In the end, really, it just depends on how enemy HP plays into the game. In some games, enemies with lots of HP are dealt with exactly like enemies with little HP; it just takes longer before the player can move on. In others, being unable to rapidly dispatch enemies creates more complicated battle scenarios which are more entertaining. However, the latter scenario is pretty dang rare, so since we're so fond of hard-and-fast rules around here, the accepted wisdom is "if higher difficulty levels only result in more enemy HP, then don't bother."
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Presumably he picked up the Proton Burst (is that its official name? I always called it the mining laser) because he thought it might come in handy, but it didn't, and he didn't want to go back and fix it, this being a test run. :)
You sure the laser weapon doesn't hit the second half of the stage-3 boss? I know he has all those balls flying around that could get in the way; maybe the shots hit those some of the time?
I didn't know that the searchlights recharged your energy; that's cool. Neat little touches like that in this game. :) I think though that he'll be best-served by simply being more frugal with energy use (and using the homing laser! Can you tell I like that weapon? :D) and not stopping along the way.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Firefox can also do the keyword search things like you describe. Right-click on a search box and select "Add a keyword for this search", give it a keyword, and then type the keyword and your search query into the URL bar. Some of mine:
en: http://en.wikipedia.org/wiki/%s (Wikipedia page)
ens: http://en.wikipedia.org/wiki/Special:Search?search=%s (Wikipedia search)
maps: http://maps.google.com/maps?q=%s (Google Maps search)
And so on. It's insanely handy.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I spent the better part of a day playing Sonic Heroes, and then when I went to bed, I kept seeing streaks of red, blue, and yellow light and generally having a very hard time getting to sleep. Generally I find that any day that's devoted nigh-entirely to a single physically-inactive activity will result in trouble sleeping.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Hm. The big problem with the flamethrower is that its projectile speed is variable -- if you're flying forward, it has practically no range. There were a lot of missed shots against the second half of the stage 3 boss because of that (though a tightly-optimized bossfight might be able to avoid that problem). If you can deal double damage with the seeking laser (by having both hits land), then it might be faster to use against the stage 2 boss/fightercraft and the targets in stage 5 at least (you might also be able to die less against the targets with the laser, since it'd destroy their shots). How much damage do the plasma shots deal? I assume it's 1/projectile.
There's also the question of energy drain; IIRC the flamethrower kills your energy pretty quickly while the laser's more efficient, and you definitely ran into energy problems with the plasma shot. If only the laser wasn't off-path!
I think plasma's definitely the weapon of choice for stage 6, though, since the laser doesn't work well against the targets (it hits the center, not the gunpods). So it's not like you could grab the laser and then skip the plasma. :\
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Ahh, Star Ocean. A massive technological feat with a game slapped onto it, and then a plot slapped onto the game.
And yeah, Milly has a tail, as does the protagonist. Otherwise they seem to be basically human. Party members can also include actual humans, all-but-human aliens, an angel, a werewolf, and a catgirl.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Awesome! I've been waiting for a TAS of this for a while. :) Glad to see someone finally made one. Here's my comments as I watch:
Stage 1a:
* You get some lag, which might be reduced if you weren't firing all the time. Then again, it might not be. Still, it looks sloppy to be firing if you don't need to (and I know the EX-UP fires when you do).
* You seem to be flying up to the targets and then dropping down to shoot at them. Wouldn't it be faster to be flying at their level in the first place? Particularly the one at ~42s in seems like it could be done faster. Of course this is a test run. I'm going to stop commenting on minor flying errors after this.
* Funny how all the enemies you skipped through the level pile up at the end. Might be some opportunities there for playing around. Or if they're contributing to lag, it might be worth killing them.
Stage 1b (first boss):
* Wouldn't it be faster to start out in the EX-UP, kill his legs, and then do the rest flying? With your path, you have to start at the top of the boss, work your way down, go into the EX-UP to kill his legs, get out, and go back up to the top to kill him. I know the EX-UP divides its fire between the different targets, but it seems like the cost of going into/out of the EX-UP an extra time (and the cost of the extra flying) would make up for not being able to use the flamethrower as much.
