Roll Eyes
Lips Sealed
1 page
List results:
Search options:
Use \ before commas in usernames
I'm currently planning to run Pool of Radiance (single-segment GOG version, IBM version 1.3).  I'm still planning and playtesting, and I just wrote this:

partly so that I had enough information to come up with a solid plan.  I'm also new to taking speedrunning seriously, so I'd appreciate advice from veterans.

I was originally hoping to get under 5 minutes, but since I can't turn off sound, and can't speed up DOSBox past the GOG standard, it looks like I'll be aiming more for 20 or so.  (Yeah, for a game that back in the day took about 50 hours to finish.)

There are only two combats you absolutely have to fight to finish the game: the two at the very end of the game.  Everything else revolves around getting strong enough to win those two combats.  The obvious step is to grab the Dust of Disappearance, which means you have to deal with Mace and his minions.

I have a route that's reasonably reliable that gets into 5 combats, and a faster, much less reliable route that fights 4.  I'm not sure yet which I'll be doing, and I haven't refined either of these yet.

Is there any interest in a 100% run?  That requires 54 combats by my count, and I'd probably end up actually fighting around 60-65.
Thread title:  
Might be magic...
I'm keen to see the speedrun of this. Not really that interested in a 100% run (not quite sure how it would be defined in this game either).

I had a quick look at your FAQ - it's quite long so haven't finished it yet but noticed your "SECTION 1.4 - Things I Want":

(1) a complete, original physical copy of the IBM version 1.0 or 1.1
- I actually did have this about 25 years ago and might still have the original 5.25" disks around somewhere - could look for them if you're interested. It might be hard to find a drive to read them now though...

(2) a more elegant way to remove the code wheel protection in version 1.3
- Isn't the version 1.3 already cracked? I can just type anything and it works. There might be a way to completely bypass the screen, but that could potentially make the run invalid. Using hacked copies of a game is generally a no-no, even though the version itself is already hacked
Edit history:
DenverCoder9: 2016-10-06 10:23:26 am
DenverCoder9: 2016-10-06 10:23:09 am
DenverCoder9: 2016-10-06 10:22:52 am
DenverCoder9: 2016-10-05 08:00:43 pm
DenverCoder9: 2016-10-05 03:55:45 pm
DenverCoder9: 2016-10-05 12:50:30 pm
Quote from gammadragon:
I'm keen to see the speedrun of this. Not really that interested in a 100% run (not quite sure how it would be defined in this game either).

I'm defining that as: finish all 16 main council quests; also defeat Norris the Gray and clear the Wealthy Area; get rewarded for all 18 of these.

I'm not including the rewards just for eliminating all monsters in an area, if the council doesn't ask, because those are nothing but random combats, and one of them is buggy.  If you want to include them, that's another 20 combats.

This is a little arbitrary, I admit, but I think doing all 16 council quests is a must for a 100% run:
1. The Slums: 24 combats (yeah, ouch -- no way around this)
2. Sokal Keep: 1
3. Mendor's Library: 2
4. Podal Plaza: 0
5. Cadorna Textile House treasure: 1
6. Temple of Bane: 1
7. Kovel Mansion: 2 (I was counting this as 3, but it can be done in 2)
8. Nomad camp: 0
9. Kobold caves: 5
10. Yarash's Pyramid: 1
11. Lizardmen keep: 1
12. Rescue heir to Bivant: 0
13. Zhentil Keep outpost: 2
14. Valhingen Graveyard: 5 (this also has basically unavoidable random combats)
15. Stojanow Gate: 2
16. Win the game: 2
total: 49 combats

and I'm throwing in:
17. Norris the Gray: 3
18. Wealthy Area: 1
total: 53 combats

(2) a more elegant way to remove the code wheel protection in version 1.3
- Isn't the version 1.3 already cracked? I can just type anything and it works. There might be a way to completely bypass the screen, but that could potentially make the run invalid. Using hacked copies of a game is generally a no-no, even though the version itself is already hacked

I got the GOG version late in the writing process, and didn't remember to fix that; I should go back and do that.  In fact, I could have just compared the GOG version to the versions I have.

