Kimball Vs. Gnolls

[Wolfgar]

In an attempt to answer the question of whether Kimball should be using his bow or melee weapons I thought I would analyze the options for Kimball vs gnolls. I may not have all of the factors correct.

option	point blank range	short range	one weapon	two weapons (1)	two weapons (2)	one weapon
weapon	18 Str bow	18 Str bow	+2 long sword	+2 long sword	+1 hand axe	+3 giant slayer
AC5 unmodified	9	9	9	9	9	9
18/50 str	-	-	+1	+1	+1	+1
17 dex	+2	+2	-	-	-	-
elf	+1	+1	+1	+1	-	+1
specialization	+2	+1	-	-	-	-
two weapons	-	-	-	+0	-2	-
magic	-	-	+2	+2	+1	+3
AC5 modified	4	5	5	5	9	4
percent chance to hit	85%	80%	80%	80%	60%	85%
---
weapon average damage	3.5	3.5	6.5	6.5	2.5	13
18/50 str	-	-	+3	+3	+3	+3
ranger	-	-	+7	+7	+7	+7
specialization	+2	+1	-	-	-	-
str bow	+2	+2	-	-	-	-
magic	-	-	+2	+2	+1	+3
base average	7.5	6.5	18.5	18.5	13.5	26
multiplier	2	1	1	1	1	1
average damage/hit	15	6.5	18.5	18.5	13.5	26
---
attacks/round	3	3	1.5	1.5	1	1.5
average damage/round	38.25	15.6	22.2	22.2	8.1	33.15

The last row is the product of percent chance to hit, average damage/hit and attacks/round. As you can see, on average, using the 18 Str bow from point blank range does the most damage (38.25) followed by attacking with two weapons (30.3=22.2+8.1). I added the +3 giant slayer column for comparison purposes.

Furthermore, since gnolls are only 2 hit dice a single blank range shot, on average, will take one down. With the ranger bonus, any melee strike is also likely to take down a gnoll; however, the difference is in the rate of attack and the percent change to hit. The melee option also requires moving into combat between targets. A well positioned archer will have many more targets without having to move. Or can use split fire to get closer and still get shots off.

I checked other scenarios and regardless of the AC of the opponent, there is currently no situation where Kimball will cause statistically more damage using a melee weapon. Without the ranger bonus, even shooting from short range is statistically better.

This has to be weighed against the average damage the opponent will do depending on Kimball's AC whether he gets his DEX or shield bonus. I'll create some tables for that as well later.

[Dead Greyhawk]

Nice chart. Gnolls are L sized creatures, so longsword damage is 1-12 and hand axe is 1-4. Changes your numbers some, but not your conclusion. I think the AC analysis will be enlightening as well.

[venger]

Damage-wise there is no doubt that the bow is superior.
Over 2 rounds it's like 10-20x6 vs. 13-24x3.

BUT: (and I'm sure your AC chart will show this)

Kimball is AC 5 against charging opponents when using the bow. AC 3 in melee. And at least AC -2 (possibly better) with sword and magic shield. It's the difference between them needing an 18 or better to hit and an 11 or 13 to hit.

In my opinion, the bow is overkill and leaves Kimball wide open to attacks. (literally the worst AC in the party)

As a first rank fighter, I'd prefer to deal slightly less damage and not get hit as opposed to taking many hits and having to beg for CLW spells at the end of combat. It's the Protection From Evil vs Cure Light Wounds argument, right?

With a longsword Kimball will go first (in his 2 attack round) and has a chance to slay the creature before it gets to attack, regardless of initiative.

Also, a quick comparison to the other fighters in the party vs. knolls for some perspective:

Kimball is 3/2 and +4 to hit / +12 to damage. (modified Thac0 = 10) (damage per hit: 13-24) (AC -2)

Basil is 3/2 and +6 to hit / +8 to damage. (modified Thac0 = 10)
(damage per hit: 10-24) (AC -1)

Kazan is 3/2 and +7 to hit / +10 to damage. (modified Thac0 = 8*) Assuming a +1 spear. (damage: 11-18) (AC 0)

Jespers is 3/2 and +3 to hit / +3 to damage. (modified Thac0 = 14*) Assuming 4th level double-specialized fighter.
(damage per hit: 4-15) (AC 2) (chain +1?, shield +1)

Kimball is no slouch in melee.

