A Random Battle Profile Interpretation

[Aethervox]

I've been reading & 'parsing' a lot of Devstrike forum posts on various subjects & this got me to thinking what, exactly, are the useful 'statistics' one can find in one's own profile?

*** Note: This is a discussion for Random Battle profiles & statistics only

How can you compare your 'stats' to what would be the 'average' or the baseline (for comparison)?

For example, is WR for any ship indicative of anything? What about average XP per battle? What about average Dmg per battle? What about the average # of warships destroyed in a battle? What about the Main Gun hit %? For torpedo armed ships, the average hit % for torpedoes? What is the baseline for the # of battles for any ship? 100? 200? What #?

What are the overall averages for any class & tier of ship? I refer here to the three basic warship classes, Destroyers, Cruisers & Battleships.

For example, what would be, say, the average damage per battle of a T5 Battleship so that anyone could compare a T5 BB of theirs to this average - whether that exceeded this # or not (per the equivalent # of battles).

Which of these profile statistics that WG offers, which of these are more valuable? Which actually mean anything (if at all)?

For example, I will list two ships of mine, both Cruisers on EU server, a T3 & a T8. Are any of my profile stats good? How would these be rated?

1) T3 St Louis 58% WR, 615 av XP/battle, 24,461  av Dmg/battle, 1.52 av Warships destroyed/battle, 33% Main Gun hit %, 303 battles played

2) T8 Hipper 50% WR, 1,006 av XP/battle, 33,409 av Dmg/battle, .5 av Warships destroyed/battle, 31% Main Gun hit %, 5% torp hit, 352 battles played

I have some ships with a better  Main Gun hit % (40%) &, ofc, some with a much less Main Gun hit % (22%). 

What are the good average Dmg per battle for any tier? I note I have a T2 Cruiser that compares well (is almost as good) with a T4 Cruiser.

What is considered average XP/battle for a T6 Cruiser? I have a T6 with 775 XP/battle (129 battles). Is this OK? Bad? Good? What is this, exactly?

Which of these above WG proflie statistics actually means anything or something at all?

 

[Verytis]

51 minutes ago, Aethervox said:

For example, is WR for any ship indicative of anything? What about average XP per battle? What about average Dmg per battle? What about the average # of warships destroyed in a battle? What about the Main Gun hit %? For torpedo armed ships, the average hit % for torpedoes? What is the baseline for the # of battles for any ship? 100? 200? What #?

The goal is to win. WR literally records that. All other statistics are just there to give you an idea of what you're doing towards that end.

53 minutes ago, Aethervox said:

1) T3 St Louis 58% WR, 615 av XP/battle, 24,461  av Dmg/battle, 1.52 av Warships destroyed/battle, 33% Main Gun hit %, 303 battles played

2) T8 Hipper 50% WR, 1,006 av XP/battle, 33,409 av Dmg/battle, .5 av Warships destroyed/battle, 31% Main Gun hit %, 5% torp hit, 352 battles played

Idk anything about St Louis. But your Hipper is not doing enough to even kill a same tier cruiser.

The fact that your Hipper still has a solid accuracy score suggests that:

• you either fire your guns rarely, or you die very fast.
• you keep firing AP against angled targets and thus do no dmg. Because your HE pen is a non-issue.

It is slightly impressive you maintained a 50% WR with that, so you're probably doing something else right.
But we can't tell because you only posted a few select stats instead of the full set.

[Admiral_Karasu]

Hmm... yes, now when you think about it, the ship WR might actually be more useful than your account WR as an indicator. The only question I have, which approach makes more sense, focusing on your lower WR ships or on those that you have the better WR on?

[Itwastuesday]

I'd say the only useful stat is the WR graph. Are you improving, declining or flatlining. It tells you if you're doing something right or not. 

[Captain_Rawhide]

Personally, I think win rate is 'overrated'.  It is virtually impossible to play a random game where 80% of your team has a WR of under 44%.  Winning is much more a matter of luck than it is of skill.  Stick to knowing when you play well vs. when you totally blew it and learn from it.  That will serve you better than chasing a meaningless statistic.  

[pew_pew_magoo]

Using any stats under tier 5 is useless as you only face bots.  
 

The best metric to judge a players skill is solo WR at Tier 10 

[Justin_Simpleton]

4 hours ago, Itwastuesday said:

I'd say the only useful stat is the WR graph. Are you improving, declining or flatlining. It tells you if you're doing something right or not. 

I agree with this.  I seem to be leveling out at 46% win rate in both an old NA account and a relatively new EU account.  

[Ensign Cthulhu]

1 hour ago, Justin_Simpleton said:

I seem to be leveling out at 46% win rate in both an old NA account and a relatively new EU account.  

Why settle for less than 50%?

Get help: https://www.twitch.tv/lord_zath 

[Wowzery]

9 hours ago, Verytis said:

The goal is to win. WR literally records that. All other statistics are just there to give you an idea of what you're doing towards that end.

I will say when I played random exclusively years ago I never looked at W/L.  For me it was all about damage done and (as a battleship main) damage per shell hit.

I did this because WR can be determined by 11 other strangers, there are times you have little to no control over that.  How much damage I do, much more under what I can control and for me a much better gauge of how well I did in battle.

For example, in a T6 battleship I set a goal of 50k of damage.  Above that it was a win for me, below that I didn't do good.  I get under 35k, I sucked.  Doesn't matter if the team won, I sucked.  It was a loss for me.  Damage per shell indicated how well my shots hit.  Get 25 hits and do 10k of damage, in a battleship, I didn't do well with my aim.  Get 50k with 25 hits, that's doing much better.

OP find something which pushes you to get better and use that.  For me WR didn't do it because there was too much randomness I had no control over for a win.  As opposed to how I play, how I aim and shoot directly effects how much damage is done.  As I got better with my aim, who to aim for, etc, my damage goes up. 

[clammboy]

The goal is to learn the basics have fun win and still hope you can keep your sanity.

Edited November 22, 2024 by clammboy

[Frostbow]

4 hours ago, pew_pew_magoo said:

Using any stats under tier 5 is useless as you only face bots.

Yes. Even at Tier V, a player can still face bots. I've seen a WG CM play a Tier V German light cruiser and he had a good time boosting his WR by fighting bots in Random Battle. 

4 hours ago, pew_pew_magoo said:

The best metric to judge a players skill is solo WR at Tier 10

I think the "best" metric should be a combination of WR plus 3 or 4 other stats. Some players are carried to victory despite their very little contribution, while some players do their utmost to carry teams, win or lose, day in and day out.

[pew_pew_magoo]

20 minutes ago, Frostbow said:

think the "best" metric should be a combination of WR plus 3 or 4 other stats. Some players are carried to victory despite their very little contribution, while some players do their utmost to carry teams, win or lose, day in and day out.

The better you play and the more positive battle impact you have, you will win more.  No player is going to carry every team they are on. We all get the same teams with the same team mates. 
At tier 10 you have the most gimmicks to be aware of and the most lethal iterations of the ship lines. If you can play well at tier 10, you understand the game mechanics and how to use them to your advantage. The most skilled players tend to play more tier 10 games than any other tier. You also have some of the worst players there as well.  

[Tricericon]

46 minutes ago, Frostbow said:

I think the "best" metric should be a combination of WR plus 3 or 4 other stats. Some players are carried to victory despite their very little contribution, while some players do their utmost to carry teams, win or lose, day in and day out.

Over a large sample. It should be Win Rate vs some baseline (probably vs an expected win rate for the collection of ships you have sailed; don't really know how you would calculate the latter). Winning, after all, is the goal, and besides, none of the other stats the game records are particularly useful. Even damage doesn't mean much (slapping all 19.2k HP off a Smaland in the first two minutes makes your team a heavy favorite, while burning a GK for the same amount is barely better than nothing).

[pew_pew_magoo]

7 minutes ago, Tricericon said:

Over a large sample. It should be Win Rate vs some baseline (probably vs an expected win rate for the collection of ships you have sailed; don't really know how you would calculate the latter). Winning, after all, is the goal, and besides, none of the other stats the game records are particularly useful. Even damage doesn't mean much (slapping all 19.2k HP off a Smaland in the first two minutes makes your team a heavy favorite, while burning a GK for the same amount is barely better than nothing).

That’s why PR on third party sites is a good metric as well. 
A player with a high WR and PR is a solid player. 

[Tricericon]

2 minutes ago, pew_pew_magoo said:

That’s why PR on third party sites is a good metric as well. 
A player with a high WR and PR is a solid player. 

True, but the same is generally true of WR. Neither is especially precise. My PR at tiers 1-5 is 350 points higher than at tiers 6-10, even though a much higher percentage of low tier games are from my "newbie era". I don't know what goes into PR but it isn't compensating for tier any more than unadjusted win rate.

[Sumseaman]

6 hours ago, pew_pew_magoo said:

The best metric to judge a players skill is solo WR at Tier 10 

Pretty much spot on.

Overall win rate can be skewed by playing powerful ships at lower tiers. Yes there are ships that outperform most others at tier 10 but it is where the most level playing field is found.

I have been a stat pervert on occasion and on doing such seen players with 65% WR overall have very large numbers in ships like Kami, GC, Belfast etc. Hell I went a little berserk in the Enterprise when the CV rework came out before realising that ship/type needed serious balancing.  

Edited November 22, 2024 by Sumseaman

[WES_HoundDog]

All of the satistics mean something.  But eac hone means more or less depending on each ship.    It's how you the player put them together and make the most out of all of them in any particular ship that makes you better or worse.   Win rate is ultimately the product of those figures albeit clouded by rng.     I've seen many players that i out damage or out accuracy in a certain ship.   But they far outpace me in win percentage.    My only explaination is they are better at putting their ship in harms way early in the game when the first team to kill a ship or 2 normally comes out on top.  While i tend to liek to conserve my hps for a important moment late in the battle.  They are better able to leverage their hps  early on to keep their team mates alive, soak up potential damage but live to tell about it.   Some ships are good at this some arn't.  Different stats may be more impactful for different ships even inside the same class.      Your ship WR tells you how effective you are at making your team win games.

And sure if you want to get down dirty on data mining.    you can say win rate after 100 games to lessen rng.  or you can augment those with your overall win rate.   You can also say what is my average tier 10 win rate as someone said, might be best for judging you against other "experienced" players with all the gimmicks knowledge.   And some ships just perform better overall or with you.   Or another thing you can do is pull up www.shipstool  and look at server satistics.  Compare your stats in a ship to what Shipstool has for your ship in each skill bracket to find where you fall and where you might be lacking.

[Utt_Bugglier]

4 hours ago, Tricericon said:

I don't know what goes into PR but it isn't compensating for tier any more than unadjusted win rate.

It is entirely and easily possible to know what goes into PR. (at na.wows-numbers.com) All you have to do is nose around the site a little. Just a little!

But the formula is:

(player ship WR-server ship WR) x (scale factor A) +
(player avg dmg in ship-server avg dmg in ship) x (scale factor B) +
(player avg kills/battle in ship - server avg kills/battle in ship) x (scale factor C)

I can’t remember the scale factors, but if someone is right on those averages for a given ship, the scale factors produce a PR of 1000.

4 hours ago, Tricericon said:

…but it isn't compensating for tier any more than unadjusted win rate.

As you can see from what I said above, PR doesn’t have to compensate for tier because a player’s PR is calculated on a per-ship basis. (At least the wows-numbers.com PR is; I don’t know if any other entity has another number they call PR).

BTW, the PRs that are displayed in the MMM that you can obtain through Aslain’s mod pack are using this formulation.

For that matter, MMM will show, for each player,  both their account WR and ship WR, both their account battles, and battles in that ship. Ship avg frags, ship avg damage. So with that information, you can see the individual components that go into the PR of every player in a match. (except for those who have their statistics hidden)

[Wowzery]

18 hours ago, pew_pew_magoo said:

The most skilled players tend to play more tier 10 games than any other tier.

And when the playstyle of T10 crept down to T6 & 7 it made those tier battles suck as well.  That was about when I stopped playing random.

[Aethervox]

On 11/21/2024 at 11:57 PM, Verytis said:

The fact that your Hipper still has a solid accuracy score suggests that:

• you either fire your guns rarely, or you die very fast.
• you keep firing AP against angled targets and thus do no dmg. Because your HE pen is a non-issue.

It is slightly impressive you maintained a 50% WR with that, so you're probably doing something else right.

I had a look at WoWS numbers EU for my Hipper & I, essentially,  have the Av Warship sunk per battle & very close to Av Dmg. Somehow, my WR is marginally (1%) better than the average, so, yes, I must be doing something right (but probably not much, lol). Oh yes, I do tend to die - I am a more aggressive type player 😁.

[Aethervox]

11 hours ago, Utt_Bugglier said:

if someone is right on those averages for a given ship, the scale factors produce a PR of 1000.

Although my PR is increasing over time recently (a WoWS numbers graph says so), most of my PRs are below average  😒 (except for a few).

[Aethervox]

On 11/21/2024 at 11:57 PM, Verytis said:

Idk anything about St Louis.

I do very well with this ship according to WoWS numbers EU. Some ships are rather easy to play, especially, if one finds the proper way to play it 😁.

[Utt_Bugglier]

1 hour ago, Aethervox said:

Although my PR is increasing over time recently (a WoWS numbers graph says so), most of my PRs are below average  😒 (except for a few).

There is a logical reason for that: if you’re like most players (including me), you stick with a smallish number of ships you’re better at, and leave a larger number of less successful ships with fewer battles played. It takes a dedicated (and fairly good player) to revisit and rehabilitate the numbers on those “less than successful” ships.

[Aethervox]

4 hours ago, Utt_Bugglier said:

if you’re like most players (including me), you stick with a smallish number of ships you’re better at,

Not me, I, basically, random the ship I play in my port these days (although, I have played a few ships more as the #s would support that). I have lists with specific ships so some do get played a bit more. Still, all will get played over time.

[SoshiSone]

In general, overall WR after a 1000 battles is the first metric.  

Then look at specific ships when players get about 300 games under their belt in those ships.

If a player has less than 1000 battles overall and less than 300 in a particular ship, then adjust accordingly.

Also, look at performance over the last 1000 or so games  Vice the totality of games.  Players get better over time..so recent stats (provided they are of a substantial number) are usually more indicative of performance than the totality of stats.