Week 1 class statistics

Data is pulled from ranked matches only. Assume all players are of platinum rank/skill. Keep in mind that there may be insufficient data, I've played with/againts the same people for multiple games. Also I haven't seen all specs (like scrapper) in my matches. As another assumption, players of a spec are using the "meta" build of that spec. There have been a total of 200-non unique players and 20 games recorded. If you're really curious to know the exact number played, multiply percent played by games played (200), so each 0.5% is 1 player (or more easily, double it).

ElementalistElementalists are doing poorly and are uncommon to see.

• Core - 1% played, 0 wins. Both unpopular and unsuccessful.
• Tempest - 1.5% played, 33% win rate. Very unpopular, and doing poorly.
• Weaver - 3.5% played, 57% win rate. Most popular ele spec, doing quite well.

EngineerEngis are currently the second least commonly seen spec, yet are averaging the highest win ratio.

• Core - None
• Scrapper - None
• Holosmith - 4.5% played, 67% win rate. While uncommon, holosmith is currently sitting at the highest win ratio.

GuardianGuardian has a decent player base but the specs success rates are not balanced at all.

• Core - 4% played, 37.5% win rate. Most popular core spec, not successful.
• Dragonhunter - 4.5% played, 22% win rate. Least successful spec with more than 5 players seen.
• Firebrand- 8% played, 62.5% win rate. Fairly popular, very successful.

MesmerMesmers a successful profession right now but a less common one. Again spec success rates aren't even.

• Core - None seen
• Chronomancer - 1.5% played, 0 wins. Mirage > Chrono.
• Mirage - 10% played, 55% win rate. Popular and only becoming more popular. High win rate considering it's skewed by me.

NecromancerHands down the most popular profession in the game at the moment, yet not doing as well as people claim. That being said, I would attribute that more to the classes popularity than it's potential.

• Core - None
• Reaper - 1% played, 0% win rate. All necros are scourges now.
• Scourge - 18% played, 50% win rate. Most popular spec with out question and the win rate is exactly 50%. ( I have faced 21 and had 15 on my team, although it's very common to have one on each side). I have played one game all season that did not have a scourge.

RangerVery popular, even success rates, but players are choosing druid almost exclusively

• Core - none
• Druid - 13% played, 54% win rate. Second most popular spec and has a decent win rate.
• Soulbeast - 5% played, 40% win rate. Uncommon, likely due to it's it's poor performance.

RevenantPeople are playing exclusively Herald if they're playing it at all. Like engineer, also doing well this season despite it's low tier in the meta. I've seen 4 revs so far that weren't me.

• Core - none
• Herald - 3.5% played, 62.5% win rate. If you see a rev, they probably know what their doing is what this means. Too little data to determine anything.
• Renegade -1%, 50% win rate. These were both me. Non-indicative of anything.

ThiefOne of the better classes this season and more popular too. Daredevil is doing a lot better than the rest of the specs.

• Core - 0.5% played, 0 wins. I've seen one so far.
• Daredevil - 11% played, 60% win rate. Daredevil is doing very well, despite the nerf.
• Deadeye - 1.5% played, 0 wins.

WarriorAll warriors have been spellbreakers and despite the nerfs, they've been doing well.

• Core -None
• Berserker - None
• Spellbreaker - 8% played, 56% win rate. It's doing well at the moment.

Something I would like to clarify before anyone says anything about scourge having a 50% win rate and dragonhunters having a 22% winrate: A bad player running scourge can still do moderately well, were as an okay player running scourge can win a game. By far the worst players I've seen have been scourges and dragonhunters, that have 0 awareness and try to take on 3 people at once. I'm not necessarily saying "if you run scourge or dragonhunter you are bad." I'd attribute it more to them being the current and previous faceroll classes and people that do pick the cheese builds typically do so because they don't know what they're doing. I'd say scourge needs a nerf, and a good dragonhunter can still carry a game, but most importantly: Don't just pick a class because you think it's stronger than others, pick the one you're good at. If you doubt any of my numbers, they're all here. I will keep updating once a week.

lul scourges with the 50% win rate because they are played both sides nearly every game.

