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
12 ->
--
--
List results:
Search options:
Use \ before commas in usernames
Game Page: Doesn't exist yet

Age of Mythology (pc) (pc) [Any %] [Individual Level]

Decision: Accept

Congratulations to 'UtterNutter'!
Thread title:  
Run Information

Age of Mythology (Any %) (Individual Level)

Verification Files

http://v.speeddemosarchive.com/ageofmythology/

Please refer to the Verification Guidelines before posting. Verifications are due by Nov. 30, 2014.

Please post your opinions about the run and be certain to conclude your post with a verdict (Accept/Reject). This is not a contest where the majority wins - I will judge each verification on its content. Please keep your verification brief unless you have a good reason otherwise.

After 2 weeks I will read all of the verifications and move this thread to the main verification board and post my verdict.
How to run Space Marine: Run fast
Hello guys. I'll be keeping an eye on the thread to answer any questions. There are some things I am aware of between this version and version submitted a while back:

Mission 1- I cannot lure the kraken out in the first wave and have to wait until it decides to haul itself towards the shore. Getting the kraken to move appears to be based on how far out into the fog of war you can see and as of yet I am unable to place a tower in the correct place to get the kraken to move early (without vision cheats)

Mission 2- I am ashamed to say this, but I can't actually do the strategy put forward by Twisted Enemy. It's not for a lack of practice, it's just an incredibly finicky teleport which then has a high failure rate after you manage to teleport across.

Mission 8- In this case I decided to record the strat for this level which is somewhat consistent, rather than Twisted_enemy's route which I failed to replicate in any of my tests.

Mission 32-I do not think it is possible to start with six villagers and two scouts. I did a search around youtube for various let's plays of the game and in no case did I find that you start with six villagers. It is therefore fairly slower than the original submission.
Can we please get any eyes on this?
Edit history:
gammadragon: 2014-12-06 03:18:46 pm
gammadragon: 2014-12-05 05:16:23 pm
gammadragon: 2014-12-05 05:16:03 pm
Might be magic...
Audio and video are good.

I've completed the game once before and have a decent amount of familiarity with it. I'm actually sorry I didn't get around to reviewing this sooner, because it's pretty smooth. The runner is a skilled player and has put effort in to optimise the strategy for each level.

There are some nice tactics used, including:
* Good use of powers (especially teleportation)
* Amanra's leap attack to jump over walls
* Speeding up slow units by pairing them with fast units
* Abusing summoning powers to trigger map events
* Nice glitch on mission 17 to teleport across the ocean

The following missions have small playback issues:
Mission 07: failed attempt at beginning - run starts at 2m 50s
Mission 08: failed attempt at beginning - run starts at 5m 19s
Mission 10: video pauses at around 1m 0s mark, lasting for around 1s (audio is fine)
Mission 13: failed attempt at beginning - run starts at 3m 50s
Mission 23: short audio outage at 9m 6s

I've watched up to mission 18 now... to be continued after some sleep. So far it's a solid accept.

Edit: finished watching all the videos now - I added one more mission note. Voting to accept.

Edit 2: It turned out I was reviewing the "high quality" videos but there are "insane quality" versions as well, which have 1024x768 resolution. My previous comments requesting higher resolution are no longer valid Smiley
How to run Space Marine: Run fast
I can't believe I didn't cut off my failed attempts at the beginning of the videos.... I'm sorry guys.

Will be quite happy to edit the videos when I get back home for Christmas so they don't have the failed attempts.
Might be magic...
No worries, it's easily fixed by trimming the video.

The mission 10 video pausing is more of an issue, but there's tolerance for a few short recording glitches like that. If the whole video is that way it can be hard to watch, but in this case it's barely noticable so should be no problem.
Edit history:
LotBlind: 2014-12-12 03:45:17 am
LotBlind: 2014-12-06 02:39:54 pm
LotBlind: 2014-12-06 12:43:15 pm
DONE (posted below)
Edit history:
Vulajin: 2014-12-06 12:50:44 pm
That's fine. Thanks in advance.
I haven't played this game for many years, but I have completed it, so seeing as it's having trouble finding verifiers, I may as well have a go.

General comments: the StatID isn't listing the scenario number, and probably should. These runs were done on Easy difficulty.