[Wolfgar]

Jun 17, 2009 6:47:22 GMT -5 Dead Greyhawk said:

Nice chart. Gnolls are L sized creatures, so longsword damage is 1-12 and hand axe is 1-4. Changes your numbers some, but not your conclusion. I think the AC analysis will be enlightening as well.

Thanks. Also fixed knolls.

[venger]

With the Giant Slayer column added it looks like the best way to kill a giant would be dual-wielding Majstang and handaxe/dagger.

[Wolfgar]

Jun 17, 2009 9:13:15 GMT -5 venger said:

As a first rank fighter, I'd prefer to deal slightly less damage and not get hit as opposed to taking many hits and having to beg for CLW spells at the end of combat. It's the Protection From Evil vs Cure Light Wounds argument, right?

Also, a quick comparison to the other fighters in the party vs. knolls for some perspective:

...

Kimball is no slouch in melee.

The issue isn't whether Kimball is good at melee combat compared to the other members of the party. It is whether Kimball in melee is as good as Kimball using a bow. Just looking at the damage analysis it is easy to see that is not the case. I would argue that we aren't using Kimball effectively by having him stand in the front row. It would be better if he was not in the front row but still able to use his bow.

For instance, in the current situation (6 gnolls in a large room with us in a door way), I think a good tactic would be for Basil, Jespers, Zinc and Kazan to charge forward to make a front line with the gnolls. At the same time Kimball should move to one side of room and shoot at the two gnolls that are closest to him. The next round he can shoot any gnolls still in point blank range or move to flank the gnolls attacking the front line.

Let the people who are only good at melee and/or have a better AC stand in the front and Kimball can pick off the sides or the enemy heavies. Kimball won't get hit as much since he isn't engaging the main group of the enemy. So he will be up longer doing the most damage. The lower damage potential characters are taking the most damage which is a better trade off.

Of course different situations will call for different tactics.

It is not clear to me how to weigh the AC analysis (I'll add the tables tonight). The gnolls have a tiny damage potential compared to Kimball regardless of whether he is shooting or using a shield. Since there is only one Kimball for the whole adventure but many combats to survive you can't take a single combat an isolation. You have to factor in how much healing is available, how easy will it be for Kimball to flank opponents, how long before we deplete our offensive spells, etc.

[venger]

Jun 18, 2009 12:45:11 GMT -5 Wolfgar said:

I would argue that we aren't using Kimball effectively by having him stand in the front row. It would be better if he was not in the front row but still able to use his bow... Let the people who are only good at melee and/or have a better AC stand in the front and Kimball can pick off the sides or the enemy heavies. Kimball won't get hit as much since he isn't engaging the main group of the enemy. So he will be up longer doing the most damage. The lower damage potential characters are taking the most damage which is a better trade off.

Believe me, those arguments fall on deaf ears... I think I say something to that effect every session.

Maybe they'll listen to you.



___
Gnolls have a 15% chance to hit Kimball when he's got sword and shield. 25% chance when they charge.

They have a 45% chance to hit him when he's firing the bow. 55% if they charge.

I think that's right.

[Ginger]

Don't forget, Kimballl's surprise bonus is a major reason for him in the front.
Certainly the optimal situation was a dwarf in the front rank with a human bow specialist behind him. Since Kimball is a short elf and we don't have a dwarf, he'll just be stuck in the front.

[Dead Greyhawk]

I'm not certain that the charge bonus should be factored into your consideration of likelihood of being hit. The +10% bonus applies equally to whether or not your are firing your bow or fighting with a sword.

I think the AC comparison is between flatfooted firing bow (AC 3), with sword and shield (AC -2?), and with sword and hand axe (AC 0).

The gnolls will always have at least a 5% chance of hitting you (a natural 20). Only when your AC is such that the +2 charge bonus still results in a natural 20 required to hit will it become germane. For example, if your shield is a +5 shield, you would have AC -6. There, a charging gnoll and a non-charging gnoll would both have only a 5% chance of hitting you.

Another argument for PfE over CLW.

Gnolls have a 16 THACO and base damage of 2-8 (avg. 5). So a 20% chance of hitting AC 0. A 10% chance against AC -2. A 35% chance against AC 3.

Without Kazan's chart mastery, it looks to me like each gnoll does 1 HP damage when you are dual wielding, 1/2 HP of damage when you have the shield out, and 1.75 HP damage when you are shooting the bow. Charging adds 1/2 HP of damage in each case.

