AFH Strategy: Moveable Feast v. LBoF (January 2010)

This forum is for threads formatted to present useful gameplay mechanics in a concise format for the community at large, and also for threads that summarize knowledge from TPTB that is otherwise impossible to spade.
Post Reply
Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

AFH Strategy: Moveable Feast v. LBoF (January 2010)

Post by Kelemvor » Sat Dec 19, 2009 11:24 pm

moveable feast v. little box of fireworks by Kelemvor

After doing a speed run using the moveable feast, I wanted to do some critical analysis on whether it's better, worse, or about the same as the little box of fireworks, the current standard familiar equipment pull in an SC speed run.

tl;dr
For a bleeding-edge speed run, the feast is about 2-5 turns better than a lbof. A wee tiny bit of powercreep, but nothing game-breaking.

Assumptions
  • I'll assume that you are going to pull some piece of familiar equipment, and you are going to pull it on day 1.
  • I've purposely omitted the damage and deleveling effects of both, assuming that for high-end runs, the player is able to fully handle combat and the marginal benefit of these is negligible. I understand there might be some MP cost savings; I'm assuming they are small enough not to matter.
  • I'm not really sure how to handle the pickpocket-enabling effect of increased initiative from the lbof, since I don't know how often the lbof causes the user to gain initiative, so I've ignored it as well. I've also omitted the ability to Fire a Roman Candle; of the 4 runs I looked at that pulled an lbof, just 10 total casts were made.
Methodology
Since we are omitting the damage effects of both items, and they both give +5 weight to all familiars, the difference comes down to:

Moveable feast: +10 weight on 5 different familiars per day for 20 combats, +1 familiar xp/fight
Little box of fireworks: +3 stats or +27% items, randomly chosen

To properly analyze the effects of the +10 weight buff, we need to find out how many familiars are typically used in a run, and for how often. To do this, I looked at the following 9 runs:

Low skill 3-day
Kelemvor, 9 skills, 3/653 - fastest <10 skills Log
Kelemvor, 14 skills, 3/641
Upstairs Frank, 18 skills, 3/587 - 2nd fastest turncount, <20 skills Log

Low-skill 2-day
Pantsless, 15 skills, 2/763 - lowest skill 2-day, fastest <20 skills Log

High-skill 2-day
Mr Crac, 2/642 - fastest muscle Log
VanDamien, 2/622 Log
KujjieKujjieKoo, 2/581 - fastest for nearly 4 months, pre-boulder Log
stupac2, 2/676 - fastest <50 skills Log
NotJim, 2/533 - fastest ever Log

Here is a table showing familiar usage for these runs:

Code: Select all

              bander   slimeling    dog    pixie/other   llama  
Kelemvor 1    253      162           *      *             *     
Kelemvor 2    265      104           *      10            *   
Frank         235      122           *                   31         

Pantsless      88       *            35    235           86
Mr Crac       275       38           85
Van Damien    288       *           104     37
Kujjie        295       *            80
stupac        289       *             *    143
NotJim        208       16           57      9

Other familiars used for <15 turns each: hobo monkey, boulder, sandworm, purse rat
* denotes that the player did not have this familiar
Looking at this table, we can say the following:
- On a three-day run with the feast, it is optimal to use the feast every day on the bandersnatch, hound dog, and one (possibly 2) other fairy; excepting a use on the hobo monkey on day 3, all other uses are marginal.
- On a two-day run with the feast, it is optimal to use the feast every day on the bandersnatch, hound dog, and possibly one other fairy; excepting a use on the hobo monkey on day 2, all other uses are marginal.

Given that usage pattern, let's estimate what the feast will do:
- 3 days: 6 runaways, ~50 mainstat (60 turns of going from ~25 to 35 pounds), ~100 turns of +20% items
- 2 day bandersnatch: 4 runaways, ~35 mainstat, ~50 turns of +18% items

Also, there will probably be 1 turn's worth of savings from everything else combined (coming mostly from the hobo monkey at the nuns, but possibly also from using the buff on a baby gravy fairy while looking for an adding machine).

