Increase your damage score with science!
Sat, 10/19/2013 - 10:43
I don't really care about racking up high damage scores, but I do find this amusing to think about. Some (a lot) of you out there might already know this and just not be sharing it with people. I'd be very much surprised if this isn't something Darkcub keeps in mind, for instance, but for those of you who want to see that large score at the end of the game and haven't already thought this through, read on!
The theory
The most efficient weapon is one that will kill your opponent. For the purposes of gaining damage, you want to be just a little less efficient than that.
Suppose your enemy has 1080 health.
You have two swords. One does 1080 damage to that opponent and the other does 1079. Which do you pick?
Most people would jump straight at the first sword. It kills your opponent in one hit, after all! Can't beat that for sure! True enough, killing opponents in one hit will help you win matches, but you're not interested in that.
A thought experiment
We have our 1080 damage sword and our 1079 damage sword. Our opponent comes at us and we attack them with our 1080 damage sword, killing them instantly. They will now respawn in 15 seconds. Against that particular opponent, we get 1080 / 15 = 72 DPS.
Instead, we pick our 1079 damage sword. We struggle a bit more and spend 5 seconds getting our second hit in. They die and respawn in 15 seconds, but we inflicted two hits worth of damage. We achieved (1079 x 2) / (15 + 5) = 107.9 DPS.
See that? More damage by using less damage. Of course, a true min-maxer would hit with the 1079 damage sword first and THEN follow up with the 1080 one.
The idea
Basically, you want to get your opponent down to 1 hit point as fast as possible and then follow up with the strongest attack you've got. For each opponent you will gain
(opponent hit points + strongest attack - 1) damage
The execution
Let's start with a dual-penta striker Skolver first. We want to add up 1079 damage to them as quickly as possible. I'll leave this mostly an excersize to the reader, but I did find this combination in the damage chart;
3 x max damage Final Flourish = 290 x 3 = 870
1 x high damage Nova Driver = 204
870 + 204 = 1074
Very close! Now just finish your opponent off with the strongest attack you have. This would probably be the second swing on your Gran Faust or something:
1074 + 527 = 1601. You milked 48% more damage from the opponent than they even had health! Not bad at all!
Be careful that you don't use Polaris for this, as the shock would finish your opponent off and ruin the extra damage you could farm from them.
To get max sword damage and high gun damage, you would wear something like Black Kat Cowl + Snarbolax.
Other things to consider
Have a look at hits required to kill a dual-penta striker.
Notice the amount of damage bonus required to need one less hit against your opponent and avoid this amount of bonus at all costs!
At ultra damage, you need 4 hits instead of 5 to kill a skolver with Final Flourish. Let's play it out:
Max damage: 4 x 290 = 1160 damage extra for your damage score
Very high damage: 5 x 263 = 1315 damage extra for your damage score
Of course, you could argue that you're only getting 13% extra damage for 20% extra hits spent to kill your opponent, but you've got to consider that your opponent will take time to respawn as well.
To milk an extra hit out of your opponent with Rocket Hammer, you need Very High damage instead of Max as well:
Max damage: 3 x 378 = 1134 damage extra for your damage score
Very high damage: 4 x 347 = 1388 damage extra for your damage score
In conclusion
I hope you found any of this helpful. Happy damage-farming!
Sat, 10/19/2013 - 13:13
#2
I have to ask even before I finish reading. Have you proven that a second GF swing on a player with 1 % hp does full damage and not just the 1 % the player has left?
Edit: Also thx for sharing as usual :)
Sat, 10/19/2013 - 14:00
#3
Of course! I did a test today before publishing this.
It's a bit icky because the Gran Faust combo ended up at 849 damage and the opponent had 840 health. I should probably redo it with a wider gap to more sufficiently prove it.
Edit: I just did 1605 damage against a single striker using three Triglav charges. I had max damage, and my target was wearing Chaos+Skolver. Thanks to Redblades for volunteering for this.