Scenario 1

As I remember, this scenario is mostly based around a number of triggers that have to be triggered in sequence (I can't remember the details, but guess it involves killing specific enemies, or all of a group of enemies). The runner's using the strategy of placing as much firepower as possible near the enemy spawnpoint; I don't think it's likely that there's a faster general strategy for this scenario (any improvements to the strategy would be an improvement in the specifics, such as training a different balance of units/buildings).

I'm a little worried that some time may have been lost in not building the towers soon enough; the towers near the starting area, especially, would have benefitted from having many more villagers building them. I'm more worried by the amount of unspent resources; it would have been possible to train more archers during downtime (which would help kill the krakens faster).
There was enough going on that I couldn't make out the limiting factor on tower building, but it may have made sense to shift villagers from favor onto wood/gold early and back the other way later to get the towers built earlier; the centaurs seem less important.

Incidentally, the reason the runner doesn't use Rain seems to be that he doesn't need anyone at all on food production, and thus it would be pointless. I'm not 100% sure I'm right on this, though.

Scenario 2

This mission is mostly about combat, and the strategies seem well thought-out (e.g. manipulating who the enemy attacks by sending different units at different times). Building a wall to prevent the enemy escaping was pretty clever (the wall only needs to be partially built, as seen here, to block enemy pathing). I'm not sure it's necessary to build the barracks in the enemy base; using a barracks in your own base and starting to train earlier would have the same effect, possibly freeing up the villagers to attack the enemy town center.

Scenario 3

This is a nice illustration of starting on one goal before the preceding goal is completed. The starting army is large enough to destroy the only buildings that actually matter; using the transport to get around the wall on the east is faster than destroying it. I don't think there's much to improve on here.

(I haven't checked scenarios 4 to 9 yet.)

Scenario 10

The only factor determining how quickly this scenario ends is how quickly you can move the heroes from the start to the end, picking up the relics along the way. There are only some minor pauses, mostly to break down walls, which seems to be done quickly (e.g. lots of units help), so I think this is going to be impossible to improve by much. The runner takes a ton of damage, but doesn't actually die, which is all that matters.

Scenario 11

The only determining factor on speed here is how many villagers you have excavating, and how soon you get them to the excavation. Thus, this is pretty much an autoscroller (I assume there's no Town Centre on this mission, because I didn't see one, and the runner would definitely use one if there was one). I wasn't watching in high enough quality to make out everything the runner said during downtime, but I assume it's amusing.

Scenario 12

I'm not sure whether it's possible to complete this scenario without the Guardian, but given how quickly the Guardian finishes it, it's almost certainly faster. Thus, the important thing is to keep the sword moving. The only possible improvement I see is therefore to knock down the gates sooner; it might be possible to train soldiers to do so, or to send the villagers in to help.

Scenario 13

The video seems to contain two separate attempts. I didn't notice at first, watched the first attempt, and thought "this loses someting like 20 seconds"; apparently the runner agreed, because it ends in a reset. The second attempt is much faster, but still possibly not perfect; the important thing is to keep the Osiris piece moving, and it spends around 10 seconds under the runner's control but not moving at the start. Micromanaging the initial battle to get enemies and allies out of its way might save some time here.

Scenario 14

It's a little ridiculous that the game's designers gave Amanra enough HP that this works, but it does. I don't think it'd be possible to do much better.

Scenario 15

Again, this scenario is mostly limited by unit movement. The runner loses around 5 seconds with bad pathing as the Osiris piece is released, and maybe about 2 from bad luck with Meteor. Was all the damage on the wall next to the gate overspill damage? I couldn't tell if some of the units were accidentally attacking it.

Scenario 16 part 1

Pretty clean; I don't see much wasted time here (i.e. pretty much all the time, the maximum possible number of available units are aiming for the next area).

Scenario 16 part 2

Go straight to the Wonder, attack it with all available units. I'd be surprised if it was possible to do this faster.

Scenario 17

It's unclear exactly what happens to Amanra near the start of the mission. Is that some sort of pathing glitch with transports? (It's similar in nature to glitches I've seen in other games, at least.) Apart from that, this is pretty straightforward.