Comparatively, the lbof is much easier to analyze:
- 3 days: ~150 turns of +14% items and ~225 total stat (112 mainstat)
- 2 days: ~100 turns of +14% items and ~150 total stat (75 mainstat)

Conclusions
The moveable feast gives significantly fewer turns of slightly more +item, along with 4-6 extra runaways. The little box of fireworks gives about 1 powerleveling turn's worth of stats. Based on these numbers and the fact that +item is most powerful in short, concentrated doses, I feel comfortable saying that for players who expect to use all their runaways every day, the feast will likely shave 2-5 turns off a bleeding edge speed run.

Other considerations
- The tower familiar test is unlikely to be affected by either item, except in extremely niche cases (only one +weight skill permed, no VIP key) that favor the feast.
- The feast also gets a little +item back from the +experience buff, since the familiars will be gaining weight much more quickly with it equipped.
- Interaction with sugar shield: if the player creates a sugar shield, that tips the scales even more in favor of the feast, since its buff stacks with the sugar shield and the lbof does not.
- Pathed runs: clear win for the feast, since the buff is usable more often and there will be likely more turns spent with a sugar shield and fewer turns with the lbof.
- RL time management: the feast loses points here; it's one more thing to keep track of every day. That being said, it's pretty easy to plan when you're using your feast buffs each day, and as long as you're buffing your item familiars and your bander, you're getting 90% of the utility.

FAQ
  • You're just trying to manipulate the market! I don't own either a feast or an lbof, and so have no stake in the price of either. Personally, I think the the relative price of the feast is kooky, but hey, that's a free market for you.
  • What if I don't have a bandersnatch? You shouldn't get either of these; save your meat until you can get a bandersnatch. I suppose if you're in a clan where you can borrow either of these, but don't have a snatch, you should borrow the lbof.
  • If I have a lbof already, should I get a moveable feast? Depends -- do you have a bandersnatch? Do you use all the runaways every day? Are you willing to invest millions of meat to cut off a handful of turns? If so, then yes. Keep in mind that this isn't game-breaking. It's not the difference between having a he-boulder and not having one. Perming another skill or just looking through your log will probably shave off more turns than switching from a lbof to a feast.
  • If I don't have either yet, which should I get? If they're the same price, I'd get the feast. If there's a significant price difference, I'd consider the lbof. They're both really, really good.
  • I have both an lbof and a feast. Should I pull both? Generally speaking, no. On a 2-day run, there's no room, and the lbof is likely to save you no more than 2 turns, given the diminishing returns of +item. On a 3-day run for a low-skilled player who has pulls to spare but no +item passives, it might be worth a pull.
  • What about for 100% runs? Virtually all of the feast's advantage comes from the free runaways, so the lbof is by far a better choice for 100% familiar runs.
  • I'm a casual player who doesn't like micromanaging everything. Which of these is better for me? Fireworks, hands down. Put it on your familiar, hit the lock icon, don't touch it the rest of the run. Switch a jub-jub bird onto your bandersnatch if you feel like it.
  • What about aftercore? Depends on what you're doing. As soon as you break the prism, feast is clearly better as it will drastically cut the time required to level your familiars. After all your familiars are leveled, fireworks are probably better, except in cases like fighting hobo bosses.
  • I buy IotMs for the flavor text. Which of these has better flavor text? I'm going to resist the very strong temptation to make a terrible pun, and just say that you are probably reading the wrong thread.
Last edited by Kelemvor on Sun Dec 20, 2009 1:52 am, edited 2 times in total.

User avatar
lostcalpolydude
Forum Maestro
Posts: 2940
Joined: Wed Apr 23, 2008 6:54 pm

Post by lostcalpolydude » Sun Dec 20, 2009 12:49 am

On a two-day run with the feast, it is optimal to use the feast every day on the bandersnatch, hound dog, and possibly one other fairy; excepting a use on the hobo monkey on day 3, all other uses are marginal.
3 should be 2.
bmaher: Softcore character
lostcalpolydude: Hardcore character (abandoned)

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Sun Dec 20, 2009 1:49 am

Good catch, lost. Fixed.

