Talk:Ant rake

From TheKolWiki

Was attacking for 10-12 hot damage with both my 30 pound Leprechaun and 11 pound ghost pickle on a stick. --xtrat

From my (limited) initial spading, it appears to give a 20% boost in meat drops. Ershatz 00:44, 1 July 2007 (CDT)

Some farming data

I spent today spading out the effect of the various ant items. I adventured in the Beanbat Chamber with a baseline meat drop modifier of +122%.

40 lb Baby Gravy Fairy
74.24 meat / adventure
6.32 standard deviation
n = 25

(Consistent with base meat drop of 28-40 and a +122% modifier, for a baseline total meat drop multiplier of 2.22)

20 lb Baby Gravy Fairy + ant rake
87.38 meat / adv.
6.02 std. dev.
n = 42

35 lb Baby Gravy Fairy + ant rake
85.16 meat / adv.
5.66 std. dev.
n = 25

35 lb Leprechaun
127.64 meat / adv.
9.59 std. dev.
n = 25

40 lb Leprechaun
132.32 meat / adv.
10.45 std. dev.
n = 25

35 lb Leprechaun + ant rake
141.84 meat / adv.
13.05 std. dev.
n = 25

35 lb Leprechaun + ant hoe
136.12 meat / adv.
8.48 std. dev.
n = 25

35 lb Leprechaun + ant pick
140.52 meat / adv.
12.11 std. dev.
n = 25

35 lb Leprechaun + ant sickle
139.64 meat / adv.
11.22 std. dev.
n = 25

By dividing the average meat per turn by the baseline average meat (in this case ~34 meat / adv), we can figure out what effect the equipment has.

From the BGF data, we have an observed baseline meat drop multiplier of 2.18 (theoretical value of 2.22). Our two data points are 87.38 and 85.16 meat / turn, which give multipliers of 2.50 and 2.57.

The 35 lb leprechaun data has a baseline multiplier of 3.75, and with the ant items we get multipliers of 4.17, 4.00, 4.13, and 4.10.

The average meat drop multiplier increase is 35.1%, which I'm guessing is 35%, even though my sample size is fairly small (~150 adventures). This appears to be independent of familiar weight.

The combat damage range is correct (10-12 points of elemental damage, depending on the type of the ant item). I had 48 familiar attacks out of 150 adventures, so a rough guess of attack frequency would be around 33%.

--Fiche 16:33, 6 July 2007 (CDT)

My own data indicates that the meat drop boost is variable. Anywhere from +0% to +50%. What the average boost is might need a little work, as Fiche and I seem to get different results.

I was using the Themthar Hills, which doesn't have its averages spaded out very tightly, but has such a high base meat gain that the game's internal rounding mechanics shouldn't play any meaningful role (and has every percent of meat drop boost result in a gain of nearly 10 meat on average). Like the beanbat chamber, there's also absolutely no variance in the monsters, so the meat drops would be consistently distributed. Making a few assumptions on average drops there, the average boost I'd get for the pitchfork is around 20-25%, which is substantially different from Fiche's numbers. This number feels nicer to me, as it sits so nicely in the middle of the range, but the hills only allowed me about 45 turns worth spading there. Still, I'm suspicious that I'm seeing such a noticeable difference after that amount of data.

Were your Leprechaun numbers done using the currently understood NS13 formulas, or the 5%/pound? I'd be wary of using Leprechauns when spading meat drops right now because of the changes. I was using a star starfish, if you're curious. --Flargen 18:34, 20 July 2007 (CDT)

I spent a few turns with an ant hoe attached to a ghuol whelp, and received the following meat drops from beanbats: 50,45,42,47,52,46,53,48,51. This is enough to conclusively show the meat bonus from the ant hoe (and thus all other ant equipment) is 25%, instead of the 35% listed here. How? Meat drops when boosted by items do not form a consecutive range, the original meat from the monster - in this case 28 to 43 - is rolled and then the meat bonus is applied and rounded down. So it's a simple matter of comparing the series of meat drops received to the series of meat drops possible due to various amounts of +% meat.

