Username:
B
I
U
S
"
url
img
#
code
sup
sub
font
size
color
smiley
embarassed
thumbsup
happy
Huh?
Angry
Roll Eyes
Undecided
Lips Sealed
Kiss
Cry
Grin
Wink
Tongue
Shocked
Cheesy
Smiley
Sad
1 page
--
--
List results:
Search options:
Use \ before commas in usernames
Edit history:
Flip: 2011-08-21 09:27:05 pm
1-Up!
Game Page: Doesn't exist yet

IL Table

Verifier Responses

Quote:
General Comments and Summary:
Audio quality fine. Video acceptable, given that it isn't too easy to get a direct capture of classic Mac stuff (I'm guessing he used S-video or composite or something? But I don't know where the 576x432 resolution comes from; maybe he cropped it down?). No cheating apparent.
The Ares missions can be loosely divided into 'transport levels' and 'non-transport levels', where the former are essentially all beaten in minimal time by building a transport as soon as possible, having the game (randomly) launch it directly towards the target base, and then have it travel and land without being interrupted by enemies.
The runner seems to do all of the transport levels optimally if not near-optimally. The non-transport levels are also all handled very well without any obvious easy improvements.
I tried running some of the levels myself this weekend, and only matched six of the times, was 1 second slower for one, and was 2 seconds slower for three. And I probably would prefer not to talk about how well I did on the others this weekend.
Overall, very slick and impressive runs for all levels. If he did these all during one weekend, or even within one week, then he is absolutely crazy. I can only imagine that during these two years waiting for verifiers, he has been incessantly improving his runs.

A solid accept.

Also, another general note is that the runner seems to not use any of the fixed zoom modes (1:1, 1:2, 1:4, 1:16) in many situations when they would seem suitable (for example, when checking the departure directions of transports at the beginning of some levels, and in some combat situations); I'm not sure what the deal was with that.



Individual Level Comments:
I wrote a set of comments for all the levels before reading any of the runner's comments, and then if I had any extra comments after reading his, I appended them at the end.

Level 1:
Looks pretty optimal; he appears to initiate the next warp as soon as he exits the previous one for the first four kills, and only missed a few shots on the final transport.
I couldn't match his time after 20 completions of the level; I got 1 second slower (33s) since in that attempt, I warped slightly past the last transport so I had to turn around.
(after reading runner's comments: Same as what I said, apparently. Wow, he tried thousands of times?!)

Level 2:
Pretty optimal.
Lucky bumping into the gunship upon arrival at the third radar dish. I suspect he wouldn't have been as ideally aimed at the radar dish if it weren't for the bump.
I matched his time on this level, after 20 completions.
(after reading runner's comments: Same as what I said ...)

Level 3:
Transport level, looks optimal; initial launch direction looks pretty much direct, and there were no obvious interruptions of the transport when it landed.
I matched his time after 3 completions (1m30s, 1m30s, 1m28s).
(after reading runner's comments: Oh, interesting; I didn't know using manual control of the transport for landing actually could improve the time. I thought that was for making sure it didn't do any evasive maneuvers near the end.)

Level 4:
Pretty slick. He commandeers the transport when arriving/departing from prison cells and the drop-off point in order to minimize the time spent turning around (the AI doesn't do such a good job).
Almost dying was interesting.
My best time was 5m44s after three tries, but I didn't think of controlling the transport.
(after reading runner's comments: Oh, so that's why he waited to shoot open the prison cells.)

Level 5:
Strictly, this is a transport level, but I think the fact that there's a flak drone at the enemy planet allows some suboptimality.
The runner builds the transport about 3 seconds into the level (he sends off the engineer pods first), but looking at the endgame, it seems that the transport pretty much enters the firing radius of the second flak drone just after the runner gains ownership of that flak drone, so realistically there probably isn't any time lost for the transport here.
A small amount of time is lost by the transport not moving straight at the very end.
I tried this level during the weekend until I realized there was a flak drone at their base, then I gave up.

Level 6:
Transport level, looks optimal. He gets lucky at the end, since an enemy ship is built and destroyed by his cruisers just before his transport arrives; in my attempts, my transport always got bumped by spawning enemy ships while it was landing.
My best time after 8 completions was 2 seconds slower (2m07s, 2m09s, 2m06s, 2m09s, 2m06s, 2m09s, 2m06, 2m07s).

Level 7:
A very slick run of a quite random level.
At the end, it looks like he only collects one of the three asteroid samples from the final medium asteroid he breaks apart; only after watching and re-watching this several times didi I realize that what looks like a fragment going off to the left is actually a rocket that he just shot, and he actually collects two samples on top of each other.

