Balance on Legendary Drops, Zakarum, Paladins, Templars and Crusaders, Are Numbers Too Large in RoS

Armory Stats - Battleground Win Ratios
Today we are taking a look at the win percentage for each faction and battleground that took place in the last month. To find these numbers, we compared the number of battles and victories a character had in a December snapshot of data to a January snapshot of data.

These numbers come from a character's statistics pane, so any bugs with character statistics will have skewed the results. This means that there isn't any way for us to verify how accurate these numbers are, so do not take them as fact.

Percentages came fairly close to 100% in most cases, so we normalized the data to 100%. Unfortunately it seems that most Eye of the Storm wins are still not recorded in your statistics most of the time. This bug has been around for five or more years now, but we normalized the data we had (~7.9% and ~9.8% win ratios for each faction) for the chart.

The data used today is a sample made up of ~7 million level 90 characters and ~7 million battles. The sample is slightly biased, as players who are not in a guild are less likely to appear in our sample.

You can see the data from the last time we did this back in 2011, which shows a slightly different picture.

Don't see a chart? Enable javascript!

Upcoming Siege of Orgrimmar Changes
Originally Posted by Blizzard (Blue Tracker / Official Forums)
I agree. I also am not sure why removing the trivial healing mechanic from Sigecrafter is even...worth it. Most guilds are 4 healing this encounter to kill it, now it's definitely going to be a 4 heal fight. Directly after your tuned for 6-8 fight, Thok. ???
Raid groups are largely making steady progress, but as I said, there are exceptions. Regarding Heroic difficulty in particular, it's pretty obvious that there is a very large jump in difficulty from Thok to Siegecrafter. Siegecrafter/Paragons/Garrosh are currently very close to equal in difficulty, with each presenting different challenges and guilds taking roughly the same number of attempts to defeat each. We'd rather see an upward slope in difficulty, which leaves guilds feeling like the next step is within reach given enough time and dedication, rather than a sudden jump up to a much higher difficulty plateau. That's why we're nerfing Heroic Blackfuse moderately, Paragons just a little bit, and Garrosh not at all.

The changes shouldn't be enough to meaningfully alter healer composition, but they will make some of those "combo" deaths to two damage sources overlapping a bit less likely.

I'm certain that guilds who have already defeated these encounters or who are on the verge of doing so will find that the adjustments make a big difference, but remember, that's because you've already refined a working strategy and have practiced the execution. Guilds coming into the fights for the first time will still need to handle the precise movement associated with the different Blackfuse phases, belt teams will have to learn to navigate the belt lasers consistently, and so forth. Both Blackfuse and Paragons will remain quite challenging. And then there's always Garrosh....

Blue Tweets
Originally Posted by Blizzard Entertainment
I've been asking to allow us to make uncorrupted orcs for a while. Any chance of that happening???
can't definitively say yes, but it's likely that will be an option in some way. It would be cool to see UO PC in Azeroth (Source)

Would it really be that difficult to update the goblin/worgen models? You guys surely still have the original LoDs, right?
I'd like to at some point, they still use our original rigging/animation system. Pandaren introduced new system. (Source)

Character / Items
Will [Ai-Li's Skymirror] ever be allowed to copy one's character illusion again? Ex: Grasp from Pre-ZF? An updated WoD version?
Those are fun, but we don't want it to become annoying, and that opens up too many possibilities. (Source)

Warlords of Draenor Stats
Will Disc at least be fixed to do double healing from a multistrike crit. Else it is 100% useless stat for Disc.
I think you may misunderstand something. Crits don't do doublehealing for Disc, as a rule. But how does that make MS useless? (Source)
Sorry, said that backwards, when a spell crit as Disc procs MS, the MS is just as powerful as non-crit, makes it feel weak
You're misunderstanding Multistrike. The healing of the hit that triggered multistrike doesn't matter. Separate healing roll. (Source)
And an MS that crits would trigger Divine Aegis. (Source)