Sat, 10/19/2013 - 14:28
#4
.
The fact that overkill is counted towards your damage score says alot about the coding behind Lockdown, doesn't it?
I noticed you missed one thing in your calculations - the increase in epeen that one gets from killing someone. I'm no master of statistics, but I think that it could make a very noticeable difference.
Sat, 10/19/2013 - 16:11
#5
That chart is a LIE!
I did a test, and with a max Combuster, it still took 3 hits on the guy. After the game, I inspected him, and he had no defense UV's. I think you need to redo that chart.
Sat, 10/19/2013 - 18:49
#6
@Gianor
That's what the chart says. At least 3 hits to kill anything you haven't deathmarked.
Sat, 10/19/2013 - 19:25
#7
oops
~nvm~
Sat, 10/19/2013 - 19:28
#8
Um, Zeddy...
The reason people go for damage all the time is because of a thing called "human pride". They like seeing their damage score higher than other people's. That very same pride will also motivate them to kill their opponents faster ("why should I spare him a turn if I can kill him now?"). I highly doubt anybody would be willing to show a little mercy just to get damage.
Sat, 10/19/2013 - 19:32
#9
@Klipik
Most of the entries have a row labeled "Chaos".
For piercing weapons, check "No armour".
There weren't any Chaos entries for Combuster, so I just added them. Thanks for reminding me!
@Cradarc
But if you manage to kill opponents using more hits to get more damage out of them, you will rub it in their face extra hard how much better you are than them. In fact, you should almost feel ashamed about needing max damage!
Sat, 10/19/2013 - 19:38
#10
@Zeddy
The difference is that you aren't going to stop in the middle of the battle and announce you are sparing them so you can farm damage. To anyone watching, it would seem you need 2 hits to kill them, implying less skill. Remember damage farmers aren't deep thinkers. They judge performance on the superficial level.
Sat, 10/19/2013 - 19:42
#11
@Cradarc
Why would you announce anything? Just switch your weapons around. Open the fight with a Nova shot, hit them 3 times with Flourish and then hunt them down with your Faust as they'll no doubt try escaping at 1% health. This can easily be worked into your rhytm if you're serious about your damage farming.
Sat, 10/19/2013 - 20:01
#12
@Crad
I don't think anyone's dumb enough to determine skill based on number of hits it takes to kill someone.
Sat, 10/19/2013 - 20:19
#13
.
Zeddy. A good chunk of these mathematics will be ruined after the gunner update. Yeah, you know which one I'm talking about...
"One spamgun to ruin them all. One spamgun to find them. One spamgun to hit them all, and in the darkness freeze them."
Sun, 10/20/2013 - 01:08
#14
Just write freeze trollaris already Arkate, I doubt anyone serious about LD doesn't know already..
All it will do is make Skolver the N1. armour in LD again, in favor of the reigning Chaos clones.
Besides, with any luck OOO does a rebalance right and decides pulsar lines only get a low chance on minor freeze/shock/fire.
But who am I kidding, it's not like the crystal balancing went that well /:
Sun, 10/20/2013 - 02:23
#15
I think/hope they will make all the pulsar line guns have 2 shot per clip.
Sun, 10/20/2013 - 10:11
#16
.
Except Feyi, Skolver only gives bonuses to swords...
Sun, 10/20/2013 - 10:17
#17
-
bkc freeze low
skolver
gg.
Sun, 10/20/2013 - 10:26
#18
BKC freeze medium
Snarbolax
Now you even have some shadow defence!
Sun, 10/20/2013 - 21:10
#19
For the sake of argument :P
@Zeddy
Nobody is perfect. Allowing the opponent to survive will always increase the risk of taking damage.
Also your DPS formula is flawed. What usually happens is that after killing one person quickly, the striker searches for another person to kill. They don't just stand there and wait for the same victim to come back. With that in mind, I don't think the total damage dealt over an entire game would be affected much.
