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
page  <- 1234567891011 -> <- 1 .. 5 .. 11 ->
--
--
List results:
Search options:
Use \ before commas in usernames
Edit history:
Gombill: 2010-06-29 06:21:25 pm
Here's a new run of Joan of Arc 4 in 12:31.

I maybe see this going about 25 seconds faster maximum.  I think the rams can get through the wall a little quicker, and it might be possible with luck for 2 or 3 of the rams to survive until the last town center is destroyed. 

http://www.filefront.com/16909755/JOA4_1231.mp4

Also here is Barbarossa 6 in 8:48.  It looks a bit sloppy, but the first few minutes don't seem to matter as far as the time (as long as the paladins are off quickly).  I've attempted new runs with getting blast furnace much quicker so the paladins have it about half the time against the wall, but none of those attempts beat this run.

http://www.filefront.com/16909939/BARB6-848.mp4

EDIT:  The tower strategy from Saladin 2 to lure villagers out has been working excellently on some other missions as well.  In particular, Joan 3 and Saladin 5 should be much easier using this.  I don't have a Joan 3 time yet, but I've finished Saladin 5 in 14:44 by destroying the two land-based enemies.  Blue is easy using the tower thing, they take about 8 minutes to kill off.  Orange may seem menacing at first, but their army can easily be lured away.  The strategy I'm using now is to sneak a few monks into the orange town while the army is away and convert one of the three villagers (they resign if they lose them all).  Then I have that villager build a stable and get a knight to kill the other two.  It would be best to just convert all three of them to avoid having to build the stable, but I can usually count on the villagers scrambling away from me once I start converting. 

I think Saladin 5 can be brought down to about 11 minutes if the stable is necessary, otherwise somewhere between 9 and 10 minutes.  I'll post a video once I'm satisfied with it.
Joke of all trades
looking good, keep up the cool work
I've been working on Saladin 5 some more, right now I have an 11:20 which had an unfortunate mistake.  I managed to kill orange in pretty short order, they resigned at 10:50, but this was the only run I've ever had in which I couldn't get blue to resign in under 10 minutes.  So this cost an additional 30 seconds on the end.  In any case, I found a few things to speed it up even more, and I hope to hit about 10 minutes flat.  If I get a time near here I'll post up the video.  The biggest problem is just distracting orange's towers so they don't target my monks on their way in.

I also started working on Saladin 4.  It didn't turn out to be as hard as I initially thought it would be.  The goal is to destroy the 5 towers in blue's town.  Two of the guard towers can be destroyed with a single well-placed castle at the start of the game.  However, two of the towers are bombard towers and need trebuchets to destroy.  I don't see the mission getting beat any faster than about 15 minutes for this reason.  In a quick practice run I got 19:36, but I think with some practice it shouldn't be too hard to hit 17 minutes.

I also spent some time finalizing Saladin 3.  I had posted a video of a 7:16 before, and I've brought it all the way down to 6:04.  Watching the 7:16, I can't believe there was over a minute of improvement possible.  The 6:04 should be within a few seconds of perfect.  Here's the video:

http://www.filefront.com/16934603/SAL3_604.mp4
I've taken a short break from the levels I've been brooding over and did some quick and easy conquerors levels.  Hamorad had already done two of them, but I found improvements for them both.

The one new video is of Montezuma 4 in 4:35.  It tells you the objectives are to destroy the spanish wonder and free your allies in the Tlaxcala camps, however destroying the wonder is the only thing you really need to do.  Your soldiers for recruitment nearby are mostly very weak, but you start with a super-strength jaguar warrior that does 32 damage per hit, so the object is to get him to the wonder as quickly as possible.  The base of the wonder is mostly guarded by castles, but there is one sweet spot where the castles can't reach, big enough for one soldier (namely, the jaguar warrior) to hit.  He takes it down in a mere 10 to 12 hits and the game ends.  The only possible improvement would be to use the lone recruited eagle warrior north of your starting position instead of going south first, but I've found it way too difficult using just the one eagle warrior for distractions to let the jaguar warrior reach the wonder.  The biggest trouble is the two yellow towers, which are strong and have ballistics. 

