A Comparison of Hero DPS

As long as I like theoretical simulations I think there are several major flaws in your test:

1) you can't have comparision of single target dps and multi target dps in one list, just can't! You should make one list of single target dps and second of multi target (with same amount of cleave targets) - maybe lists of dps to 1 target, double target, three targets...

2) I didn't find your method of the actuall measuring. I think the ideal way (not sure if it's possible in dota2) should be letting the hero attacking for couple of minutes (minimum of 5, ideal is somewhere around 1 hour) and counting the dmg hero did.

3) I didn't find if you used spells, if you used them on cooldown and you didn't touched some mechanics that may make difference in dps (such as clinkz ulti or precasting clinkz strafe and wait before engaging etc)

4) You should (if it's possible) make test for different amounts of armour and then you can make comparision on how the armour changes dps of specific hero, some graphs etc (i suppose AM and PL will go up with increasing armour)

Now it maybe looks nice but has no value as research... sorry

These maths, I like them.

ANOTHER!

Meepo apparently was bugged and only the DPS of Meepo Prime was counted. I'd be willing to try it again, as Meepo would do a considerable amount because, well, 100% stat sharing. Imba hero.



1) Checking the multi-target DPS of heroes that only do single target damage, ie most of the ones on this list, is pretty pointless.

2) It varied from hero to hero since their damage is spread out - Chaos Knight went over the duration of his ult, Terrorblade over the duration of Meta, and Clinkz over two Strafe cycles. Essentially, with the exception of Medusa and Luna because perfect scenarios for them are highly unlikely, the tests were conducted with an attempted realistic teamfight situation.

(For those wondering, Luna with only two glaive bounces still did 2.2k, and Medusa loses 700 DPS for every secondary target she isn't hitting. I did take that into account)

3) Spells were included, such as Death Pact and Adaptive Strike. Morphling, for example, began with a shotgun and then used Adaptive Strike on cooldown. This was the highest DPS over two durations of Manta Style illusions.

4) Sadly, I can't do this with the target dummy feature. I'd set up more complex and reliable tests, but it'd be a hell of a lot of work. The armour was the reason I tried to shy away from heroes with lots of -armour like Slardar and Razor.
______________________________________________

Sub-optimal builds come down to me not really knowing what items to build on some of the heroes. I was tossing up between BKB and Satanic for Medusa's last item, and in the end decided to go with Satanic. Despite this, the fact that there's literally 1 raw damage difference between them should have basically no impact on the test. I made a similar note with Troll Warlord. Buying a BKB ****s on your damage output, unsurprisingly.

10/10, would math again
XD

I believe the point of comparision is to give all compared subjects same conditions in order to recieve results that are actually comparable. So all heroes must have same amount of targets and attacking for same time period (the longer, the better) or your results are just rough estimations and not comparable results.

Sorry if I'm offensive but doing research has "laws" everybody has to adhere so it's actuall research and not waste of time because your comparision has the potencial to be research and not just some random numbers without specifications of your experiment so noone can repeat your experiment etc...

There's one thing I'd like to point out - if you're dead, you can't deal damage.

Which is why, although some heroes can deal a lot of damage, they don't seem to deal that much - either they are kited out, or the just plain get stunned to death.

Leshrac has a thing or two to say about this

Which is...?

One of the main ways to counter Lesh (as demonstrated in the patch where he was OP) is to restrict his movement and stun him to death...

Usually this is true, however Dota works a bit differently. If every hero is hitting one target, that's going to restrict some heroes whose entire thing is their AoE damage, like Sven and Luna. The purpose was to test the DPS 'under ideal conditions'. Ideal conditions are different for different heroes. Comparing the single target DPS of Medusa and Terrorblade tells you nothing, because Medusa does area damage. Setting up a multi-target situation for the both of them is also pointless, because TB doesn't deal area damage.

And yeah, of course kiting and stunning reduce the DPS of a hero. That's obvious, and why you don't ever see Sven actually doing that much damage because you can kite him outside of Warcry (because if there's something 6.83 taught us, it's really hard to kite a hero who moves close to max movespeed a la Tiny and Troll Warlord, hence why SnY Troll is still a thing). It's why the tests include BKB on heroes you'd commonly see it on, exceptions being highly elusive heroes like Morphling. But it's also not really relevant. These tests were conducted, as I've stressed, 'under ideal conditions', which means that naturally they're going to be limited in their usage and applications. Were the available research tools I had more effective, I'd be doing more complex tests that are more useful and reliable. This was more a case of testing the efficiency of damage steroids, more so than actually measuring a hero's damage output in a real game.

Addendum: I might do a video of this for better transparency. It's a fairly easy experiment to replicate.

I really like this thread, I started playing with some of my friends and I'm the position 1, so I think this is a good way to select the best carry according to the situation. For example if I'm going to figth a Phantom Lancer I'm not going to pick a Terrorblade, I will pick a Sven and then proceed to wreck his face. Every hero is strong in some way and weak in another.

In a separate note, I'm playing 5 heroes with my friends right now: Phantom Assassin, Sven, Slark, Troll Warlord and Wraith King. Should I try another heroes or those are good enough? I will accept any advice.