Stage 2a:
* Is there a significant difference between flying forwards and backwards?
* Have you done any damage checks for the guided laser vs. the other subweapons? It's a big help in normal play for augmenting your machinegun since it fires straight forward if turboed and will destroy bullets (especially useful against the plasma turrets in the last two stages).
Stage 2b:
* It looks like two of the "fightercraft" things lagged significantly behind the others. Might just be their flying speed, but it might be that you're getting too far away from them and something weird's going on?
* You use the flamethrower against the fightcraft and then the single machinegun against the boss since you're out of energy. Seems like going into the EX-UP for doubled machinegun would be worthwhile for one or the other (if the EX-UP doesn't track the boss, then of course use it on the fightercraft). How does the other subweapon (rolling landmine?) compare damage-wise?
Stage 3a:
* You use up your energy early on and then hover along the top of the map; hovering has a slower horizontal speed than straight flying, though, which means you don't get to some of the targets as quickly as you could. You also use the birdmissile against some mooks, which means it's not available for use against targets.
* Picking the right route here's gonna be tough. I almost wonder if it'd be faster to fly all the way over to the right and then head back to the left, picking off targets as you go. Probably not since the birdmissile would get confused.
Stage 3b:
* I always used the rolling mine on the first half and the seeking lasers on the second half. You don't have the lasers, but you should have time to switch at halftime. Properly-timed, a rolling mine will make him close his eye as soon as he opens it.
Stage 4a:
* Stylistically I think it looks better if the only window you break is the one with the mining laser.
* Now I know why this level was impossible in normal play -- I didn't have a turbo controller! Thanks a lot, game developers.
Stage 5a:
* This is lagtastic with all the ships flying around as soon as you break stealth. It's probably worth going undetected for at least a little while.
* Oh, you died! Bah! I can understand why, given how little time it takes to continue, but it still rankles.
Stage 6a:
* When you do this for real, it's worth investigating if the boss dies faster if he's at the bottom of his up-and-down motion.
Stage 6b:
* Any way to dodge those shots aside from using the jets? Or do you fall for a fixed amount of time, as opposed to a fixed amount of distance?
* You start firing early; do those shots deal damage?
* The death here might actually cost more time than it saves.
* You never did use the mining laser. Oh, well. Cool weapon, but the power draw means it's probably not worth it.
All in all, looks very solid! I think the only major question is if it's worth picking up the seeking laser weapon.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Along similar lines, Monopoly. Circus Charlie, too. And Dungeon Magic, for that matter.
And then there's the category of "I don't expect to see improvements because I don't expect anyone to want to TAS this game again." Games include Castle of Dragon, Cobra Triangle, etc.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Video looks fine to me, but man this game has terrible music! It resets on every screen transition! The art's not so great, but I'm used to looking at ugly art.
At least the TAS is good. Nice work!
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Per the rules for this site, you need a good justification for using a non-English ROM, e.g. for better music (Castlevania 3), glitches that don't exist in the English version (Chrono Trigger), higher difficulty, etc. Shorter text bubbles isn't a good reason because it doesn't actually contribute anything to the gameplay.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
To clarify, all of the upgrades that count towards percentages in Metroid Fusion are E-tanks or ammo upgrades (and you have a small but non-zero initial capacity for both missiles and powerbombs). Since taking damage is not generally helpful and Samus' beam weapon outperforms her missiles damage-wise, there's no point in getting any of the upgrades if your goal is speed. Hence the any% TAS is also a 0% TAS.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
The trilogy port to the SNES is really pretty lousy, in the area of music especially. I don't consider the port to be an improvement in any way; the NES versions are better.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Super Mario World (1 Star)Super Mario World "small only"
Super Mario World any%
Super Mario World 96exit
Super Demo World 120exit
"Small only" gets my vote for this one. As noted, the other two SMW TASes suffer from the player flying over much of the level. And I don't feel that SDW should be starred, just as a personal bias.
