Check out this reddit thread regarding the best performing builds for this last iron banner.

https://www.reddit.com/r/DestinyTheGame/comments/3pn303/with_iron_banner_now_over_are_you_curious_which/

Here's a terse version of the stats (hit reddit for more details & the builds):

Hunter
Bladedancer (60.67% win rate, 1.39 KD)
Gunslinger (55.13% win rate, 1.18 KD)
Nightstalker (56.36% win rate, 1.05 KD)

Titan
Defender (64.04% win rate, 1.23 KD)
Striker (53.46% win rate, 1.02 KD)
Sunbreaker (61.29% win rate, 1.40 KD)

Warlock
Stormcaller (57.29% win rate, 1.22 KD)
Sunsinger (60.87% win rate, 1.31 KD)
Voidwalker (56.76% win rate, 1.17 KD)

The interesting points for me are:

• Sunbreaker has the best KDR, but it's TOTALLY on par with the Bladedancer being only one hundredth of a percent lower.
  
• Defenders have SIGNIFICANTLY better Win/Loss ratios at 64% win rate.

Interesting. That bladedancer build is the exact one that I landed on for 2.0. Skip Grenades are the new hotness for sure.

I know a lot of people hate the sunbreaker, but I don't really think it's going to end up with a nerf after Bungie looks at the stats. That's probably why the hammer is so loud to begin with.

I know a lot of people hate the sunbreaker, but I don't really think it's going to end up with a nerf after Bungie looks at the stats. That's probably why the hammer is so loud to begin with.

Heh. I had fun last night - but saw both the good and the bad with Sunbreakers. Had a total Hammer game (in Mayhem Clash, of course) - I had 17 kills, and 11 of them were hammers. I also played a couple of games of Rumble - and had one Super kill in each. (Every time my hammer popped, people ran. I didn't die much - but I didn't kill much, either.)

I had a game where I was killed by a bladedancer, as I was popping. He heard my hammer activation, activated BD (which is a faster activation), and killed me before I could throw even one hammer.

I really gotta think about running the unkillable Defender build. :)

I had a game where I was killed by a bladedancer, as I was popping. He heard my hammer activation, activated BD (which is a faster activation), and killed me before I could throw even one hammer.

I had the exact same situation but I managed to get one hammer off, which so happens isn't enough to kill a blade dancer...

The data isn't normalized, so there is lots of noise in these stats.

You need to take into account the number of players of any given subclass, since that will weight the results. Likewise K/D and win/loss are not the only stats one should look at, since it is a team event and different classes can compound on each other as we've seen in trials.

A better examination could be a look via score, as opposed to just K/D, since that will take into account captures and other factors.

I read this:

I had a game where I was killed by a bladedancer, as I was pooping.

I was like there's the problem. Your subclass is so OP that you can walk away, use the bathroom and come back with the reasonable expectation that you wouldn't be dead. :-)

On topic, I found gunslingers to be the most viable class for hunters. The amount of tripmines you can place is staggering in mayhem if you use ahamkara's spine. The entire map was my own personal deathtrap and I'd get kills popping up on my screen a minute or two after I laid the mine. It was glorious.

The data isn't normalized, so there is lots of noise in these stats.

You need to take into account the number of players of any given subclass, since that will weight the results. Likewise K/D and win/loss are not the only stats one should look at, since it is a team event and different classes can compound on each other as we've seen in trials.

A better examination could be a look via score, as opposed to just K/D, since that will take into account captures and other factors.

what would be nice to see is what they succeeded (or failed) against. I don't know how fine scale the data goes (i remember halo 2 stats that tracked everywhere you died, and where the person was standing when they killed you, and with what weapon; so i would imagine that they still have that capability). Did hammer bros fail only when they went against another hammer bro or gunslinger (ranged supers).

The Defender numbers are interesting. I suspect they have less kills and instead lead their teams to victory by providing increased zone control multiplier. It might not have much to do with the Defender's powers at all and is them hanging back because they know they can't compete head on.

Quit backpeddling. You're preferred subclass is OP and needs a serious nerf! Defenders are are clearly the scourge of PvP.

The Defender numbers are interesting. I suspect they have less kills and instead lead their teams to victory by providing increased zone control multiplier. It might not have much to do with the Defender's powers at all and is them hanging back because they know they can't compete head on.

I think you're right. The subclass is definitely geared towards strategic play which is a good change of pace for the ecosystem of the game.

I had a game where I was killed by a bladedancer, as I was popping. He heard my hammer activation, activated BD (which is a faster activation), and killed me before I could throw even one hammer.

I remember there being a lot of bladedancer hate, and whenever I hear of a bladedancer in my area I still try & get out of dodge. Why isn't there much bladedancer hate? Was there a nerf, or have people just gotten used to it?

I really gotta think about running the unkillable Defender build. :)

Yeah, I really want to put together that build and give it a try.