I keep seeing promises that the item squish won't affect soloing capabilities, but is this for all characters or only level 90s?
All. (Source)
That is to say; the squish won't affect soloing capabilities, but other changes may. Soloing at low levels isn't a balance focus. (Source)

Does this mean that we'll start seeing the new stats (possibly mastery even) on gear at very low levels?
Mastery is staying lvl80+. New stats are still being determined, but most likely not available from level 1. (Source)

I hope this wouldn't replace Warstomp. I know it's not the best for PvE DPS but great for questing/soloing as a caster druid.
No, we wouldn't replace Warstomp. (Source)

That is good news! Are you planning to increase base resilience with the new season?
Yes (Source)

How big in game scale wise is Ashran? Like are we talking AV big or Wintergrasp big
From base to base is similar to AV but it's wider. (Source)
Would you say Ashran is a spiritual successor to AV? Or more like Timeless Isle?
More like AV (Source)

back to gear being the major determining factor in PvP success. Guess the accessibility experiment didn't work out?
I don't really see how you came to that conclusion. (Source)
dialing back resilience "as much as possible" says bursty PvP and emphasizes PvP power. Sounds like old times 2 me.
In WoD, health pools will be relatively larger and PvP Power is being redefined. It will end up less bursty. (Source)
Is this true for low lvl PvP too? Some spells are too potent in heirloom gear. Aimed Shot and Shield Slam can *literally* one-shot.
A lot of spells have base damage which leads to this issue. We're going through these spells to reduce the base damage. (Source)

so I was wondering if you guys could start doing realm rankings/leaders instead of region? A lot of friends and I are curious.
We've gotten the request a lot so we'll have to look into it. (Source)
"Instead"? Why not "in addition to"?
Certainly what I meant (Source)

It also feels lousy to work hard to acquire full grievous gear only to be killed by a dragon slayer in heroic pve gear
Yup, we've mentioned we don't want to end up in this situation again. (Source)
but I give you credit in admiting that what occurred was not what you intended.
Certain systems (like warforged and upgrades) that inflate ilevels didn't make much sense for PvP. (Source)
PvP not having them meant gear falling further behind. (Source)

I've already given you a good solution. Nerf dmg/healing by 65/55% and roll those numbers into a raid wide buff for players who pve
It isn't a good solution for the same reasons base resilience and battle fatigue aren't good solutions. (Source)
The entire reason why you needed BF/Resil is because abilities do too much damage/healing.
It's lousy to have a spell hit for 100k against a creature and 20k against a player no matter where you increase/decrease it. (Source)

as horde I frequently spend more time in random BG queue then in the actual games. Fix it please. Faction conflict !> fast queues
I wish it could be fixed quickly. I really do. (Source)

Warlords of Draenor
Didnt know where to ask. Blizzcon mention quest items gone from bank in WoD. Will my [Amulet of Draconic Subversion] disappear?
Don't know yet. Could be considered a Toy, perhaps. If you still have the quest, may be kept. Very up in the air, know important (Source)

Dark Legacy Comics - #424
DLC #424 has been released!

This article was originally published in forum thread: Battleground Win Ratios, Upcoming SoO Changes Feedback, Blue Tweets, DLC #424 started by chaud View original post
Comments 219 Comments
  1. Mavarax's Avatar
    I understand why they've slight decreased the difficulty of the last bosses but I still feel like I have cheated if I don't get to kill it as it was made.
  1. HerrDrache's Avatar
    Quote Originally Posted by Miuku View Post
    And why is that exactly?

    I can't expect to clear normal raids with PVP gear, why should we let PVE'ers do "normal" BGs with their gear?
    I would liken normal BGs to "normal/heroic dungeons" and not "normal raids". That being said, I would fully expect you to be able to clear any LFR in mostly PvP gear as DPS or one of the healers.

    I would not expect PvE gear to work well in rated BGs and arenas, though, just as I would not expect PvP gear to work in normal and heroic raids.
  1. quizzlemanizzle's Avatar
    Just look at Strand of the Ancients.