Sonic (1 Star)Sonic 3 & Knuckles
Sonic Advance 2
Sonic The Hedgehog
Sonic The Hedgehog 2
Here I'd say either Sonic 1 or S3&K. I'm biased a bit against S3&K because I don't like the zip glitches -- IMO they're similar to flying over levels in SMW -- but the rest of the TAS is top-notch.
I'd actually say that the Metroid 2 glitched TAS could go onto this list too; it's highly entertaining for just how badly broken the game gets. I don't think we should have more than 1 100% starred TAS, though, and of those two SM 100% is the clear winner, if for no other reason than that, because of damage boosts and the fact that the shinespark drains health, resource management is much more crucial in SM than in ZM.
Megaman (1 Star)Mega Man 1
Mega Man 2
Tough choice. MM2 is a much better game on the whole, but MM1 is hilariously broken. I'm leaning towards MM2 here, though.
Zelda (1 Star)Legend of Zelda
A Link to the Past
Links Awakening DX
LoZ 2nd quest is my vote here. I'll grant that I haven't seen the LttP TAS lately, though. Link's Awakening is simply too long.
Castlevania (2 Stars)Castlevania
Castlevania: HoD
Castlevania: CotM
Super Castlevania IV
I think SCIV is an underrated game, but sadly it isn't as obviously challenging to play as the original Castlevania. I don't think CotM makes the cut in either short or long versions; the long version uses a glitch to skip much of the power curve in the game (by making use of magic spells that shouldn't be available) which I don't like much, and both versions have very boring boss battles. All that said, I really don't know which of these games should get starred; the only one I think is pretty secure is the original Castlevania.
Special concept (2 Stars)MMX+MMX2
Kings Bounty
Nightmare on Elmstreet “4player”
Monopoly
River City Ransom
I'd go with Monopoly and RCR here. Reason being that while King's Bounty is very impressive if you know what's going on, it's not if you don't, while Monopoly, as noted earlier, is obviously behaving strangely. And the RCR TAS is just hilarious.
NES games (3 Stars)Ninja Gaiden 2
Darkwing Duck
Mega Man 5
Gimmick!
Super Mario Bros 2
Journey to Silius
Batman
Arkanoid “Warpless”
Battle of Olympus
Legend of Zelda 2
Battletoads “2 player warpless”
A.k.a. the miscellaneous section. I'm not going to explain my choices here because there's too many games in the list, but I'd suggest SMB2 warpless and one of Battletoads 2P warpless, LoZ2, or Battle of Olympus.
SNES games (2 Stars)Super Mario World 2: Yoshi’s Island
Rockman and Forte “100 CDs”
Sparkster
Donkey Kong Country (which run?)
Biker Mice from Mars
The Mask
I actually prefer the R&F Forte run (with the insane lightning zipping) over the 100 CDs run; Forte's a much more dynamic protagonist than Rockman is, and watching Rush dig for CDs really kills the pacing. Sparkster suffers from information overload -- there's so much going on at any given moment that the viewer gets tired. BMfM has camera issues. So I'd suggest DKC1, on the general theory that Rare got it right the first time and each sequel was slightly worse.
Genesis (1 Star)Gunstar Heroes “ 2 player”
Ecco (which run?)
Dynamite headdy
Golden Axe
Comix Zone
Gunstar Heroes 2P should be in the "no question" category, IMO. I don't know about the others. Brawlers tend to make poor TASes and Golden Axe is no exception. I haven't watched Comix Zone. Dynamite Heady was interesting but not that great. And the Ecco runs are neat but a bit incomprehensible.