I no longer think my 5-combat route (for a regular run) is viable.  It involves completing Sokal Keep, for the XP and to get access to the boat.  I don't think this works, because the required combat against 34 orcs and 15 hobgoblins takes too long.  I end up spending a lot of time watching orcs and hobgoblins putter around.  Doing this first is still probably best in a 100% run, since it has to be done anyway.

So now I'm heading west through the Slums and never returning to New Phlan.  That means never training, and winning the game with level 1 characters causes unusual problems, but I think I have a way around all of them.

The first combat is against Mace and his minions, and winning this combat is actually easy compared to the rest of the run; cast three fireballs from scrolls, you're basically done.  I'll probably only cast two now, because the third probably helps more in the final combat than here.

The current plan involves casting these spells: Bless, Cure Light Wounds, Protection from Good (no, really!), Hold Person, Prayer, Shield, Sleep, Mirror Image, Stinking Cloud, Blink, Fireball, Haste.  Any spells that aren't level 1 are getting cast from scrolls.

I now think it's best to have everyone multi-class to include cleric, just so that everyone can cast Protection from Good.  I can't just have magic-users learn it, because I'm now never training; and everyone has to know it, because you can't target another invisible character with a spell in combat.

edit: Stojanow Gate requires 2 combats, not 3; Kobold Caves require 5, not 4
Also, as far as runs for later games go, would transferring characters from previous games be interesting?  This is actually how many people play the games, but the problem with doing this is that you can ridiculously front-load your characters and not have it count against your time.

If so, I'd be doing a lot of that; if I did this for Curse of the Azure Bonds, I'd have 5 elf fighter/magic-users who start at level 7/6 and 107 hit points (and 1 level 6 cleric with 78 hit points).  That's far stronger than people would expect, and it would look like cheating so I'd have to explain how I even did that.

(If it helps, a few games do require you to transfer characters.  I suspect I could beat Wizardry II in less than a minute, and that game requires character transfer from Wizardry I.)

I have another question: is it correct that game timing does not actually start until after I create my characters and start the adventure (so the second I hit "B" to begin adventuring), but I also have to show character creation anyway?
Everyday is puppies and sunshine...
Quest for Glory has a similar mechanic. There is no multigame gauntlet precedent, so if you are doing that, HEAVILY frontload them. Max stats and level if possible. It'll be a New Game + category.
Yeah, Quest for Glory done that way is going to be full of stuff like "break world record for time spent nose-picking", "cast spell X umpteen zillion times", and "let's fight the multigoblin combat, again!"

Wizardry II would require raising a samurai to comically high levels so that you can one-shot the animated magic sword naked, for no reason other than to shave a few seconds by not buying equipment.

For Curse of the Azure Bonds, I may have to look more into the wacky world of Pool of Radiance done with Constitution 23+, so that my transferred characters have about 240 hit points.  If I'm not transferring, my bet is currently on 1 paladin, 4 elf F/MUs, and 1 human cleric.  I haven't analyzed this game in detail yet, but I suspect it's not worth the time to mess around with dual-classing here.

Secret of the Silver Blades: transfer 5 dual-classed ranger/magic-users 11/11, and 1 dual-classed paladin/cleric 10/10.  If I'm starting this from scratch, I probably do get into dual-classing, since there's so much XP available.

Pools of Darkness: transfer 5 dual-classed ranger/magic-users 15/15, and 1 dual-classed paladin/cleric 15/15.  For 100% you'd need to replace one of the ranger/MUs with an elf F/MU/T 7/11/18, since you need a thief for Dave's Challenge.  If I'm starting this from scratch, it's probably best to use at least one of the pregenerated characters, because a couple come already dual-classed.

I haven't really planned anything else yet; I'm working on the Pool of Radiance FAQ and speedrun primarily.  And yeah, transferred characters should basically be their own category.  It's not the same as New Game+ though; in Curse, 5 human dual-classed ranger/MUs 10/11 loaded with all the best equipment is going to be far better than the 5 elf F/MUs without any equipment, even though only the latter ever gets the benefit of Constitution 22+.

