Solo question - When do you stop trying to slow the mob?

[remen]

This also could be posted in the tank section, but my question basically boils down to at what mob % do you generally give up on trying to slow with weapon proc and switch to a dps weapon?

I know this is highly variable based on which class we are talking about, what type of mob it is, and whether or not it flees. I'm just wondering if people have any general rules of thumb they go by?

Some examples might be ranger swapping out earthcaller for a higher dps main hand, I usually do this at about 40% if the mob still isn't slowed, or 30% on a mob that doesn't flee. Or warrior swapping willsapper for silver whip of rage (better ratio and rune proc) which I do at about the same percents as the ranger. I also wonder if anyone has actually run any math on the tipping point of trading off reducing incoming damage vs. finishing the mob off more quickly. These would be situations where the idea is soloing many mobs in a row as possible for xp before needing to stop and med / regen.

If you have other class / weapon combos that you use to slow and solo with, I'd be interested to hear your strategy on how you use those.

[bcbrown]

I've never done the math before but I've been curious too. After running some numbers just now I don't think it's ever optimal for a ranger to swap from a slow-proc to a pure-dps weaponset. The 50% slow is just too powerful.

I'll do some of the math for my ranger against Basher Nanrum, as that's the numbers most readily at hand. He's a bit of an odd duck (high health and low-ish dps) but hopefully it will at least illustrate how to go about answering the question.

With swarmcaller I do 33 dps against him (with 34% haste), and with sboz/swiftwind I do 47. He does 11 dps to me, and has 15235 hp. I've got about 130 dex, so I'll assume 1 proc per minute. With a delay of 41 and 34% haste that's 19.6 swings/minute or about a 5% proc chance per swing.

I think what we want to solve for is minimizing the damage taken. The formula for "time to kill" is mob_health / player_dps, and the formula for damage taken is mob_dps * time_to_kill or mob_dps * mob_health / player_dps.

Let's imagine a scenario now. The mob's at some health, and we're about to swing. We can either swap to sboz/swiftwind or keep the swarmcaller in. No matter whether it procs or not we're going to swap to sboz/swiftwind after this next swing (because we think we're close to the threshold where it's better to swap to dps weapons even if we don't proc a slow).

Every time you swing the swarmcaller, you extend the fight by a bit. For 3.06 seconds you did 33 dps instead of 47 dps, for a net loss in damage done of 47*3.06 - 33*3.06 or 42.84 damage. At 47 dps that's 1.1 second. So each time you swing the swarmcaller you "spend" 1.1 seconds of extending the fight for a 5% chance of cutting damage in half. 1.1 seconds at 11 dps is 12 damage. So you "spend" an extra 12 damage taken for a 5% chance of cutting all future damage in half.

A 5% chance of a 50% damage reduction is equivalent to reducing expected damage by 2.5% (or 0.5 * 0.05). 480*.025 is 12, so until you expect to take less than 480hp damage it's a better bet to keep the swarmcaller equipped.

[Naethyn]

Warrior: Start the fight with a snare weapon and wait for it to proc. From there you swap to trunch and arrows. Run in, turn on attack, get a slow proc, run away, fire arrows, repeat until slowed. Once the mob is slowed equip the blood points and finish it off.

[Drueric]

Originally Posted by remen [You must be logged in to view images. Log in or Register.]

This also could be posted in the tank section, but my question basically boils down to at what mob % do you generally give up on trying to slow with weapon proc and switch to a dps weapon?

I know this is highly variable based on which class we are talking about, what type of mob it is, and whether or not it flees. I'm just wondering if people have any general rules of thumb they go by?

Some examples might be ranger swapping out earthcaller for a higher dps main hand, I usually do this at about 40% if the mob still isn't slowed, or 30% on a mob that doesn't flee. Or warrior swapping willsapper for silver whip of rage (better ratio and rune proc) which I do at about the same percents as the ranger. I also wonder if anyone has actually run any math on the tipping point of trading off reducing incoming damage vs. finishing the mob off more quickly. These would be situations where the idea is soloing many mobs in a row as possible for xp before needing to stop and med / regen.

If you have other class / weapon combos that you use to slow and solo with, I'd be interested to hear your strategy on how you use those.

If u are a solo ranger, you snare, run away and shoot them with a bow. If its animal you can just fear kite it.

[bcbrown]

I fucked up some of the arithmetic in my first post. I'm still not 100% sure it's correct but the python code in the spoiler tags should tell you at what mob remaining health you should swap out of your proc weapons to your dps weapons. I'll do a little more testing, probably tomorrow.

 

def remaining_health(low_dps=33, 
                    high_dps=47, 
                    mob_dps=11, 
                    proc_delay_seconds = 4.1/1.34, 
                    procs_per_minute = 1, 
                    slow_reduction = .5
                    ):
    damage_spent = proc_delay_seconds * (high_dps - low_dps)
    print("By swinging the proc weapon once you lowered damage dealt by " + str(damage_spent))
    time_extended = damage_spent / high_dps
    print("By swinging the proc weapon once you extended the fight length by " + str(time_extended))
    extra_damage_taken = time_extended * mob_dps
    print("By swinging the proc weapon once you took extra damage of " + str(extra_damage_taken))
    swings_per_minute = 60 / proc_delay_seconds
    procs_per_swing = procs_per_minute / swings_per_minute
    slow_expected_reduction_percent = slow_reduction * procs_per_swing
    damage_needed = extra_damage_taken / slow_expected_reduction_percent
    print("Damage needed to take to outweight extra damage taken: " + str(damage_needed))
    damage_to_do = high_dps * damage_needed / low_dps
    print("Mob health remaining before die/flee when you should swap out of proc weapon: " + str(damage_to_do))
    print("at that remaining health, possible outcomes:")
    print("using dps weapons")
    length = damage_to_do/high_dps
    damage = mob_dps * length
    print("length of fight: {}, damage taken: {}".format(length, damage))
    length = proc_delay_seconds + (damage_to_do - low_dps * proc_delay_seconds)/high_dps
    damage_no_proc = mob_dps * length
    print("swing slow weapon and then switch to dps weapons, no proc")
    print("length of fight: {}, damage taken: {}".format(length, damage_no_proc))
    print("swing slow weapon and then switch to dps weapons, proc")
    length = proc_delay_seconds + (damage_to_do - low_dps * proc_delay_seconds)/high_dps
    damage_proc = slow_reduction * mob_dps * length
    print("length of fight: {}, damage taken: {}".format(length, damage_proc))

Anyway, the way I've been treating it on my ranger is that if the mob is easy (Grobb Bashers, hill giants) I skip the swarmcaller. If the mob is challenging (geos), if I've been using the swarmcaller with no proc and I'm at ~30-40% I probably won't be able to win a fight toe-to-toe even if I get a proc, so I'll swap to sboz/swiftwind and when I hit ~20% I'll root the mob, back up and switch to archery until the root wears off. By that point I've probably regenned enough health that my dps weapons will be able to finish the fight.

[Drueric]

[You must be logged in to view images. Log in or Register.]

[kjs86z2]

until the mob is slowed