Level 8:
Fast run of a random level. Nice luck on the last two medium asteroids (which each launched two samples almost directly towards a ship that just shot it apart).
The best I got was 1m09s after six completions.

Level 9:
Transport level, looks optimal. In this level, I found it was much harder whenever the enemy chose to build a defense drone.
My best time was 2 seconds slower in four completions (4m45s, 4m45s, 4m42s, 4m43s).

Level 10:
Looks optimal. The target carrier in his run does the short warp jump before he even arrives, which means he can start the final attack phase sooner; sometimes, only after you have reached the carrier does it warp forward, at which point you may have to warp forward a second time.
I matched his time after 22 completions. (In my 16-second completion, I passed the carrier in warp mode before it started its warp, but I got a few shots in on it as I passed, and then it exited warp right in front of my new position.)

Level 11:
Pretty fast. Could have been faster by perhaps two or three seconds if he initiated warp while turning around after his second, third, and fourth returns to the UNS Ares to drop off people.
Good optimization by staying in warp mode for the second pass of the second prison cell (and this is probably necessary given that he would be too low on electric power to start the last warp, I think).
I got 2m42s in one completion and then gave up.

Level 12:
Transport level, reasonably optimal. The transport departure looks slightly over 45 degrees off-axis. Other than that, no other obvious interruptions or imperfections in the transport's journey.
I don't have a completion time written down for this level, so I must have given up for some reason.
(after reading runner's comments: I was wondering if he intended to die to the flak drone. I didn't recall that the energy balls counted as enemies. Nice!)

Level 13:
Looks reasonably fast. I think the level time is essentially fixed once you collect the last escape pod, assuming the tractor tug isn't interrupted (I think the runner's tractor tug gets shot a few times once it enters via the jump gate, but I don't see any performing of evasive maneuvers which cost time).
I got what I thought was a time which should have been competitive (collecting two escape pods in a single warp twice, if I remember correctly), but its time was 3m15s.
As it turns out, the initial velocity of the disabled liner is random (I started the level twice just now, and once it was moving away from the gate and the other time it was moving towards it), so maybe that's why my only completion was so much slower.
Though the runner only collects one escape pod per warp at the start, the fact that you have to reroll for the initial velocity of the liner to get a good time probably excuses this.
(after reading runner's comments: Okay, so what I said, but he didn't mention multiple collections per warp. Perhaps it is easier if you use 1:16 fixed zoom, though still really hard.)

Level 14:
Transport level, looks near-optimal. Transport departure is about 30 degrees off-axis, I guess? Landing is fine; the Cantharan fighter shoots at the transport, but it doesn't evade.
The runner chooses to engage enemies while the transport is in transit, which works quite well. (In my attempt, I made all my built ships go to some rally point near their base, and then before the transport arrived, I sent all the ships in to clear their base of any enemies there at the time. I completed the mission with three warping Cantharan carriers en route back from my base.)
I matched his time in two completions.

Level 15:
Transport level, near-optimal. Very slight off-axis departure. He gets bumped once at the end of the landing, but I don't know if this costs any time.
In this level, often three carriers spawn when I pass the middle of the level, or something like that. I'm not sure what triggers the spawning, but I restarted the mission whenever they spawned.
I got 2 seconds slower after 3 completions (1m43s, 1m39s, 1m46s).
(after reading runner's comments: Oh, so that's what triggers the carriers to spawn.)

Level 16:
Pretty fast; this level is insane. The combat at the end of the level is possibly the most variable, time-wise.
Normally, I aim to exit warp with the massive device still in front of me, so I can get a large initial burst of shots out on it before the three Audemedon ships swoop back in to start firing.
The runner manages reasonably good orbiting-and-shooting of the massive device (I normally focus on staying out of the gunship's line of fire). The carrier starts far away from the device, making the first few seconds of shooting easier.
Sometimes, the game will end the mission saying all your ships were eliminated, when in fact all you lost was the Eljeetan crusier. I found that this happened if I completed the killing of the massive device quickly, and then died right after; the message popped up after I changed control to a Heavy Destroyer and then was clicking a lot and pressing Order to Go. Only later when I happened to not accidentally dismiss the dialog box did I read what that message was.
I think the runner waits for automatic transfer of control to the heavy destroyer in the fleet of reinforcements (at least that's what I did), and only then holds down Target for a moment and releases it to target his new ship.
Tip for the runner: you can hold down your Target key before control is transferred, and release it once you gain control of your new ship to target your new ship immediately (the blue "Heavy Destroyer" text can appear on his new ship essentially right after control is transferred, or at least as soon as he can react to it).
Even if you transfer control manually (using the keyboard command), you can still use this trick.
The combat at the end happens to be much faster than anything I managed.
My best time for this was 9 seconds slower (1m39s) in 24 attempts.
(after reading runner's comments: Oh. I didn't know that you needed the research vessel to die to gain control of the first two heavy destroyers.)