User avatar
Mr_Crac
Spy vs. Pie
Posts: 57
Joined: Fri Jul 18, 2008 7:31 am

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by Mr_Crac » Mon Dec 21, 2009 11:30 am

Kelemvor wrote: I don't own either a feast or an lbof
Feel free to ask me if you want to borrow either of them. And thank you for putting up this comparison.

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Mon Dec 21, 2009 5:06 pm

Thanks for the offer! I'll keep it in mind.

Anyone else want to offer suggestions on things I should add, remove, or modify from the post before I put it up on GD?

User avatar
Mr_Crac
Spy vs. Pie
Posts: 57
Joined: Fri Jul 18, 2008 7:31 am

Post by Mr_Crac » Mon Dec 21, 2009 8:57 pm

Just let me buy one more movable feast in the mall first. >.>
<.<

Damn, that price is ridicuolus. 50% more than a LBoF? People are nuts. You know what, you could repeatedly point out that the feast is only marginally better, under certain circumstances, and much more complicated to use. Maybe that'll help.
Last edited by Mr_Crac on Mon Dec 21, 2009 8:59 pm, edited 1 time in total.

User avatar
stupac2
Oh my! Guy with Pie!
Posts: 3027
Joined: Mon Dec 08, 2008 10:04 pm
Location: Stanford, CA
Contact:

Post by stupac2 » Mon Dec 21, 2009 8:58 pm

Kel, I'd wait until the weekend to post it. I rushed mine because I didn't think you'd be ready until this week and I wanted to see a 1/week frequency at most. Up to you though.

User avatar
Mr_Crac
Spy vs. Pie
Posts: 57
Joined: Fri Jul 18, 2008 7:31 am

Post by Mr_Crac » Mon Dec 21, 2009 9:23 pm

Oh yeah, and for the purpose of gaining weight on familiars in aftercore, the flaming familiar doppelg?nger should be a better (if too expensive) right?

Wait a minute. Umlauts don't work here? That is not cool.

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Mon Dec 21, 2009 11:13 pm

Eleron's promised some feedback for me, so I'm going to wait until I hear from him before I post on GD. Be happy to hear others' thoughts in the meantime, though.

Tissen
Inscrutable Pi
Posts: 229
Joined: Tue Apr 24, 2007 3:19 am
Location: Sweden, baby!

Post by Tissen » Tue Dec 22, 2009 2:19 am

I'm not sure, but doesn't the +famexp of the feast push the bander over 20lbs over the course of a run, making yet another runaway possible?

Quick look:
Assuming no white candy hearts, corsicans blessing or love song of icy revenge, the extra familiar experience of the feast will grant extra runaways after the following number of combats:

5lbs:
LBoF: 25
Feast: 13

10lbs:
LBoF: 100
Feast: 50

15lbs:
LBoF: 225
Feast: 113

20lbs
LBoF: 400 (not gonna happen in a speedy run)
Feast: 200

Calculating the average number of familiar uses after day 1 (and 2), it is possible that the feast grants more than 1 extra runaway (up to 1 extra every day). Still not much or game-breaking.

I'm just throwing it out here, and I don't have any other numbers to back this up. Feel free to prove me wrong.

salien
Pie in the Sky
Posts: 115
Joined: Fri Jan 11, 2008 1:50 pm

Post by salien » Tue Dec 22, 2009 2:29 am

Tissen, the biggest thing you're missing is that you don't ever actually equip the feast (or the LBoF) on the bander, because that's locked to the jub-jub. The feast and the LBoF are only used on other familiars (basically your +items fam(s)).

Tissen
Inscrutable Pi
Posts: 229
Joined: Tue Apr 24, 2007 3:19 am
Location: Sweden, baby!

Post by Tissen » Tue Dec 22, 2009 4:31 am

Of course. I didn't think you were puling two equipments. That's what I get for getting into a speed SC discussion.

