|
Post by johannhowitzer on Mar 30, 2016 6:05:29 GMT
I think I've finished with level 1-10 routing for now, and I streamed a test of it to make sure it works. You can see the video here: www.twitch.tv/johannhowitzer/v/57534769Level 10 was achieved in just over an hour. Money route works, XP route hit a momentary snag at the end but it's easily corrected. So that gets me to the level threshold for the Kuo-Toa Chief, Hive Mother Beetle, and the Black Dragon Disciple; I've already routed most of that trip, so I should have more to show soon. Level 10 is also where I get Greater Magic Weapon, Freedom of Movement (very important for the underdark), and Divine Power, which I won't get to use liberally for a while because of its low duration.
|
|
|
Post by johannhowitzer on Mar 31, 2016 0:25:46 GMT
Does anyone have any insight about how Appraise works on the shopkeepers? The prices fluctuate a little bit from visit to visit - I see the Summon Creature VI scroll in the Item Shop vary between ~3600 and ~3900. However, casting Fox's Cunning and Prayer on myself does not seem to affect the shop prices at all.
|
|
|
Post by desocupado on Mar 31, 2016 1:13:25 GMT
Well, I think it follows regular NWN rules no? (I'm probably wrong on this)
|
|
|
Post by johannhowitzer on Mar 31, 2016 1:37:05 GMT
Interesting. I had not considered the possibility of using Charm Person for this; I have Enchantment focus, too. Thanks for pointing me to that article. It makes me want to consider increasing my Appraise by 1 at a time, though, to reroll as many times as possible until I get something very favorable. The article makes me think this will also strongly affect sale prices. Also, the Underdark trip is looking verrrrrry interesting. I think I'm going to be able to do everything I want to do with only two rests - one in town, one halfway through. And the Kuo-Toa Chief is the star of the show. I won't spoil it though. EDIT: The town vendors seem to be unaffected by Charm Person. They will fail a save, but have nothing in the Examine window indicating they've been charmed, and the prices don't seem any lower.
|
|
|
Post by johannhowitzer on Mar 31, 2016 16:31:04 GMT
After some limited testing, I have come to the following tentative conclusions about the shop prices. Note that I have not done any testing on the Druid Grove vendor, the Lathander shop, or the Rogue shop, as the first two don't sell anything of significant value to the speedrun, and the Rogue shop is very difficult to enter on a freshly created character and probably also doesn't have anything of value - I'll find that out later.
- Rips' Sales Associate, the Clothing Shop vendor, and the three vendors in the Imports Shop either have static prices, or their Appraise is too high for a reasonable Appraise skill at level 3 to affect the prices. At the very least, a large investment in Appraise and/or Intelligence would be necessary. I will thus be counting these prices as static until I find otherwise or decide to include Appraise skill in the build. So Boots of Sprinting will always cost 13502, and Phalanx Armor will always cost 19484.
- The Item Shop, Alchemist, and the four Emporium merchants' prices all fluctuate based on Appraise skill and can be re-rolled. My character starts with 18 Intelligence, and will get Fox's Cunning and Prayer early enough in the run to allow re-rolling a few times to set a solid price foundation for the rest of the run. Once a sufficiently good price is found, care should be taken over the rest of the run to be sure the shops aren't accidentally re-rolled.
- The Buy and Sell sections of each vendor are linked, they are not separate rolls. As the prices in the Buy section went up, the Sell price for a given item in my inventory would also go down proportionally. Thus, I can judge a shop's price quality based on a selling price instead, for a better idea of how well I've rolled.
- It also means there will be one shop that I will be selling all my items at, to get the best price for them. Among the eight shops tested that have a Sell option, the Imports should be avoided at all times, as the sell price there was usually a THIRD of sell prices elsewhere! Never sell anything to the Imports dealers. Of the remaining five, the Emporium Melee Weapons merchant was significantly behind the curve, while the Ranged Weapons merchant was a little ahead of the curve and will likely be the shop I pay the most attention to for selling.
I haven't checked the Weapon and Armor shops to see if they have sell options; there's already an item in my route that will be bought from the Armor shop, the normal Tower Shield (since I have Magic Vestment to play with).
EDIT: I also noticed the item I was testing selling - Potion of Cure Light Wounds x50 - could only result in multiples of 10 gold. Going to take some math to figure out what this means, beyond the obvious conclusion that there are "steps" in the price spectrum - prices can't be just any old value between p and p'. It's more like they can be any value q, where q = p + (((p' - p) / 10) * n), n always being a positive integer. The question is how many steps, and how far apart. NWN Appraise article states a 30% cap in either direction from the base price, and the % is the difference between the two rolls, so that tells me 61 possible values, 1% incrementally. But it's an opposed roll of two d10s, which means 19 possible outcomes - from the player winning by 9, to the player losing by 9, taking the dice without the appraise skill into account.
|
|
|
Post by madzapper on Mar 31, 2016 19:18:48 GMT
I haven't checked the Weapon and Armor shops to see if they have sell options; there's already an item in my route that will be bought from the Armor shop, the normal Tower Shield (since I have Magic Vestment to play with). In my experience, these two are the best merchants for selling goods.
|
|
|
Post by johannhowitzer on Apr 1, 2016 1:44:19 GMT
Alright, I've done a lot more testing and yeah, the weapon and armor shops are much, much better for selling than the other shops. Also I noticed selling prices are affected a lot more than the prices to buy items. For example, though the price to sell a Rez Pack to the Item Shop merchant has so far varied from 10509 to 16241 (a 54% increase), the price to buy a Summon Creature VI scroll from the same merchant has varied from 3708 to 3993 (a 7.6% increase). So it will be much more important to get a solid roll on the sell prices in the Armor or Weapon Shop than it will be to get a good roll in any of the other shops for buying equipment. Especially as two of the more important shops - Rips and the Imports - don't even change with Appraise at this point.