@Klipik
Suppose in a hypothetical 1v1, Person A wins over Person B. Would a typical spectator be more impressed if Person A killed B with two quick hits or two glancing blows and a final killer one? The quicker, more efficient kill is just innately more satisfying. As I mentioned to Zeddy above, doing more hits also runs more risks. There's a reason GF and DA are so common in T3. People like them because they like the hitting power. Is it stupid? Yeah. But are people stupid? Yes, apparently lots of them.
Sun, 10/20/2013 - 22:16
#20
@Cradarc
"What usually happens is that after killing one person quickly, the striker searches for another person to kill."
So what you're saying is you're too chicken to do this.
(Or alternatively care about other things than the damage score at the end of the game. Which is fine by me.)
"What usually happens is that after killing one person quickly, the striker searches for another person to kill."
Keyword: "searches". That takes time. You could be milking 40% more damage out of your current target right there on the spot.
Sun, 10/20/2013 - 22:47
#21
"I think/hope they will make all the pulsar line guns have 2 shot per clip."
Such a good idea!
And give the iron slug/callahan 3 shots per clip and the ability to move while firing like the alchemers.
Balance ftw!
Mon, 10/21/2013 - 03:09
#22
you know I always wondered
why I did more damage with swords, even though I kill almost just as quickly with guns. I just assumed overkills weren't possible. :| Can't really get decent overkills with guns (unless it's a charge shot I guess).
The most common Sword combo for quick-quick kills is a 1-2 with a FF followed by a DA/GF combo. A lot of good swordies will use this to destroy people and if their timing and positioning is on point there's little to be done to counter it. The overkill with that is pretty decent so obviously it hugely inflates the final damage score they get.
Your final score aint damage done but swings swung.
Mon, 10/21/2013 - 18:35
#23
@Zeddy
1) I'm saying a more important factor in average DPS over the whole game is you, yourself, surviving. Your opponents are not always guaranteed to be complete noobs. Just because you are a damage hoarder doesn't mean you can avoid getting hit all the time. When your health gets low you are forced to waste time healing or risk death and respawn delay.
2) Sure, you can potentially take 40% more out of the current target, but what % is added to the time fighting the current target? Using your supernova/flourish example, landing those hits take far longer than a quick GF combo, especially if your enemy is not just standing there.
A quick two hits may cost ~ 30% of potential damage (40/140), but it will save you more than 50% of the time, and some good percentage of death risk.
3) I want to reiterate that your perfect condition analysis is flawed. In a fair duel (both full hp, 1v1), which is the only context in which your strategy is viable, both adversaries are cautious. One guy is not just going to stand there while he gets ripped.
Scenario 1:
You go in with your flourish, hit him twice. During that time, he manages to graze you with a GF. You hit him again with a flourish and pull out the supernova. He tanks one Supernova hit but succeeds in GF comboing you. GG.
Scenario 2:
Alternative scenario is he runs away to heal after you land 3 hits. Oops, you went from a potential 140% to <100% because the your enemy is smart enough to look at his hp.
Scenario 3:
You manage to kill him, gaining 140% the damage you would have otherwise, but giving him two extra hits to survive allows him to land a single GF hit. Let's say you fight another guy like him. Then you have to use ~10 seconds of your time healing instead of searching for the next target. This is added on to the extra time you already spend fighting each person for that extra damage. There's also a chance that someone may ambush you before you reach the health pad, delaying you with respawn.
4) If we assume your perfect conditions then wouldn't it be even better to use a Dread Venom Striker on someone until they are at the lowest hp possible before hitting them with a Triglav charge while they're poisoned?
Mon, 10/21/2013 - 20:02
#24
Very interesting. I didn't know that overkill actually counted toward your damage.
I shall now use this to my advantage for my quest to 40K.
