Sunday, December 16, 2012

League Analysis: Liandry's Torment

In my League Analysis series so far I've mostly gone over items that are/were considered underpowered, underused, or needed some quality of life changes. In this article, I'm going to go over something a little different and suggest changes to an item implemented recently, is pretty strong, and hasn't been nerfed in the patches following the Season 3 patch so far (thus Black Cleaver is ruled out). The item in question, as you probably can tell from the title, is Liandry's Torment.

The Problem(s)

The first issue with the item is the passive's doubled effect against slowed targets. While the item's passive (which deals 5% of the target's current health over 3 seconds, or 2.5% over 1.5 seconds) deals a fair amount of damage, it increases to a staggering 10%/5% (for AoE and damage over time) against slowed targets. To me, this seems like a very strange conditional for increasing the damage of the effect. This is especially so since it's slows are a common mechanic in League of Legends. This is because numerous champions can apply it through their abilities and certain items, like Frozen Mallet and Rylai's Crystal Scepter, can slow. This effectively means the passive might as well always deal the doubled damage effect (at least by later stages in the game, when it is more useful anyways due to larger health pools).

The second issue with this item is how it behaves with damage-over-time mechanics. First, let us look at how the passive on this item works mechanically, as quoted from the League of Legends wiki:
"Liandry's Torment procs every half a second. If your spell is single-target, Liandry will tick 6 times. If your spell is AOE (e.g.Poison Trail), Liandry will tick 3 times. And each tick does 5% divided by 6 (0.833333% of enemy's current hp) magic damage every half a second. Of course this proc damage doubles if enemies are slowed (e.g. Rylais Proc), dealing 1.6666666% magic damage every half a second."
In short, when a damage-over-time ability applies this effect, the duration is 1.5 seconds (the ticks still deal the same amount). However, this effect is refreshed each time it ticks (while the debuff is not already active on that target), potentially causing Liandry's Torment to deal more damage with damage-over-time effects (even AoE damage-over-time effects). Because of this, Liandry's Torment deals a lot more damage than intended with certain abilities, such as Teemo's Noxious Trap (which causes a total of 7 ticks, or 11.62% of the target's current health, not accounting for the fact the damage recalculates each tick, which causes this effect to typically do less damage than the listed amount).

The Solution(s)

I believe there are multiple ways to handle Liandry's Torment in its current state. To bring it in line it may need more than one of the suggested changes, which I will list based on priority:
  • I think the most important change is to remove the doubled damage against slowed targets. The base damage value may need to be increased to compensate (not by double, but to like 6-7.5% current health), but this still turns into a net nerf due to the large number of slowing mechanics available.
  • The next change (which is nearly as important) is to put some sort of internal cooldown on the effect or prevent subsequent damage-over-time ticks resulting from the same spell from causing. The Blackfire Torch's charge system could even be inherited to attempt to resolve this issue. The point is that with how damage-over-time effects are reapplying the Liandry's Torment passive, the damage done by said passive can (and usually is) more than the amount dealt by a normal application due to a single-target spell.
    • I'm especially partial towards having each given spell apply the effect only once. I would also add a charge system which consumes one charge per application of the effect (with a recharge time of one second and a maximum of five charges - charges are not consumed when this effect is applied to minions and monsters) to prevent some AoE damage-over-time effects (especially ones like Poison Trail, which can be applied multiple times over a short time) from applying the effect too frequently (as mentioned in the problems above).
  • Make a second version of the debuff that is applied by damage-over-time and AoE spells that deals half damage over the same duration of 3 seconds and reword the tooltip to match this change. This may even out the damage done by the item's passive since even though a damage-over-time effect refreshes the debuff, it now deals less damage overall. It would be very hard to program this in comparison to the other suggestions, however, because it would require numerous conditionals.
Here's to hoping that something is done about Liandry's Torment. For the record, I do play a lot of Singed and use this item along with Rylai's Crystal Scepter because of how absurd the synergy between the two items is.

8 comments:

  1. I accidentally stumbled over this blog while trying to find out if there is actually a champion where Liandry's is a viable option.