Level 17:
Transport level, reasonably optimal. Maybe 45-degree off-axis departure. Clean landing.
I matched his time after 9 completions. In my attempts, I followed the transport to make sure it didn't get hit by asteroids. I don't think you can make the transport exceed its maximum speed by bumping into it with asteroids.

Level 18:
Sort-of transport level; seems reasonably optimal. Presumably the first two transports landed without much interruption. Maybe 30 degrees off-axis departure (of the built transport). Clean final landing.
I matched his time after 9 completions. My attempt had about a 110-degree off-axis launch of the transport I built, so probably one or even two seconds can be saved with extreme luck.

Level 19:
It's not clear what the best strategy for this level is, but the runner's method seems to work well; he builds a transport to take the main enemy base immediately, and then takes out the remaining normal enemies while he captures the remaining two bases. When the majority of normal enemy ships are gone, he sends everyone after the gate ship.
When attacking the gate ship, he seems to switch to building cruisers. I suspect the runner chose cruisers on the basis that he can build 11/5 cruisers per gunship, and he thought that he would benefit from more decoys for the gate ship to target with its jammer and seeking shots.
However, I noted that the AI-controlled cruisers never fired at the gate ship. And although the AI-controlled gunships also didn't seem to use primary fire on the gate ship, at least they seemed to always use their laser turret when they were within range.
Finally, when the player is jammed, the player can immediately just transfer control to a different gunship adjacent to the gate ship to resume firing, so there is little advantage to having many decoys. I think he built ten cruisers in the level.
So, my recommended improvement to the runner for this mission is to use gunships only, and to switch control to a different gunship when being jammed by the gate ship.
Other than this, the level is handled extremely well.
(I didn't try to complete this mission during the weekend.)
(after reading runner's comments: I still think my gunships-only idea stands.)

Level 20:
This level was very well done. I have problems doing the initial asteroid-launching bit (the gate ship moves as well), so it's nice to see it done reasonably quickly with the first two.
The fact that the orientation of the level and the initial position of the player and some other ships are slightly randomized makes it hard to determine where to wait for the asteroids to start appearing.
The final combat against the carriers and cruisers (and some fighters) is slick with few re-passes (one, or even zero?) needed when getting carriers to commit suicide by crashing into the asteroids.
The only improvement I can suggest to the runner is that since carriers only launch fighters when you get close to them (or they're low on shields, I think?), if he successfully kept his distance during the combat with the two carriers, he could have probably prevented two if not all three of the fighter launches.
Also, the pulse gun provided in that mission (and some previous ones) significantly knocks back your enemy, so you could have possibly kept attacking the first of the two carriers there instead of shooting at the first one and then switching targets, while keeping it a safe distance away from you.
These three possibly extra Cantharan fighters are certainly something to think about for any planned improvements to this level, but is certainly not a run-breaker.
(I only completed this mission once in the weekend with a time of 6m31s after a long asteroid juggling session.)
(after reading runner's comments: So, what I said, except without mentioning the induced launching of fighters.)


Quote:
Wow! He makes these ships look quite easy to pilot, dropping out of warp within firing range of his targets most of the time. That's a lot of precision maneuvering... maybe one or two things he could improve on, but it would take me quite a while to beat any of those times. It surprised me that he didn't use any of the fixed zoom modes, but he definitely handled the dynamic zoom modes expertly.

The other verifier did a very thorough job examining the details of the runthroughs, and I pretty much agree with him and have nothing to add. No cheating apparent, a stellar performance! GG


My last verifier never responded Sad

Decision: Accept

Congratulations to Nathaniel "NMS" Stalberg!
Thread title:  
Balls jerky
Another success story from the community verification topic!
Even obscure Mac-only games that are almost called 'Arse' can get verified here!

Good job whoever verified this.
1-Up!
Quote from dballin:
Another success story from the community verification topic!

Keep it up!!
This was on the waiting list for almost two years, wasn't it? Nice to see it finally get accepted.
Thanks a lot for verifying! It's been a long wait, but I'm looking forward to seeing it on the site. Hopefully Invisible War won't take quite so long to get verifiers.

Quote from ExplodingCabbage:
Good job whoever verified this.


Yes, it's almost like the first verifier was that guy who always goes over runs in insane, nitpicky detail... Well, whoever it was is certainly thorough and good at the game. Matching or coming close to so many of my times in relatively few attempts is very impressive. Although a few transport missions were fairly trivial once I'd figured out how to optimize the takeoff and landing, I spent several hours at least on each of the other levels.