Mon, 10/21/2013 - 21:56
#25
heres something to think about
Chaos set with piercing max on both. You play with ancient plate set. You have a Ff and a Gf, the gf does more damage, but is too slow with this set, This is where combo comes in. While the ff will only do about 2 pips, you attack with it to make him back off, then you pull out the slow gf and they run into it because its slow, so even if they ran 9 blocks, your gf is slow so they will run into it. So the math here is 80+720=800, so if you use this combo, in 3 hits, youve gained 2k and also WARNING: Tjis combo is meant for ancient plate, with anything else it will go too fast and you will get slaughtered
Mon, 10/21/2013 - 23:03
#26
^
The amount of wrong things is that post is TOO DAMN HIGH!
Tue, 10/22/2013 - 01:25
#27
@Cradarc
1) 2) 3) and 4)
When people flaunt their damage score, it's not usually under the implication "Look at how lucky I got with this damage score! I got sooo lucky!"
Instead, they usually imply "Look at how much better I am than other people. Including you! I'm much better than you!".
These people, if they are indeed better than anyone else, should have no trouble using this method. They're better, after all! Risk vs Reward.
4) No. You will notice how I didn't give an example of how you can reduce your opponent's HP to 1 before dealing the final blow, even though could devise several such methods: All of them are too slow. Opening a fight with a Nova Driver shot before you close the distance for 3 FF hits and then finish with your heavy sword, this is a quick method. You can train yourself to act in this manner without even thinking about it.
2) and 3) This is how you'd do it under ideal conditions. I'm not telling you to perform every single kill you make like this. Indeed, in practice that's nearly impossible. You would have to calculate and have ready a method to optimally milk strikers without pentas. Recons. Guardians. With and without pentas. Mabe some of them only have trihearts! Maybe some of them have unexpected defence UVs! Maybe most people you meet are hurt in advance? Maybe they're hazebomber! How are you gonna FF hazebombers with any kind of expediency?
I can't tell you how things are working out in practice. All I can do is describe an ideal, theoretical scenario and then you can worry about bridging the gap between those two.
And yes, sometimes killing fast is going to be the most efficient solution. If you're surrounded by 6 guys, you can be damned sure that taking them all out in a couple of GF combos is going to net you more damage over time than whittling down each one in advance. (Unless, perhaps, bombs will work better? I dunno, man! Crazier things have happened!)
It's a pointer. If you're looking to increase your damage, keeping this theory in mind can help you work out a method for whittling enemy HP down and setting them up for overkill (since you are now aware that it is a factor, whereas previously most people responding here weren't). Whether that starts when they have 1, 50 or 300 HP left is entirely left to your own skill and imagination.
@Terrty
http://forums.spiralknights.com/en/node/90606
You should not get Chaos with piercing max. There are more useful things to throw your money away at.
Zee-cores, for instance.
Tue, 10/22/2013 - 19:59
#28
@Zeddy
1) Just because they don't say it doesn't mean it's not true. Now, I'm not saying skill is irrelevant, I'm just saying luck is quite a significant factor in T3. This is supported by the fact that all "pros" die more often in T3 than in T2. Every mistake a striker makes is more costly.
Your skill compared to that of your opponent determines the probability that you are likely to land a hit before he does. It's like flipping biased coins. You flip one with 80% chance head. Your opponent flips with 50% chance heads. However you choose to get heads 5 times to win while he chooses to only get 3.
^This is exactly why trinkets are so valuable. A more skilled opponent that doesn't have as much health essentially has his skill nerfed because he needs more hits to kill.
2) How fast does a single Nova bullet travel? How hard is it to dodge compared to a GF rush?
What's easier to land? 3 consecutive flourish swings without getting hit or 1 GF swing without getting hit?
What's easier to land? A GF second swing after a GF first swing that hit or a GF second swing after 3 consecutive flourish swings that hit?
No matter how much more skilled you are than your opponent, they won't just stand there while you spam 4 consecutive hits at them. Two consecutive hits is much more easy to pull off.