Final word - 25%. This lines up nicely with what flargen wrote. And I doubt Fiche had the NS-13 leprechuan data when he did his spading - that information was just added to the wiki recently.--QuantumNightmare 00:12, 23 July 2007 (CDT)

Thanks for the feedback. I'm kind of new to this spading thing and probably don't have as good a grasp on the underlying game mechanics as others. However, I'm not convinced that 25% is the correct value.

There has been some discussion about whether I used the pre- or post-NS13 leprechaun meat drop formula and how that influenced my results. I was actually closely following the initial work to spade out the new leprechaun bonus after NS13, and that is what prompted me to look into the ant equipment. Specifically, I wanted to know if ant equipment was better than a meat detector for increasing meat drops, given that boosting leprechaun weight no longer caused meat drops to scale linearly. The way I arrived at the 35% value should be independent of the bonus given by leprechauns. In fact, if anything, the data presented above validate QuantumNightmare's new leprechaun drop bonus. The average meat drop modifier using using a baby gravy fairy with ant rake was 2.535, and the average meat drop modifier a 35 lb leprechaun with ant equipment was 4.1, which makes the meat drop bonus for a 35 lb leprechaun 4.1 - 2.535 = 1.57 => 157%, which is quite close to the 151% given by the new formula and obviously not the expected pre-NS13 bonus of 175%.

So, to look into what is going on more closely, there is no better solution than more data. I took an unascended level 39 turtle tamer with no familiar and no buffs or effects that influence meat drops (base meat drop bonus of 0%) and adventured in the Beanbat chamber. 100 turns with no familiar gave me an average of 34.03 meat per turn, with discrete meat drop values from 28-40 (all values in between included). I then equipped a 25 lb baby fairy gravy with an ant rake and adventured for 100 more turns. The average meat drop was 46.68, and the meat drop values ranged continuously from 39-53. 46.68 / 34.03 = 1.371, which gives a meat drop bonus of ~37%. Going from 3403 meat in 100 turns to 4668 meat in 100 turns with the only difference being an added ant rake sure looks like about 35% more meat to me. I recognize that this is a small data set and we may have sampling issues as well. Is there something else obvious that I'm missing here? Any feedback from more experienced spaders would be greatly appreciated.

In any case, we still need to work out how often ant equipment attacks, and that's a lot easier to figure out. In my 100 turns with the ant rake equipped, my familiar attacked 33 times. That seems to confirm my original idea that the attack rate is ~33%.

--Fiche 15:54, 23 July 2007 (CDT)

Solved

You guys were approaching this the wrong way. It's simply an extra 10-14 meat at the end of battle. (or maybe 10-15, I don't have enough sample points to rule this out) I went and fought fluffy bunnies, and received a continuous range from 17 to 23 meat, when the base rate is 7-9.

This should produce either a +12 or +12.5 average meat take, which is exactly what was observed by Fiche. It also explains why the meat drops are continuous rather than discrete, and thus QNM's usually valid method failed.

To doublecheck, I ventured into the Themthar hills, and didn't see any large boost to meatdrops in a handful of adventures: 914, 931, and 862 meat. If it was even +10%, I don't think I'd have seen that last point.

I also tried both 1lb and 20lb familiars, and didn't see any diff. in meat drops.

Disclaimer: this was done with an ant hoe, not an ant rake, but it seems pretty unlikely they're different. --Starwed 16:31, 23 July 2007 (CDT)

Just took a lot more data points on the bunnies: it seems to be +10-15 meat. --Starwed 16:51, 23 July 2007 (CDT)

Tested with an extra +90% meat drops; didn't affect the extra meat, so it's added after the multiplier. --Starwed 17:09, 23 July 2007 (CDT)

Good work solving this problem, Starwed. I'll double-check after rollover, but your idea explains quite well why we aren't seeing big meat gains in other areas. It looks like all I have done is convincingly show that 12.5 is 35% of 34. :)

--Fiche 19:16, 23 July 2007 (CDT)