User avatar
lostcalpolydude
Forum Maestro
Posts: 2940
Joined: Wed Apr 23, 2008 6:54 pm

Post by lostcalpolydude » Tue Dec 22, 2009 6:31 am

Tissen wrote:Of course. I didn't think you were puling two equipments. That's what I get for getting into a speed SC discussion.
You aren't pulling two familiar equipment items, you're pulling 3 (quadroculars). Or if you drop one for a two day run, it's fireworks/feast that you don't pull.
bmaher: Softcore character
lostcalpolydude: Hardcore character (abandoned)

User avatar
Eleron
Has a thing!
Posts: 870
Joined: Tue Jun 03, 2008 1:52 pm

Post by Eleron » Tue Dec 22, 2009 6:57 am

lostcalpolydude wrote:You aren't pulling two familiar equipment items, you're pulling 3 (quadroculars). Or if you drop one for a two day run, it's fireworks/feast that you don't pull.
I pulled jub-jub and the feast. Maybe that's why I ran out of time? =P

User avatar
KujjieKujjieKoo
Spade Ninja
Posts: 2240
Joined: Mon Apr 30, 2007 8:50 pm
Location: !Chicago
Contact:

Post by KujjieKujjieKoo » Tue Dec 22, 2009 7:58 am

A few things I'd like to see discussed... I'll throw my thoughts in tonight, but maybe someone gets there before I do.

1) How does the analysis change for players without a Bander?
2) How does the analysis change for players doing 100% runs?
3) Comparing the Feast / LBox to a Mr Jr (or other item drop pull) for analysis in a 2-day setting.
4) Noting the feast's additional benefit on Starfish types (GGG, Hobo, Slime)... extra weight = more MP.
Image

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Tue Dec 22, 2009 1:36 pm

KujjieKujjieKoo wrote:A few things I'd like to see discussed... I'll throw my thoughts in tonight, but maybe someone gets there before I do.

1) How does the analysis change for players without a Bander?
2) How does the analysis change for players doing 100% runs?
3) Comparing the Feast / LBox to a Mr Jr (or other item drop pull) for analysis in a 2-day setting.
4) Noting the feast's additional benefit on Starfish types (GGG, Hobo, Slime)... extra weight = more MP.
1 - I address this in the FAQ: since the feast's advantage is largely in the runaways, if the player has no bander, the lbof is a better pull.

2 - Also in the FAQ: again, since the feast's advantage is in the runaways, for 100% runs, the lbof is a better pull. Also, for a 100% run, the lbof will be equipped the whole time instead of just on the non-bander combats, while the feast is limited to just one cast per day.

3 - I purposely omitted this, because it seems like a much bigger discussion: should I pull familiar equipment at all on a 2 day run?

4 - While there clearly is some benefit to the feast gaining additional starfish MP, the familiar usage patterns of the runs I analyzed seemed to indicate that this was a highly marginal use; players seemed to use MP-gaining familiars very little. I have a particularly poor perspective on this issue, since I have very few buffs to run (less benefit) and a smaller MP pool (slimeling tends to fill me up every fight almost regardless of its weight).

In any case, since the feast is already better, I figured it wouldn't change the analysis much. I'd be happy to put it in if someone wants to crunch numbers and see if it's worth a full turn or more, though.

User avatar
Mr_Crac
Spy vs. Pie
Posts: 57
Joined: Fri Jul 18, 2008 7:31 am

Post by Mr_Crac » Tue Dec 22, 2009 3:04 pm

Kelemvor wrote:should I pull familiar equipment at all on a 2 day run?
At the very least one pull for familiar equipment seems a given to me:
a.) jub-jub-bird if you have a bander
b.) LBoF or feast if you don't have a bander.

As others have said, you could even pull up to three familiar equipments on a 2-dayer (it is not necessary though if making the daycount at all is your main goal).

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Tue Dec 22, 2009 7:42 pm

Sorry, what I meant is:

Given that I have all the ascension-relevant IoTMs at my disposal, should I pull one of the Mr. Store familiar equipment items (maypole, ppb, lbof, feast, whatever)?

I think the jubjub is a given if you have a bander, and probably quads too.

