Post:Box Difficulty For Mid to High Level Critters - 08/29/2012 - 11:36
|Box Difficulty For Mid to High Level Critters · on 08/29/2012 11:36 PM CDT||1481|
| Over the last few days I've checked pretty much all the critter boxes that Kaxis is able to get accurate numbers on. It's everything starting from Dragon Priest Intercessors down to Orc Reivers. Someone else will have to do the next section of critters since none of the people I can use have box skills in that range. I really should have done this a couple years ago when Kaxis's skills were much lower. Would have been able to get to a few more critters. To approximate this I disadvantaged Kaxis as much as possible: full armor, no khri, wilderness, base bonus, etc on the theory this would slide him down the effectiveness scale a bit.
I stopped his testing roughly around the level of where Elanthipedia says critters start at 300 ranks in skill levels. By coincidence this is where critters started dropping below 2000 on box scores, which is where statistical accuracy begins becoming an issue. To drop below 2000 means you're starting to pick up some minimum (1) level results, and there's no way to know if it truly is a 1 score or if it would go below 1 to negative except that the scaling won't let it. So I included the few I did below 2000, and being at the top of the lower chart they should be okay.
The method used to calculate boxes was this: when a box is disarm IDed a bunch of times, it doesn't give certain results. It gives a cluster result centered on where the certain result should be. This means you need to disarm ID a bunch of times and then average in order to get a represenation. I disarm IDed each box 100 times. With Kaxis's stats and skills I found this to give a possible error of less than .33% or so.
Boxes from the same critter aren't all created equal. Some are harder or easier than others. Ideally a 100 box sample from a critter would be perfect, but that would mean me not doing the project since I do want to get back to training eventually. I settled for a 10 box sample from each critter, which at least makes the math easy, not that Excel cares about that. This gave me 1000 data points to use per critter.
There are 17 difficulty level messages you can receive when disarm IDing a box. These were numbered from easiest to hardest. When I would disarm ID a box, whatever level it registered was recorded as a numerical value. Each of the ten boxes were done 100 times and added together to give a final total score for that critter type. So the higher the scores per disarm ID, the higher the total score at the end. This procedure is nearly identical as was done with crossbow hits several years ago. It has been also done with bows and arrows by other players in the past.
The idea for doing this was to improve the Locksmith Union listing. I don't like the way the list is set up now with levels. It's better to just have an accurate ranking order of critters and just say the highest you can do is, for example, Sky Giants. Someone told me just today that the chart levels make it convenient, but I responded that it doesn't do that at all. At some point every method requires a chart to reference. Rather than impose artifical and arbitrary level divisions, just have the correct order of critters and say I can do every thing below this one but not above. Someone else said maybe add more levels to fine tune it. I said that's what I'm doing, except that I'm taking it to the point that each critter represents its own unique level. The Locksmith Unionn can handle their business any way they wish, I'm just providing the information.
Unless there's some compelling reason not to, Vile Plague Wraiths, Black Goblins, and DP Sentinels need their drop rate increased. I was in Black Goblins 1 1/2 hours before I got ten boxes. Vile Plague Wraiths were 2 hours. Sentinels were nearly 3 hours. Assassins were very quick.
Gargantuan Bone Golems and Vaporous Blood Wraiths meant going into the Abyss. I'd never been in there before. That puzzle was frustrating, and when I got in the first time I thought I would just walk into the hunting ground I needed. Found out I needed to actually do the Cleric Infusion quest, part of it anyway, to get where I needed to go. So I had to come out and start over. I added a note to elanthipedia about these critters in case anyone is interested. One thing for sure: you won't have to worry about lack of hunting rooms! Especially for Gargantuan Bone Golems and Vaporous Blood Wraiths, which are in the black Pit and really aren't in what one would normally think of as a hunting ground. Wraiths do drop boxes, something that wasn't known for sure on elanthipedia, and have a Snaer Hafwa style spirit attack, and I changed those entries as well.
Orc Clan-Chiefs were not practical to do since they spawn so rarely. They're probably not far above Orc Raiders, my guess would be likely somewhere around Dragon Priest Juggernaut level at most, possibly a bit lower.
Trekhalos are supposed to have boxes, yet in the 2+ hours I hunted there I got my full sample of Hunter boxes without ever getting the first Trekhalo box. I killed large numbers of each. Unless I got very unlucky it looks like Trekhalo box dropping is broken.
Didn't go to Hara'jaal, which I doubt many box hunters do either, so bypassed Corsairs and Telga Moradu, which I'm not sure are even active right now. All other critters I got from Elanthipedia's listing were all too low to check.
Here's the listing from the hardest, Dragon Priest Intercessors, no surprise there, in order to the lowest:
And here's the ones below 2000:
It is important to remember the total score numbers are specific to Kaxis and his stats and skills. And when his skills change these numbers will no longer be relevent to him either. If you look at the post a few previous where I posted only four critters from long ago, you'll see how his number results have shifted.
So anyone with different stats/skills would get a different looking set of numbers. The gap pattern should be the same, but the actual numbers would be different. I included them only to show how close together some critters are and the gaps between others. In particular note the huge jump from Resuscitants to Assassins. But they're in no way meant to be used as any kind of universal scaling. Having said that, it remains possible to perhaps devise some kind of multiplier to correlate data from two different investigators, but leaving that thinking for now until there is a second set of numbers to examine.
Anyone who wants to continue this method on down to the critters below these can do it this way. It would be best to first collect numbers on as many critters below mine as you wish to do so that they're in proper order. Then do the lowest one on my list to see if that tells you where the top of your list seams into the bottom of mine. If my lowest critter is still higher than your highest critter, then that's a cinch! Your list picks up where mine leaves off and you need do nothing more. Any new critter that is added one day can be found where to be placed with only a little trial and error.
For reference so we're all on the same page here is the scaling I used:
So the maximum score a box could have using this system is 17000. Minimum would be 1000.
Yes, I scripted this to avoid losing my mind.
And for even more fun here's the remaining critters from the list I got off Elanthipedia, in descending order:
Icy Blue Ghast
|This message was originally posted in Abilities, Skills and Magic \ Survival Skills - Economic, by URUTU on the play.net forums.|