Jump to content
You need to play a total of 20 battles to post in this section.
Moggytwo

What is the most powerful class at winning the battle for your team?

119 comments in this topic

Recommended Posts

1,412
Members
663 posts
20 battles

I thought it might be nice for us all to have a look at some of the latest data on the relative strengths of each class, after it was discussed a little bit in other threads.  What we're trying to do here is assess the relative power of each class when played at the absolute limit of what the class is capable of.  Fortunately we have publicly available data sets that give us this info:

SEA serverNA serverEU server.

These data sets are the top 100 players of each class, over the last two months, and cover 400 players and tens of thousands of battles per server!  That's a whole lot of data to give us a fairly accurate picture of the relative power of each class on each server (each of which have different metas and class population percentages).  Let's have a look at the averages of this data: 

   SEA
                 Win    Damage
DD          76.1%    61.0k
CA/CL   73.3%    105.0k
BB          72.2%    130.2k
CV          72.7%    115.9k


   NA
                Win    Damage
DD          75.7%    74.2k
CA/CL   71.9%    113.7k
BB          69.6%    131.5k
CV          70.1%    109.8k


   EU
                Win    Damage
DD          75.5%    72.7k
CA/CL    74.1%    117.3k
BB          73.1%    132.9k
CV          72.1%    111.0k

So what can we see from these results?  CV's are second on average damage on SEA, and third on both NA and EU.  CV's are just ahead of BB's in win rate in third position on both SEA and NA, and a distant last in win rate on EU.  DD's are an absolute mile ahead on win rate on every server.  BB's are first by a significant margin for average damage on each server, but this extra damage vs other classes does not translate into being more likely to win the battle - in fact BB's are at the bottom of effect on the win alongside CV's despite their superb damage.

So what conclusions can be drawn from these results?  CV's and BB's quite clearly have comparatively limited battle influence, especially when compared to DD's, who are by far the most powerful class in the game when it comes to influencing the win as a good player.  Despite being difficult to play for poor to average players, a well played DD absolutely dominates.  We can also draw the conclusion that damage does not really have the greatest impact on winning the battle, it is clearly more all those other aspects of game play that DD's are so brilliant at (capping, spotting, screening, counter DD work for example) that really wins battles for your team.

Overall balance does seem to be pretty good, there is no real outlier, except perhaps the DD win rate is a little high.  It would be tough to nerf DD's based on this though, since they do generally have it harder than other classes, and it would be hard on struggling poor to average DD players to nerf their class because the best DD players are overperforming.

It does make for interesting reading though!  A few facts are always good for improving our analysis of the game.

  • Cool 10
  • Thanks 1
  • Boring 2

Share this post


Link to post
Share on other sites
1,625
[-DF-]
Beta Testers
3,497 posts
6,798 battles

Stats are as stats do.

Deviation from the mean is the issue here and from that perspective they are all basically in the same general bar park.  What that tells you is that skilled players are able to do just about as well in any of the classes.

DDs take caps.  DDs in the end game are exceptionally powerful because they take caps. BBs and CVs are the worst at taking caps.  That's where the few extra percent comes in. 

Beyond that, I think you have to read in more than the stats are actually telling you.

  • Cool 4
  • Boring 2

Share this post


Link to post
Share on other sites
1,114
[-TRM-]
[-TRM-]
Members
3,421 posts

Little Anklebiter DD can be run off a cap by 4 ships coming to take it all.

I just had a coop where a 620 point hull Yorck chased a half health DD across the map. I sunk the yorck using the range distance to the ship and laid the guns at about that range using the counter next to the center. The rounds had to go over two ridges and reach the Yorck.

I will never understand why DD ran. A few taps would have remove the yorck.

  • Meh 1

Share this post


Link to post
Share on other sites
Guest
Guest
0 posts

I was having this discussion the other day with a clan member about DDs vs CVs. DDs absolutely have the highest impact on the match outcome but that comes with a caveat of if they don't get taken out by another class. Not sure where this land philosophically but if another class like a CV takes out a DD then the balance of power gets shifted. For that game the class that takes out another high impact class becomes the highest impact.

So it makes sense for everyone to target the DDs first if you want to disrupt the balance of power. German CVs should then have less impact on game outcome than their counter parts. I wonder where their win rate will settle.