On topic, I found gunslingers to be the most viable class for hunters. The amount of tripmines you can place is staggering in mayhem if you use ahamkara's spine. The entire map was my own personal deathtrap and I'd get kills popping up on my screen a minute or two after I laid the mine. It was glorious.

You are so right! I maxed my regular Crucible discipline and get a 25 second cool down on Tripmines in normal modes... In Mayhem... Oh my goodness. They last for days and you always have 2. So much fun. I am surprised more people don't shoot them.

I wonder what a sunbreaker with Armamentarium and simmering flames would be like in mayhem. Because that would be double grenades with halved cooldown with simmering flames and then whatever mayhem does....

Some classes may have a high learning curve, that weeds out lower skill players. You could filter the class data by crucible game time to look a players with low amounts of time vs. high amounts of time to identify a "veteran effect".

We could also look at the kill breakdown to see if it is weighted to weapons over abilities.

Nice info. I'd say it makes sense. I pulled 72 games from my IB run past week using DTR's query tool from weds thru sunday, and my k/d was 1.43 and win% was 69.44. I'd say over 80% of my games were as a defender with Armor of Light, pulse rifle, shotty, and sword. I was also running with Helm of St. 14 but I would have preferred to run with No Backup Plans if they were leveled up.

One thing that has really improved for defender is that the bubbles cannot be penetrated by a Titan smash or BD wave anymore. They have to come inside to get the kill or kill you after they have destroyed the bubble. Also I saw fewer strikers and DB's in general. Strikers are the worst for defender bubbles.

Some classes may have a high learning curve, that weeds out lower skill players. You could filter the class data by crucible game time to look a players with low amounts of time vs. high amounts of time to identify a "veteran effect".

We could also look at the kill breakdown to see if it is weighted to weapons over abilities.

This is a good point. For bladedancers, for example, I didn't land on skip grenade as the best option until day 4 of the event, so only my day 4 and later K/D would factor into that specific build's numbers, well after I established my primary weapons for the event and got a feel for the "flow" of things.

I feel like I'm okay at PvP but those stats make me wonder. Damn hunters and titans got all the phat stats.

W/L ratings on Defender, Striker, Bladedancer, and Sunsinger are probably inflated due to the low population of those classes. Sunbreakers, Stormcallers and Nightstalkers are probably a little suppressed, but overall it may just be a percentage point or two, I haven't run the numbers.

The top IB ranked players all have high 310 gear, (unsurprising) with Pulse Rifle/Scout and Snipers dominating.

This would imply the mid-long range domination is key to high KD and performance, which is logical. As someone who frequently tried to sneak up on those people, I can attest that the guy jumping forward to take a control point usually is going to die, but those backing him up can rack up the kills as he attracts attention.

I'd like to see some score based stats, along with some kills/ability and points/ability. Points per weapon archetype, by class would be a complicated chart but also interesting.

Since they have broken each player into ranks, the same metrics confined within rank, and compared across rank, would be interesting as well.

W/L ratings on Defender, Striker, Bladedancer, and Sunsinger are probably inflated due to the low population of those classes. Sunbreakers, Stormcallers and Nightstalkers are probably a little suppressed, but overall it may just be a percentage point or two, I haven't run the numbers.

How do you come to that conclusion? I don't see how having a low population would inflate the Win-Loss ratings of those classes.

- No text -

Yeah, I could go w/ less representative, but not inflated. Over 250,000 games with defenders accounting for 15% of the players that's still 37500 samples, quite a few. Defenders accounted for 2.2%, that's 5,500 samples, still a lot.(I have no idea where I cam up w/ 15%) The other classes would certainly be more representative with more samples.

edited text in gray/strikethrough.

Lets say you only have one class. In a 6v6 matchup you will have 6 wins, and 6 losses for that class, and it counts as two matches (one for each team) , for a perfect 50% W/L.

This acts as a forcing function. Classes that are played more frequently have a higher chance of being on both the winning and losing team, so their W/L will be pushed to 50%. Infrequently used classes won't be paired against themselves, so they are not subject to that mechanism.

It really becomes noticeable when you have a class a breakdown like Sunbreaker/Defender. Odds are you won't see two defenders on opposite teams very often, if at all.

Math to the rescue!

Ok, gotcha. I hadn't thought about the fact that more frequently used classes are matched up against themselves more often than infrequently used ones & thus are pushed towards a 50% w/l.

I used to play World of Tanks, and the US tanks were only about 5% of the total population, German tanks were somehting like 15% and Russian tanks comprised around 80% (there were a few one off chinese, japanese and french tanks at the time).

There were a handful of US tanks that would have 60% win rates, and the Russian players constantly called for nerfing them for being OP based on that number, but the truth was that there were so many Russian tanks, and Russian players, that when I did the calculations their tanks were having their W/L number suppressed by about 20%. The IS-7 (Russian premier heavy tank at the time) was better then both the Maus and the US T-30.