Next is an improvement of Hamorad's Agincourt time by about 25 seconds to 3:09.  Most of the improvement comes from finding out that about 1/4 of the time (and entirely by luck) the green player sends a scout through his south gate just as henry is arriving there, so you can take the fastest possible shortcut through green's town.  The rest is pretty self explanatory.  It is noteworthy that the victory condition is Henry landing anywhere on 'Britain', not getting through the walled area there, so I land him on the closest possible point.  This saves a few more seconds.

Finally, I improved Hamorad's Tours time from 6:10 to 5:31.  I think another 13 seconds or so can be saved getting the trade carts out of the gate - looking at it now, I can't figure out how that one trade cart got stuck so far behind the rest of them.  It's harder than it looks to keep the saracens from destroying a trade cart in the process, though, and took me many tries just to get the 5:31.  I destroy my own buildings in the city to give the trade carts a straighter path, and my own farms so that enemies aren't encouraged to camp near the trade route's path to the city.

http://www.filefront.com/16945725/MONT4_435.mp4

http://www.filefront.com/16945745/AGIN_309.mp4

http://www.filefront.com/16945751/TOURS_531.mp4

Over the next 2 weeks I'll be away from the speedruns so don't expect to see much more of me for a while.  I will be back at them afterward, though!  No giving up with only 7 levels to go in Age of Kings.
we have lift off
Lots of progress which is nice to see; this will be great when it's done. You should update your table or ask to get control of the first post and do it there. How many of the levels do you need to redo when the last 7 levels are done or are they already done?
Edit history:
Gombill: 2010-08-14 10:06:37 pm
Gombill: 2010-08-14 10:05:42 pm
Gombill: 2010-08-11 09:43:46 pm
Updated Table
Best Time ; Runner ; (My goal)

Total Time So Far:  5:45:33        Total of Goals:  5:44:46

WW
1:  2:05  Gombill
2:  3:07  Gombill
3:  1:33  Gombill
4:  3:38  Gombill
5:  5:14  suga  (on standard)  5:44  (on hard)
6:  3:43  Gombill
7:  4:46  suga  (on standard)  5:33  (on hard)
JOA
1:  5:01  Gombill
2:  5:25  suga  (on standard)  8:20  (on hard)
3:  26:54  Gombill    (24:00)
4:  12:31 Gombill (12:10)
5:             (15:00)
6:  7:21  Gombill (6:00)
Saladin
1:  7:48  Gombill (7:20)
2:  22:37 Gombill (20:00)
3:  6:04  Gombill
4:  9:11  Gombill  (8:30)
5:  11:20 Gombill (10:20)
6:  34:03  Gombill  (33:30)              
Genghis Khan
1:  11:30 Gombill
2:  2:57  Gombill
3:  49:53 Gombill  (30:00)
4:  21:44 Gombill (18:00)
5:  24:26 Gombill (22:00)
6:  21:12 Gombill  (18:00)
Barbarossa
1:               (10:00)
2:               (18:00)
3:  4:29  Gombill (4:10)
4:  13:28 Gombill
5:  10:43 Gombill (10:30)
6:  8:40  Gombill (8:30)

To answer Ridd3r's question, I'm not sure when I'll end up submitting this - as you can see from the table above, I have plans of redoing almost every level I've already done to different extents.  It's just a matter of how close to perfect I want them to be when I submit them.  I'll probably just play until I get sick of these levels and then send in whatever I happen to have at the time, perfect or not.  As I said when I first started posting here, the primary goal is to just get something decent for all levels, so that it can get submitted and others can pick apart individual levels from there.