As a matter of fact, the only items I purchase in the route so far that will be affected at all by Appraise rolls are a few scrolls and potions from the Item Shop, the Bone Phoenix from the Emporium, the starter quest items, and the Tower Shield from the Armor Shop. Almost all of those are very small expenses anyway, the exceptions being a few Summon Creature VI scrolls, the Bone Phoenix, and a handful of speed potions. Also, without random drops during an actual run, I'm not even selling that much. Most of my money so far has come from quest rewards.
|
|
|
Post by Werehound Silverfang on Apr 1, 2016 3:32:10 GMT
I haven't checked the Weapon and Armor shops to see if they have sell options; there's already an item in my route that will be bought from the Armor shop, the normal Tower Shield (since I have Magic Vestment to play with). In my experience, these two are the best merchants for selling goods. How do they compare to the sell chest and bulksell? I thought those two were 100% item value. EDIT: Some numbers. I learned something today. [CHAT WINDOW TEXT] [Thu Mar 31 21:42:47] Properties of Blazebolt's Eye (Ring): Armor Bonus: +15 Damage Immunity: Electrical 71% Damage Immunity: Fire 43% Freedom Immunity: Mind-Affecting Spells Quality: Good Skill Bonus: Parry +14
Vendor | Before Song (25) | After Song (38) --------------------------------------------- Alchemist | 1722553 | 2153191 Armor | 2392435 | 3301560 Weapons | 2918771 | 3110166 Wonderous | 1913884 | 2775225 Emp Armor | 2248889 | 2679527 Emp Item | 2057494 | 2918771 Emp Melee | 2535981 | 2870922 Emp Ranged | 2248889 | 3062317
Massed for 1913948
EDIT2:: Couple more items are consistently higher at the dwarven Armor Merchant both before and after song/buffs.
|
|
|
Post by johannhowitzer on Apr 1, 2016 17:54:13 GMT
Song and buffs will cause you to re-roll, and will not increase sale price by a consistent percentage. I have not tested sell chest or bulksell yet. Here's some of my data collection in image form: s23.postimg.org/4ndksb7ln/shops.pngChecking the sell price of a Rez Pack allowed me to judge the "absolute" quality of the opposed Appraise roll. I collected this information using three Appraise rolls at each merchant over 6-7 servers... notice that the Armor and Weapon shops' sale prices cover about the same range. It's reasonable to assume they would cover exactly the same range if I were to manage to collect data for all possible roll differences; they very likely have the same Appraise skill. If the mass sell chest yields 100% of the item's value, then that should be a very good reference point for the exact Appraise skill of all merchants that have a Sell option. Also, note that an "unfavorable" floaty message should mean you're getting less than 100%, while "favorable" should mean you're getting more than 100% - those should represent losing the roll and winning the roll, respectively. Also, see those 477s and 478s at the top? That's the difference between adjacent encountered sell values for my test item, the Rez Pack x10 (which is awarded from Sirina's first quest). 477.x is the smallest amount the sell price of the Rez Pack can change, presumably if the opposed roll result changes by 1. For example, if the vendor rolls 10 and you roll 4 and the price is 18152, then if the vendor rolls 10 and you roll 3, the price will change to 17674. Knowing the item's 100% value will probably tell what percentage change in the sale price a difference of 1 in the roll causes. More testing today hopefully! EDIT: Mass sell chest yielded 19108, which is waaaaaay toward the best sell prices in the Armor / Weapon Shops. I will try this a few more times to see if it gives that price every time; I may also test the Appraise rolls a bit more in the Armor / Weapon Shops to see if the floaty message becomes "favorable" when the price is 19585. Assuming 19108 is the "100%" sell price for Rez Pack x10, then 477 is ~2.5% of that, which reciprocally means the increment for each change of 1 in the Appraise roll is 477.7. Note that the NWNWiki article claims the maximum change in price in either direction is 30%, so the maximum and minimum price of Rez Pack x10 should be 24840.4 and 13375.6. This is NOT THE CASE since I've seen as low as 8120. I'm not sure yet what this means.
|
|
|
Post by johannhowitzer on Apr 5, 2016 23:10:10 GMT
I've been going about this all wrong. A complete re-route will be needed, but the result will be much faster.
|
|
|
Post by johannhowitzer on Apr 6, 2016 22:19:06 GMT
In general, what I'm finding is that I need a lot of bulk information collection before I can really proceed from here. So here's what I'm collecting right now, and since I'm leveling several characters at once to do this, it'll take a while:
- Level caps on every quest: leveling a character to 40 with grinding alone, then doing every quest to see the maximum XP yield
- Checking every loot spot for set drops, and selling every set drop to check gold yield
- Checking every important enemy and boss's stats - HP, AB, AC, saves, ability DCs, elemental weaknesses, etc.