    While your assumption seems correct at first glance, you fail to factor in both item cost, magic resist and the diminishing returns the passive provides as opponents health sinks. In most situations, you're better off with quite a few other items, especially on Singed.

    ReplyDelete
    Replies
    1. You're right to a point. I am aware of how the effect diminishes. However...

      On the point about Magic Resist, it can be negated to a point by building Void Staff, similar to how AP champions would build such an item to "counter" high amounts of Magic Resist or AD champions "counter" high amounts of Armor with Last Whisper. While this passive doesn't exactly scale with Ability Power or other stats you can build besides Magic Penetration, it still has some form of scaling (i.e. enemy health).

      On the diminishing returns, that doesn't exactly negate the fact the item's passive is perfectly capable of inflicting quite a bit of damage (it can really add up, especially if you start factoring in more than one target). In addition, this doesn't exactly negate the fact the passive effect deals a lot more damage than it probably should in a multitude of cases because of its design, which I highlighted in the problems section.

      On the topic of cost, I didn't think it was particularly relevant to the analysis (incidentally, I believe it falls somewhere in the average cost of the highest tier Ability Power items). The fact of the matter is the passive just seems way too strong with the way it works (intentionally or unintentionally, especially on the damage-over-time part). Besides, if the item was seriously under-budgeted in terms of cost/strength of the item, that's more of a balance issue in that sense (i.e. Giant's Belt gives less health/gold than Ruby Crystal).

      As for Liandry's being a "poor option" (I'm assuming you're at least implying towards something like this - correct me if I'm wrong), I wouldn't say that's true. If anything, it might be one of his strongest buys later because Poison Trail can pretty much affect the entire enemy team and afflict the debuff constantly, dealing gradual damage that is essentially the equivalent of casting single target spells on the enemies repeatedly. Liandry's passive might as well be bonus damage added to Poison Trail (especially so since Rylai's Crystal Scepter is a decent buy on Singed).

      I find this works well with Singed's lategame presence since he's going to be running around the team fights with Poison Trail up (obviously protecting carries, picking off enemy carries, and the like), keeping uptime of both his Poison Trail and the Liandry's debuff on one or more targets. You can make it even stronger if you wish with a Void Staff (though you will sacrifice a bit of tankiness due to less effective health - your other two items should probably provide armor/magic resist).

      I will agree, however, that there are other fantastic options on Singed, especially with Season 3 (Ohmwrecker is a personal favorite of mine, even if it may be suboptimal to build).

      The point is that Liandry's works mechanically in a fashion that I described in the problems section. I feel this causes the passive to deal disproportionate amounts of damage in many cases, and even more so when applied to specific abilities (which are also pretty common). While this may be intended, I do not believe it is regardless of the factors given.

      Delete
  2. While I agree that Singed may be one of the characters that can pull off Liandry's, I'm awake enough to crunch numbers now (The numbers are taken from Leaguecraft):

    A "naked" Singed at L18 deals 70 magic damage/sec with poison trail. If we assume a medium MR of 50 on the opponents, which is very low, imho. It will be reduced by roughly 1/3rd. Result: ~45damage/second

    Liandry's (2.9k cost):
    Bonus through 70 AP: +21 damage
    Assuming an opponent with 2k life, we get 100 damage per proc. AoE malus and slow bonus counter each other out, so 100 damage over 3 seconds, ~33 damage.
    Total pure dps: (70+21+33) 124/sec
    Mres factored in (50-15=35): ~91/sec
    And we get +200 life

    Doesn't sound that bad, but let's look at other options.

    Abyssal (2,650):
    Bonus through 70 AP: +21 damage
    Total pure dps: (70+21) 91/sec
    Mres factored in (50-20=30): ~70/sec
    And we get 45Mres, plus a -20mres debuff our team profits from.