It's been interesting to me how little actual RTS skill was involved in most of these.  It almost feels more like a puzzle game, figuring out how all the pieces should fit together to get the best possible time.  There aren't many 'kill everything' missions in Age of Kings, but Conquerors should have much more base-building.
Quick update:

Finished Saladin 4 without the imperial age, it turned out to be a lot easier than I thought.  Here is a video of Saladin 4 in 9:11.  It was merely a recorded practice run with many things left to be optimized, but it turned out great anyways and I think only 30 to 45 seconds more can be taken off of it.  It should represent the basic overall strategy for the level.  Blue won't start training his army until about the 6 minute mark, so that's almost enough time to get the job done by itself.  I made a bad mistake with my rams at about 5:40, I thought I was controlling my troops at that moment for some reason.

http://www.filefront.com/16965753/SAL4_911.mp4
Back to speedrunning...  here's the updates from today.

Genghis Khan 2  -  2:57 <  3:02

Not much to say here, just tried it a bunch more times and got lucky once with some of the timing.

Genghis Khan 5  -  24:29  <  27:18

Can still be done a few minutes faster I think, but it's an improvement.

Genghis Khan 6  -  30:07  (NEW)

This was only my second time trying the level.  The current strategy is nothing but trebs and mangudai.  The 'saboteurs' you are given at the start of the level are ridiculously strong, have a huge blast radius, and (best of all) the enemy ignores them with all units.  Due to this the first 15 minutes are spent building an army of trebs and upgraded mangudai and defending with nothing but well-placed saboteurs, then I wipe out the town pretty quickly.  I will have to toy around with which upgrades save time.  25 minutes should not be hard, but I feel like 20 will be a difficult time to get to. 


I also began playing Barbarossa 1 and getting a strategy together.  I believe I have a pretty good strategy but executing it will be painful.  It will require attention in 4 different enemy bases at the same time, 2 of them with heavy micro involved and no room for error whatsoever.  The longest of the four relics should take about 10 minutes to capture, so I will shoot for somewhere around there.  In a practice run I got 3 of the 4 relics in about 11 minutes, the 4th didn't make it because the monk carrying it got converted  Angry 


No videos in this update.  I'll wait until I have some good stuff to put up.
we have lift off
Great to see more progress, I look forward to some more videos.
Am looking forward to seeing more segments. By the way, you should find you're able to upload attachments on the forum again now if you prefer that to filefront (nate increased the upload folder capacity).
Edit history:
Gombill: 2010-10-08 12:22:03 am
I don't have anything fabulous on the new levels yet, but I'll put up videos of what I'm working on and critiques are welcome.

First is an improvement of Genghis Khan 6 to 27:31, but I still believe sub-25 will be possible.  A minute of this is just awkward trebuchet placement at the end.  I kept alive a lot of my original units for this one, which is crucial so the castles can be creating trebuchets instead of fighting units.  It is, of course, nice to keep as many saboteurs alive as possible to reduce building destruction time, but I think it is more important to keep fighting units alive and use a few saboteurs early.  The file I created was just barely over 50 MB, so I'll have to put it on Filefront:

http://www.filefront.com/17138159/GK6_2731.mp4

The second (attached here) is another attempt at Barbarossa 1.  This attempt only picks up 1 relic of the 4.  The northeast one should have been captured as well, I wasn't thinking and didn't get the decoy villagers in place in time.  The west relic is mostly up to luck, since I don't have enough attention to do anything more than put my units in place in a defensive stance and hope the relic has dropped a few minutes later when the monk arrives.  The south relic is usually not bad for me but I got distracted at the wrong time in this attempt and my units all got away from me.  My idea is to attack with 3 teutonic knights, one of them will get converted, but I convert it back and destroy the monastery.  It helps if some of the knights from the west relic survived for the south one, but none did in this try. 

EDIT:  The video attempt at Barbarossa 1 has been removed.
Edit history:
gia: 2010-07-23 12:03:36 am
I love YaBB 1G - SP1!
it didnt seem to me like the mill was a good investment, you should build a market, there are markets in aok right?, also, fuchsia is weaker, better suited for a rush, although maybe its so you can go to yellow from the back, and I only know from normal difficulty so there could be differences

use more ctrl groups, at the yellow base you had the horse but the monk was the one out of sight, you could have controlled him with the numbers


hmm on GK, from a similar game I know that when hunting you want groups of three per deer, four if they are like right next to the drop point, unless its different in aok. You had groups of 2 there, and others with a lot of villagers. Then why build the town center there when it could be next to the gold plus lumber and be south of the deer, that way you dont spend on the drop points.