I don't need a ruling on transferring any time soon, but it's something to think about.
Edit history:
mrprmiller: 2016-10-05 09:04:30 pm
mrprmiller: 2016-10-05 08:57:20 pm
mrprmiller: 2016-10-05 08:56:06 pm
Everyday is puppies and sunshine...
Whoa whoa whoa... I have to defend QFG.  My thief route never picks a single nose, let alone grinds any stats.  The Fighter route definitely does grind, though, in game 2 and 3 (although no goblins are ever harmed, and only accounts for about 5 minutes total of the 1:35:00 for all 5 games).  But I do know from my own experience when I proposed that for QFG, an admin answered me that there weren't any game categories for the multi-game gauntlet like that for SDA purposes, even when it makes the game more difficult, so the transfer was more of a gimmick and less a category.  And if a character swapped in with bigger (or even lower) stats from a different game comes in, it will probably be New Game + even if that isn't technically right.

That being said, I personally would LOVE to see a route with transferred characters through the Gold Box games.  The order of the games in QFG might be set, but that's not the case with the Gold Box games, and picking your order might become much more important. Smiley
Oh, heh, sorry ... I was just recalling my days of maxing out my Quest for Glory characters for the next game, just in case.  That was a LOT of grinding, as I'm sure you're aware.

Hmm.  If New Game+ and transferred characters count as the same category, then it doesn't pay to actually transfer characters in the Gold Box games; heavily loaded endgame characters from the same game will always be superior.

Except I don't think people really do that, but they do transfer characters from one game to the next.  That seems counter-intuitive to me, since the rules don't correspond to how people actually play the games.  Both are very much possible in this case; I suppose the question just never came up before, because there aren't many games where that is the case.

There's also marathoning an entire series from start to finish.  In the case of Pool to Curse to Secret to Pools, it'd be best to rotate characters in and out.  I'm pretty sure this could be done single-segment, no reloading of any kind allowed, and not allowing difficulty lower than Champion once that becomes an option.  I don't think I want to actually speedrun that, though.  Smiley
Might be magic...
Looking forward to the run, my memories of Pool of Radiance involve me dying stupidly and reloading hundreds of times. Would be fun to watch a pro break the game!
Well, the run as now planned isn't going to be of much use for a normal playthrough, because it involves grabbing consumables and burning through them with basically nothing left over.  You know, the stuff most people are afraid to touch in normal play and end up never using.  Smiley

I could really use another fireball or two, except there just isn't a time-efficient way of getting my hands on another I can see.

I did realize there's a pretty serious practical issue with a 100% run: the fastest way of doing that is probably going to be doing Ohlo's quest immediately, and rolling a Necklace of Missiles as the item awarded.  That's a very low-probability event, and I doubt it's going to be practical for single-segment; since single-segment is all I want to do, I'm going to let someone else worry about that.  Or maybe I'll worry about it if competition shows up.  Smiley
Now that I will never be training characters, it's clearly best to just use 6 half-elf cleric/fighter/magic-users.  My current plan is also apparently overkill, which is fun to say with level 1 characters; I don't actually need to buy nearly this many arrows.

I'd like a rules clarification.  "M"odifying my characters doesn't count against my time:

Quote from SDA rules:
Starting point: when the player first gains control of the game's character, timing begins. This is defined as the frame before the character starts to move (in case that's hard to define, a reference point, such as fade-in can be used instead). If the game starts by going through menus that have an impact on the gameplay (adjusting stats or equipment) or an overworld map (or level selection screen etc), then those are defined as ”gain of control”. However, if character creation is done as an independent activity before the game starts, it's not considered as part of the run time, even when it involves defining stats, equipment or similar. Having camera control (for example Half-Life), but no means of impacting the character's movement, is not considered gain of control for the purpose of SDA-timing.

so not only do I not need to do that with optimal speed, it sounds like timing doesn't actually begin until the tour of New Phlan ends.  The game even helpfully says "you are on your own now".  Is that right?