    Deathfire Grasp (3,000):
    Bonus through 100 AP: +30 damage
    Total pure dps: (70+30) 100/sec
    Mres factored in (50): ~66/sec
    We get 15% cdr, but, most important, the insane active:
    Initial damage: minimum 300pure/200 after MRes, assuming the target was at max life, and buffs the trail, letting it deal ~80 damage for 4 seconds.

    Void Staff (2,295):
    Bonus through 70 AP: +21 damage
    Total pure dps: (70+21) 91/sec
    Mres factored in (50-40%=20): ~76/sec

    Now, while the item may still seem good in theory, bear in mind that the damage it generates is still -highly- situational:

    -You need a target that has low MR, the higher the MR gets, the more negligeble the bonus damage is. The number used in my calculations is rediculously low for an endgame situation.
    -You need an ability that slows, barring this, you need an additional item for ~3k gold.
    -The target needs to have high health, which is unlikely in many situations, especially considering the bursty AoE and poke metagame seen on high ELO.

    With item slots limited to four (Shoes+Wards taking up the other two), the highly situational manner of Lilandry's will let it rarely see play. The only champions it seems viable on are casters with a long range slow (Zyra comes to mind) in a poke/siege team composition, or strong AoE initiators who would benefit from the AP (Amumu yes, Sejuani maybe).

    While I can understand your issue with the mechanic by itself, I feel you should consider other game mechanics, the available item pool, champions, and even the metagame situation, as well.

    My feeling is that Lilandry's is an interesting item, but there are very few champions that can actually benefit from it on such a consistent basis that the pick is justified. As the game evolves, this may change.

    Just think of enervating locket. The item was deemed interesting, but useless on most champions. With the appearance of Sona and a general tendency towards low cooldown champions, the item was discovered as extremely unbalanced to the point that it needed to be removed after rebalancing attempts failed.

    ReplyDelete
    Replies
    1. Thanks for the response. Your math is pretty logical given your example, but I would like to clarify some points.

      On the math for Liandry's, the damage is the same whether a DoT is ticking down someone's health (because it continuously refreshes the effect), an AoE spell constantly hits them at least every 1.5 seconds (which also effectively refreshes the DoT), or they are hit by a single target spell every 3 seconds. Since Poison Trail meets the first criteria because it is a DoT, despite being "AoE" in a sense, Liandry's actually deals full damage for as long as Poison Trail keeps ticking down their health plus 1.5 seconds (approx 33.3... damage every 0.5 seconds if target is at 2000 health).

      In your example, for the sake of argument, let us assume that the target dies at an equal rate (note - this is very flawed logic) and thus deals damage at the rate you mentioned (16.66... damage per 0.5 seconds - the true damage per time dealt will likely either be lower or higher depending on how attentive other team members are at identifying you weakened a target and rapidly execute them given how some abilities, including the 21st Offensive Mastery point passive work to greater effect on killing off lower health/%health targets). While this may seem your math is right, note how I approached the conclusion in a different manner.

      I find Abyssal to be rather situational. Funnily, I probably should build it more often but there are numerous other champions like Diana who benefit from its usage. Because the aura does not stack, having another is redundant (unless the champion holding it dies, I guess). Unlike stat increasing aura items, I don't think debuffing auras stack for the holders of the items. I won't disagree that this item is probably one of the biggest damage increasing ones otherwise.

      Deathfire Grasp isn't bad for doing something like dueling, but it is worth keeping in mind that while you gain an enormous amount of offensive stats, you do sacrifice some tankiness. While the 200 health Liandry's provides might not seem like much, it does increase to a sizable amount of effective health with some defensive stats (the same also applies to Abyssal Scepter).

      In addition, the damage per time greatly diminishes - even more than Liandry's in my opinion (unless the target has an inordinately low current health %), dealing less than Liandry's would after 11 seconds (deals 252.8 damage, which is the sum of the DFG damage plus bonus after accounting for Mres, every time DFG is used -> 252.8/11 = ~23 (rounded up) -> 23 + 66 = 89 DPS by 11 seconds). Even if DFG was somehow used again after 60 seconds, the damage per time is really low ((252.8 * 2) / 60 = 8.4266... DPS)

      So pretty much it is great for bursting, which is not something Singed particularly excels at even given Fling (the damage per time still decays rapidly even if you use Fling) and InvertedComposer's secret tricks like the Sky Auto Attack (which deals physical anyways). Also, it costs a little more than Liandry's, so I can't even give DFG the benefit of it costing less, thus granting more damage per gold.