User avatar
Pantsless
Fie the Pie
Posts: 1011
Joined: Fri Aug 15, 2008 4:23 pm

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by Pantsless » Wed Dec 23, 2009 1:35 am

Kelemvor wrote: Low-skill 2-day
Pantsless, 15 skills, 2/763 - lowest skill 2-day, fastest <20 skills Log

Here is a table showing familiar usage for these runs:

Code: Select all

              bander   slimeling    dog    pixie/other   llama  
Pantsless      88       *            35    235           86
I'd strongly discourage reading anything into this specific data from that run, as all those green pixie turns were a complete mistake, and I would have been much better served by the bander. I ran the pixie in a lot of zones I shouldn't have (like the airship), in part because I thought I'd need all the absinthe I could get (I didn't; I actually finished the run with ~7 of them left over), and in part because I hadn't done a run yet with the bander and didn't think the jub was worth a pull.
[quote="Serra725"]What did you use for the structural part of the bear?[/quote]

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Thu Dec 24, 2009 1:00 pm

P -- your run was a clear outlier in terms of familiar usage, but when I aggregated them to draw conclusions, I mentally converted some of your pixie turns to hound dog turns, and shifted the rest to bander. Something like this:

Actual use
Bander: 88
Dog: 35
Pixie: 235

Adjusted use
Bander: 225
Dog: 98
Pixie: 35

...which would have gotten you the 3 absinthe you needed, ran the bander for a more conventional number of turns, and taken advantage of the extra item drop of the dog.

User avatar
KujjieKujjieKoo
Spade Ninja
Posts: 2240
Joined: Mon Apr 30, 2007 8:50 pm
Location: !Chicago
Contact:

Post by KujjieKujjieKoo » Tue Dec 29, 2009 12:23 pm

Ok, to touch on some of the points I raised previously...

Assumption #1: 600 turn run, at 70% average combat rate. This gives 420 combats over the run.
KujjieKujjieKoo wrote:2) How does the analysis change for players doing 100% runs?
Since 100% runs will have a single familiar type, we need to look at this for a few scenarios:
Volleyballs, Fairies, Other. If we assume the "others" would rather have some stats and items, the LBOX would seem to be a significant winner here.

Feast gives +1 familiar exp per combat, in addition to the once-a-day buff. Over the course of a run, you'll certainly be at 20lbs by the time the run is over, but the feast gets us there far quicker. To look at a few misc experience gains....

Code: Select all

Exp	Weight	Volley Fairy 
25	   25	    5.0	59%
50	   27	    5.2	63%
100	  30	    5.5	68%
150	  32	    5.7	71%
200	  34	    5.8	74%
300	  37	    6.1	79%
400	  40	    6.3	84%
A volley would gain about .3 mainstat more per combat, or roughly an extra 125 mainstat over the run. Additionally, the 10 bonus weight on a 3-day run would be another 60 turns at additional weight, or roughly another 40 stats or so. So... a 100% volley run with feast gains 160 bonus stats, with LBOX gains 630 stats and 13% item drop. Clear win for LBOX.

A fairy gains about 6% on average, plus gains roughly 25% when using the buff. If a "normal" run uses a fairy about 100 turns, we could expect to have about 30% extra item drop during 60 of those turns, and 6% extra item drop during the other important 40. An LBOX fairy would gain 630 stats and have an average 13% item drop over the run. This is a little tougher to determine, since missing an item drop can cost turns, and a rough 15% extra item drop could be the difference between a missed drop or not. If we use 100 mainstat per turn as our rough powerleveling rate... the LBOX saves about 3 turns for leveling. It's difficult to quantify the turn savings for 15% item drop in a given area, but if we assume that allows us to receive one quest drop we would otherwise have missed, a 3 turn savings feels reasonable. I'd guess that on a 100% fairy run, a feast and a LBOX are roughly equivalent.
KujjieKujjieKoo wrote:3) Comparing the Feast / LBox to a Mr Jr (or other item drop pull) for analysis in a 2-day setting.
As noted above, using the weight gain buff on a fairy adds roughly 25% item drop for 20 turns. If we assume 2 fairies are used (perhaps a hound and a slimling?), we could reasonably assume 25% extra item drop in most critical areas. This is equivalent to a Mr. Jr pull, and stacks with any other accessory based item drop. Additionally, the feast will provide some weight gain for other familiars used, and an extra meat bonus at the nuns. All things considered, I'd pull a feast over another item drop accessory on a 2-day setting.
KujjieKujjieKoo wrote:1) How does the analysis change for players without a Bander?
4) Noting the feast's additional benefit on Starfish types (GGG, Hobo, Slime)... extra weight = more MP.
I think these two points tie together nicely, and will cover them together. Without a bander, we'll need to rely on other volley types for stat generation. A reasonable run might use some combination of a llama, sandworm, volleychaun, ggg, hobo, hound, slimling, hobo monkey, dandy lion, green pixie, or fairychaun. In the absence of an overpowered Bander, players have more options, and aren't penalized as much for non-bander turns.