So I can take the time to do things like lower unneeded stats all the way to the minimum, which helps cut down on random encounter size.  A forced random combat isn't necessarily a run-killer because they scale so small, though it's still not good.
Edit history:
mrprmiller: 2016-10-09 08:31:32 am
Everyday is puppies and sunshine...
You are correct.  During the verification process for QFG1, it was asked why I didn't use a stat glitch that happens during character creation. Firstly, I thought it was cheating, but an argument was made that it wasn't. When I added that it would take about a minute to set up, I was informed that because it was before the actual gameplay started, it doesn't count as part of the time.
Thank you!

That does mean I should keep in mind the dummy scam of creating and dumping characters just for the starting gold ... but I think this would save something like 1 second on average, maybe not even that.  The reason to do this is to have the starting gold to buy two-handed swords instead of long swords.  This only ever helps very slightly; a two-handed sword does an average of 1 extra damage, except for the most part I'm relying on fireballs and coup de graces to win combats.  This also doesn't help against Tyranthraxus (where the two-handed swords make more of a difference), because I can take all the two-handed swords +2 I need from the combat with his guards.

I probably won't actually do this, as I expect that in a majority of runs it won't help at all, and it will cost me a half a second or so of menu time to buy two-handed swords instead.
Everyday is puppies and sunshine...
As it's a function of the game, I would think that would be allowed. I'd love an admins input, just to be sure. It's a dirty tactic, but not a cheat code.
This used to be the gold standard in "If you're stuck and can't find enough supplies" tips for starting RPGs. Create a bunch of characters, steal their stuff, buy what you need, carry on. I know I'm the new kid hereabouts, but I'd at the very least be looking funny at anyone who said that this behaviour wasn't okay.
Yeah, the main reason not to do that isn't because it's cheating, it's because it's not actually worth the time it takes, even if it's only about half a second on the timer.

Another thing I was seriously considering and decided not to do: getting a Wand of Lightning.  Getting another fireball isn't happening, but I can get a wand of lightning fairly easily, and in fact there are not one but two places in Valjevo Castle to get one, and they're also both almost right on my path.  Getting one takes extra time, though, and I run a significant risk of getting multiple characters killed by the enemy mage using the wand on me, so I very much doubt this is worth it.

I added another shortcut to my plan: I'm crashing through one of the hedges in the Valjevo Castle Hedge Maze.  The chance that this kills an actually important character is well under 50%; this is another reason to include cleric as a class on everyone, as that gives a 55% success rate of saving vs. poison.  A single cut saves a significant number of steps and takes out the random step of using the maze teleporter, so I'm doing it.  I can in fact recover from an important character getting killed by the hedge's poison thorns (this is another reason it helps to have everyone the same class), though it will cost me 10-15 seconds as I have to juggle equipment and marching order.
Okay, I can do this start to finish in about 20 minutes, even with significant mistakes and being forced into one random combat.  An unwanted combat adds around 10-20 seconds, so it's likely though not necessarily a run-killer.

My goal is now 15 minutes.

I still need to refine my plan further -- all the buffing I do for the final battles wastes a significant number of seconds, so I'll have to cut down on that.

I do have a question on recording.  DOSBox's native video recording does so at a rather strange framerate.  Do I need to worry about that, or is it fine for more experienced encoders or Anri-chan?
Edit history:
gammadragon: 2016-10-13 02:32:47 am
Might be magic...
Nice - are you able to upload a practice run video? Should be fun to watch the progress.

Use Yua to encode the final video before sending to SDA. It should be able to handle the DOSBox recorded videos.
Everyday is puppies and sunshine...
DOSBox works great with Yua.
Everyday is puppies and sunshine...
Not to revive an older thread, but I'd love to know if there is an update on this.  Cheesy
Might be magic...
Me too, hoping for an update before Christmas Smiley
Everyday is puppies and sunshine...
"Me too, hoping for an update before FOR Christmas :)"