      Void Staff in this example is pretty iffy. It's more effective against higher Mres targets. In fact, it does a lot less damage in this example (even accounting for the fact it's a lot cheaper).

      (More in next post)

      Delete
    2. In response to the rest:

      For your first statement, there is at least one slot open for building some Magic Penetration (I even suggested putting Void Staff and Liandry's Torment together in my earlier reply. This is especially potent given that flat Magic Pen applies after the % reduction). Even if there isn't, Abyssal Scepter is situational and Liandry's will always be an option for increasing damage (or at least is a more available option).

      Regarding your second statement, Rylai's is essentially a core item on Singed. It effectively increases his own movement speed by slowing enemies. It also increases Health, which helps Singed's tankiness by a bit and the item is pretty cost-effective due to how Singed benefits from all the stats so well (including the passive). In addition, as I mentioned, there are many slows available (even Singed's Mega Adhesive) when a teamfight breaks out (unless your team comp for some reason doesn't have that many slows).

      I wouldn't 100% agree with your last statement. While I am aware of this meta, that doesn't mean one completely ignores tank stats like health. In fact, to a point, tank stats have a good chance of negating the meta to a point if these %health damage items didn't exist/weren't that strong. Also, I have strong reason to believe that meta is going to wane to some degree (due to the Black Cleaver nerf, etc). Besides, in the examples given, where a target has pretty much the bare minimum stats of a level 18 champion (i.e. ADC stats), Liandry's STILL seems to deal more/a high amount of damage per time. (Also, Singed technically can "poke" enemy champions or at least zone them with good Poison Trail usage, even if everyone is in the midlane waiting for their minion waves.)

      On your note about item slots available, while you certainly should be getting wards up to a point, it isn't worth sacrificing the slot by the later stages of the game. You really need that full item build to become tanky. Besides, this is why items like the Sightstone exist (supports won't go broke anymore trying to get wards) and with the Oracles nerf, you probably won't need to have to provide wards as the tank by the late game. So in fact, by late game, (which is when Liandry's is most effective), you will have 5 items slots at your disposal (I guess it would be four if you counted Rylai's, but then it would be three if you accounted for always having wards). It's funny you mention Amumu by the way as that was a champion I considered for Liandry's as well (which probably helps the argument for how strong the item is).

      Incidentally, I did consider just about everything involving the game. Liandry's isn't the only item I have a problem with but it did rapidly shoot to the top of my list, right below the Black Cleaver. I can think off the top of my head of several champions that would benefit from this item greatly (even compared to other items and all that), so it isn't just like the Innervating Locket where it was OP on champions like Sona and Udyr because they could cast so frequently, thus constantly proccing the effect (it would apply to way more champions now). Even if Liandry's was in the same position as Innervating Locket, it really still should see some changes to the passive effect (probably a little easier to resolve than Innervating Locket, though I was thinking of writing something about that too since there was some mention of that particular item making its way back into the game soon).

      Delete
  3. I guess I might as well drop this here because it's relevant:

    http://www.reddit.com/r/leagueoflegends/comments/150cwy/the_top_5_champions_with_the_highest_win_rates/

    ReplyDelete
  4. Oh, nice find. Thanks for the link.

    I can see your logic as well as as you see mine, I think the diversion is mostly point of view. I guess only patch notes and S3 will prove who's assumption was right.

    ReplyDelete
  5. Protip: Nunu snowballs have a 3 sec cooldown at max CDR, apply a 3 sec slow, and scale at 1 to 1 with AP. Liandryballs on AP Nunu can hit for >500 initial damage and THEN burn 10% of current hp. Every 3 seconds. Torment the yeti at your own risk. :)

    ReplyDelete