To throw out some assumptions for the sake of numbers...
15 - Llama (for gongs)
15 - Sandworm (for aqua)
15 - Hobo Monkey (Nuns)
80 - Hound Dog
20 - Green Pixie (for absinthe / secondary familiar)
275 - Volleyball (ggg or hobo)

LBOX: 630 stats, plus 13% item drop over the course of the run.

Feast:
-Llama moves from a base 20 lbs to 30lbs, adding roughly 10 stats.
-Sandworm moves from 20lbs to 30lbs, adding roughly 20 stats.
-Hobo Monkey adds 35% meat drop from extra weight, potentially saving a turn.
-Hound Dog adds ~35% item drops, and maxes combat rate for beach.
-Pixie adds ~25% item drops
-Volleyball gains an extra 82 stats from exp rate, plus another 60 from feast gain. Additionally, using a volleystar may provide enough MP to switch from a Katana to Sickle, or perhaps drop an additional MP pull. Alternatively, running a volleychaun would provide enough meat to run learning pills the entire run, generating stats in that way.

Net gain: ~170 stats, plus more item drop in critical times, plus meat boost for nuns and additional MP or meat from the new volleyball. The 2 additional powerleveling turns required for this equipment should easily be offset and greater turn savings provided thanks to the variety of other bonuses.

While the feast requires more micromanagement, I believe it comes out ahead for bander-less runs thanks to the ability to increase weight and desired bonuses at critical times during the run. The numbers can certainly be tweaked above, but keep in mind the exp gain cannot be overlooked for non-bander runs, where it is largely irrelevant when the bander is the primary familiar.
Image

User avatar
Mr_Crac
Spy vs. Pie
Posts: 57
Joined: Fri Jul 18, 2008 7:31 am

Post by Mr_Crac » Tue Dec 29, 2009 12:42 pm

Hmmmm... in speed softcore, the bander is the primary familiar. And in Hardcore (where it would be absolutely great) you can't use the feast. Take that, mall speculators!

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Post by Kelemvor » Wed Dec 30, 2009 12:48 am

Kujjie, great analysis as always. I think I'll try and sum up your conclusions and mine for the GD post, and link to this thread for anyone who wants the gory details.

One thing, though -- your non-bander analysis splits the non-bander turns between llama, sandworm, hobo monkey, hound dog, pixie, and volleyball. Seems a bit hard to imagine someone who has all those familiars, plus their choice of a feast or lbof, but no bandersnatch.

User avatar
KujjieKujjieKoo
Spade Ninja
Posts: 2240
Joined: Mon Apr 30, 2007 8:50 pm
Location: !Chicago
Contact:

Post by KujjieKujjieKoo » Wed Dec 30, 2009 7:45 am

Kelemvor wrote:One thing, though -- your non-bander analysis splits the non-bander turns between llama, sandworm, hobo monkey, hound dog, pixie, and volleyball. Seems a bit hard to imagine someone who has all those familiars, plus their choice of a feast or lbof, but no bandersnatch.
That's true, and is a good point. I was more trying to reinforce that the bander runaways aren't the only reason to use the feast. Once multiple familiars are brought into the picture for niche uses (spleen filling, starfishing, etc)... the extra daily feasts give some additional marginal utility which adds up. Also, the familiar exp gain on non-bander familiars wasn't really touched on previously, and seemed worth noting as well.