Make the castle closer so your attack can have more backup, make the market and castle/towers(?) asap, it looks to me that way you could trade the petard's loss for your castle's health a bit. I'm not sure and dont have the game installed to tell for sure but I think you should start right away with the town center plus a market, "bloom" then castle. If I am not correct then you bloom first then market and castle, but you leave upgrades last.
Quote from gia:
there are markets in aok right?


Yes, but the upgrades available at them are different to in AoE 3. The only upgrades available at AoE 2 markets are to reduce the penalties for tribute and market trading, and to share LoS with allies; upgrades to improve your resource gathering rate are all located at the relevant drop point buildings.
Edit history:
gia: 2010-07-23 11:24:55 am
I love YaBB 1G - SP1!
eh? no who cares about that I mean you put a market asap and trade resources to what you need :P, on barbarrosa you start with stone and dont use it, you can place a market and do a resource swap from stone to gold and end up "earning" resources despite having spent wood on the market, if you do need the wood you can even buy it back and you still end with extra gold

Gombill I know you dont intend to get the best times ever but you have already "warmed up" with these runs and at the cost of a speedrunning day you can learn a lot to get the best times you can on the remaining levels. And it will not cost you extra stress while playing, actually it will reduce your stress since you will be following certain rules automatically instead of coming up with something at the moment. Im not sure about aok but it should have these:

1. I assume you record in that resolution because a bigger one doesnt give good fps with fraps.
2. control groups, abuse them
3. shortcuts, the ones that focus your view on buildings, the ones that produce units, the ones that change your units' formation or fighting stance, etc, if you notice you are clicking/moving your mouse too much to do something then learn its shortcut
4. Read the comments of these runs: if it is too much work then at least read this one:
5. economy planning for the particular mission, this might be too much if you want to keep your work light, but at least include market abuse in your testing and also reducing your expenses on drop points and upgrades if unnecessary
Ok, I think next time I play Barb 1 again I'll try getting a market, kill the west monk with my starting troops then send them south, send the teutonic knights to the west, and do everything else roughly the same.  I am considering saving my wood for building a monastery close to the northeast relic, it is the one that takes the longest. 

On GK6, building the second castle any closer is going to be really difficult.  Definitely it can't be built on the other side of the river until the two towers are gone, and I could try to get a quick treb to destroy them but the enemy responds with some military and my army isn't strong enough.  Building it slightly closer on my side would save 5 seconds tops.  I probably should be smarter with my TC placement though.  I agree the market/castle would be good to get up as soon as possible, upgrades can wait since my mangudai play backseat to the saboteurs at the start anyways. 

I got the idea of garrisoning all my injured units inside the castle halfway through the game I referenced before, this would make monks less necessary.  That game was still kind of a preliminary practice run to see how things play out, but I think now it is time to really start optimizing things and figuring out exactly where to park my trebs, put my buildings, etc.

For gia's points below:

1)  You are correct (and because larger resolution eats memory way faster)
2)  I attempt to when it makes sense.  Mostly on unit production buildings and market.  If I assign too many hotkeys I'll get confused and click the wrong ones.
3)  This I probably should work on, especially for gather point buildings...
4)  Will do
5)  Too much for the time being.  As I've mentioned, the plan is to get something pretty nice but not perfect for each level, then once they are all decent I'll spend a few days at a time on each particular level and plan the economy.
I love YaBB 1G - SP1!
if you build a second monastery you could also get yellow from there but facing some archers with the knight, on the other hand you dont have to dance to the onager/mangonel whatever its name

about 2) I can help a bit, this is what I'd use in general and edit a bit depending on the mission or its state, roll your own:

Ctrl+1: melee units
+2: ranged units
+3: Rams
+4: Monks
+5: Trebuchets
+6: Hero unit if any
+7: Villagers

You dont have to group the buildings because they already have shortcuts (Ctrl+ the key that orders to build them, Market Ctrl+M) and IIRC if you group, lets say, many barracks you cant order them all to build with a single click anyway.
Edit history:
Gombill: 2010-07-24 11:53:03 am
I came up with some better ways to attack Genghis Khan 6, and the results are pretty nice.  Here's the summary:

- One town center, early market, less villagers total.  This saves 275 wood, 100 stone, about 400 food
- No mill or hunting, just food from sheep (since I don't need nearly as much now)
- No blacksmith or university (so no mangudai upgrades).  This saves 350 wood, 600 food, and 400 gold.
- Early castle, before the "Charge!" message at 6:20

A note on this - I found out that if the castle is built close enough, then enemy units that it hits trigger the computer's defensive response.  This lures all of his units (including trebuchets, which is important) out of his base so you can kill them all with 3 or 4 well-placed saboteurs.  Because of this finding:

- Build some early trebuchets, my goal is to be across the bridge with my army by about the 10 minute mark
- Send my cavalry immediately at the siege workshops to intercept scorpions (which is all they seem to build after my crossing, those and monks)
- Best trebuchet placements so they have to pack/unpack the least number of times

I've completed a run with a time of 21:40, but this was even with some bad luck with villagers scattering and rebuilding 2 buildings.  With some better villager containment this will easily be under 20 minutes, possibly even as low as 19 or 18.  The endgame screen said I lost a total of 8 units.
I've attached another run I just completed of Genghis Khan 6 in 21:12.  Once again the villager containment was an issue - I had all the relevant buildings destroyed by 18:30, and if my containment was better I would have had it when the villager died at 18:52.  Oh well, more attempts I guess.  A lot of other things went perfectly on this attempt.

A look at the last 6 levels left to do, and my progress on them so far:

JOA 3 - The main enemy is the british (red) but they can be dealt with using the tower tactic and eliminated from contention within 12 minutes or so.  Orange starts out very weak and can be killed with just a few archers.  My tries so far have been attacking the two leftmost red castles (including the one in his town) with rams, and the middle one with units including La Hire and Joan.  This needs to be done quickly (under 25 minutes) or purple gets enough units to attack.  No successful attempts yet, mostly because the tower thing only seems to work about 25% of the time on this level.

JOA 5 - I hate this level so much, just because of the refugees you have to escort.  You start with 10 and 6 have to make it, but enemy targeting tends to target weaker units, so you have to kill EVERYTHING nearby the villagers or else they will get picked off.  This means taking out towers, archers, onagers, and sometimes ships.  Lots of slow maneuvering with siege units, and to those not aware of the villager targeting situation it will just look like an unnecessarily slow run.  I've already tried just running with the villagers and ignoring enemies, but even with save/load I can't even get the villagers across the river that way, much less all the way home.

SAL 6 - I've already posted some comments on this.  Haven't given it any new tries in a while.

GK 3 - Destroy 4 enemy bases completely from a start of almost nothing?  This will be long.  The three closest enemies all seem pretty weak, I believe I can take them out one by one within the first 15 minutes or so.  The trouble will be yellow, which builds lots of chu-ko-nus, has a huge economy, and is walled in.  I believe the answer will be the tower strategy once again, maybe even a close castle.

BARB 1 - Comments already posted.  No completed runs yet, but the 11 minute one I referenced earlier with getting 3 of the 4 relics was my closest attempt.

BARB 2 - Not going to attempt this until I reinstall AOK 1.0 so I can use the early surrender glitch.

Attachment:
I love YaBB 1G - SP1!
The start looked great, during the invasion it wasn't that great but not too bad either, You had extra gold, maybe less gold workers and more at lumber so that it is more balanced.