Another factor not added in is bow breakage and the subsequent devaluation of the bow damage. The breakage factor is 1 in 10. I'm not certain how to calculate the probability of an event occurring in a small set of events. (While it is a 1 in 10 chance, getting hit 10 times doesn't guarantee the bow breaking.)

[venger]

The only reason I included the charge was because it tossed the to-hit right up to 50/50.

They hit AC 5 on an 11 or better. Which is 50% of the time.

Jun 19, 2009 7:14:31 GMT -5 Dead Greyhawk said:

Gnolls have a 16 THACO and base damage of 2-8 (avg. 5). So a 20% chance of hitting AC 0. A 10% chance against AC -2. A 35% chance against AC 3.

I'm not a maths person... so my confidence is shaken. But I was wrestling with these % yesterday and I was sure I had it right.

If your Thac0 = 20 you have a 5% chance of hitting AC 0.

19 = 10%
18 = 15% (gnoll attack against -2)
17 = 20%
16 = 25% (gnoll attack againt AC 0)
15 = 30%
14 = 35%
13 = 40% (gnoll attack againt flatfooted AC 3)
12 = 45%
11 = 50% (gnoll charge against AC 5)

Jun 19, 2009 7:14:31 GMT -5 Dead Greyhawk said:

I think the AC comparison is between flatfooted firing bow (AC 3), with sword and shield (AC -2?), and with sword and hand axe (AC 0).

Not that it's particularly relevant, but dual-wielding is a luxury Kimball is rarely afforded in a combat. That AC 0 is usually longbow in one hand, longsword in the other.

And how useful is this "surprise bonus"? How many segments of attacks would Basil (100hp) and Zinc (AC -2?) have to absorb if we were actually surprised?

Kimball, in his 7 levels of experience, has never surprised anything or been able to use those extra segments of bowfire (like Morvan did against the bugbears) even with his ranger/elf surprise.

[Dead Greyhawk]

You're right. My math's bad. Recalculating with the correct percentages leads to:
Each gnoll does 1.25 HP damage when you are dual wielding, 0.75 HP of damage when you have the shield out, and 2 HP damage when you are shooting the bow. Charging still adds 1/2 HP of damage in each case.

[Dead Greyhawk]

Maybe one way of looking at this is in a worst-case scenario.

With a bow, Kimball kills 3 gnolls a round, on average. With the longsword 3/2 rounds. Dual wielding 2/round (3/2 + 1/2).

If 8 gnolls charge Kimball wielding a bow, and Kimball loses initiative such that he gets no shots off as they charge, he is attacked 4 from the front (AC 5 with charge), 2 flank (AC 7 with charge), 2 rear (AC 7 with charge). Kimball takes 4 x 2.5 + 4 x 3 HP damage on average (22 HP). Kimball kills 3 gnolls.

In the second round, again assuming the gnolls win initiative, Kimball is attacked 1 from the front (AC 3), 2 flank (AC 5), 2 rear (AC 5). Kimball takes 2.0 + 4 x 2.5 HP damage on average (12 HP). Kimball kills 3 gnolls.

In the third round, again assuming the gnolls win initiative, Kimball is attacked 2 from rear (AC 5). Kimball takes 2 x 2.5 damage (5 HP) and kills 2 gnolls. Combat over.

Kimball takes 39 HP damage on average in the worst case scenario. In this scenario, Kimball takes 15 attacks. For every point his AC is lowered (like with PfE) he, on average, suffers .125 HP fewer damage per attack or 1.875 HP for this scenario.

If 8 gnolls charge Kimball wielding a sword and shield and Kimball loses initiative, he is attacked 4 from the front (AC 0 with charge), 2 flank (AC 2 with charge), 2 rear (AC 7 with charge). Kimball takes 4 x 1.25 + 2 x 1.75 + 2 x 3 HP damage on average (14.5 HP). Kimball kills 1 gnoll (set for charge).

In the second round, again assuming the gnolls win initiative, Kimball kills a gnoll (pre-initiative attack) and is attacked 2 from the front (AC -2), 2 flank (AC 0), 2 rear (AC 5). Kimball takes 2 x 0.75 + 2 x 1.25 + 2 x 2.5 HP damage on average (9 HP). Kimball kills a gnoll.

In the third round, again assuming the gnolls win initiative, Kimball is attacked 1 from the front (AC -2), 2 flank (AC 0), 2 rear (AC 5). Kimball takes 1 x 0.75 + 2 x 1.25 + 2 x 2.5 HP damage on average (8.25 HP). Kimball kills a gnoll.