- Finding all potential grind spots and evaluating their XP yield at various levels. This route will require grinding since it isn't possible to get level 40 through just quests and the level up book. Level 3 is gained via the Rowan Guardian, 1.5 levels from the book, and there are 90 quests for - theoretically - 25% of a level each for a total of - theoretically - 22.5 levels. That leaves 13 levels through grinding, and then subtract the XP gained while doing bosses.
- Testing every tag's required kills. If I can avoid wading through hordes of low XP enemies to get a boss tag, I'm going to do it.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 6, 2016 22:52:25 GMT
- Level caps on every quest: leveling a character to 40 with grinding alone, then doing every quest to see the maximum XP yield This thread has some information about the quest experience mechanics (they appear to have caps for every 5 levels?). On another note... not a speedrunner here, and of course you are free to go about it how you want, but my understanding (based on videos like Cosmo's OoT highlight) was that *in general* speedrunning is a slow 'improvement over time' process, with optimisation occuring over many runs. From the little that I have seen so far, I feel that you might be over-optimising to achieve an amazing first full attempt. Seems to me like a recipe to burn out. Perhaps it would be better/more rewarding if you just ~naively (only a basic route plan) attempted the entire run first, and then gradually optimised it with research (which could be done mid-run) into stuff like XP/gold yield, enemy stats/weaknesses and so on. It also might help you avoid burning time dedicated to fine-tuning that becomes redundant later, eg: I've been going about this all wrong. A complete re-route will be needed, but the result will be much faster. What if it turns out the current build isn't able to kill MOAD?
|
|
|
Post by johannhowitzer on Apr 7, 2016 2:59:54 GMT
Excellent points / concerns, I'll address each in turn. My testing and Acaos' own words to me on this tells me it's a regular level interval, but it's different for each questgiver. Those with eight quests exactly should be every five levels. Yes, every speedrun I've done has been like this... but every speedrun has also had an initial routing that went as far as reasonably possible with current knowledge. OOT is a special case, since it was one of the first speedruns routed and researched by many people at once, and its improvement truly was that slow, crawling gradual process. I have also been speedrunning Freedom Planet, which is a very new game; HDL laid down the first basic route as he was an alpha tester, and then the rest of us gradually trickled into the community and discoveries were made slowly, one at a time, over the past year and a half. Don't worry, this isn't my first rodeo... I also routed Final Fantasy X Nemesis, which took me nine months from concept to first full run. I pace myself and work on other projects here and there. I haven't been pulling all day marathons on this project, just chipping away at it when I have the time and desire. I also burn out pretty slowly. I'm also not trying to achieve near-perfection on the first attempt - this game is just so complex that I want to make sure I do plenty of information gathering first. My first forays into the start of the speedrun have been an important step, as I re-familiarize myself with how leveling happens in HG. Mostly right now I'm trying to avoid huge oversights, by knowing what items are available, what glaring weaknesses each boss has, etc. - why spend time whacking a boss to death when there's a sweet one-shot kill asking to happen? This isn't far from what will soon have to happen. HG does not afford me a huge advantage other games do - save files. I have to re-create the character from first level every time I want to try something new, in most cases. While I can sit and kill the same boss over and over to try tactics, I can't re-complete the same quest, or re-test the same boss's kill count mechanics. So as much as I would like to be able to focus on each leg of the run in painstaking detail... you're right, that can't happen here, HG just isn't built for it. After the initial passes of information collecting, I will have no choice but to do exactly as you suggest and do run after run, taking notes on changes to make each time. Well, not a lot will be lost in that event. All my information will still be intact, I'll just have to use that information in a different way. However, among the run-throughs I'm doing to gather information will also be a test of all boss kills using the build I've designed, so I'll know pretty soon. When I said "I've been going about this all wrong" I meant the level 3-10 leveling process. Instead of working with very little gold, mostly quest rewards, now I'm just grabbing over 300,000 at the very start and going into an area I would not normally be able to tackle, to grind XP much faster and save the opening quests for later. While the quests won't increase in value, it will still be a large boost over how I was doing the leveling. Bonus points if you can figure out where that 300,000 gold is coming from.
|
|
|
Post by simpetar on Apr 7, 2016 5:26:16 GMT
Bonus points if you can figure out where that 300,000 gold is coming from. A wild guess would be randomization. If you !randomize every piece of junk you find in rat cellar or Road to Town and sell them for profit, your wealth will probably be well over million gold by the time you hit level 5 (quests excluded).
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 7, 2016 6:08:00 GMT
Bonus points if you can figure out where that 300,000 gold is coming from. No clue, when I (rarely) make a new toon the first thing I do is !wallet transfer 100000000 Anyway, cool - sounds like you've got it all worked out + have the dedication to see it to completion eventually!
|
|