    This BG is the exact same for Horde and Alliance, yet in this case Horde has outplayed Alliance which should tell us that in the polled players the average Horde player was simply a better PVP player otherwise it is hard to explain the gap for a BG that is identical for both factions.

    Outside of AV and IoC all BGs are reasonably close, Horde has outperformed Alliance a little bit but not as bad as they did during Strand of the Ancients which is identical.

    Everyone knows how imbalanced AV and IoC are and it's time Blizzard fixes this and don't give me this "take this away from alliance". That is total crap, everyone wants fair and fun BGs and not BGs that are simply looked at free win grinding.
  1. helmaroc's Avatar
    I guess people are afraid that if they balance IoC and AV that the Horde will start to dominate and the alliance wins will be unbalanced compared to now. AV I understand because of chokepoints but with the damn Isle, Really Horde? All you have to do is destroy the docks and disregard the stupid ass workshop. Why does the Horde always go for the damn workshop when you can just use superior manpower and take the docks away from the alliance?
  1. quizzlemanizzle's Avatar
    Because the docks are a losing battle.

    Horde can't get there before it is already tapped and then you can't win there simply and even if you kill the glaives there are still the catapults that brings alliance into the fortress quickly while the Gunship... well the Gunship is awful.
    If Docks were not so good why are Alliance players afraid of moving the flag position on docks and gunship so that they are equally far from each spawn point? It's a simple solution.
  1. Royalite's Avatar
    The problem with AV isn't AV. The problem is being able to blacklist battlegrounds. In 2011 the win ratio was a lot more balanced: 54 to 46 compared to ~20ish to ~80ish now.
    Most horde blacklist AV so it is just filled with bots.

    The blacklist is why the win ratio is so unbalanced now not the map.
  1. Frankyice's Avatar
    The simple fact that Horde shows such a major advantage in every BG that is entirely or nearly identical for both factions shows that Horde is better at PvP.

    We can disagree about the reasons. Personally, my opinion is that the average Horde player is probably older and more experienced. They are also probably more likely to look at all aspects of their character and options before making a race decision (instead of simply choosing someone that looks like them anthropomorphically, like Alliance races).

    These factors generally favor Horde, because the minor advantage in racials (yes, Horde's are better, but this is a very minor actual BG difference) causes better players to join Horde more often and THAT is what makes the huge difference.

    Alliance PvP racials: Every Man for Himself (strong, frees up a trinket slot), Stoneform (weak), Gift of the Naaru (weak/medium), Escape Artist (medium), Shadowmeld (weak), Darkflight (weak)

    Horde PvP racials: Blood Fury (weak), Hardiness (weak/medium), War Stomp (strong, probably the strongest PvP racial), Berserking (weak/medium), Da Voodoo Shuffle (weak), Will of the Forsaken (medium, would be strong without shared CD), Touch of the Grave (weak), Arcane Torrent (strong, probably the 2nd strongest PvP racial), Rocket Jump (weak, probably the weakest PvP racial, Goblins probably worst race for PvP)

    So out of that list, War Stomp and Arcane Torrent make great differences. But Human is probably 3rd best race for PvP. The difference I suppose being that overall, there are just MORE in the Horde list. But again, this doesn't explain the UNGODLY disparity between Horde/Alliance on mirror maps.

    Therefore, Horde is better at PvP (on average).