N64 (1 Star)Majora’s Mask
Goldeneye
Mischief Makers
Super Mario 64 120 star
Hm. "MM" is an ambiguous acronym here. :) Majora's Mask is frankly too long, and while Mischief Makers was neat, in my opinion we have better platforming TASes available. SM64 120-star is also very long. And that leaves Goldeneye, which has rendering bugs. :\ I do think Goldeneye is the best of the lot here, especially considering that SM64 0-star is already on the auto-list.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Personally I think that Umihara Kawase has much better wire physics than any of the other grappling games we have on the site. The only problem is that most of the other grappling games feature a main character who naturally moves much more quickly, which means that the grappling sees less use. So I'm not certain that "games with grappling hooks" is a good "category" to use here, since there's really only two games that take that element seriously (Umihara Kawase and Bionic Commando) and they're so totally different that comparisons aren't really feasible.
I do think that UK is done well, and personally I think it's worthy of a star, but I've played it quite a bit so I'm biased.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I never played this game...but man, it looks just like the Gamecube version!
And probably a whole bunch of other Star Wars games. Ahh, Lucasarts, what happened to the rest of that universe you guys made?
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I do tend to feel that after the first two Zelda games, the appeal of TASes sharply declines. Majora's Mask may be a fantastic game (it's still my favorite of the 3D Zeldas) but in the end it's also still hours long, which makes it nigh-impossible to keep up what I like to call the "WTF pace" (that is, the rate at which the viewer says "WTF?!") that IMO is one of the requirements for a star.
If we're going to star a Zelda, NES Zelda second quest is the clear winner. I'd prefer first quest since more people have played that and there's more leeway to do things in weird orders, but the first-quest run's currently pretty outdate and the second quest is still pretty awesome. Bomb and health management, luck-manipulation of drops, and in particular rapidly clearing rooms of powerful enemies, are all standard features of the run.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Excitebike: I do feel this run keeps up its pace well. It's not just an accumulation of speed -- it's the insane jumps that the player performs, and particularly the way the biker's vertical position wraps around to zero again!
While I feel that Super Mario Kart and Biker Mice from Mars are both very well-done, the latter has camera issues (much of the time you can't see what's going on) and the former extensively abuses a glitch I don't find very entertaining (driving in small circles to complete laps) -- and when it doesn't use that glitch, most of the laps are the same.
Personally, I think that Excitebike just strikes the right balance between novelty and length, which means it "keeps up the pace" better than the runs you compared it against do.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
For clarification: the game timer doesn't advance when paused, which means that if you've fixated on improving the in-game clock you may do things like pause at the top of a vertical shaft room to remove the high-jump boots. The "no pausing except when necessary" condition basically says that pauses made solely for purposes of improving the game clock (as opposed to ones to enable glitches, like the Gravity Jump, or special abilities, like the special charge beam abilities that are only available when only two beam weapons are equipped) are not in the game. To me this just highlights how silly it is for TASers to worry about the in-game clock in Super Metroid. The only reason people do so is because of the speedrunning community that existed before TASVideos came in existence which only had the in-game clock available. But for TASes, framecount is way more important.
In any event, the discussion between mmbossman and Saturn does highlight how confusing these conditions are for anyone who's not actually involved in Super Metroid speedrunning/TASing. It's like its own little sub-community, and I'm not certain it's big enough to warrant yet another officially-published TAS.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
If you look at the videos earlier in the thread, it's done there too. From what I can tell, there's a glitch that sets your speed higher if you do that.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Well, it's pretty clear that there should be a maximum, to avoid dilution of what a star "means". Obviously there's no point if every movie gets starred (or, less extreme, if every good movie gets starred; then it just becomes a marker of what movies one person thinks are worth watching, and we have other better ways to do that). I don't think we need to set in stone what the maximum is, though. If whoever gets appointed is handing out too many, there's plenty of ways to deal with it.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
I find the SMB series' plots much more entertaining when I imagine that Peach is some Machiavellian genius who's manipulating everything from behind the scenes, and maintains an airheaded princess persona just to throw people off.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
It definitely increases your horizontal speed when you jump. And it seems to last a decently long while. The question is really whether or not the level allows you to jump enough for it to be worthwhile.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
