wo_9]

 

And yet their damage output was still on part with their peers. 

 

This^  

 

now take that away, and you shouldn't be surprised why Taihou/Hak have stats the way they do now.