    - - - Updated - - -

    Quote Originally Posted by helmaroc View Post
    I guess people are afraid that if they balance IoC and AV that the Horde will start to dominate and the alliance wins will be unbalanced compared to now. AV I understand because of chokepoints but with the damn Isle, Really Horde? All you have to do is destroy the docks and disregard the stupid ass workshop. Why does the Horde always go for the damn workshop when you can just use superior manpower and take the docks away from the alliance?
    Because Alliance is closer to Docks flag. Period. And it's significantly closer. Alliance can cap that flag BEFORE Horde arrives, and then it's just such a cluster so that it doesn't matter that Horde can eventually take it because it takes longer than a minute to down the 20 Alliance that went Docks and glaives will be up by then. And saying that all you have to do is down the glaives doesn't negate the fact that you get them. Basically, the Horde is forced to play D no matter what while Alliance doesn't ever have to play D to win. And that's what makes that BG unbalanced. If they just moved that Docks flag about....15-20 yards north, that place would be dramatically different. The other 2 nodes are basically equidistant.
  1. Osmeric's Avatar
    Quote Originally Posted by Frankyice View Post
    Because Alliance is closer to Docks flag. Period. And it's significantly closer. Alliance can cap that flag BEFORE Horde arrives,
    If alliance does that, and the horde don't suck, the fight will still be going on when the glaives spawn. At that point, the glaives are indefensible and will die.
  1. Lemonpartyfan's Avatar
    Quote Originally Posted by Warlord Booty View Post
    I'm shocked the SotA is such a landslide.

    We all know AV is Alliance slanted, we are 30 seconds slower, and that's why we lose. Those 30 seconds are killing the archers/being slowed by the archers.

    IoC... I honestly have no idea why we lose.

    It comes down to Alliance avoiding PVP and the Horde looking for it. AV/IoC is a PVE zerg? Nah. AV, not only are we 30 seconds slower, but alliance ALWAYS backcaps and horde does this 5% of the time. Even when we do guard the towers, those 4-5 people get overrun by the 12-15 alliance that are backcapping.
    Where does the Horde/Alliance wave meet?
  1. Necrotix's Avatar
    Ah yes, typical Horde. Whining as usual.

    The stats just show that generally, the alliance are better at PvE. Also, most players are alliance. We are a majority. Not everyone wants to be playing as ugly orcs, trolls or cows.
  1. Vaelkyri's Avatar
    Issue is IOC has ZERO to do with map balance.

    Here is the problem.

  1. Ryme's Avatar
    Quote Originally Posted by Khwaj View Post
    i agree, so let us have those 2 while u have the rest
    I'd rather they were all balanced, which they nearly are, the outliers are problems - I'm sorry if you can't see that.

    Quote Originally Posted by Khwaj View Post
    for crying out loud plzzzzz stop using this as an excuse
    all it takes is 2 ppl out of 40 to kill both glaives, 1 if ur exceptionally skilled, they only have 1.2mill hp, thats pretty weak as it is

    same can be said abt av, i havnt been to drek in ages as i always go with 3-4 ppl including myself to back cap towers
    80% of the time there is only 1-2, 3 tops defending the towers
    the times i lost av is when horde have like 5-7 players back capping and wipe us out and our tower defenders
    I know, that's why I put in casual/pug groups - I've never played IoC outside of this format, but I imagine getting your group to send over a small division of stealth/burst DDs is a lot easier when the group isn't made up entirely of people with a split division of tactics.
  1. Frankyice's Avatar
    A lot of this is also about inertia. Alliance have won AV / IoC so much that it's a self-fulfilling prophecy on Horde side.
  1. kamelioe's Avatar
    bg's are good for melee only. they ruin them with all the stun b.s, and the general education of bg's needs to be explained to alliance mostly. something like in this battle the object is to get a horde flag have it carried and protect from your entire group when the flag carrier is safe have a team go retrieve the alliance flag and please notice the healers healing the flag. also explain do not stand in the middle attacking anything they can. education is a huge must in bg,s
  1. sidhebane's Avatar
    Quote Originally Posted by Rhelyo View Post
    That is wrong dude.Lets see how your statistics are at 50/50 w/l for both factions.
    AB-H60% vs A40% (that is 10% in favor of horde to the mark of 50/50 that the ideal w/l should be for both factions),i dont know where u pull the 20% crap
    WG-H47,4% vs A52,6 (that is a 2,6% in favor of Alliance) not that big deal (u called that 6% remember ideal w/l should be 50%-50% for both factons
    EotS-A44,4% vs H 55,6% (5,4% in favor of Horde)
    Strand-A31,7% vs H68,3% (18,3% in favor of Horde)
    AV-A78,2% vs H21,8% ( 28,2% in favor of Alliance)
    IoC-A84,9% vs H15,1% (34,9% in favor of Alliance)
    Twin Peaks-A42,7% vs H57,3%(7,3% in favor to Horde)
    BfG-A42,1% vs H57,9%( 7,9% in favor to horde)
    ToK-A41,9% vs H58,1% (8,1% in favor to horde)
    Silvershard mines-A35,5% vs H64,5% (14,5% in favor to horde)
    Deepwind-A42,1% vs H57,9% (7,9% in favor to horde)