The video quality and the odd size are indeed due to recording via S-video and a DV bridge which unfortunately downscaled it a bit. Fortunately MPEG Streamclip did a nice job deinterlacing it with minimal artifacts. I haven't done any more speedrunning of Ares since, due to the complexity of the setup and problems with the computer I used. (I fixed it, but it took a while. Taking the hard disk out of an eMac involves basically disassembling the entire machine.) The next version of the OS X port, Antares, will have significant improvements in the efficiency of the graphics code and maybe the option to play in higher resolutions, so I might do some more when that's ready.

I find that closest hostile and all objects are the most useful zoom modes, but I do use fixed zoom in a few places. I'm not used to using the hotkeys for fixed zoom modes and it takes some time to cycle to the one I would want with the zoom in key. Also, some of my runs were done before I realized that a keyboard hardware limitation makes many of my keys not work while I'm holding command and space together, which is particularly relevant with a heavy destroyer.


About some of the level comments:

5: It should be possible to get the minimum time here by distracting the flak drones between when the transport enters their range and when I gain control of them, but it would require the enemy to be incredibly incompetent. I was never quite that lucky.

10: I got 16 s several times. Although some seemed faster than others, I don't think it's possible to get a better time due to trigger timing.

11: Due to triggers only being checked every so often, it's unpredictable how long it takes to drop off the prisoners. And the warp sound doesn't start playing until a little after the key is pressed. So in most cases I was actually taking a risk of hypering away before dropping off all the prisoners. While the execution could theoretically be better, it's very hard.

13: Picking up the pods without dropping out of hyper is extremely unreliable. Not only is it hard to aim at that speed, but you can fail to collide with them even if they're directly in your path. Since the starting direction of each pod is also random, it's sometimes possible for two to be right next to each other when you get to them, but it didn't happen in this attempt, which was the best of many.

16: I did transfer control manually, actually. I was a bit lucky that my escape pod didn't expire before the reinforcements arrived. I probably should have thought of holding target before transferring, but it didn't really matter here because the main factor is how well the battle goes. Rushing in as fast as possible is a good way to waste your heavy destroyers. This attempt was so good because my forces stayed concentrated and the enemies arrived gradually. They were mostly distracted by the carriers and fighters, which let me pound their carriers as they showed up and kept my heavy destroyers alive to help out with their smaller ships.

19: I love human gunships. No other ship has the combination of toughness, expendability, short range firepower, and poor maneuverability to make slugging it out nose-to-nose with a carrier seem like a good idea. Whereas human cruisers are not really on par with the tougher enemy fighters. So believe me when I say that I was reluctant to consider building cruisers here. The problem with the gunship only strategy is that you just don't have the time or resources to build enough of them by the time I want to be engaging the gate ship. Without a certain critical mass of ships, the gate ship will just jam each ship and hit it with a high-impact pulse as soon as it gets in range, making it impossible to engage effectively. While it's true that the cruisers are almost totally ineffectual, the AI gunships aren't that much better against the gate ship. Because of the way the AI decides when to start and stop firing, it rarely fires short-range weapons at anything close to the maximum rate. So the one human-controlled gunship actually does the majority of the total damage. Therefore it's critical to have enough ships around the gate ship that I'm rarely the closest one. Transferring control can help, but it's difficult to always be ready to transfer to a gunship that's close but not about to get hit itself. While this attempt is obviously suboptimal in places, it's very tough to do better.

20: I was able to consistently get close to where the asteroids appear by traveling away from Earth for a certain amount of time, then in the perpendicular direction for a certain amount of time. This was reliable enough that a few attempts failed due to the asteroid appearing on top of me. Aiming the first two asteroids at the gate ship went very nicely and I was astonishingly lucky at crashing the carriers into the rest. I'm not exactly sure why I switched targets. Maybe I thought that circling around the closer one would bring me too close to the farther one. It is unfortunate that I caused them to launch fighters, but I'm pretty sure it still saved a bit of time compared to waiting for one more asteroid, especially since I found and killed the stealthed cruisers while killing the fighters.
Now a hit show on the CW
Just curious...does this game hold the distinction of taking the longest to verify so far?
I was verifier 1.

I should say that although I saw this run on the verification list since the end of 2009, I only offered to verify for it this June, seeing that it had been there for so long. Sorry for the hold-up!

Also, when I started writing the comments, I did not expect them to be so big, but I guess that could be said about my Raven Shield comments as well...

Congrats on the great run!
Oh. Well, thanks for verifying. Assuming you used Antares, you wouldn't have been able to verify it when it was submitted anyway. I was hoping some people from the Ambrosia forums would sign up, but I'm not sure if they did.

And I really liked your Raven Shield run.