You must be logged in to post messages.
Please login or register

Rome: Total War Discussion
Moderated by Terikel Grayhair, General Sajaru, Awesome Eagle

Hop to:    
loginhomeregisterhelprules
Bottom
Topic Subject: Collective Data on Charge Bonuses
« Previous Page  1 2 3 ··· 4  Next Page »
posted 27 March 2005 18:50 EDT (US)   
I have begun a large series of tests involving charge bonuses. I spent about 90 minutes earlier taking a collection of 78 test battles. This is my story...

After reading arguments from both sides that involved the "Charge Bonus Bug", I decided that I would run some tests myself. First, of course, I tested Companion Cavalry, the infamously underpowered cavalry. For this test, I used 5 different charge bonus values to test the difference it made when charging. For each value, I ran six test battles. In each battle, I used a unit of Macedonian Companion Cavalry to charge a unit of Carthaginian Iberian Infantry. The values below represent the number of casulaties the Companions inflict between impact and the time when the "Charging" message disappeared. I always paused as soon as this messaged disappeared and recorded my data. Asterisks (*) indicate a rout on impact. The charge bonuses range from 1 to 63. 63 is the highest value that the game will allow for a Charge Bonus. The number at the end of each round is the average amount of casualties inflicted.

Round 1:
1 Charge Bonus: 66, 52, 70, 59, 50, 46= 51.17

Round 2:
16 (default) Charge Bonus: 48, 44*, 51, 50, 43, 49*= 47.5

Round 3:
25 Charge Bonus: 58*, 53, 59, 37, 58, 42= 51.17

Round 4:
40 Charge Bonus: 61*, 63, 52*, 58*, 49, 44= 54.5

Round 5:
63 Charge Bonus: 63, 63*, 75, 54, 53, 57= 60.8

For the following tests, I used Scythian Nobles. The tests were the same in every way to the Companion tests, except for the Scythian Nobles. I also removed the 25 charge bonus round. I felt it was unneeded.

Round 1:
1 Charge Bonus: 37, 43, 54, 41, 36, 48= 43.17

Round 2:
17 (default) Charge Bonus: 53, 37, 38, 33, 42, 55= 43

Round 3:
40 Charge Bonus: 25, 28, 45, 30, 27, 35= 31.67

Round 4:
63 Charge Bonus: 46, 35, 53, 57, 66, 58= 52.5

For the following test, I did exactly the same as the other two tests, except used Praetorian Cavalry (SPQR). Again, the 25 charge bonus round is removed because I felt it unnecessary.

Round 1:
1 Charge Bonus: 54*, 61*, 45, 70, 82, 61*= 62.17

Round 2:
9 (default) Charge Bonus: 60, 61, 70, 56, 62, 66*= 62.5

Round 3:
40 Charge Bonus: 57*, 56*, 74, 44, 60*, 53= 57.3

Round 4:
63 Charge Bonus: 53*, 52, 86*, 79, 68, 62= 66.67

My conclusion: Charge bonuses do not make the difference as previously believed. I always thought that the charge bonus was added to the base attack value to deliver a more powerful blow when charging. Clearly, this is not the case. The game gives Praetorian Cavalry and Companion cavalry the same mount (heavy), giving them the same weight. The Companion's base attack value is 10, plus the 16 charge bonus, is 26. The Praetorian Cavalry has a base charge attack value of 12, with a charge bonus of 9, which comes to 21. Now, as my tests show, Praetorian Cavalry give a considerably more powerful charge. 9 out of 24 routs on impact for the Praetorians, 7 out of 30 for the Companions. The Scythian Nobles don't even compare, even though they should deliver the most powerful charge in the game. Scythian Nobles get 8+17 (25), almost as high as the Companions, but failed to acheive a rout on impact against Iberian infantry once in 24 attempts, with increasing charge bonuses!

My guess is that any real improvement upon increasing the charge bonus is a fluke. Even if it isn't, the difference is next to negligible. As has been mentioned, the seeming only way to make up for the deficiency here would be to increase the base strength of the two most underpowered cavalry in the game- Companions and Scythian Nobles. I also believe it is noteworthy to say that the Praetorians never suffered more than 8 casulaties (averaging about 3), while Companions averaged about 10, reaching 19. Scythians were usually somewhere near 13 or 14, reaching as high as 21.

Anybody have any test results they'd like to post? Well, do so here! I want to see some evidence that can disprove this. Also, it would be nice to have some light cavalry tested out, as Light Lancers have been a major point against the inactive charge bonuses. I wouldn't really bother with Cataphracts, as we already know they're devastating and nearly impossible to kill. But hey, I'm not stopping you. Maybe if we get enough data CA will notice, and possibly fix this? Here's hoping!

(Oh yeah, the point of this thread is for argument on the charge bonus issue, especially with data from tests you've done. We're trying to actually get to the bottom of this...)


We are all in the gutter, but some of us are looking at the stars.

[This message has been edited by DoitzelKaiserIII (edited 03-28-2005 @ 06:45 AM).]