    Now if you look at all statistics it is true that horde win ratio is better than alliance on most bgs,but u have to admit that the difference in w/l is not that great.I mean come on in 10 games played on many of those bgs alliance wins 4 and horde wins 6 (w/l <10%)
    It's true that Strand 18,3% and silvershard 14,5% which favors the horde are a bit high and they should take a look at them(but its mostly player skill and such).
    Now take a look at AV/IoC (28,2% and 34,9%).Don't you see something wrong there?
    My point in all this is that most bgs that people say its horde dominated,the w/l ratio between both factions are <10% and are close to the 50/50.But IoC and AV w/l ratio is too overwhelming.

    If i were to blame anyone i would probably blame the hk premades which are made especially by the russian servers.
    Not sure how your math works. A horde/alliance win ratio of 60/40 doesn't mean a "10% in favor of horde". It means that the horde win 20 percentage points of the time more than alliance (60 minus 40). It also actually means that the horde win 50% more often than alliance (60 is 150% of 40). These numbers mean slightly different things, but they are both correct, and in neither case is "60 is 10 more than 50" actually relevant to anything
  1. VeraeliaBlack's Avatar
    In AV, Horde used to be able to mount inside the gate at the starting tunnel. Alliance wasn't able to, and I suspect that largely evened out the distance/time that the two sides had. Now, Alliance is across and almost at Galv before Horde are even off the GY at Iceblood. Couple that with the jumping-the-fence thing, and it's going to be Alliance dominated. Glaives are overpowered, and most Horde go Workshop in IoC - most of them trying to do the cheapshot of getting Demos and sitting at the keep GY to one-shot people rezzing, which Alliance figured out by just steamrolling them with Glaives. If you bust the Glaives, it's almost certain Horde is going to win it.

    The real issue isn't the fact Alliance dominates those two BGs, the issue is that Horde absolutely crush Alli in any BG that doesn't have a PvE standpoint/can't be zerged. I've got Strand and Temple blacklisted, because I can't stand either of them; Strand, it's usually whoever's on the Offensive first wins. Horde just seem to start on O more than we do. Not sure what controls that, but I imagine balancing that out would make that one a lot more even. Over a third of all the Strands I've ever done, end up being a turtle at Yellow with people camping the stairs to stop bombs. Not very fun, just an attrition war (like most of the non-zerg AVs and IoCs turn into). People have been saying that Horde racials/PvP elements are way too strong, ever since Vanilla, and maybe now Blizz will pay attention to that fact. I'm seconding that they do comparisons with data pre-blacklisting option to get a better handle on what's adjusted what where, and how the changes between those times have reflected.

    But I honestly doubt even with this, the only thing that will happen is a hard nerf to Alliance in AV and IoC. If nothing changes as far as making it more even so that Horde doesn't absolutely smash Alliance in the other 9 BGs anymore, I doubt I'll be even bothering with PvP post getting my Battlemaster/Khan finished. I play Horde as well, but it's not fun when your team sits and GY camps the other side. Might be good for a laugh every once in a while, but when it happens all the time, it's hardly a thrill, or very sportsmanlike conduct.