Scenario 18

This is basically the same as scenario 11, except this time, the runner does have a Town Center. So at least you can deploy villagers to speed things up. It'd be still faster if the enemy wall didn't get in the way towards the start, but that is unlikely to have made a huge difference. One other possibility is using the Pharaoh to motivate the working villagers, rather than the Town Center, towards the end; I don't know whether or not this will help, and am not 100% sure it's even possible, but it may be worth looking into.

Scenario 19

Yet another example of the "go directly for the goal" strategy. Any improvement here would involve better micromanagement in the combat, but it was already pretty good.

Scenario 20

And this is yet another example of a "wait for a slow unit to reach the goal" mission. This wastes maybe 3 seconds as the second Osiris piece struggles with its pathing; micromanaging units near the Fortress, or positioning the Fortress a little differently, might help here. Apart from that, it's pretty fast.

Scenario 21

I understand how it saves time to free the compounds holding the villager pigs (you need some to build things). However, the main heroes are standing around for 10 seconds at one point (just before building the Fortress), with the exception of a few soldiers walking to become un-pigged; even though waiting for more soldiers might be necessary, they don't spend all that time arriving, so you could probably have started the attack during that time. This scenario looks like it could be somewhat improveable, as a result. The end is pretty clean, and that use of Bolt is hilarious (despite being completely obvious).

Scenario 22

I find this scenario hard to verify because so much is going on at once, and there's a lot of jumping around of the camera. I didn't notice any obvious mistakes, but there are plenty of large mistakes (such as having units idle for long periods of time) that I wouldn't be able to catch. Sorry about that.

Scenario 23

Weird to see a scenario where the limiting factor is resource-gathering, but here we go. The waiting periods are due to getting enough resources to advance to the final Age, then waiting for the Town Center to research it, in order to get Ragnarok, which is probably the fastest way to take down the base. It would be possible to save some time by sending in the army a little earlier and casting Ragnarok just before the enemies started to attack them; maybe about 5 seconds. Apart from that, this is mostly waiting.

Scenario 24

I'm somewhat dubious about what happens with the economy on this mission. There's quite some time when the Hill Fort isn't deploying units (presumably due to lack of resources), and the Town Center isn't training more resource-gathering units to potentially speed that up (the runner doesn't use up the entire supply of food). Also, is there really not a cheaper/faster building deployment building for siege? Apart from the economic section in the middle, this mission is done well.

Scenario 25

Pretty straightforward. The only thing I'm unclear on is if Amanra was wasting time running in circles, but even if she was, I don't think it mattered.

Scenario 26

The limiting factor here seems to be combat, in which case I can forgive the weird economic decisions because they didn't matter. (It might have been possible to train a few more troops, but they probably wouldn't have got into combat soon enough to matter.) Exploiting the enemy AI with Amanra at the end was pretty clever.

Scenario 27

I can't see how you'd do anything different here, short of training units (which has to be slower). So this is pretty straightforward.

Scenario 28

Presumably, it's only due to playing on Easy difficulty that this actually works. It might be possible to clear it faster via better micromanagement in combat to prevent the heroes dying so early, but I'm not sure.

Scenario 29

The only factor determining completion time here is sending all the dwarves to the taproot ASAP and leaving them there. This is entirely an autoscroller; freeing the dwarves in the north section was very slow, but they don't matter at all, so the completion time here is still basically perfect.

Scenario 30

The forced half-hour wait in the middle means that this scenario is just about the start and end. What trigger starts the timer? If it's building the Town Center, it might be possible to save a second or so there. The ballistas see to a perfect time otherwise.

Scenario 31

I don't quite get what happened here. Clearly a plot trigger was screwed up, but is this method really faster than just using a transport to get there? Or isn't it accessible?