In the fourth round, again assuming the gnolls win initiative, Kimball kills a gnoll (pre-initiative attack) and is attacked 2 from the flank (AC 0), 2 rear (AC 5). Kimball takes 2 x 1.25 + 2 x 2.5 HP damage on average (7.5 HP). Kimball kills a gnoll.

In the fifth round, again assuming the gnolls win initiative, Kimball is attacked 2 from the front (AC -2), 1 rear (AC 5). Kimball takes 2 x 0.75 + 1 x 2.5 HP damage on average (4 HP). Kimball kills a gnoll.

In the sixth round, again assuming the gnolls win initiative, Kimball kills a gnoll (pre-initiative attack) and is attacked 2 from the flank (AC 0). Kimball takes 2 x 1.25 HP damage on average (2.5 HP). Kimball kills a gnoll.

In the seventh round, again assuming the gnolls win initiative, Kimball is attacked 1 from the flank (AC 0). Kimball takes 1 x 1.25 HP damage on average (1.25 HP). Kimball kills a gnoll. Combat over.

Kimball takes 47 HP in the worst case scenario. In this scenario, there are 29 attacks on Kimball. For every point his AC is lowered (like with PfE) he, on average, suffers .125 HP fewer damage per attack or 3.625 HP for this scenario.

Some things to think about, and I'm certain that you can fiddle with these numbers too.

[Wolfgar]

Very nice analysis. From the worst case scenario it is clear that is is better to remove opponents as fast as possible at the risk of taking a lot of damage up front. I think the 2 weapon scenario would be somewhere in the middle. You would remove an additional gnoll every other round but take an additional .25 HP per attack.

Maybe I will write a program to look at all possible combinations (or at least a large number) of dice rolls to see what on average would happen.

I am also wondering if there is a blended solution that minimizes damage. Like have sword and shield out for the charge, get two attacks in the second round and then switch to the bow in the third round. From the numbers and the loss of attacks in the third round I don't think it is better.

[Ginger]

Jun 22, 2009 8:36:15 GMT -5 Wolfgar said:

I am also wondering if there is a blended solution that minimizes damage. Like have sword and shield out for the charge, get two attacks in the second round and then switch to the bow in the third round. From the numbers and the loss of attacks in the third round I don't think it is better.

Not that when doing that, you'll need to factor in dropping weapons. If Kimball's using a sword and shield, he'd need to drop both of them to fire his bow.

[Ginger]

I think we've always assumed "no" but I want to check. Can you split fire with a fighting withdrawal if you move a quarter of your move? That seems to be the solution to the "Kimball surrounded by gnolls" situation, where he'd shoot them all in one round, move from combat more than 10 feet, and then shoot them while they close to melee. This is good even if you can't split fire, given that we're assuming Kimball goes last, where he wouldn't provoke attacks of opportunity.

[venger]

Jun 21, 2009 12:09:31 GMT -5 Dead Greyhawk said:

Maybe one way of looking at this is in a worst-case scenario.

With a bow, Kimball kills 3 gnolls a round, on average. With the longsword 3/2 rounds. Dual wielding 2/round (3/2 + 1/2).

If 8 gnolls charge Kimball wielding a bow, and Kimball loses initiative such that he gets no shots off as they charge, he is attacked 4 from the front (AC 5 with charge), 2 flank (AC 7 with charge), 2 rear (AC 7 with charge). Kimball takes 4 x 2.5 + 4 x 3 HP damage on average (22 HP). Kimball kills 3 gnolls.

In the second round, again assuming the gnolls win initiative, Kimball is attacked 1 from the front (AC 3), 2 flank (AC 5), 2 rear (AC 5). Kimball takes 2.0 + 4 x 2.5 HP damage on average (12 HP). Kimball kills 3 gnolls.

In the third round, again assuming the gnolls win initiative, Kimball is attacked 2 from rear (AC 5). Kimball takes 2 x 2.5 damage (5 HP) and kills 2 gnolls. Combat over.

Kimball takes 39 HP damage on average in the worst case scenario. In this scenario, Kimball takes 15 attacks. For every point his AC is lowered (like with PfE) he, on average, suffers .125 HP fewer damage per attack or 1.875 HP for this scenario.

In this scenario Kimball is attacked 15 times. He kills 8 gnolls in 3 rounds.