While I agree it would be difficult to imagine a scenario today where a bander wasn't present given the other hoard of familiars... that scenario would be far more likely to exist 6 to 12 months from now with newer players. For their cases the familiars being used would be different based on new releases, but the utility of the feast should remain as strong thanks to swapping.
Image

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by Kelemvor » Mon Jan 04, 2010 5:03 pm

Kelemvor wrote: I've purposely omitted the damage and deleveling effects of both, assuming that for high-end runs, the player is able to fully handle combat and the marginal benefit of these is negligible. I understand there might be some MP cost savings; I'm assuming they are small enough not to matter.
At Eleron's behest, I did some more thinking about this to see if my ignoring these effects is justified. First, the combat comparisons:

LBOF: 1.5-2.0 x fam weight, split between physical and hot damage; deleveling
feast: .5-1 x fam weight, random elemental damage

I assumed that the LBOF combat effects are equally likely (33% chance of each) and that the deleveling effect was equivalent to doing about 15 damage (3 rounds at -5 level each). Also, deleveling will save some MP on healing. How much? Looking at Kujjie's 2/581 and !Jim's 2/533, they each ran non-banders about 80 turns and spent about 900 MP on healing, with an average of 360 combats. I'll give it a very generous WAG and say it's 300 MP saved -- that's less than 2% of the total MP spent and about 2.5k meat in seltzer.

I also assumed that the feast's elemental damage would do double damage about 1/6 of the time (assumed 40% of the monsters you fight are elementally tuned, and take double damage 40% of the time). That makes our comparison as follows:

LBOF: .6 (weight) damage, 2.5k meat
feast: .8 (weight) damage

So...yeah, looks like it's ok to call this a wash.
I'm not really sure how to handle the pickpocket-enabling effect of increased initiative from the lbof, since I don't know how often the lbof causes the user to gain initiative, so I've ignored it as well. I've also omitted the ability to Fire a Roman Candle; of the 4 runs I looked at that pulled an lbof, just 10 total casts were made.
I still think Firing a Roman Candle is completely trivial and can be ignored, but initiative is fairly important, especially for a moxie class. How important? I'll take a guess. The best case scenario for the LBOF is that you're running enough ML and have no +init otherwise, so that you never get the jump. Assuming a 33% chance of triggering the init effect, you're either saving one hit from the monster or giving yourself a chance to pickpocket. However, you tend to be able to max item drop on any quest items that are over 30% drop rate, so say we're looking at a 25% drop rate (this is also highly skewed in favor of the lbof, since this assumes only one item drops from the monster).

Then we're talking about an 8% increase in item drop, best case, plus perhaps 1k in meat savings from not having to heal that first hit, and this is with some very favorable assumptions. I think it's probably more likely that we're looking at a 3% +item drop and 700 meat savings.

So yeah, I think it's safe to ignore these effects, but my reasoning is awfully handwavy.

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by Kelemvor » Mon Jan 25, 2010 12:26 am

Summarized the combat effects, cut some of the methodology details, expanded the FAQ, posted on GD. Thanks to everyone who provided feedback!

Not sure it merits being moved into the Archived Threads of Awesomeness, but if so, now would be a fine time to move it.

User avatar
stupac2
Oh my! Guy with Pie!
Posts: 3027
Joined: Mon Dec 08, 2008 10:04 pm
Location: Stanford, CA
Contact:

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by stupac2 » Mon Jan 25, 2010 12:27 am

I'll move it.

Also, I think the bonus on the LBoF is 25%, not 27%. That's just what I've always heard.

Kelemvor
AFH
Posts: 980
Joined: Fri Aug 28, 2009 11:36 am

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by Kelemvor » Mon Jan 25, 2010 12:33 am

Wiki said 25-29%, so I took the average and said 27%.

25% sounds more likely, though, and wiki says it doesn't know for sure. Doesn't change the analysis much if any, obviously.

User avatar
stupac2
Oh my! Guy with Pie!
Posts: 3027
Joined: Mon Dec 08, 2008 10:04 pm
Location: Stanford, CA
Contact:

Re: AFH Strategy: Moveable Feast v. Little Box of Fireworks

Post by stupac2 » Mon Jan 25, 2010 12:35 am

Ah, ok.

Post Reply