Good job recording all that.

@Master Ketsu.4569 said:lul scourges with the 50% win rate because they are played both sides nearly every game.

I'm so god damn sick and tired of scourges.

Herald has highest win rate, nerf it, nerf it fast!

Side note, nice work but I feel like 20 games is too small of a sample size to make many conclusions from. Admittedly I posted a similar thread based on last monthly AT with around the same sample size, but soloq is a whole other beast when it comes to statistical variability.

@ArthurDent.9538 said:

I feel like 20 games is too small of a sample size to make many conclusions from. Admittedly I posted a similar thread based on last monthly AT with around the same sample size, but soloq is a whole other beast when it comes to statistical variability.

totally agree. I'm going to continue updating it. I started this more out of curiosity than anything.

Reaper - 1% played, 0% win rate. All necros are scourges now.

And that was probably me shortly before ragequitting this scourge fest.

...Not really, I never got outta Gold 3 cause I don't play Scourge.

Needless to say the sample is incredibly small to draw conclusions. One thing always to take in consideration, the stronger something is it’s win percent will get very close to 50% cuz everyone plays it. Scourge is in almost every game on every team, so the winner has it and the loser has it. Good players play it and bad ones as well.

Same analogy for weaker classes. The weaker something is, the more likely its win percentage will be high, cuz only highly skilled players will play it.

I know you tried to normalize player skill as a factor, but let’s face, that is not possible, unless we get close to the top end of the competition.

One thing I remember the devs in another game mentioned is winning composition. Let’s say that the highest 10 winning composition consistently have scourge and FB. That is a much stronger indicator that they are over performing versus winning percentages Individually.

Appreciate the effort though.

How many of the scourge losses were when fighting another scourge team? Can you post the scourge win% when facing a team without one (if said team even exists).

I play core Mesmer, I never see another one of me around but this season I got 6 wins 1 loss (still in the placement 10 games, I don't play a lot)why don't ppl show core specs some love lol

I think the small amount of data on engineer and rev is skewing the statistics you present. People who represent these classes in ranked tend to be some of the best at those classes, thus they are naturally more likely to have a higher win rate. Not to mention, I like to run holosmith builds, but they don't influence the outcome of a match quite like a scourge can. Scourge... you can do a lot even if your team is underperforming.

@Vagrant.7206 said:I think the small amount of data on engineer and rev is skewing the statistics you present. People who represent these classes in ranked tend to be some of the best at those classes, thus they are naturally more likely to have a higher win rate. Not to mention, I like to run holosmith builds, but they don't influence the outcome of a match quite like a scourge can. Scourge... you can do a lot even if your team is underperforming.

Its a very small sample in general, hard to draw much conclusions (that the community isn't blatantly aware of already via personal experience) from it.

@Hooglese.4860 said:ThiefOne of the better classes this season and more popular too. Daredevil is doing a lot better than the rest of the specs.

• Core - 0.5% played, 0 wins. I've seen one so far.
• Daredevil - 11% played, 60% win rate. Daredevil is doing very well, despite the nerf.
• Deadeye - 1.5% played, 0 wins.

That said, i'm hovering in Gold 3 as P/P Deadeye right now, lets see if i can be the 1% among the endless amount of Female Human Daredevil clones :>

@xDudisx.5914 said:How many of the scourge losses were when fighting another scourge team? Can you post the scourge win% when facing a team without one (if said team even exists).

Of all the games all but 4 have scourges on both sides, and that's a 50% win for scourge.If I add in the games where there are more scourges on one side than the other, it becomes 4/9 wins.

@Hooglese.4860 said:I'm so god kitten sick and tired of scourges.