    Tl;Dr - this needs to be fixed, but I highly doubt that it will be as far as balancing the other 9. Alliance is just going to see a hard nerf on the two we don't constantly lose, and non-arena/RBG PvP will continue to suck for Blue, and be amazing for Red (when they're not sitting in 45 min queues).
  1. PlutoEjecto's Avatar
    There were so many issues with AV map being pro-alliance, and its only recently that Blizz has decided to address it is sad.

    If these balance issues were addressed 'sooner' I probably still be playing.
  1. stochastics's Avatar
    Can the article poster please post the Mathematics used to calculate the battleground win percentages for the past month?
    How was the percentage calculated for a sample of 7 million players? What was/ how was the math done - please post your fumula or calculation used. I am also highly doubtful of the accuracy of posted win ratios if they are taken armory statistics because these statistics reflect a characters entire bg data since character creation which could have been 7 years ago.

    The win ratio's that i have arrived at are:
    A horde players win probability when joining a random bg queue is ~80%( + or - 5%)
    An Allaince players win probability when joining a random bg queue is ~24% ( + or - 5%)

    These win probabilities i posted above are derived from the data aggregation and its subsequent calculation using two different methods. One method is using a wow addon called reflex. The other method is a manual aggregation of statics of new character's statistics, meaning a new character created since the release of the MOP expansion. Both the data which was compiled and gathered by reflex, when calculated matched up with the win probabilities i have calculated by manual aggregation of bg statistics for a new character. My calculations are explained below. Different mathematical calculations and different data aggregation based on probability theory arrive at similar win probability ratios. Read the rest of the of the post to find out how to do these calculations.

    Here are my calculations and my understanding of how win ratio's should be calculated - BG stats that are not recent should not be used. Win probability calculated from over 1,2 , 3 or more years worth of PVP or more should not be used.
    If you just created a new character in MOP and these were your statistics, this is how you would accurately calculate your Alliance random BG win probability for this season:

    Alterac Valley battles 25
    Alterac Valley victories 18
    Arathi Basin battles 13
    Arathi Basin victories 5
    Battle for Gilneas battles 8
    Battle for Gilneas victories 2
    Eye of the Storm battles 15
    Eye of the Storm victories --
    Strand of the Ancients battles 29
    Strand of the Ancients victories 7
    Twin Peaks battles 15
    Twin Peaks victories 6
    Warsong Gulch battles 12
    Warsong Gulch victories 2
    Silvershard Mines battles --
    Silvershard Mines victories --
    Temple of Kotmogu battles 2
    Temple of Kotmogu victories 1
    Isle of Conquest battles 29
    Isle of Conquest victories 28
    Deepwind Gorge victories 4
    Deepwind Gorge battles 17

    Since you have no stats for silver shard mines, maybe because you blocked it, the total number of battlegrounds is 10 otherwise it should be 11. What this means is that you have a probability of 1/10 of getting into any of these battlegrounds from the random bg queue. This is important because we will now use this to calculate your overall battleground win probability when you join the random bg queue for PVP.

    your random bg win probability is:
    (1/10) * (18/25)-Alterac +
    (1/10) * (5/13)-Arathi +
    (1/10) * (2/8)-Gilneas +
    (1/10) * (0/15)-Eye of +
    (1/10) * (7/29)-Ancients +
    (1/10) * (6/15)-Twin Pks +
    (1/10) * (2/12)-Warsong +
    (1/10) * (1/2)-Kotmogu +
    (1/10) * (28/29)-Isle of +
    (1/10) * (4/17)-Deepwind +

    = .1 * .72 + .1*.384615 +.1*.25+.1*0+.1*.241379+.1*.4+.1*.16666+.1*.5+.1*.965517+.1*.235294
    =.072 + .0384615 + .025 + 0 +.0241379 + .04 + .016666 + .05 + .0965517 + .0235294

