I’m fine with requiring success - it’s pretty difficult to fail that mission at all - IIRC, the only way to do it is to fight all the buccaneers and then surrender to the captain. What would your take be on refusing to help the nomads (which is very unheroic, but completes the mission in the council’s eyes as they obviously can’t join Tyranthraxus if they all get slaughtered), or opening Cadorna’s treasure, meaning you get a threat from him instead of a reward?

You can get rewards for clearing the library, Podol Plaza and the textile house, but they’re separate from the main quests in those areas and don’t come up on the roster. You also get rewards for killing the major monsters in the graveyard even if you don’t clear it. You’ll get token rewards for books from the library the council doesn’t find useful. Any minor reward is not preceded by the phrase “the clerk speaks”, which is why I suggested that could be a useful criteria to say what quests are compulsory. And I admit, that is entirely self interested as I don’t particularly want to grind my way through dozens of random encounters if I can avoid it!

Actually, re the 100% category, here’s a thought: since a true 100% category would be a nightmare both to do and to verify, (every random encounter that exhausts, every treasure, every planned encounter), what about an “all quest” category that requires you to receive every reward that begins with “the clerk speaks”? That way, you’d minimise the number of tedious grinds against repeated random encounters

So that would be:

1. The slums
2. Sokal Keep
3. Kill Norris
4. Get all five useful books from the library (extra books and clearing optional)
5. See auction in Podol Plaza (clearing optional)
6. Retrieve Cadorna’s treasure (clearing optional - getting threatened for stealing the treasure counts as completing)
7. Clear mansion
8. Clear wealthy area.
9. Clear temple.
10. Resolve nomad camp (foul means or fair)
11. Kobold caves
12. Clean river
13. Lizardman keep
14. Clear graveyard (individual rewards for clearing out particular undead optional)
15. Rescue Bivant heir (failing mission also counts)
16. Escape from Zhentil keep
17. Clear gate
18. Kill Tyranthraxus.

Thoughts?

I'd agree if it was only vorpal blades that made it through - for a long time, I thought it was. Elminster said something like "only the most powerful magic items can pass", but when I found out that a bog standard ring of fire resistance gets through as well, I was like, hmmm.....

I got the sense that it was just a way of nerfing your ridiculously high levelled characters - and a relatively cheap one at that.

I know the disintegration glitch. It also works if you dual class your characters and any items get automatically unequipped because their new class can't use them. Vorpal long bows ahoy! I think it only works on status effects that appear under the "Display" menu, so maybe that's why it doesn't work on the mace.

Incidentally, Pools of Darkness lets you take vorpal long swords and any ring that has a status effect attached (not rings of protection) into the outer planes without destroying them - I wonder if avoiding that permanent effect was the reason.

"Force Reaction" is an awesome name and I'm going to use it in future. :)

I tend to see it as glitch on the grounds that it was clearly something the programmers did not intend, but you could also make the same argument about any number of mechanic exploits, so it's probably splitting hairs. I also agree that there's probably not much point in a glitchless category, at least not in this game. I was more interested in how you'd do it. There are a few easy sources of experience lying around Kuto's Well and the library, and you could kill the basilisk in the library by just equipping mirrors.

100% would be fun, but I think I'd struggle to find the time - it's one thing to zip through the game in 15 minutes when you get a moment, a several hour long grind is something else. I would like to give it a go at some point though, off the top of my head, I can think of a few good ways to really shave some time off it. Grabbing the necklace, wand and dust early and duplicating the hell out of them would make most of the paths pretty smooth.

In the meantime, I think I've optimised my route as far as I can. I've tweaked the 2F/2M route a few times since the video, including:

1. Selling the wand of lightning instead of using it - it gives you 15,000 gold, which is more than enough to pay for everyone's training, and stops you having to go and appraise jewelry all the time.
2. Not training after Sokal Keep, but going straight to the temple - the experience you get there gives you enough for 3rd level, and the reward for Sokal puts you over the line for 4th, which tightens up that path.
3. Using the necklace more often against the orcs and hobgoblins, which reduces the time you wait for everyone to have their turn.