My example is just an extreme of what you are proposing. If you acknowledge that my example would be less effective, how can you be sure yours would be more effective? Since it's all theoretical, how can you claim one method fails but the other method excels when they come from the same vein of thought?
3) Your original argument was based on the fact that if you kill them too quickly the respawn/search-for-new-target time will decrease your DPS. However because this strategy only works for rare ideal situations, your damage dealt may actually decrease because you devoted more time on that 1v1 rather than killing quickly and finding a distracted mob to spam at.
My Claim: While this seems good on paper, it would actually hurt damage performance in practice. Hitting hard and fast is the best way to get damage. If you had a choice between two weapons that both kill in the same number of hits with relative ease, then obviously more overkill is better. However, using weaker weapons just to maximize overkill in certain situations will hurt overall performance.
Wed, 10/23/2013 - 00:53
#29
@Cradarc
1) I'm sure if you asked Feller, he would tell you he could pull this off 100% of the time using Auto-Target because Auto-Target lets you win everything.
2) Here's the kicker. You don't have to hit with the Nova bullet. Your opponent will die after 3 FF + GF with the bullet or without it, but the Nova shot will add to your damage. Open the fight with the Nova bullet, maybe it'll hit, maybe it won't.
Like you said, your opponent is probable to retaliate as well. While your opponent is doing that, I assume you want to be outside of the range of their sword hit, and you could fire off the Nova at that point, maybe? There are a lot of ways to pull this particular combo off, and, more importantly, it doesn't have to be this particular combo. That one just hit really close to 1 hit points so in terms of damage per kill it paid off pretty well. You could go lower risk with something akin to 2 x Gran Faust + Polaris + a third Gran Faust or 2 x Flourish + Gran Faust Combo.
Your example with DVS isn't extreme at all, it's just inefficient. Against which opponent and at what level of damage bonus will DVS take your opponent down to 24 HP or less? (You need less for the poison to pay off.) Why are you using a Triglav charge when Divine Avenger's charge is easier to land and hits harder? Why are you poisoning your opponent when you could get a teammate to deathmark them?
There, that's an extreme example.
3) Honestly, though? I have no idea. Any given member posting on this subforum will have orders of magnitude more experience damage-farming and being a striker than I do, including you. I've been hoping for someone to provide feedback on the practical aspect of it.
I can run numbers, however. Wanna run numbers?
Let's say you kill all of your opponents using 3x Gran Faust. (Is this what you guys usually do?): 3 x 413 = 1239 per kill
In a seperate match, you kill all of them using 3 x Flourish + Nova + Gran Faust. 3 x 290 + 204 + 413 = 1487 per kill
In the first match, you get 14 868 damage. That's 12 opponents. A Lockdown match lasts for about 7 minutes, so that's a kill every 35 seconds.
In order to match that score, you will need 10 kills in the second match. Assuming a 7 minute match, that's a kill every 42 seconds. In other words, the second method must not add more than 7 seconds to your fights in order to be efficient.
We'll up the ante. You're aiming for 40k damage.
That's 33 kills (40887) with the first method and 27 kills (40149) with the second.
We're looking at 12.7 seconds between kills for the first instance and 15.5 seconds for the second. In this instance, the second method is not allowed to take more than 2.8 seconds longer than the first one.
Does it? I'm sure you or someone else will let me know if it does. Admittedly, 2.8 seconds is not a lot of time to squeeze in 3 flourish hits and a gunshot in place of 2 heavy sword hits. And ideally you want it to be even faster than that for this method to actually be the superiour method to use. It does look grim for this method at higher levels of damage farming.
One thing to keep in mind is that in matches with ridiculously high damage numbers around the range of 40k, the farmer is very likely to be facing a team full of really inept opponents and would probably not have any problems at all pulling the combo off.
Wed, 10/23/2013 - 00:56
#30
A skilled player or an AT noob could easily do overkills. Your wall of text might be true for you though.