While you had a castle early you tried to protect it and still spent saboteurs I think, just focus on the ranged and leave the melee alive so they keep calling the enemy.

The monks are annoying, try to send two scouts to the monastery? or are they targeted by other units. Maybe the scouts dont auto target the monk hmm, try to have one guard the other. Maybe you can defend your trebs and kill lost villagers in the meantime with the patrol command.

Before blowing the town center, do an incursion with your mangudai/scouts and kill the villagers and hit the town center, they'll go to the town center and then you blow it and kill them. You can even leave it in flames so that they are forced to repair it.
Edit history:
Gombill: 2010-07-27 09:50:12 pm
These are good ideas, I will work on them.  I spent some time yesterday also taking down gathering rates for all resources and building rates for buildings.  There were some surprising results, like that the sheep gathering rate is the slowest in the game (and has no upgrades to improve it).  In light of this I don't think I'll even bother with the sheep at all, just send everyone to gold and convert to food.  I realized after looking at the endgame screen I only gather 150 food anyways, it would be much more worth it to just convert once from gold (or maybe get by without the extra villagers).

To be honest, the scorpions are a bigger threat than the monks, so I make it a priority to destroy siege workshops first.  A few unchecked scorpions can decimate my mangudai army if I'm not careful.  The monks aren't that bad, just kind of annoying.

The spending of saboteurs isn't for the purpose of protecting the castle, it's more for just killing his stuff so I can start my attack sooner.  If I wait for my castle to deal with all those units, my attack would have to wait for 2 more minutes.

UPDATE:  I Improved some of my William Wallace times to be a bit closer to suga's.  I managed to beat him on WW3 by 1 second, and also I think his WW2 can be improved by 1 or 2 seconds but I haven't gotten the luck yet.  The other times still need work.  WW5, 6, and 7 are separate categories entirely because the difficulty actually matters in those missions.

Also I had an idea about Barbarossa 4.  I suspected that it would let you build the wonder somewhere outside of one of the towns and still have the trigger count it, and I was correct.  You can build it directly adjacent to any of Venice's walls and it will count, which is nice because Venice doesn't build an army so I don't need to do any fighting at all before the wonder starts, and very little fighting afterwards.  I can forget about my entire army except for the onagers, which I need to destroy ships.  So far I've gotten 13:42 < 16:28, but there's another wonder location I'm testing and it should trim another 20 to 30 seconds off if it works.
Two times and one video in this update.

WW2 in 3:07 < 3:12

As it turns out, getting a super-fast time on this is entirely up to luck.  Namely, one of the deer close to your town center has to be practically standing on it when you get the extra shipment of 4 villagers.  I got pretty lucky this time and killed a deer very close.

BARB4 in 13:28 < 13:42

This is due to the better wonder placement I mentioned in my last update.  I don't see this improving by more than a few seconds.  The only way it could be significantly improved is to keep the TC alive longer and train more villagers out of it before it is destroyed.  Then there would be the housing issue, though.  Bonus points for the flaming wonder at the end!

Attachment:
No real running today, I just spent a lot of time planning tricks and routes for Genghis Khan 3.  Here is what I found:

Yellow, which starts on an island, will get to the imperial age and turtle with cavaliers and chu-ko-nu if given the chance.  So he is the top priority for destruction.  Fortunately, the tower tactic works on him and effectively eliminates him within 15 minutes or so.  Unfortunately, it takes almost my whole army to get this to work since he starts with a decent army of archers.  Thus I have almost nothing for the other 3 enemies.

Green and Light blue start out very weak and I could kill either one of them individually within about 15 minutes.  My army is over with yellow, though, so the best I can do is set up towers next to both of their towns.  They both create only archers, so the towers can kill every unit they create without allowing them to build up large armies.  As soon as some siege equipment and a few archers arrive to finish off the villagers, these towns are both easy to kill.

The big difficulty is Red.  He has a huge economy and makes lots of archers, rams, and scorpions.  The rescuable bombard cannons and a few military units are technically enough to wipe them out, but it's really tricky and I foresee lots of restarts because of the red player. 