Scenario 32

I'm dubious about the economic route on this mission. It starts with Age advancement, putting all available villagers on that, but waits a few seconds between becoming eligible to advance Age and actually starting. Later on, we discover that the limiting factor is resources; Plenty from an early Age advancement helps, but resource gathering doesn't seem to be optimized either, with the limited number of villagers. Would it not be faster to build more villagers before advancing Age, so that you could collect more resource while it was researching? Or is Plenty so good that it's faster? The Town Center is left idle for some time, too (just before starting to build the Wonder), which possibly implies that resources have been gathered in the wrong proportions. Combat seems good, at least (unless it's possible to get units other than Arkantos to damage the statue, but that seems unlikely).
Edit history:
UtterNutter: 2014-12-10 03:03:24 pm
UtterNutter: 2014-12-10 02:52:09 pm
UtterNutter: 2014-12-10 02:22:15 pm
UtterNutter: 2014-12-10 12:55:58 pm
How to run Space Marine: Run fast
Ais- You are the hero I've been looking for, but not the one I deserve right now. The criticism is all really insightful (and I wish I had it when I was running in the first place)

Some replies:

M1- You are correct. Archers require wood and gold, as do towers. I'm given enough food to get as many villagers as I'd ever need.  Will do some testing using additional archers to kill the krakens. (Training additional Centaurs would save about 5-10s if done optimally)

M11- I got very bored and it's a bunch of bad jokes and comments about it being a Let's Play.

M12- You can finish without the guardian, but the guardian is much faster. Good suggestions

M13- Correct. Probably the segment I dislike the most. It's not very fun to run either.

M15- Meteor has a pretty large blast radius, so I'm assuming it was that (I haven't checked the video)

M17- The transport has nowhere it can unload, so it tries to unload at the closest location. For some reason when you position the transport there, it teleports directly where you click. Unfortunately it's not useful in most other missions.

M18- It's been a long time since I did this level so I'll have a test. (I tested, and 10 villagers cut 100hp every 10s whether empowered or not)

M21- Good points. I don't remember spending long on this segment so that may be the reason for the sloppiness. Definitely a re-record. (Test: I can save definitely record a sub 5 minute segment)

M22- I'm slightly confused on the wording here. Are you saying there are large mistakes?

M24- The only building the norse can train siege units from is the hill fort. Good point about resources. I agree.

M25- This has an interesting feature where the mission won't complete until they have stopped talking. So I need to focus on killing the clan with the longest death speech first.

M28- I don't recall if you get freeze at the start of higher difficulties, but if you then the strat is still possible.

M30- The countdown begins when the TC is built.

M31- I'm limited by how quickly I can rescue the prisoners. I get as far inland as I can with my ships and then split the units according to which ones can take the lightning storm by the temple. Everyone else deals with the appropriate task to get me 18? prisoners.

M32- I admit to not doing my own testing on this level. I used a route that involved having the extra 3 villies at the start. I will probably have to reroute this.

(Bearing in mind this comment, is it possible to put the verification back in the vault for now while I record new segments?)
Quote from UtterNutter:
The criticism is all really insightful (and I wish I had it when I was running in the first place)

I feel for you bro ;_;
Re M22, I'm not saying there are large mistakes. I meant that I find it so hard to follow, that even if there were large mistakes, I wouldn't be able to notice them.

I'm offline a large proportion of the time (and don't think I noticed this game when it was in the verification queue; that said, I might not have volunteered as a private verifier anyway because I haven't played it for years, so am verifying entirely on memory). It may be some time before I can fill in the verifications for the remaining segments, if necessary.

Oh, and for Vulajin, my verdict: most of this run is good but some of the individual levels have large enough scope for improvement that they need to be redone. So I'm going to have to reject this IL set as is, but it'll be acceptable with a few improvements to a few of the segments (it's definitely not a "reject the entire run" situation, and luckily IL is a category where you can just go back and patch out mistakes later).

As for what causes the M17 glitch, there's a similar glitch in Neverwinter Nights (which we never found a reliable setup for, annoyingly). I think it happens when programmers program something as "path to location X, then do action Y", and forget to put in a special case for what happens if the pathing to location X fails.
Edit history:
LotBlind: 2014-12-11 04:00:48 am
Mission 1 seems to be the only one with no high quality encodes amongst the verification files but I don't think I'll need it.

Mis 8: I wonder if the restoration really should have been used that early on in the fight. In case the enemy A.I. tends to target down one unit at a time I suppose it doesn't matter.
Mis 12: Couldn't you have tpd more of your army to the final battle to make it that much faster? You could have sent them to where the tp happens to meet up with the guardian.
Mis 17: Could the glitch be caused by something like transports being able to offload units even when the path to the shore is blocked (by other ships)?
MIS 30: I'm guessing building enough ballistae to 2-shot the guy isn't doable?