I'm going to have a few goes at it then upload my best one, rather than keep uploading every time I lower the time slightly, to save you having to verify each one, but my best so far is 12:34 and it's got some way to go yet.

You’re quite right. I stand corrected.

I first played Pool back on my old C64. I recently played it on an emulator and even speeded up, the loading times were unbearable. Looking back, it’s a miracle I ever had the patience to finish the game to begin with.

Very interesting glitch there. I wonder how exploitable it would be. I tend to agree that they’re hard to define. Maybe they should be defined as not doing anything too obviously cheesy. :) Duplicating items and walk up walk back are definitely glitches. Hiring NPCs to kill and loot them is cheap, but not a glitch.

I seem to remember there’s a glitch around the Bag of Holding which can give you backpacks full of various goodies, which could have some potential if used properly.

I also like the idea of a 100% category in theory, although I’m not sure I’d have the patience to sit through a multiple hour run. I’d simply define it as doing all the quests - although that in itself can be hard to define, because not everything you can be rewarded for comes up in the roster. Doing all the groves and caves in the wilderness would be a nightmare - they come up as very rare (1 in 32) random encounters, so it would take a very long time to get them all.

Well, speaking of new strats, I've just submitted a run (pending approval) that uses roughly the same route, but has two human fighters and two human mages, and duplicates the necklace after picking it up in the castle. I then mow down the fighters with the four necklaces and use a combination of the Wand of MM and the two fighters to take out Tyranthraxus on Quick, finishing up in 14:24.

As you'll see from the lack of comparison splits, this was the first time I tried this route while recording, and so while I was lucky with random encounters, only having one battle with goblins in the Slums, it was pretty sloppy in places, especially the duplication, so this definitely has some scope to come down, probably to below 14. But I'm pretty happy with the route - it allowed me to skip the graveyard visit, and the trip to grab the treasure in Sokal Keep.

Incidentally, as an academic exercise (not sure it makes too much sense to add categories with only a handful of people running this game), I'd be interested to see how fast you could do a glitchless category. I'd be inclined to do something similar to my first run, although obviously, you couldn't fight the mage in the castle that way - without the walk up walk back glitch, he'd wipe you out with his wand.

Here's the link:

I also used a few game mechanics to manipulate the size of random encounters and make sure the monsters surrender rather than run, so that minimises the time impact. The game uses an algorithm to determine your party strength, which is used to determine strength of most random encounters, among other things. Giving the F/C 10 hit points instead of the max of 11 keeps the score just low enough to stop the random encounters in the slums including leaders. And not buying armour for one character means the orcs and hobgoblins will surrender as they have a movement rate of 9 and intelligent monsters will only run if they’re at least as fast as your fastest character.

Very often at the moment TBH, although that’s more because I’m experimenting with some alternative strats, including using three characters and duplicating the necklace. If I’m playing the route I used on the video, there will be a death maybe 40% of the time, mostly against Tyranthraxus, but sometimes, it’ll be something like Mace or the 5th level cleric winning initiative and getting off a Hold Person spell, or maybe the mage in the castle knocking out a character with his wand.

Once you’ve got past instant deaths and the execution is a bit tighter, getting a decent pace is dependent on your luck with random encounters. You can avoid 50% of the encounters in the slums and Kuto’s well and most of the encounters in the temple and the wilderness. There’s only one spot in the graveyard on my route that’s actually in the random encounter zone, so you’ve got a good chance of getting through that unscathed. On my run, I managed to largely avoid random encounters, so I definitely got good rolls of the dice there. Each random encounter will cost you 15-45 seconds, depending on when and where, so getting sub 16 depends a lot on luck.