I've been playing this game on and off since launch, and meta Scourge is the easiest build I've ever tried. I feel guilty using it, personally, but you're putting yourself and your team at a disadvantage if you play with damn near anything else. I can't believe ANet started a PvP season with Scourge in its current state.

I think win rate is a poor metric for such a small sample size. For one thing this early in the season some people rise and fall dramatically initially so all those people you are facing might not really be platinum level players I believe the minimum games for leaderboard is already well over 20. The class population % is probably more accurate and probably woulnd't change too much with additional games.

Yay, can we have updated/expanded version of this later on ? Thanks for the effort by the way.I'm imagining all those calculations really not being the funniest thing to do though.

I think if you analyzed the comps of winning teams only it would at least make better conclusions than "Scourge has 50% Winrate :trollface: " If a class is present on 99% of winning teams, or 0% of winning teams, that's more valuable a conclusion

May be a small sample but more than ANet are offering. One way of stopping a lot of 'this is OP and that is OP' is to produce stats.

Good work. Although it does tend to clearly indicate 'no PoF uuuggghhhh PvP'.

I think core anything should be made competitive.

@Wesa.3580 said:Yay, can we have updated/expanded version of this later on ? Thanks for the effort by the way.I'm imagining all those calculations really not being the funniest thing to do though.

Yep, Im posting results weekly (or i hope to be) and i update as i play. I don't hate doing it though cause I am curious about the results.

@Chaith.8256 said:I think if you analyzed the comps of winning teams only it would at least make better conclusions than "Scourge has 50% Winrate :trollface: " If a class is present on 99% of winning teams, or 0% of winning teams, that's more valuable a conclusion

Scourge is in 88.5% of winning teams in games I have played. I lack the data to do comps for right now, that's end of season stuff for me or if other's make and record shit.

@Chaith.8256 said:I think if you analyzed the comps of winning teams only it would at least make better conclusions than "Scourge has 50% Winrate :trollface: " If a class is present on 99% of winning teams, or 0% of winning teams, that's more valuable a conclusion

That is not how this works. The way you do it is to look at the top 10-20 team comps in terms of winning percentage, and notice which elites are consistently showing up.

Theoretical example, where everything pulled out of my tin foil hat :# . let say the top winning comp is: Scourge, DD thief, FB, ranger and SB, with 63% win rates.

Next one is: Scourge, FB, Mirage, SB, power herald, with 61.5%.

You go down through the top 20. Then you notice scourge shows up 17 times in the top 20 comps, where the median elite build shows up 5 times. This is the strongest indicator that it is out performing, cuz it accounts for synergy and minimize skill factor impact on skewing the data.

Class participation is a good indicator, but alone is not sufficient. Cuz sometimes you may enter in popularity contest, where it maybe popular cuz average Joe thinks its cool versus effective. Same also regarding class winning percentage, where class cuteness and perception of it being strong can easily skew the data.

OP, you should try to contact other PvP players to provide more data. This is probably what Anet looks at the most when time for the nerf hammer and are afraid to release data which people would probably misinterpret all the time.Pretty sure scourge and weaver shouldn't be doing that well in lower tiers since they do take quite some skill po pull off. You should state your average MMR too.Amazing data btw, <3.

@otto.5684 said:

Class participation is a good indicator, but alone is not sufficient. Cuz sometimes you may enter in popularity contest, where it maybe popular cuz average Joe thinks its cool versus effective. Same also regarding class winning percentage, where class cuteness and perception of it being strong can easily skew the data.

Necromancer will probably always be the most popular class unless it's utter garbage because edgy kids love their death and darkness.

@Tomiyou.3790 said:OP, you should try to contact other PvP players to provide more data. This is probably what Anet looks at the most when time for the nerf hammer and are afraid to release data which people would probably misinterpret all the time.You should state your average MMR too.

My MMR is sitting around 1500. I am tempted to pull from streams/ask others but I haven't seen most of my friends on due to The Scourge of the Scourge, and pulling from streamers will heavily skew their profession's win rates but i am going to look at some