My two guesses were "Pillar" and "Palm" and I didn't mean "rosy palm" like you seemed to think so that one's on you.

Couldn't you do something like "stand ground/stop" and re-issue the attack command to make Arkantos' war shout activate a bit more efficiently?

My spell checker says there's mistakes in the run comments
REJECT

EDIT: Oh wait apparently we've stopped judging submissions solely based on that.
ACCEPT (based on execution not strategies used, also didn't have time to watch 32 but apparently it's gonna be redone anyway)

PS did you consider doing Ground Control?
How to run Space Marine: Run fast
Thanks for all the advice guys! I'll record new segments for Mission 13, 21, 24, 30 and 32 over the christmas break (I'll get access to recording software again on the 20th).

Regarding M1, I do not believe that training additional toxotes or centaurs will significantly change the time, given that Krakens ignore 69% of their damage and I can't reorder construction to get more than one extra centaur out before the third kraken arrives. Perhaps better catapult micro would help here.

While it's not guaranteed, I'll put a day or two of grinding into Mission 12 as I believe that can have better execution. Additional troops will do little other than tickle the champion elephants and will most likely get killed.

I think I may be able to get 5-6 seconds off of Mission 24. I'll try for that as well.

I hope you can all come back when I resubmit the new recordings, and when I submit my recordings of the Titans expansion, which I am rerecording 2 segments for, because I have found route improvements.

LotBlind-

Mission 8- The AI attack 2-3 units at a time, but given they only have 50hp and about 10% armour, they get killed very quickly. The restoration is to maximise my damage at the beginning before units begin dropping.
Mission 30- Crap... It is.

Before this is closed- Are there any other segments that you guys believe are in desperate need of rerecording, besides the ones listed at the top of this post?
Nothing from me.
Might be magic...
They look pretty decent to me. Maybe a few seconds can be shaved off here and there, but they are a pretty good benchmark for future runs.
UtterNutter: if you think that you'll be coming back with new IL submissions within a short period of time (say, a couple of weeks), I'll just leave this submission open until you have the new videos ready to go. Otherwise, let me know and I'll close this one and you can resubmit later.
How to run Space Marine: Run fast
I can start practicing/trying for IL recordings on the 20th. I plan on having them all done by the 11th January so I can work on my exams after Christmas. I assume that's the former category?
Yep, I'll leave it open for now. Feel free to ping the submission PM when you have the updated videos.
Finishing off my verification comments:

Scenario 4

The limiting factor here is Age advancement (for Underworld Passage). I can't see a faster way to do the map with this strategy, and I'm guessing that marching across the map "manually" would be slower.

Scenario 5

This is basically all about combat, but I think the combat could be done a bit faster. There's something of a mixup with destroying the gate; the units don't focus their fire very well, and leave the gate to attack the oncoming cavalry (it'd be faster to wait next to the gate until the enemy cavalry arrived, kill them while remaning adjacent to the gate, and then turn attention back to the gate, than it would be to waste time running back and forth). I also think I caught a group of units standing idle for a while during the destruction of the town (but I can't be sure, because they were off-camera for most of that time).

Scenario 6

A scenario that comes in three parts. The start is all economy. The basic problem is that you need to gather a bunch of wood, and you can spend food for villagers and wood for fishing ships, and you can use fishing ships to collect food or villagers to collect food or wood, so it's a question of how much of your resources you spend gathering resources. The strategy used here was executed quite well. I'm far from certain on what the fastest strategy here is. One thing I note is that putting an extra villager on food would almost certainly be faster if villager production was only capped by food, but it's also capped by how fast the Town Center can build (this is the only real reason the Fishing Ships are worth it: because they have a different deployment facility). I take it there isn't another Settlement somewhere to capture? (Even if there is, it's probably slower.)

The second part of the mission seems to have been done as fast as possible. The only theoretical potential improvement is getting more units attacking the gate, and the heroes are so much slower at damaging down the gate than the siege units that this would only save a fraction of a second, if anything.

The last part is combat, most of which is done via god powers. This thus mostly comes down to luck with where Meteor hits, and what I saw on-camera of the luck was pretty good.