The strategy I'm looking at would avoid going into the castle age, since it wastes a ton of resources.  It's always best in speedrunning to avoid aging up if possible.  No resource gathering either, just the units I start with.  So here's the summary:

1 Tower, 2 Villagers + Most military for yellow.  The ram doesn't need to come. 

3 Towers for green

1 Tower for Light blue - this one needs to be surrounded by walls or the villagers will come after it and knock it down (which I don't want in this case)

Bombard Cannons + the rest of my military for red.  I haven't decided yet what units to train to be most effective.  Upgraded skirmishers are tempting, since the bombard cannons can deal with the rams and scorpions.  Red has so many villagers it seems best to tower them as well to get their villagers to congregate.

The ram can take out green and light blue's buildings, and my armies from red and yellow can finish them off once they're done with their respective towns.

Putting this all together in a single segment is going to be a real headache.  I think 30 minutes might be possible with great execution.
Edit history:
Gombill: 2010-10-08 12:25:45 am
Gombill: 2010-07-31 12:06:30 pm
I tested a few other various strategies and approaches to GK3, then on a whim decided to try it all in one go while recording.  The result:  Success!  To a degree...

For yellow I brought my whole army minus one cav archer and the ram.  I took 3 villagers also.  They built two towers, killed yellow's army, lured all the villagers to their deaths, and the mangonel destroyed the TC.  Yellow resigned at 24 minutes, no problem.

Green, I found out in a practice run, is the easiest of the bunch.  I towered his gold mine with 3 towers, destroyed his TC with the ram, and over the course of the game he ended up sending every villager to its death at the towers.  No other military intervention necessary.

For Light blue I set up a tower with wall around (to prevent villagers knocking it down).  This kept their army busy until my bombards could arrive later.

For red, I found out today that it's pretty nice to build towers behind his town.  This brings the archers and scorpions to attack the towers immediately, and the rams begin a long journey all the way around the water.  They get ambushed halfway around by my bombards and are all destroyed.

The remainder of the mission is cleaning up red and light blue, which I do with the bombards and an army of scout cavalry.  Total mission time should have been about 34 minutes (that's where I cut the video off).  But you might notice the video's name is GK3_4951.  This is because light blue's villagers scattered when I destroyed his town, and it took me ages to find them all.  Some of them were chopping random trees almost all the way across the map.  Once again the villager containment issue. 

So, in conclusion, I still think sub-30 should be possible.  And I was pretty surprised the whole thing actually worked on my first practice run!  Comments are great, but don't be too harsh, I know there are some issues to fix.

EDIT:  GK3 Video removed because it sucks.
There will be no new updates until somewhere around the 20th.  I have a lot of important things going on right now, including a legal battle with my landlord, so slightly less important things (like speedrunning) will have to wait. 

A few days ago I tied Suga's 3:38 for WW4 and beat my own WW6 time, improving it to 3:42.  I think both can be improved by another second or two.  So for the WW levels, I still have to make it to Suga's time on WW1, I've beat or tied him on 2, 3, 4, and the difficulty makes 5, 6, 7 a separate category anyways. 

I'll be back in about 2 weeks!
Quick update:

On WW1 I suspected that the enemy spawn block-out trick that suga uses for WW4 would apply, and I was correct.  The three spawning militia can be blocked out by three of your own units at just the right time to shave off some time hunting them down.  As a result, I achieved WW1 in 2:05 < 2:13. 

Also I put together a video that strings my WW 1, 2, 3, 4, 6.  I beat Suga's time on all of these but mission 4 (which I tied), and as I mentioned before 5 and 7 are a separate category anyways.

After I get back to speedrunning in earnest, the priority will be on missions that haven't been finished yet.  Once every mission has a run I will start chopping times in order of easiest missions to improve.

Left to Run:  JOA3, JOA5, SAL6, BARB1, BARB2
Easiest to Improve:  JOA2, SAL2, GK3, GK4, GK5, GK6

Attachment: