Posts for zidanax


1 2
12 13 14 15
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I was working on Fillmore, and I got that house-on-fire event that Sir VG got in his run. Kind of annoying, because I wasn't sure what caused it. I tried to find out what my latest attempt and Sir VG's run had in common. I noticed that in both runs, enough enemies were killed before the first construction phase to cause not just a corn field to be constructed, but also a house. So I did a little experiment: I saved state, then waited until the fire event happened. I then waited again, but this time I destroyed the suspicious house with lightning right before the fire event would normally happen. That time, it didn't happen, and it didn't happen even when I waited long enough for another house to be built on the same location. I reloaded the save state, and tried destroying the offending house immediately after it was built instead of just before the fire event would normally occur. No fire event happened, even when another house was built over the same location. So I'm guessing the way to prevent the fire event is to not kill too many enemies before the first construction phase, so that that house doesn't get built. The location of the special house is in this image EDIT: Looks like the maximum number of enemies you can kill before the first construction phase and still not get the house is 7.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
A Yes vote here, too. I also think that when an all cups done is run, it should obsolete this, but this is pretty nice for now :)
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Well, I can always check. Yet another test, but this one shouldn't take too long. EDIT: Looks like knowing the soul counter system will be helpful. Before the first lair in a town has been sealed, all of the souls you've freed come from the enemies you kill. So if you want to get as fast growth as possible going up to the first lair, you want to be careful to kill as many enemies as possible. I did some quick testing with watching the soul counter (the author of the max pop. FAQ actually gave the memory addresses!) going up to the first lair in Fillmore. and I got one more house than in my previous WIP. Not much, but every little bit helps. This also explains why my population growth wasn't as good when I tried to seal the upper-right lair first: I wasn't able to kill enough enemies to keep the growth rate as high as possible, because it takes longer to reach the upper-right lair. Depending on how long it takes to kill enough of the stronger enemies in later sims, I may end up getting Angel Arrows in order to keep the growth rate high in the beginning of those later levels.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
At least I won't have to test my strategies frame by frame. It doesn't really make a big difference when testing sim strategies whether or not you're playing frame by frame or in real time. Also, the slow rate of growth for towns you're not in limits the possibilities I have to test for somewhat. EDIT: It's better to wait in Casandra? I would think it would be better to wait in Aitos because then you'd have the growth in Aitos + slow growth in three other towns instead of just 2. Am I missing something?
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Did some testing and... it looks like the rate of construction in a town when you're not there isn't as fast as I would've liked. I may end up waiting anyway, but hopefully that can be minimized by building those extra pieces of road in Fillmore, and maybe some in Bloodpool too. Perhaps I will wait in Fillmore and pick up the scroll in the process, but it's unlikely. Even with some waiting to build up the population, it may still not be worth it due to the total of about 450 frames spent clearing the bushes, which is still more than I could save with any single scroll in Deathheim. In any case it looks like I'll have to spend a LOT of time testing sim strategies if I want to do this right. Depending on how long that takes, you may not see another WIP for awhile.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
With these extra pieces of road, Filmoa's maximum population would be even greater than the high 300's. I did a test in Filmoa where I went ahead and sealed the last lair, going from that lower bridge I built in the demo. I then stopped road construction and let the town grow on the roads I'd already built. The maximum population of Fillmore had increased to around 500, in contrast to a maximum population of 390 in my previous run, and it took less than one construction phase worth of time to build those extra pieces of road that allowed that. And yes, I noticed that you waited a while in Cassandra to complete that big square of road. Was that to have a high enough population to reach Aitos/Maranha? If it was, then.. that's the waiting I'm hoping this strategy will eliminate. I'm hoping that Filmoa will have grown to the new max of 500 or whatever it will turn out to be by the time I've sealed all the lairs in Cassandra. Then my waiting time in Cassandra would be greatly reduced/eliminated. Or did you have to wait for something other than building up the population? I noticed that you stuck around for awhile after you already had 950 people.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
OK, now I'm going to start messing around with sim strategies. I thought of one possibility for Fillmore that I didn't see in OgreSlayeR or Sir VG's run (or at least not in Fillmore). When houses have already been built on a square with roads, you can branch off pieces of road from that square into other squares without waiting for another construction phase, yet new houses can still be built on those quickly constructed pieces of road. I'm wondering if, since the people in sims build even when you're away, this might reduce time spent waiting for population growth in later sims? Here's a rough demo of what I'm talking about. EDIT: Since what's done up to the end of Fillmore Act 1 doesn't really affect how the Fillmore sim should be done, I've gone ahead and redone Fillmore Act 1 with the jumping trick. I was also able to make the boss somewhat faster. Overall I saved 30 frames up to the end of Act 1.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Yeah, I wasn't thinking about the various alternatives for weakening/defeating the enemy. I generally didn't have as much success with things like plots, but since luck manipulation could be used, you could probably use such methods quite often. IMHO, you'd want to be careful to use as many methods of defeating/humilating the enemy as possible to keep things interesting.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I've played a little of ROTK III and IV, and a whole bunch of VI and VII. They're fun to play, and you could probably make a TAS that would impress fans of the series through its technical quality, but I have to wonder if they'd be at all fun to watch for people who aren't familiar with the series. I always turned off watching other warlord's battles, because I found them to be terribly boring, which makes we wonder if people watching a TAS of one of these games would be bored by the battles. True, you could play the battles WAY better than the computer does, but it would still essentially look like a bunch of little sprites running around on a map with swords occasionally appearing between the sprites (for ROTK III, at least, IIRC), which always bored the snot out of me if I wasn't participating. Other than that, I'm guessing it would be lightning fast menu manipulation and some diplomacy. If you do end up making a TAS: I can't really give an informed opinion on which game to TAS, because like I said, I've played more of the later ROTKs. I generally play Historical mode, but if the sort of luck manipulation you mentioned would make the TAS more interesting, I'd go for that. I don't know about the earlier ROTK games, but in VI at least, you generally can beat the game faster in later scenarios. You can choose a warlord who already has a lot of territory and resources that way. But starting with a weak warlord may be interesting just to kick everybody's ass really quick in spite of the warlord's weakness. And yeah, multi-player would be REALLY cheap. If you did that, you could just pick a three kingdoms scenario, control all three of the kingdoms, and have two of them ally with the third (or is it submit? I can't remember).
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Yeah, I only had to experiment a bit before finding a place where I could hit the dragon four times as he swoops down. Interestingly enough, I'm having trouble saving time off of the Maranha boss using stardust now that the no scrolls version is so much faster. I tried the Dragon without scrolls, and was able to beat him in 901 frames instead of 1142, still without scrolls. Once I use stardust on him, I should be able to save even more time. Boss 6 EDIT: I didn't save quite as much time on the Northwall Boss with stardust as expected, but I did save some. I think my plan will be to use 1 scroll on each of the first four bosses, no scrolls on the 5th boss because I can't seem to get non-negligible savings there, and two scrolls on the Northwall Boss. That may change if for some reason the stardust patterns change to something quite different when I actually get to Death Heim in my TAS.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
OgreSlayeR: Your trick for an extra boost while jumping is really coming in handy. It gives the hero just enough of a boost to hit the Maranha Act 2 boss. Here's an SMV of my current best attempt at the boss with no scrolls used. It's actually 138 frames faster than OgreSlayeR's attempt with two scrolls thanks to the jumping trick. As for the first four bosses: The only boss where two scrolls saved time compared to one well-used scroll was the Bloodpool boss. Apart from the first boss, where I saved 420 frames from using stardust, I haven't been able to save more than 350 frames per boss from usage of scrolls. Also I was able to defeat the Maranha boss in 1156 frames, and that's without scrolls. Given that OgreSlayer was able to defeat the Dragon in 1142 without scrolls, and that that could certainly be improved now that Snes9x has frame advance, I can state with certainty that getting the 7th scroll would not be worth it, since I would have to save about 3400 frames from usage of that one scroll. It might be nice for a normal speedrun, but it's a waste of time to get the 7th scroll in a TAS. I will, however, finish finding out how much time scrolls would save on the last two bosses.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Looking at it through frame advance: Is it that the boss appears to be undecided about which way to go before slamming into the wall? I didn't notice it at first when looking at it in real time. Now I can see it, but I wonder if most people would even notice? Anyway, I don't think it really makes a huge difference in entertainment. Personally, I think the boss actually slamming into the wall is funnier than the boss being undecided before finally slamming into it.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Question for OgreSlayeR/SirVG: Do you know of any ways to manipulate how the stardust falls on the enemy apart from having the enemy on a different part of the screen when you cast the spell? I can cast the spell again to change the falling pattern, but I'm not noticing any real changing of falling patterns from just walking around.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Perhaps having the SRM would help?
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Probably because the SMV is started from a saved game. I wasn't sure whether that would make a difference. I guess it might. Which version of SNES9x are you using?
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
The first boss was surprisingly quick. Measuring from when the hero is fully materialized to when the boss died, it took me 775 frames using no scrolls. After many tries, I got one scroll to bring the boss down to 2 life, after attacking the boss twice at the beginning (that's neccessary because the boss goes offscreen very quickly at the start). I then just attacked the boss twice before it was able to get offscreen. Time: 355 frames , resulting in 420 frames saved. Using two scrolls does not incur any savings on this boss. Here's an SMV if you're interesting in watching the boss get owned :)
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Yeah, I agree that it's all about Deathheim and not about the houses. I guess I just wanted to say that the population gain from the new houses would likely not be enough to compensate for not saving quite enough time in Deathheim. I'm not sure I'd be able to shave off any frames on Bloodpool Boss 1 with a second MP scroll. I checked on how long the fireball animation takes: the hero stays in that sword-in-the-air position for 63 frames, if I'm counting correctly. That's 3 frames longer than the time to execute 3 sword attacks. So where the fireball comes in handy is when the boss is about to attack you, otherwise you're probably better off doing regular attacks, unless you can somehow make the fireball do more than 3 damage against that boss. I would save some time on the 2nd Fillmore boss, but it probably wouldn't be more than half a second thanks to the amount of time you stay in the magic casting pose. Anyway, even several seconds saved on those two bosses wouldn't be nearly enough time to make a 7th scroll worth it. Remember, to make this worth it, I would have to save about one minute , so it's all about Deathheim. EDIT: I'll test this stuff, if for no other reason than to get a good "feel" for each boss, but I'm seriously doubting that even Deathheim would save me enough time with a 7th scroll. It looks like it would be the Maranha and Northwall bosses where the most time could be saved, but Ogreslayer manages to kill even those two bosses in about 3300 frames from the appearance of the Maranha repeat to the death of the dragon. And he only had 5 scrolls at the beginning of Deathheim, not 6, and he only had 2 by the Maranha Act 2 repeat. EDIT: What I think I'll do is test which bosses I should use stardust on in Deathheim, and how many scrolls for each boss. To a certain degree, it's obvious, e.g., the Maranha repeat and the Northwall repeat could use stardust. But to be certain, I'll see how long it takes me to kill each boss given up to 3 scrolls per boss (possibly 4 in the case of the Maranha repeat). So I'll see how long it takes to kill a certain boss sans scrolls, then test with one scroll allowed, then with 2, then with three. I'll repeat that process for each boss. Yes, that'll take a long time to test, but it's probably the best way to find out how many scrolls I'll want for Deathheim, and how to balance out the scrolls between bosses. In the process I'll get a very precise measure of how much time a 7th scroll would save. So I'll go ahead and test things out that way. This could take a while, but I think it will be worth it.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I did a little testing to see how long it would take to get back to building towards lair three. It looks like the delay would be roughly 3400 frames. Actually, the delay was 3561 frames, but I'm giving a conservative estimate because some of my maneuvers were slightly imprecise, and the route change could have caused subtle changes of some sort. Are you guys sure one MP scroll could save 56-57 seconds? I realize that building a road would allow for more houses, but because of the forests in that area, the maximum number of houses in those three squares is limited from the usual 15 to just 8. If I wanted to raise my population, there are other places where I could just build roads into 3 unobstructed squares, and get 15 houses. EDIT: What am I saying, I should just go ahead and test using 7 scrolls compared with using 6 scrolls in Deathheim (I'm not sure why more than 6 would come in handy anywhere else). I'll report the results back when I get that done.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Sir VG wrote:
Wait wait wait...there's TWO scrolls in Filmoa?