    = .3863472721 * 100 = 38.6% win ratio in the random bg queue since MOP expansion

    Now let us calculate alliance win probability based on your stats without the big battlegrounds. We should be leery of the large PvE centric battlegrounds, which i believe have been adjusted to balance or rather i should say skewed to create a false sense of balance between factions in PvP. Now let us remove the two big PvE centric bgs from our calculation to find out a player's real win probability. If we remove the big battlegrounds which alliance has a high probability of winning evident by the stats posted; alterac valley and Isle of conquest. So then when we remove these two bgs from the queue we have a total of 8 possible random bgs that we will queue up for, hence 1/8 (this should be 1/9 but no stats given for mines). I will also remove the Kotmogu bg from this calculation because the sample space is too small to be accurate - basically this bg was played too few times only twice. Therefore the total number of battlegrounds are 7, hence 1/7. The following calculation should be the true measure of an alliance players win probability in the random bg queue:

    (1/7) * (5/13)-Arathi +
    (1/7) * (2/8)-Gilneas +
    (1/7) * (0/15)-Eye of +
    (1/7) * (7/29)-Ancients +
    (1/7) * (6/15)-Twin Pks +
    (1/7) * (2/12)-Warsong +
    (1/7) * (4/17)-Deepwind +

    = .14286*.384615 +.14286*.25+.14286*0+.14286*.241379+.14286*.4+.14286*.16666+.14286*.235294
    = .239707 * 100 = 24% win probability for alliance players in the random BG queue.

    This 24% win probability for alliance feels right.

    Here is some data from a horde player who just recently created a fresh 90 in the MOP expansion. To find out the win probability in the Random bg queue of a horde player i used the following calculations ;

    (1/11) * (8/11)-Alterac +
    (1/11) * (6/10)-Isle of +
    (1/11) * (29/39)-Warsong +
    (1/11) * (19/20)-Gilneas +
    (1/11) * (17/21)-Arathi +
    (1/11) * (17/19)-Twin Pks +
    (1/11) * (9/11)-Ancients +
    (1/11) * (7/8)-Kotmogu +
    (1/11) * (4/4)-Mines +
    (1/11) * (8/9)-Deepwind +
    (1/11) * (7/25)-Eye of +

    =.090909*.727272+.090909*.6+.090909*.74359x+.090909*.95+.090909*.8095+.090909*.894736+.090 909*.818181+.090909*.875+.090909*1+.090909*.88888+.090909*.28

    =.78065398 * 100 = 78.06% probability of winning in the random bg queue for horde

    We should be leery of the large PvE centric battlegrounds, which i believe have been adjusted balance or rather i should say skewed to create a false sense of balance between factions in PvP. Now let us remove the two big PvE centric bgs from our calculation to find out a horde player's real win probability.

    (1/9) * (29/39)-Warsong +
    (1/9) * (19/20)-Gilneas +
    (1/9) * (17/21)-Arathi +
    (1/9) * (17/19)-Twin Pks +
    (1/9) * (9/11)-Ancients +
    (1/9) * (7/8)-Kotmogu +
    (1/9) * (4/4)-Mines +
    (1/9) * (8/9)-Deepwind +
    (1/9) * (7/25)-Eye of +

    =.111111*.74359x+.111111*.95+.111111*.8095+.111111*.894736+.111111*.818181+.111111*.875+.1 11111*1+.111111*.88888+.111111*.28

    = .806657 * 100 = 80.7% win probability in the random BG queue for a horde player

    When you compare this horde win probability to the alliance win probability i calculated earlier which was equal to a 24% win probability for alliance, both probabilities roughly adds up to ~100 with a + or - 5% error / deviation.

    Therefore i conclude that the horde are steam rolling alliance in battleground. My Statistics calculations show that they have a 80% win probability when joining a random bg queue. While alliance players have at best a 24% chance at winning a bg in a random bg queue. Please note that my statements are contingent to both of the large PvE centric battlegrounds being blacklisted and removed from win probability calculations.

    I have a theory that blizzard is aware of the steam rolling of alliance players in PVP and is therefore tweaking the PVE aspects of the large bgs to balance PvP between the factions. Because the statistics do not add up i suspect this. What i mean is that if the horde are steam rolling alliance in 9 out of 11 bgs by hudge margins then that begs the question why are they only losing in the big BGs that are PvE centric.

