Research on battle map - post test results here

[Barry Fitzgerald]

Well I pretty much got the bugged part...still would have expected at least some reasonable casualties on the cavs part just at the charge point!

29! I almost fell off my seat!

[Frantz]

My "studies" (lol) on crossbowmen shootouts and swords fighting

Tried ( 3 round each ) French crossbowmen against HRE PAvise crossbowmen
( always 91 men each )

1) the french ( AI ) went loose which alway negate the advantage of pavise ( if they remain tight ) ... the french were about 10 men under when the Human controlled HRE pavise crossbowmen but they win because it seems that the AI have always more rounds than you ... i will not count that so its 3 HRE slight victories (just because of the first volley ) .
Is a DRAW , the pavise is countered by getting a loose formation

2) i put the pavise also in Loose formation , that give a +20 kills for the pavise ... that always win ...

3) giving a 3 bronze experience to the french help NOTHING , experience seems have nothing to do with shooting and thats is really BAD ...should be modded or patched .... surely count in hand to hand but come on ....

4) with 3 gold experience MAYBE they have a slight advantage but they act very strange , they always try to get very close , they routed the pavise once getting very close too but two times they got heavy casualties
So a 3 gold unit can be countered by a zero experience unit while shooting

5) giving armour and weapons bonus dont help ... i can agree on that , the crossbow is so powerful that wont be stopped by an heavier plate ...

6) tryed Dismounted MAA vs Dismounted MAA
the lessons : the charging unit win always ...except if the defende go in loose mode ... even if some of you dont i will tend to agree on that , the loose formation negate the advantage of the charge and allow to surround the enemy unit compensating the first losses BTW go on loose is a tactical disadvantage on the field cos expose the unit to multiple contacts and the morale will suffer alot ...
Amour and weapon improvement HELP alot , the stand the charge ALOT better ... EXPERIENCE also help not enourmosly but a 3 bronze experience unit will always beat a zero experience unit with an advantage of 7-15 knight

[econ21]

A link to a thread by Reapz that includes battlemap tests:
https://forums.totalwar.org/vb/showthread.php?t=73777

And a link to a thread with tests on gunpowder units by Whacker:
https://forums.totalwar.org/vb/showthread.php?t=73599

Interesting tests by Altien showing Genovese crossbow militia owning most other missile troops:
https://forums.totalwar.org/vb/showthread.php?t=73597

[Kraxis]

Regarding some rumours that the padded upgrade to units might in fact give them 4 points of armour (as per the findings in an unpacked file), I decided to test this out.

Sadly teh discussion was closed as people began to go off topic. So I can't it there, thus this is the secondbest place to do it.

"Ok, to see if the padded armour truly does give 4 points of armour I tested it out in custom.

Militia Pikemen (to get as many in a single unit as possible) against peasant archers (weakest possible missile to get the least chance of hurting the pikement too much) over 7 volleys (even the PA begins to edge closer around 7 volleys, which will spoil the test), over at least 5 rounds (did 8 with the bronze).

Here are the results.

After 7 volleys Peasant Archrs killed on average:

27,6 vanillla pikemen

23,1 bronze pikemen

18,4 silver pikemen (included silver to make certain that the difference from vanilla to bronze was larger)

The results speaks for themselves. We should have seen a considerably larger jump from vanilla to bronze, than bronze to silver (which is a single point if I'm not mistaken) if padding truly gave 4 points. However there is a larger jump from padding to light mail which shouldn't happen.

Two possible reasons for this in my mind.
Custom Battle uses the MP units. And perhaps there the upgrades are 1 point at a time to make sure it isn't too unbalanced with upgrades (though they only cost 30 a piece).
The other option is that there is no 4 point upgrade and what we see is a remnant from a time where there was."

Perhaps a stronger missile factor could be used to get more kills, and perhaps a stronger melee unit as well (Tercios perhaps). But I think these finds are enough for now.

[Reapz]

A comment and observations on battle-testing:

In this thread we have been discussing and analyzing the mehcanics of cavalry charges and that led me to do some battle-testing that I posted in this thread. I was surprised by the results not just in their implications for cavalry tactics (for which I would refer you to the threads) but for the implications on the value and accuracy of battle-testing in general.

Econ21 suggested I post a summary here also. I particularly wanted to share the observation that the results of custom battles to test unit combat vary greatly even if the conditions are identical, when you repeat the test.

Here as an example are ten results from the identical unit combat scenario:

Mailed Cavalry vs. Noble Pikemen, both experience zero, no armor or attack bonus
Flat grassy plain
AI managed the Pikemen
I activated the cavalry attack with a double right click one time on the target
Then let the units do their thing:

Survivors (Knights/Pikes) - Victory or Loss

0/32 - Loss
21/13 - Victory
19/0 - Victory
18/0 - Victory
1/21 - Loss
13/0 - Victory
12/0 - Victory
1/25 - Loss
1/39 - Loss
22/0 - Victory

Now the spread of results here is from a victory with almost 60% of the knights surviving to crushing defeat with no survivors and 60% of the Pikemen surviving. If you just did one test and relied on the single result you could arrive at wildly inaccurate conclusions.

I saw this spread with every custom battle scenario I repeated ten times - some losses, some wins - despite the conditions being absolutely identical

So I would suggest that people be very wary of single test results and go as far as to say I don't think you can generalize conclusions without doing a series of identical tests until a trend is clear. I am not a statistician but I can see from the results above that if ten people did the above test once each, four people would conclude that Pikemen slaughter charging knights, a couple would conclude knights win but at the cost of most of the unit and the rest would conclude it is a fairly safe bet for the Knights to come out with a win and about half the unit surviving.

caveat emptor

[FactionHeir]

Bullethead, try your scenario with spearmen walking away vs running away from the cav when it is about to or starts charging, as that changes them to pursue mode and then turn and attack the cav once they go melee onto your spears.

[Whacker]

And a link to a thread with tests on gunpowder units by Whacker:
https://forums.totalwar.org/vb/showthread.php?t=73599

Per Mr. econ's suggestion, I'll post a few summary items as I see them here for this thread's purpose. Testing wasn't nearly completed as much as I'd wanted to, got sidetracked and eventually gave up, as I'd like to wait for the first patch to see if some of my major gripes are fixed and give the game a fair chance. Incidentally, most of my tests were actually done on horse archers, then we were intending to move onto gunpowder units. Here goes:

1. I firmly believe in reading other's posts and from my own experience that the new "blobbing" of units is probably to blame for a large number of our issues and gripes, in terms of charging, movement, pathing, cohesion, etc.

2. Experience in chevrons does indeed increase the kill rate. It does NOT appear to affect anything else, such as quicker unit response to commands or cohesion at all.

3. Distance from target does *seem* to affect accuracy, but in my tests it does not affect kill rate. The "trail of dead bodies" as enemy pikemen moved towards my HA's was rather uniform at all distances. Thus more or less arrows may hit at closer or longer distances respectively, but it doesn't seem to affect the kill speeds.

4. For HA's, block vs 2 row formation doesn't seem to affect kill rates. If there are penalties for firing in deep ranks, my testing doesn't show it.

5. Moving HA's around most definitely DOES impact killing rates. Stationary has the most accuracy and best killing speeds. Walking is second, and still elicits decent kill rates. Running is terrible, and you'll hardly kill anything. If you can keep your unit together that is. The "blobbing" effect while running is horrid and I found it almost impossible at times to keep my unit in a cohesive formation.

6. From the little testing I DID do on gunpowder units, it does seem that 2 row formation works best, as opposed to a block formation.

That's all I can think of for now!

[FactionHeir]

Yup, in experience, blobbing/collision AND automatic wait for all units in a group to get close before attacking (except when in pursuing-mode) is the root of all problems in M2TW.
Oh, and of course stopping because it "might" hit a stray unit from one of your own troops. In RTW they'd just run it over, which would make much more sense given that your archers can friendly fire too.

[Reapz]

Link to thread comparing longbow units vs. crossbowmen and demonstrating strong effects from unit experience on missile combat outcome.

https://forums.totalwar.org/vb/showthread.php?t=74553

[Rothe]

To comment on the earlier testing of peasant bows against upgrade armored pikes, I would like to see a test that compares two units that have as many men, but the other has no armor and the other has padded armor. I think there are units in the game for the test.

My idea is that you'd have to also compare how much the padded armor protects compared to no armor, when you are actually sure that the padded armor is there (via unit type) as opposed to padded armor via upgrade.

It could be that padded armor does not protect that well even if you compare unarmored vs. padded in arrow fire.'

The previous test did show that the armor has significant effect on the survivability of the pikemen in any case, even if it is upgrade armor.