Yep, I tested it. When you build a road on the square highlighted in OgreSlayeR's picture, the people of Fillmore tell you that they found an MP scroll.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I've found a document on GameFAQs that could end up being very helpful in determining how to go about doing simulations. It's a maximum population guide, but it could be very useful because it has an extremely in-depth analysis of how the population part of the sims works. For example, it has a table of supporting structures (i.e., structures that feed people). Apparently, those purple buildings in Fillmore are mansions. They can support 72 people each, in contrast with 48 for wheat fields and 32 for corn fields. This seems to support my hypothesis about the Fillmore people no longer needing wheat fields when they're at a high enough civilization level. In fact, the author explicity states this sentiment a little later. I will be reading the whole document before I redo the Fillmore sim. At a quick glance, It looks like it has a lot more quantitative info than most FAQS. It touches on all sorts of elements of how population works that most people wouldn't notice, such as the soul counter system, and how the scores you get in the action segments actually influence the maximum population you can get. I'm not sure just how much of this will actually apply to a TAS, but with the depth of analysis in this document, there will probably be some applicable info. Sir VG: The FAQ was very recently submitted to GameFAQs. It was submitted on March 14,2007, after you submitted your latest 1:10:06 SDA run. If you ever decide to try and make an improvement, I would suggest looking at the guide now that it's available, because it looks really detailed.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Sir VG, I'll do some further testing to make absolutely sure that what I said in my previous post about wheat is correct, but I should mention something else. When I tried destroying all the crop fields in Fillmore, then building more roads, the people of Fillmore never built another crop field. Instead, they built more of those purple houses. So my hypothesis as to why giving them wheat appears to make no difference is that they are fed by something being produced in those purple buildings, possibly fish, since they learned how to fish at the highest civilization level. I will also do some testing to try to find the minimum number of Magical Stardust castings needed to kill each boss. This will give me a much better idea of when to get those magic flags, or whatever you call them, and when not to. EDIT: Correct me if I'm wrong but it looks like the MP scroll in Fillmore would take two construction phases to reach. There is a period of 720 frames between each construction phase, plus some lag before actual construction, plus the time for people to actually come out and build stuff before the countdown begins again. So that MP scroll would have to save me at least 1440 frames of time in the boss fights, possibly a lot more thanks to the time it takes for the actual construction. To me, 24 seconds saved throughout the various boss fights from one more MP scroll sounds like a tall order. What do you guys think? Would one more MP scroll save that much time? Perhaps this should be part of my testing.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Sir VG wrote:
Filmoa Sim: Good, though I wonder why you skipped waiting the turn at the end. Building the road to the end connects Filmoa to Bloodpool, allowing Filmoa to have Wheat. Wheat allows better population growth than Corn.
Thanks to fast forward, I did some testing in Bloodpool to see how much the people of Fillmore would reproduce behind my back when they weren't given wheat and when they were given wheat. It turned out that giving them wheat made no difference at all to their maximum population, given that no extra roads are built or obsolete housing destroyed. I would have to build more roads or destroy more housing before getting wheat to the people of Fillmore would be useful. EDIT: Given Bisqwit's comments about the downtime in the Fillmore sim and the fact that I might be able to save some time in the Fillmore acts now that I know how that that jump trick works, I think I'm going to restart this run. I will post a WIP of the Fillmore sim when it's only partway done, because I will definitely want comments about how I'm doing entertainment wise. I will admit that my sense of entertainment is not very good. I will also look at your guy's other suggestions and see if they'll work out (If not, I'll say why).
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
With sims, I can test different strategies in real time/fast forward, so I can mess around a lot with stuff and see what works best. As mentioned, I did make some strategy changes in Fillmore from what Sir VG did based on my analysis in Fillmore, such as the different route to lair 3 that lets me move the construction pointer back to the shrine faster. I may find more changes like that in Bloodpool. If my strategy changes turn out OK, I wonder if Sir VG will be interested?
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Well, that was quick. Bloodpool Act 1 is not as hard to TAS as Fillmore Act 2. I saved 165 frames, mostly in the boss fight. The WIP is here
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I just found a new trick to make the hero move faster off of jumps, although I'm not sure if this is already known. Depending on when you start a jump, the hero will move faster than normal for a few frames. Perhaps this has something to do with why I had such a hard time matching the times you had in some segments of Fillmore Act 2. EDIT: wow, just knowing that trick really helps me know what I'm doing in these platform stages. I'm not having to fumble around nearly as much to get good segment times in Bloodpool Act 1.
1 2
12 13 14 15