    In regards to my calculations I did not cherry pick the data for my calculations. I also have not biased the results of the calculations by using my own data. In my calculations that i have posted i have used only the pvp stats posted by others in the battle form thread called "Dispelling a myth - Alliance win ratio". I have also written out all my calculations and the reasoning behind them. I have specifically excluded the large PvE centric bgs from both the horde and alliance win probability for reasons i explained in my posts on this thread. I used only the recent PvP stats/data that was posted in that thread, This means recent stats, explicitly mentioned by posters which correspond to MoP expansion, not 3 or more years of pvp stats. I also used only the data from the first alliance player and the first horde player who posted their pvp stats in that very thread on blizzards forum.

    A horde players win probability when joining a random bg queue is ~80%( + or - 5%)
    An Allaince players win probability when joining a random bg queue is ~24% ( + or - 5%)

    An alliance players win probability when including the 2 PvE bgs is about 38%
    A Horde players win probability when including the 2 PvE bgs is about 78%

    There is almost no difference in a horde players overall random bg win probability, a 2% difference when including or excluding the 2 big PvE centric bgs. However there is a big difference for an alliance player, a 14% difference in the random bg win probability for an allaince player if the 2 large bgs are included or excluded.

    How come a horde player only has a 2 percent difference in their win probability from the inclusion or exclusion of the two big bgs(alterac & isle of..) while alternatively the alliance player has a huge percentage difference of about 14%?

    If you use the reflex wow addon can also accurately calculate your win ratio . The following reflex data was taken from a horde poster in the same thread. If the reflex calculations are 8 to 31 or 8:31 then you can calculate your win ratio like so:

    When odds of losing an event are A:B
    Then the probability that you will lose the even is P(E')= (A)/(A+B)
    Where P(E') is the probability that the event will not occur.

    So in your case the odds given by reflex were;
    8:31 = (8)/(8+31) = (8)/(39) = .205128 * 100 = 20.5% chance of losing in the random bg queue.

    In order to get your random BG win probability you would need to subtract the probability of losing the random BG from the total sample space. By the first axiom in probability your calculated win or losing probabilities must be between 0<P(E)<1. By the second axiom of probability theory the total sample space is 1, P(S) =1. There fore:

    P(Probability of Losing bg ) - 1 = |P(probability of wining BG) |

    .205128 - 1 = .7948 * 100 = 79.5% chance of winning a bg in the random bg queue

    An alternative way of calculating the win ratio is by If reflex shows that you won 31 bgs and lost 8 bgs, then you have played 39 total games

    31/39 * 100 =79.49% wins in BGs recently, this would be your most up to date win ratio with bg sample space data gathered and compiled by the addon reflex.

    The win probabilities compiled and gathered by reflex when calculated match up with the win probabilities i calculated using different mathematical calculations based on probability theory. Hence i believe that my predictions about alliance and horde win probability are arguably valid because similar probability is arrived at by the add-on reflex's bg win/played data aggregation which is then calculated and computed differently from the calculations and data aggregation i used based on probability theory with different sample spaces of different players.

    A horde players win probability when joining a random bg queue is ~80%( + or - 5%)
    An Allaince players win probability when joining a random bg queue is ~24% ( + or - 5%)
  1. chaud's Avatar
    Quote Originally Posted by stochastics View Post
    I am also highly doubtful of the accuracy of posted win ratios if they are taken armory statistics because these statistics reflect a characters entire bg data since character creation which could have been 7 years ago.

    Quote Originally Posted by chaud View Post
    To find these numbers, we compared the number of battles and victories a character had in a December snapshot of data to a January snapshot of data.
    As for the rest of your post, this isn't something we talked about at all, just individual battleground ratios. If you wanted to combine win ratios in this post with weights from these numbers, you could repeat your wall of calculations.

Site Navigation