How well can gates block trebuchet projectiles? Some data. The setup:
(My AoE2 is bugged, the terrain is not showing up)
How projectile blocking works in AoE2 is that some shots “miss”, based on the firing unit’s accuracy. If a shot “misses”, it uses the firing unit’s dispersion and the distance to target to get a destination: more distant targets lead to higher dispersion. If nothing blocks the shot, it can still hit the target if it lands close, but “missed” shots can be blocked by other units and even buildings. For example, fire ship projectiles are blocked by palisade walls because there is a 100% miss rate.
Trebuchets are very inaccurate, with 15% accuracy. However, vs 1v1 buildings they are 80% accurate, according to the wiki. This testing confirms that gates provide little or no benefit to towers against trebuchets.
Data collection: I counted the number of volleys it took to kill each trebuchet. If a firing trebuchet was lost to counter-fire, I replaced it with an idle trebuchet from the second group, so each volley had 4 shots. The range is only about 10 tiles, since you can’t zoom out in AoE2 HD and I wanted to fit everything on one screen, so at maximum range gates might block slightly more shots.
I also counted the number of hits sustained by each gate (hits doing half damage when they hit a different target).
The first test was very unlucky, since from left to right it took nine, three, and one volley to kill the trebuchets, but I think it’s fine to average the results. The damage sustained by gates is obviously correlated with the number of volleys it took to kill the trebuchets, so I hope just taking the average of both series is statistically fine. If I tried to screen capture all this I would get maybe 5~10 FPS and my CPUs already averaged 89°C without this, so no video or SotL-type narration.
Trebuchet in gate tunnel took 5.8 volleys on average before being destroyed (max 9, min 3, 11 tests). Front gate took 11.3 hits (max 22 = destroyed, min 4). Middle gate took 0 hits. Back gate took 3.6 hits (max 8, min 0).
Unprotected trebuchet took 2.2 volleys on average (max 3, min 2).
Trebuchet with single gate in front took 4.5 volleys on average (max 10, min 1). Gate took 10.4 hits (max 22 = destroyed, min 1).
Gate in front of tower took 1.1 hits (max 4, min 0).
This doesn’t necessarily mean that the middle gate is useless. The trebuchet was positioned in the scenario editor; in a real game, you would build the gates, then move the trebuchet inside of them. The collision calculation might be different if the trebuchet is slightly closer or further away.
The second data-collection test involved how fast a garrisoned castle can kill units. Setup:
A total of 49 monks were present, with Sanctity disabled (since this was originally a test of whether castles retarget immediately).
With no Mangudai present, the castle took 132±3 sec to kill all 49 monks, or 2.7 sec per monk. (Some volleys would fail to kill a monk.)
Castle fire rate measured (against another castle) at 25 volleys/min ungarrisoned (2.4 sec/volley), BUT looking at a YouTube video I think I measured a lower value of 0.05 sec/extra arrow so it might depend on game framerate, maybe arrows fire faster in DE. This is data for AoE2 HD.
With Mangudai garrisoned for 20 extra arrows in total, the castle took 114±1 sec to kill all 49 monks, or 2.3 sec per monk. Measured against a castle as ~17.5 volleys per minute or 3.43 sec/volley. After killing a monk, the volley would stop for a little bit, then resume against another monk, but at maximum range all arrows would go towards one monk.
With Mangudai garrisoned and manual targeting of monks to eliminate the switching delay and reduce overkill, it took 95, 84, 70, and 72 seconds to kill all monks, minimum 1.43 sec per monk. Lower times are because it seems like switching targets just before a volley begins can cause a volley to be skipped, so I learned not to do this.
With Mangudai ungarrisoned and vulnerable to conversions, it took 39±4 sec to kill all monks, or 0.8 sec per monk, with Mangudai losses not counted.
Other data of less interest: TC with 10 arbalests (10 arrows) fired 21 volleys in 60 sec. With 4 arbalests (5 arrows), it was 24 volleys in 60 sec. DE might be different.
Tl;dr: a gate in front of a trebuchet absorbs over half of the hits from another standard trebuchet (not Huns, or Britons with Warwolf).
Are you going to try putting gates in front of your trebuchets during trebuchet wars?
- Yes
- No
- I always fully wall my trebuchets when melee units or other trebuchets threaten them
Original post:
My AoE2 is broken, anyone want to test something?
I’ve heard that gates can block Trebuchet projectiles, because their hitbox is high. But I never see expert players doing this. I wanted to test it, but I only have AoE2 HD, which I hadn’t tried opening in over 3 years, and when I tried it now it gives a bunch of errors and then opens a window with no text on the buttons since I use Wine and random OS updates broke things.
So this is what I wanted to do:
Targets:
1) Gate directly in front of (touching) a Keep or Bombard Tower for the high HP.
2) Gate directly in front of (touching) an unpacked trebuchet.
3) Three connected gates, forming a tunnel, with an unpacked trebuchet in the middle gate.
4) (control group) A tower with no gate
5) (control group) A trebuchet with no gate
Make four trebuchets to attack each target. Then use the copy function to make 10 copies of this setup.
Optional: set both players to post-Imperial age, make target player Chinese for extra HP on walls and towers, make a scout for the target player if AI wants to mess things up by scouting with a trebuchet or turn AI off if that works.
Start scenario, pause game, set trebuchets to attack targets, and then record damage to targets after one or two volleys.
It would be nice if someone could test this! I’m curious whether gates actually help in trebuchet wars.