Average damage from a gnoll is 5.

If Kimball is hit 40% of the time with an AC 3, that means he is hit 6/15 times for 30 damage total after these 3 rounds.


The next section I had to read carefully because it wasn't adding up to me:

Jun 21, 2009 12:09:31 GMT -5 Dead Greyhawk said:

If 8 gnolls charge Kimball wielding a sword and shield and Kimball loses initiative, he is attacked 4 from the front (AC 0 with charge), 2 flank (AC 2 with charge), 2 rear (AC 7 with charge). Kimball takes 4 x 1.25 + 2 x 1.75 + 2 x 3 HP damage on average (14.5 HP). Kimball (1) kills 1 gnoll(set for charge).

(KILLS 1 GNOLL)

In the second round, again assuming the gnolls win initiative, Kimball (2) kills a gnoll (pre-initiative attack) and is attacked 2 from the front (AC -2), 2 flank (AC 0), 2 rear (AC 5). Kimball takes 2 x 0.75 + 2 x 1.25 + 2 x 2.5 HP damage on average (9 HP). Kimball (3) kills a gnoll.

(KILLS 2 GNOLLS - GNOLL TOTAL 3)

In the third round, again assuming the gnolls win initiative, Kimball is attacked 1 from the front (AC -2), 2 flank (AC 0), 2 rear (AC 5). Kimball takes 1 x 0.75 + 2 x 1.25 + 2 x 2.5 HP damage on average (8.25 HP). Kimball (4) kills a gnoll.

(KILLS 1 GNOLL - GNOLL TOTAL 4)

In the fourth round, again assuming the gnolls win initiative, Kimball (5) kills a gnoll (pre-initiative attack) and is attacked 2 from the flank (AC 0), 2 rear (AC 5). Kimball takes 2 x 1.25 + 2 x 2.5 HP damage on average (7.5 HP). Kimball (6) kills a gnoll.

(KILLS 2 GNOLLS - GNOLL TOTAL 6)

In the fifth round, again assuming the gnolls win initiative, Kimball is attacked 2 from the front (AC -2), 1 rear (AC 5). Kimball takes 2 x 0.75 + 1 x 2.5 HP damage on average (4 HP). Kimball kills (7) a gnoll.

(KILLS 1 GNOLL - GNOLL TOTAL 7)

In the sixth round, again assuming the gnolls win initiative, Kimball (8) kills a gnoll (pre-initiative attack) COMBAT OVER

(KILL LAST GNOLL BEFORE IT STRIKES - GNOLL TOTAL 8)

________________________________________________

oh no here comes TWO ADDITIONAL GNOLLS as Kimball has already slain 8 __________________________________________________

and is attacked 2 from the flank (AC 0). Kimball takes 2 x 1.25 HP damage on average (2.5 HP). Kimball (9) kills a gnoll.

In the seventh round, again assuming the gnolls win initiative, Kimball is attacked 1 from the flank (AC 0). Kimball takes 1 x 1.25 HP damage on average (1.25 HP). Kimball (10) kills a gnoll. Combat over.

Kimball takes 47 HP in the worst case scenario. In this scenario, there are 29 attacks on Kimball. For every point his AC is lowered (like with PfE) he, on average, suffers .125 HP fewer damage per attack or 3.625 HP for this scenario.

Some things to think about, and I'm certain that you can fiddle with these numbers too.

In this scenario, with sword and shield, Kimball is attacked 26 times and kills 8 gnolls over 6 rounds

(the 6th and 7th round attacks you calculated come from gnolls (9) and (10) as Kimball already dispatched 8 gnolls over 6 rounds.)

Average damage from a gnoll is 5.

If Kimball is hit 15% of the time with a -2, that means he is hit 4/26 times for 20 damage total after these 6 rounds.



In conclusion, if Kimball were alone:

A) It would take 3 rounds to kill 8 gnolls with his bow and he'd take 30 damage. (6/15 hits, avg. damage 5)

B) It would take 6 rounds to kill 8 gnolls with sword/shield and he'd take 20 damage. (4/26 hits, avg. damage 5)

20 damage over 6 rounds vs. 30 damage over 3 rounds.

BUT---

During those 6 rounds I can imagine at least one of the party assisting to kill the 8 gnolls ganging up on Kimball.

So each round Kimball is in combat using sword and shield he can expect to take 3.3 damage. Each round he is in combat using the bow he can expect to take 10 damage.


*** I guess a certain percentage of those attacks are rear/flank attacks and should be accounted for... but that's complicated and I don't know how ***

*** OK I got it. So worst case scenario if Kimball loses initiative every round and the gnolls get optimal flank/rear attacks:

Kimball w/bow vs. 8 gnolls
15 attacks over 3 rounds
(4) AC 7= 60% to hit = 2.4 = 12
(6) AC 5= 50% to hit = 3 = 15
(5) AC 3= 40% to hit = 2 = 10

37 damage over 3 rounds.
12.3 damage per round

Kimball w/sword/shield vs. 8 gnolls
26 attacks over 6 rounds
(2) AC 7=60% to hit = 1.2 hits = 6 damage
(7) AC 5=50% to hit = 3.5 hits = 17.5 damage
(2) AC 2=35% to hit = .7 hits = 3.5 damage
(10) AC 0=25% to hit =2.5 hits= 12.5 damage
(5) AC -2=15% to hit =.75 hits= 3.75 damage

43.25 damage over 6 rounds.
7.2 damage per round


Bearing in mind this is Kimball completely surrounded and at the mercy of gnolls.

And that damage would be much less if his back was to a wall.

9 rear attacks in 6 rounds of combat is pretty unrealistic.

[Wolfgar]

Note that if you switch around the initiatives so that Kimball gets to strike first (which is more likely given his DEX) there is a big difference for the bow scenario but not much with the sword and shield scenario. Damage reduction would be 7.5, 7 and 5 HP in the respective rounds during the bow combat for a total of only 17.5 (=37 - 19.5) HP. I haven't done the math for the other scenario but it looks like it would still be close to 40 HP. Kimball could readjust how many where attacking his rear/flank but would only take two less attacks.

[Wolfgar]

What this analysis also shows is that the most effective attack for the gnolls is to gang charge the character with the worst AC. And it is usually the guy who doesn't look like a tank which they have to worry about the most. So it probably makes sense to attack him first. Unfortunately Kazan looks the least armored of the party.

[Dead Greyhawk]

Jun 23, 2009 8:54:12 GMT -5 Wolfgar said:

What this analysis also shows is that the most effective attack for the gnolls is to gang charge the character with the worst AC. And it is usually the guy who doesn't look like a tank which they have to worry about the most. So it probably makes sense to attack him first. Unfortunately Kazan looks the least armored of the party.

Good point. Poor Kazan.

[Dead Greyhawk]

Jun 22, 2009 12:57:30 GMT -5 Ginger said:

I think we've always assumed "no" but I want to check. Can you split fire with a fighting withdrawal if you move a quarter of your move? That seems to be the solution to the "Kimball surrounded by gnolls" situation, where he'd shoot them all in one round, move from combat more than 10 feet, and then shoot them while they close to melee. This is good even if you can't split fire, given that we're assuming Kimball goes last, where he wouldn't provoke attacks of opportunity.

No. You can split fire, but it is simply moving from combat. They'd get their attacks on the segment you break melee distance.

[Ginger]

Jun 23, 2009 19:05:37 GMT -5 Dead Greyhawk said:

Jun 23, 2009 8:54:12 GMT -5 Wolfgar said:

What this analysis also shows is that the most effective attack for the gnolls is to gang charge the character with the worst AC. And it is usually the guy who doesn't look like a tank which they have to worry about the most. So it probably makes sense to attack him first. Unfortunately Kazan looks the least armored of the party.

Good point. Poor Kazan.

That'll become a major advantage once he's a fighter, where he'll look just like an unarmored weakling, except he'll have an AC in the negatives.

[Ginger]

Jun 23, 2009 19:14:26 GMT -5 Dead Greyhawk said:

Jun 22, 2009 12:57:30 GMT -5 Ginger said:

I think we've always assumed "no" but I want to check. Can you split fire with a fighting withdrawal if you move a quarter of your move? That seems to be the solution to the "Kimball surrounded by gnolls" situation, where he'd shoot them all in one round, move from combat more than 10 feet, and then shoot them while they close to melee. This is good even if you can't split fire, given that we're assuming Kimball goes last, where he wouldn't provoke attacks of opportunity.

No. You can split fire, but it is simply moving from combat. They'd get their attacks on the segment you break melee distance.

I think this is something Kimball should do more often. He's already flat footed, so he'd just be giving the ones who aren't already flanking him an extra +2.