Scenario 7

(There are two attempts in the video for this segment.) The start of this scenario is mostly about AI manipulation, as far as I can tell; you have a decently-sized army, but the enemy has a tendency to run away. The second attempt (the one that wasn't reset) does this pretty well. Past that is marching to the enemy base as fast as possible, and I see very few moments where the runner doesn't put all available units on this task (and those all have reasonable justification, such as killing enemy myth units who would probably kill some of the runner's generic soldiers and slow things down).

Scenario 8

(There are two attempts in the video for this segment, too.) I could tell this was going to be Underworld Passage cheese from the start. I have three main concerns. One is about the balance between the east and west villagers; the east villagers are mostly left farming, which is one of the slower methods of resource gathering. Is there really nothing better they could be doing? (You could put all the west villagers on food, and the east villagers on everything else. I'm not sure if there's a gold mine over the east, but it'd likely be worth mining there even if you had to walk some distance from town and build a new dropsite.) My other concern is about the unit mix you use to take down the forts. Are villagers the most cost-effective anti-building units you can recruit? You have time to build other units while waiting for Age advancement (e.g. you could put the east villagers on wood and build pretty much any deployment building you wanted over the west side). This might or might not be faster, but I think it's at least worth mentioning. (Also, the Passage itself survives some time after you send units through it; you can continue building and sending units through while you're capturing the forts). Finally, some of the archers seem to have got somewhat distracted while taking the Forts, although it shouldn't make a large difference in the grand scheme of things. Even though killing off enemy units makes sense to start with, they should switch to attacking the Fort itself once it reaches low HP.

Scenario 9

It's always amusing to see a speedrun of a mission whose gimmick is a time limit. The other gimmick here is that deployment is done for you, which mostly just makes routing much easier. The only potential improvement I see here is better micromanagement during the battles; AoM has a rock-paper-scissors thing going on where heroes are more time-efficient at killing myth units, myth units at killing regular humans, regular humans at killing heroes. Some of the fights you were going the wrong way round the triangle, meaning they're likely to have taken longer than normal. This is unlikely to have wasted more than a few seconds, though. (You also have to take into account how time-effective various units of yours are at destroying buildings like the ram, compared to in combat, to choose which units to take off ram-destroying duty for combat. I have no idea whether this was done correctly or not.)
Edit history:
UtterNutter: 2014-12-18 02:54:16 pm
How to run Space Marine: Run fast
Discussion:

M5- It's a good idea. I'll try and lure them in while continuing to attack the gate. I really don't want to lose any units though, as I need them to kill the TC and fortress.

M6- The only other TCs are inside Troy.

M8- Villagers are the most consistent, but this probably isn't the best IL strat I could use. I'll do some messing around with this segment to see if there is a better way of doing it. (I tried an all-in with the right army and a minotaur which got a PB of 4:18. I'll try and record a sub 4:25 as proof)

M9- There is a good reason for me generally using the centaurs to kill incoming soldiers. The ram has 95% resistance to piercing attacks (arrows) which means their attacks are only doing 2-3 damage a shot. If I want to divert damage away from the ram, I want to use the soldiers that are doing the least damage, in this case the centaurs and chiron.
How to run Space Marine: Run fast
Good news. I've cut 2 minutes off the run through endless repetition and level changes.

Mission 8 is now 3:54 thanks to (I can't believe I'm typing this) YOBO Minotaur strats (You only build one)

Mission 21 is now 4:49 because I didn't screw up this time.

Mission 24 is 7:27 (I think). I'd like to submit this (although improvement is possible) as it has a very good execution of the first 6 minutes and 30 seconds, with the RNG god power at the end causing me to lose about ten seconds. I have yet to get an attempt anywhere near this good since.

Mission 32 is 10:18 (I think). I changed the opening to: Build two villies-> Age up -> Plenty -> None-stop villagers while sending two armies out to secure plenty vaults around the map. The god power grants you 15 of each resource every 5 in game seconds. The two I capture bring this up to 20 rp5 (resources per 5s)

So if it's possible, I'd like a submission link to update my videos? Or should I use the old link and name them as NEW_AoM_8
You're not redoing 30? Or you did already?
How to run Space Marine: Run fast
I've now redone 30. Unfortunately it seems like there is a significant lag period if you kill Gargarensis in a single hit, meaning there is only a single second difference between the runs despite me doing a lot better in the latest recording.