Share this post


Link to post
Share on other sites
1,625
[-DF-]
Beta Testers
3,497 posts
6,798 battles

Let's look at it from a game design perspective:

The game is capture the flag.  DDs have been designed to capture the flag.  That's why you need your CAs around to stop the red DD from capturing the flag.  And that plays into why your DD wants your BBs around to blap the enemy CA.  So the DD can capture the flag and not get blapped themselves.  

So, the beauty of the game design is that all three classes are absolutely integral to capturing the flag.  But, it's designed for the DDs to do much of the capturing.  Or perhaps, DDs have been designed to be the best capturers.  And so that necessarily puts them ahead of the other three classes.

But, given the beautifully intricate rock paper scissors class design.  All three classes are absolutely vital.

  • Cool 3
  • Boring 2

Share this post


Link to post
Share on other sites
2,837
[SIDE]
Members
4,939 posts

Destroyers. Hands down. No doubt. Good ones are patient, able to overcome, able to outwit, think strategically and ultimately win games by capitalizing on the ships strengths. Bad ones die in first 3 minutes.

Edited by thebigblue
  • Cool 7

Share this post


Link to post
Share on other sites
7,363
[WORX]
Members
13,291 posts
20,135 battles

I object to the conditions/parameters to measure/test the original theory to which is the powerful class.

If you only have 2 parameters which dont quantify the definition of "power." Therefor, to conclude with this

28 minutes ago, Moggytwo said:

So what conclusions can be drawn from these results?  CV's and BB's quite clearly have comparatively limited battle influence, especially when compared to DD's, who are by far the most powerful class in the game when it comes to influencing the win as a good player.  Despite being difficult to play for poor to average players, a well played DD absolutely dominates. Not even close, You first have to quantify/define, what is determine what is power. Having only 2 parameters, makes your conclusion to have a high deviation of error. In other words, its an opinion based on swampy numbers, not on concrete mathematical foundation..

Furthermore I object to this statement as well

33 minutes ago, Moggytwo said:

 We can also draw the conclusion that damage does not really have the greatest impact on winning the battle, it is clearly more all those other aspects of game play that DD's are so brilliant at (capping, spotting, screening, counter DD work for example) that really wins battles for your team.

  • Damage farmers
  • Those who play the objective

These two philosophy, are a never ending dual for who is the top tactic to use in WOWS.I would conclude use both not one.

As much as you can have an opinion... The math practice you use to support your claims, I pretty much object to...

 

  • Cool 2
  • Boring 2
  • Meh 2

Share this post


Link to post
Share on other sites
3,914
[PVE]
Members
7,342 posts
39 minutes ago, Moggytwo said:

DD's are an absolute mile ahead on win rate on every server.

Do you really think there is any real practical difference between a super unicum DD player winning 75 out of a 100 games to the BB super unicum player only winning 72?

  • Boring 1

Share this post


Link to post
Share on other sites
1,009
[TIMT]
Members
1,736 posts
5,623 battles
54 minutes ago, Moggytwo said:

These data sets are the top 100 players of each class, over the last two months, and cover 400 players and tens of thousands of battles per server!  That's a whole lot of data to give us a fairly accurate picture of the relative power of each class on each server (each of which have different metas and class population percentages).  Let's have a look at the averages of this data: 

Interesting, though with this data you would need to adjust your question/hypothesis to: Which class can unicums carry in most?

Don't get me wrong, I like the data-driven aspect this brings to the old discussion. Important distinction though as you only look at the top players; I would like to see the data for the bottom players and the overall average to see where this is going. However, WR and damage won't help you with these as WR naturally goes to 50% for every class since the teams are symmetric in composition. While I tend to agree that DDs can win you the game (and bad DDs can loose (a single or double 'o', always confuse this) it for you, I don't think the data is suitable to definitively give an answer.

For the overall stats we would need to find a metric that can be measured, and most likely such a metric would require looking at individual battles instead of player/global averages. One indicator could be the WR/damage distribution for ships/classes to show the min-max of what players can do in it, and how 'consistent' the average person does in them. Consistent meaning not how a player does from game to game, but how large the variance in average performance is.

Interesting to see the differences between servers though.

Share this post


Link to post
Share on other sites
1,256
[WIB]
Alpha Tester
3,978 posts
2,472 battles

I think DDs  can "potentially" have the biggest impact on the outcome of a game.  The have the concealment to get out front and spot.  The speed to get the caps or change flanks when needed.  They can also singlehandedly hold up the push of an entire enemy flank at times.  The mere suspicion of their presence in the vicinity of the enemy tends to force them to maneuver in ways that might not be beneficial or optimal.   They also can delete some of the most powerful ships in the game within a split second with torpedoes.  

On the other hand, they can also be the most inconsequential ship in the match if and more likely, when they get taken out early.   Their lone wolf style tends to put them at massive risk of this most of the time as well.   

Overall though, because there generally is only these two outcomes with a destroyer, I think it balances them out to having no more influence over the match then any other class.  In fact you could almost say that DDs are the RNG of ships.  Sometimes RNG is with you and others it isn't.

As far as the rest of the classes, I don't even really think any of the others, including CVs have the least bit more impact than the other.  I will go as far as saying none of them stand out as having even more "Potential" over the other to impact the game.  There might be some question about CV's being able to impact the game especially when it comes down to a last minute resetting a cap attempt or the game being down to a single BB vs CV with the BB being unfortunately located on the other side of the map from the CV but I don't see that happening nearly as much as I see the lone DD pulling out a crazy win.   

Share this post


Link to post
Share on other sites
1,412
Members
663 posts
20 battles
25 minutes ago, Navalpride33 said:

I object to the conditions/parameters to measure/test the original theory to which is the powerful class.

If you only have 2 parameters which dont quantify the definition of "power." Therefor, to conclude with this

It's right there in the thread title: "What is the most powerful class at winning the battle for your team?"

This is about the win.  The damage data is more for interest.  The entire point of the game is to win, every other statistic merely serves to provide background data to that one stat.

Therefore the thread has more than the data it needs.  If you want to analyse more of the available data to further illustrate why the win rates are the way they are, the data set also gives XP results, survival rates, kills per battle, capping data, spotting damage, hit ratio etc that you are more than welcome to go through.  That is all very interesting, but the only actual important stat is the win rate.

I also don't want to muddy the waters too much, and take away from the important data.  If I gave all those other data points, some players may get a bit confused as to which data actually mattered.  Win rate is what matters.  Hence the thread title, and the whole point of the thread.

  • Cool 1

Share this post


Link to post
Share on other sites
Members
609 posts
17,654 battles

It’d be better to filter this down to solos I think (if possible). The top 100 players are almost certainly Unicum divs, which likely colors your results - a dd’s win rate will be a _lot_ higher with coordinated support for example.

Edited by Telastyn

Share this post


Link to post
Share on other sites
1,625
[-DF-]
Beta Testers
3,497 posts
6,798 battles
Just now, Midnitewolf said:

There might be some question about CV's being able to impact the game especially when it comes down to a last minute resetting a cap attempt

DD/CA/BB all have a positive and negative impact on the capping rock paper scissors game.  CVs mostly just have a negative impact on that capping game as they really don't have a "positive" role to play in that game.

And so, since CVs are basically able to keep up with surface game classes (at the high level WR wise), that essentially means that their negative impact is right around the total of the positive and negative impacts that the capping classes bring to the table.

This is why they are perceived by their targets mostly negatively.  Because they aren't up their on the front lines trying to win the game.  The players on the front lines trying to win the game mostly feel the impact of the red CV.

But, you are 100% right that a CVs influence scales geometrically the farther the game gets in and the AA is mostly dead and the few ships left are on low hp.  I'd say they are on average more influential than a DD in that situation.

  • Cool 1
  • Boring 2
  • Meh 1

Share this post


Link to post
Share on other sites
1,256
[WIB]
Alpha Tester
3,978 posts
2,472 battles
12 minutes ago, shinytrashcan said:

Interesting, though with this data you would need to adjust your question/hypothesis to: Which class can unicums carry in most?

Don't get me wrong, I like the data-driven aspect this brings to the old discussion. Important distinction though as you only look at the top players; I would like to see the data for the bottom players and the overall average to see where this is going. However, WR and damage won't help you with these as WR naturally goes to 50% for every class since the teams are symmetric in composition. While I tend to agree that DDs can win you the game (and bad DDs can loose (a single or double 'o', always confuse this) it for you, I don't think the data is suitable to definitively give an answer.

For the overall stats we would need to find a metric that can be measured, and most likely such a metric would require looking at individual battles instead of player/global averages. One indicator could be the WR/damage distribution for ships/classes to show the min-max of what players can do in it, and how 'consistent' the average person does in them. Consistent meaning not how a player does from game to game, but how large the variance in average performance is.

Interesting to see the differences between servers though.

Personally I think a Unicum can carry just as easily in any class and that class really has no bearing what so ever.   Unicums tend to have such a dramatic impact on the match that the advantage they create becomes overwhelming and snowballs very quickly.  Doesn't matter if they are in a BB, CA, CL, DD or CV, that impact will always be there.  

However, if your wanting solid data then you really need to throw out the data involving Unicums.  Unicums are outliers and will tend to corrupt the that data.  It is kind of like a situation where in school and the teacher grades on a curve.   The curve works great as long as everyone's test scores fall into line with expectations, i.e. a 70-100 score on the test.  But when you have a couple outliers answering the extra credit bonus questions and ending the year with a 120 average, all of a sudden, a 80 score, which would normally be a B average,  can turn into a 60 which is a failing score and you find 80% of the class failing.  Honestly it is NEVER a good idea to look at the Unicums for any sort of statistical analysis and even worse to use their performance for balancing purposes.   They make up a small, small amount of the players, consisting of the 3-6% of player that are actually good enough to actually influence their own win rate.   The vast, vast majority of players can't do this even with years and years of practice and will never pull this much performance out of a ship so what is OP in the hands of a Unicum, may very well be weak in the hands of 95% of the rest of the players. 

Share this post


Link to post
Share on other sites
7,363
[WORX]
Members
13,291 posts
20,135 battles
13 minutes ago, Moggytwo said:

It's right there in the thread title: "What is the most powerful class at winning the battle for your team?"

This is about the win.  The damage data is more for interest.  The entire point of the game is to win, every other statistic merely serves to provide background data to that one stat.

Therefore the thread has more than the data it needs.  If you want to analyse more of the available data to further illustrate why the win rates are the way they are, the data set also gives XP results, survival rates, kills per battle, capping data, spotting damage, hit ratio etc that you are more than welcome to go through.  That is all very interesting, but the only actual important stat is the win rate.

I also don't want to muddy the waters too much, and take away from the important data.  If I gave all those other data points, some players may get a bit confused as to which data actually mattered.  Win rate is what matters.  Hence the thread title, and the whole point of the thread.

IMO you already muddied the water :Smile_veryhappy:. .

I am not in disagreement with the title of the thread..

I object to the math used to come to the conclusion you stated. ( I pointed out the discrepancies in post #7). If the math or the process is wrong, its conclusion will be discarded because. It has a high deviation based on the numbers and method used..

 

 

Edited by Navalpride33

Share this post


Link to post
Share on other sites
1,412
Members
663 posts
20 battles
16 minutes ago, shinytrashcan said:

Don't get me wrong, I like the data-driven aspect this brings to the old discussion. Important distinction though as you only look at the top players; I would like to see the data for the bottom players and the overall average to see where this is going. However, WR and damage won't help you with these as WR naturally goes to 50% for every class since the teams are symmetric in composition. While I tend to agree that DDs can win you the game (and bad DDs can loose (a single or double 'o', always confuse this) it for you, I don't think the data is suitable to definitively give an answer.

This is an interesting point.  Firstly the average is actual about 48-49% rather than 50% - that seems counter-intuitive, but we are looking at each players win rate, and there are a lot more players at the bottom end that have less battles, while the more successful players obviously tend to have more battles.  This is why the average win rate is actually well below 50%.  My point there isn't super relevant to the overall discussion, I just think it's an interesting thing to note.

As for the data for the bottom players, we do have some, but not as a direct comparison to this data set since it is done by ship rather than player. 

Per battle though (rather than per player), for every player in the above data set who is getting a win, there is a corresponding player of the same class who is getting a loss.  This means that you can draw a bit of a mirror to the figures, and effectively say that the ships with the most influence on the win in the stats above (so DD's in this case), have a correspondingly just as significant influence in the loss when played badly.

If you look at the results this makes sense.  DD's have a massive influence on the win partially because of how easy it is to die early in them - a good player will survive to the end regularly, while a bad DD player will die early more often, and thus have zero impact on the battle from then on.  A good DD player can also have a more direct and rapid influence in the death of their opposing DD as well.  The inverse goes for BB's and CV's - these are the two most survivable classes, and will last longer when played by a poor player than DD's and cruisers, thus a poor player is able to have more influence on the win by simply being alive a bit longer than a similarly skilled DD player, thus bringing the win influence of the best players of those classes down as their mirror on the other team is more likely to be alive.

Share this post


Link to post
Share on other sites
1,412
Members
663 posts
20 battles
19 minutes ago, Telastyn said:

It’d be better to filter this down to solos I think (if possible). The top 100 players are almost certainly Unicum divs, which likely colors your results - a dd’s win rate will be a _lot_ higher with coordinated support for example. 

Although divs very much influence the win rate of the best players, they do not have any particular influence when comparing between classes.  The reason for this, is that every player on that list is just as likely to be in a div as the next player, regardless of class.  Thus the influence of divs on the results is nullified.

If for some reason BB's on this list rarely divved, but DD's divved all the time, then the data set would be skewed, but there is no reason that there would be different div rates of the best players between classes.

Share this post


Link to post
Share on other sites
1,625
[-DF-]
Beta Testers
3,497 posts
6,798 battles
Just now, Moggytwo said:

Although divs very much influence the win rate of the best players, they do not have any particular influence when comparing between classes.  The reason for this, is that every player on that list is just as likely to be in a div as the next player, regardless of class.  Thus the influence of divs on the results is nullified.

If for some reason BB's on this list rarely divved, but DD's divved all the time, then the data set would be skewed, but there is no reason that there would be different div rates of the best players between classes.

Divs bring all sorts of comparative issues to the table.  Divs can be tailor made to complement each other on a "random" team (particularly at that level) and should out perform non-div players (at that level).

So one could conclude that since solo BBs can keep up with tailor made unicum divisions, perhaps BBs have an outsized impact on a per class per player basis.

But again, it would really depend on the makeup of the data.  I'm assuming very few BB divs and most DD divs.  And why it's a good reason to compare solo to solo and div to div.

Share this post


Link to post
Share on other sites
1,412
Members
663 posts
20 battles
3 minutes ago, CommodoreKang said:

But again, it would really depend on the makeup of the data.  I'm assuming very few BB divs and most DD divs.  And why it's a good reason to compare solo to solo and div to div.

But why would there be different divisioning rates when comparing per class?  There's no reason to consider that this is likely.

Don't get me wrong, I'd love to compare solo to solo and div to div, but we simply do not have that data available to us.  For the data we do have, there is no reason to expect that divs will be skewing the class data, as div rates are likely to be very similar across all classes at this level of play.  Therefore, we don't have to consider this as an issue that has any meaningful effect on the data set, and thus can safely ignore it.

Share this post


Link to post
Share on other sites
1,625
[-DF-]
Beta Testers
3,497 posts
6,798 battles
9 minutes ago, Moggytwo said:

But why would there be different divisioning rates when comparing per class?  There's no reason to consider that this is likely.

Don't get me wrong, I'd love to compare solo to solo and div to div, but we simply do not have that data available to us.  For the data we do have, there is no reason to expect that divs will be skewing the class data, as div rates are likely to be very similar across all classes at this level of play.  Therefore, we don't have to consider this as an issue that has any meaningful effect on the data set, and thus can safely ignore it.

Because in a div you can ensure that you have 2 or 3 players with a good mix of utility and (at this level) voice comms, one would expect them to have a much higher WR.  Typically, my theory is that 60-65% is about the max you can get to as a solo player on a team of 12 (over the long haul).

Anything above that, I generally consider either stat padding, a selection bias streak (we all have those streaks where we get those big numbers).  Or div'ing. 

I'm kind of shocked that the non div players were able to put up those numbers.  I wonder how much "streak" bias you're picking up in that data.  In other words, if you filter out by a certain win rate over a certain rather short period, you could pick up players with a dozen games and a 78% WR but they are actually 50ish% players (over the long haul).

Edited by CommodoreKang

Share this post


Link to post
Share on other sites
393
[D33P6]
Members
572 posts

You're probably going to get a lot of hate and/or random excuses as to why the data you provided isn't accurate or doesn't paint the real picture.  It goes against their belief that CV's are the most OP class out there.

My bet is that they'll go with the "these stats go against what I always say so they mean nothing now"

2 hours ago, Moggytwo said:

These data sets are the top 100 players of each class, over the last two months

I'm curious what it would say if you were to do the same for the bottom 100 players and one for the mid-level, average player.  You don't have to do this but I am curious.

I suspect the bottom/lower players will have a higher win rate in CV's and BB's than compared to the other classes.

Edited by HeadSplit120
  • Cool 1

Share this post


Link to post
Share on other sites
1,005
[BOTES]
Members
2,307 posts
9,094 battles

Comparing only the best players makes no sense because they're playing against the rest of the server. All the data really shows is that they win a lot with pretty much everything. The differences in WR are minimal.

The "power" of each ship class to win is best understood by comparing win rates at different skill levels for each ship within its own tier. If we're trying to decide if T10 CV matter more than T10 DD, then we would expect to see really low win rates from the worst players and really high win rates from the best players. The difference in magnitude suggests that the class is so important that playing it poorly really hurts your chance to win and playing it well really helps your chance to win.

Here's an example:

T10 CV on NA, minimum 20 battles: overall average WR of 49.61%

T10 CV on NA, minimum 20 battles: top 5% of players average WR of 72.58%

T10 DD on NA, minimum 20 battles: overall average WR of 52.05%

T10 DD on NA, minimum 20 battles, top 5% of players average WR of 73.75%

Difference in magnitude: 22.97 for CV, 21.7 for DD

Relatively, CV is more important than DD. This fits what we know about the game; CV are an extremely hard counter to DD, so their ability to counter DD matters more.

Edited by awildseaking
  • Boring 1

Share this post


Link to post
Share on other sites
642
[O7]
Supertester
424 posts
13,498 battles

There is more to winning than damage.

DDs generally get most of their damage by shooting other DDs. Whoever wins the DD fight gets to take the cap and provide vision for the team. The team that loses all their DDs early in the match almost always loses simply because they have no vision or cap control. Most experienced players know this and will instintively target DDs first to ensure this doesn't happen. The DD skill gap is also a major factor as skilled players can crush other DDs early on, leading to a snowball effect. 

CVs are also considered a high impact class due to their amazing mobility (in planes) and ability to effectively strike almost any target in the game. They are also very effective counters against DDs and can have a huge early game impact before switching to farming cruisers/BBs in the mid to late game. 

Cruisers are highly varied in their specialties, so a general analysis of the class as a whole will probably be misleading at best. 

BBs are low impact as they typically shoot at other BBs. Unlike DD fights which tend to be short and vicious, BB fights are very drawn out due to their HP pools and heals. BBs struggle vs. DDs as their lack of detection tools, dispersion, and BB AP overpens limits their impact against the most influential class. In addition, the skill gap for BBs isn't very high as BBs have the lowest skill floor; i.e they are the easiest class to play. Even a badly played BB usually takes a while to kill whereas a DD can literally die from an unlucky torp. 

 

  • Cool 3
  • Boring 1

Share this post


Link to post
Share on other sites
1,412
Members
663 posts
20 battles
25 minutes ago, CommodoreKang said:

Because in a div you can ensure that you have 2 or 3 players with a good mix of utility and (at this level) voice comms, one would expect them to have a much higher WR.  Typically, my theory is that 60-65% is about the max you can get to as a solo player on a team of 12 (over the long haul).

Divs are absolutely very powerful and skew win rates for individual players, you're absolutely correct.  

However, there is no reason why divs would skew this particular comparison, because the div rates will be very similar across the different classes.

Divs simply have no bearing on the inter-class comparison we are discussing here. 

The correlation of the data between each server with their different populations also goes a long way to demonstrating the accuracy of the results. These three data sets are all completely exclusive to each other, and yet they all show exactly the same thing. 

Share this post


Link to post
Share on other sites
1,625
[-DF-]
Beta Testers
3,497 posts
6,798 battles
1 minute ago, Moggytwo said:

Divs are absolutely very powerful and skew win rates for individual players, you're absolutely correct.  

However, there is no reason why divs would skew this particular comparison, because the div rates will be very similar across the different classes.

Divs simply have no bearing on the inter-class comparison we are discussing here. 

The correlation of the data between each server with their different populations also goes a long way to demonstrating the accuracy of the results. These three data sets are all completely exclusive to each other, and yet they all show exactly the same thing. 

They all have the same selection bias. 

And players aren't classes.  So we could be looking at 10,000 DD games and 100 BB games.  I'm sure it's much more evenly distributed than that but you still have to account for it since you selected by players.

And, imo, over the long haul and ignoring selection bias, you typically have to div to get above 65%.  So diving is playing a big part of these numbers.  I'm guessing.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×