Wed, 10/23/2013 - 01:31
#31
@Zeddy
Chances are that if your opponents are a bunch of really inept opponents, many might not have double pentas and stuff.
Wed, 10/23/2013 - 02:47
#32
@Krakob
All the more important to overkill them for extra damage!
Wed, 10/23/2013 - 03:09
#33
@Zeddy
Indeed, but I believe it means you need to use a strategy that is not Nova+3 FF+SS combo! Some might have one penta, some might have tetras, etc. You said it yourself earlier, IIRC. In the end, you'd basically need a whole match to optimise your strategies against that certain team.
Wed, 10/23/2013 - 07:09
#34
Hey Zeddy,
Just need fast confirmation (Thats why I posted another thread)
Does Polaris' bullets move faster than it's 4* version?
Thanks!
~Tech-Knight-Die
Wed, 10/23/2013 - 07:09
#35
~
First, the Lockdown Science Megathread; now, the Lockdown Science MegaArgueWithZeddyOverTheLineBetweenIdealityAndActuality!
Wed, 10/23/2013 - 16:42
#36
@Triglav charge
Did you get both hits of the triglav charge to hit? Under normal circumstances, the second hit gets nullified by invincibility frames (though you can get around this by making your target hold a weapon charge while you hit him).
Wed, 10/23/2013 - 19:07
#37
Yay arguments!
For the record, I don't really have a legitimate reason for challenging Zeddy. It's not like he's forcing people to do this. I just like forum debates about meaningful topics.
Thu, 10/24/2013 - 06:07
#38
You and me both, Cradarc.
Sat, 10/26/2013 - 03:55
#39
Mind=Blown
wow....i didnt know that XD
now i know to use my Final Flourish instead of my Gran faust :D
Sat, 10/26/2013 - 14:49
#40
Coliseum Recipes.
Need to be cheaper.
Wed, 10/30/2013 - 00:28
#41
That math
If you hit someone who has 10hp left with a, let's day, 760dmg swing. Then I'm pretty sure it'll be counted as 10dmg.
Wed, 10/30/2013 - 01:54
#42
@Scirio
Then how did I get 1605 damage out of killing a single striker when I tested this?
Wed, 10/30/2013 - 07:11
#43
@Zeddy
Cardboard servers.
Thu, 10/31/2013 - 05:37
#44
Okay. Now that you've destroyed my argument with an admittedly pretty good counter-argument, do you have any evidence to support your claim about how damage works?
Thu, 10/31/2013 - 06:00
#45
We all have the same cardboard server so it doesn't matter? Unless you were joking and I missed it.
Thu, 10/31/2013 - 06:28
#46
I think Scirio gave up there.
Fri, 11/01/2013 - 04:31
#47
Back again.
Cardboard servers was a joke, seeing as the matchmaking seems to be handled by a 5 year old with a crayon up his nostril. But that's a discussion for another day.
I am fully convinced that dmg scores in this game are not spot on. There have been more matches where my score differed from what I had deduced than I care to admit.
That aside, killing somene who has 10hp left with a 720dmg attack will score you 10dmg simply because it's not logical to give someone points for something they did not do i.e. you did 10dmg, not 720, although the attack had the potential to inflict 720, it did not simply because the enemy had 10hp left. You can't get points for something you did not do.
Fri, 11/01/2013 - 12:57
#48
@Scirio
But do you have any actual proof derived from tests like Zeddy does?
Fri, 11/01/2013 - 14:21
#49
There's logic, and then there's OOO.
Fri, 11/01/2013 - 23:15
#50
At this point, I'm fairly sure your counter-arguments belong in the suggestions subforum, as you're not arguing about how things are, but how they should be.
how a single stat in a generic minigame with no significant reward can change the economy of the game so substantially. I doubt anyone will actually puts this much thought into killing someone in LD. Most of LD is essentially fight or flight type kinds of battles