The last battle is one where you're definitely more at the mercy of RNG, but that would be the case no matter what route you take. I've experimented with using wands of lightning, but they're just too slow - the animation takes a while. Duplicating the necklace is a possibility that I'm exploring though.

Interesting you should say that, I tried that exact thing in one earlier today, actually. Some bad RNG luck and the execution on the duplication was a bit shaky, so I wound up at about 16:45, but it’s got potential. I also remembered there’s an extra Potion of Speed in the NW castle corner, which gives about the same XP reward as the sword and is much less of a detour, so that’s a small time save, and another character with 2 attacks per round doesn’t hurt.

That's weird. When I tried it on him, it just told me he was unaffected. Ah, the eccentricities of 30 year old games.... :)

Hmmm. I didn’t realise Stinking Cloud behaves more like later games in later versions. Might need to invest in a version 1.3 from GOG to make it fair.

I mean, it’s an advantage and disadvantage - you make a saving throw every round in the cloud and if you fail it, you’re helpless for that round, rather than just one for the duration of the spell, but consistency is still important.

I do want to experiment around with some different strats for the final battle, because that seems to be the most obvious area where there's real room for optimisation. I like the idea of hiring the 4th level cleric (I remember that from your run), problem is that they don't come up that often. But it would also give a few extra 2nd level slots that could be devoted to Resist Fire, which would definitely be useful against his flaming aura. And the idea of having meat shields is definitely appealing. The 4th level thief would be good for that purpose too.

I haven't had a Stinking Cloud work on him yet, but I do have a spare one left over on that scroll, so there's no harm in trying. It's possible that there's a difference between versions there (I'm playing 1.0 from the old WizardWorks CD from the 90s). I know Magic Missile works on him, and there's a Wand of Magic Missiles that I've got left over from the temple, so I suppose I could give that to one of the characters, put him on Quick and that's a guaranteed 10 or so points of damage a round. Every bit counts.

I think the biggest predictor of whether you win or not in the last battle is whether his fear aura works. 4th level characters don't get paralysed, but it does seem to weaken them, which I'm guessing from the names of the codes, mean that you'll only do half damage. Since the last battle is a damage race, you won't make it if he effectively has 160 hit points.

Oh, and I remember that thing about neutral NPCs from the module, so I experimented with that. I'm pretty certain it's only the evil aligned fighters that will turn on you.

And I’ve done three runs today that were on target for under 16 minutes and all came unstuck against Tyranthraxus. Oh well, I suppose that’s the risk you run by going low level. :(

I've just submitted a new run based on a strategy of getting experience from as many treasure hoards as possible, and an early trip to the castle to grab a Wand of Lightning and Necklace of Missiles to mop up any time consuming encounters like the orcs and hobgoblins on Sokal Keep.

I was pretty lucky with RNG on this one (a battle with a displacer beast in the wilderness notwithstanding), but there were quite a few stutters at the beginning and next time, I'm going to start looting some bows and arrows so that I can mop up the 8th level fighters a bit more quickly. I'd say with some luck and some tighter execution, this one could go down to close to 15m.

Thanks for uploading that, I enjoyed watching it and there were some really creative use of the game mechanics there. :)

I've been experimenting with a route which I'm hoping to upload in the next few days - basically, it involves avoiding as many combats as possible and mostly getting experience through treasure hoards. It also uses a 4 person party, which boosts the experience awards by 50% and reduces the strength of random encounters, which scale by party strength. It's a low level run through that aims to get just enough experience to get everyone to 4th level, which is the bare minimum needed to beat Tyranthraxus. Any lower than that and your characters run the risk of getting paralyzed by dragonfear. It's still a tad dicey then, and I've found that if the run comes unstuck, 95% of the time, it's fighting Tyranthraxus, which is a bit frustrating.

I agree that setting a standard cycle rating would be a good idea, because it can make a huge difference. I like the idea of setting it at 6000. I find 3000 a bit cumbersome, although that's mostly because I've practised at a much higher rate. As long as everyone uses the same one.