Replies:
posted 27 March 2005 19:06 EDT (US)     1 / 81  
Heh, we have a winner. Good job and good work man!


Hopefully someone at CA will notice this or one of the other threads at other sites on the same subject.


Ex-Seraph Cheesewiz - Former WICH Webmaster, AOE3H Webmaster, & RTWH Staff, HeavenGames LLC
World_in_Conflict_Heaven || Age_of_Empires_III_Heaven || Support_HeavenGames || The_Playpen || Do_The_Right_Thing
posted 27 March 2005 19:19 EDT (US)     2 / 81  
Some Ideas

Perhaps the charge bonus is not set, but is a random number, its value affected by the bonus. It could be that the randomness of the number is just not as random as we think, perhaps generating the same number more often than not.

And are we absolutely sure that charge bonus affects attack? What if it has an effect upon morale instead/aswell?

Just some of my thoughts...


Oopsey daisy, dowsed in flames survive....

www.maniacproductions.co.uk

[This message has been edited by Zygodactyl (edited 03-27-2005 @ 07:20 PM).]

posted 27 March 2005 19:23 EDT (US)     3 / 81  
If it had an effect on morale, then Companions and Scythian Nobles would have broken the Iberians more often. Their charge bonus is considerably higher. As for the random number thing- I doubt it. There'd be much more variation in the results. That, or they have a fairly small range of numbers.

We are all in the gutter, but some of us are looking at the stars.

[This message has been edited by DoitzelKaiserIII (edited 03-27-2005 @ 07:23 PM).]

posted 27 March 2005 19:30 EDT (US)     4 / 81  
Meh. It just seems that whatever we throw at this problem, it always comes back the same. Anything logical makes no difference... if CA could even hint whether we were going in the right direction, that would be good.

edit

Was looking around game files for hints, and found this:

attack factor
attack bonus factor if charging

Note how it says factor. This could mean actual value, or it could mean a factor of something. I could just be reading too much into things...


Oopsey daisy, dowsed in flames survive....

www.maniacproductions.co.uk

[This message has been edited by Zygodactyl (edited 03-27-2005 @ 07:35 PM).]

posted 27 March 2005 20:16 EDT (US)     5 / 81  
This seals the deal for me. The charge bonus does not work.

I have ran several tests myself, and they all indicated the same. I am rather dissapointed in this, but at the same time it explains why cavarly are out of whack in terms of what is good and what is not.

posted 27 March 2005 20:27 EDT (US)     6 / 81  
Very good. As long as charge bonus doesnt make much difference, then my strat is still good.

Michael Jackson
posted 27 March 2005 21:03 EDT (US)     7 / 81  
Briliant tests, DoitzelKaiserIII, I'm making links to your post in the other 3 biggest RTW communities, so everyone will see your story. Hopefully CA will be aware of this problem, unless they are completely blind.
posted 27 March 2005 21:31 EDT (US)     8 / 81  
So this is another error. Isnt it still true the units in question are still Good units? They may not be incredible, but arent Companions and Preatorians still damn good compared to the unit below them? I've never been interested in the Romans or the Macedonians, so i cant say i care to much

That is all I have to say.
posted 27 March 2005 21:32 EDT (US)     9 / 81  
If I can find time sometime this week I'm going to test Sacred Band Cav. and maybe Cataphracts, too (then again, why bother? ) Oh yeah, and Gothic cav as well.

We are all in the gutter, but some of us are looking at the stars.
posted 27 March 2005 21:50 EDT (US)     10 / 81  
Wow, i used to be a firm believer in the charge bonus working. But now I don't. Great job

Ya sigpost #1111: 1111(minor) - Gadz13
±±÷÷±± ±±÷÷±± ±±÷÷±±
I BELONG TO THE NEW AGE OF REASON!!
*doesn't put post number at bottom*
posted 27 March 2005 21:51 EDT (US)     11 / 81  
CA does watch forums, but they don't seem to watch this one. They watch these ones:

http://forums.totalwar.org
http://pub133.ezboard.com/bshoguntotalwar

To Heresiarch: your post in forums.totalwar.org was too brief, people don't know that you are in fact representing, officially, rtw.heavengames.com. You should've wrote a more formal

"I am from... and I represent ... and we are here to request a co-operation regarding a bug that we have found recently...".

Your post in pub133 was better though.

I don't know, never liked Greek factions, so underpowered companions doesn't really bother me, but a bug is a bug.

And bugs NEED TO BE FIXED!!!

posted 27 March 2005 21:52 EDT (US)     12 / 81  
Thank you for all the work that you are doing to make this game better, it is appreciated.
posted 27 March 2005 22:40 EDT (US)     13 / 81  
Well, it's nice you took the time to figure this out, but your margin of error is so huge that the data you've collected is really not much help. Even if the charge bonus IS bugged, I doubt that scythian nobles would do LESS damage as their charge bonus increases. I think that in order to get more meaningful data you'd have to do at least a couple dozen tests with each charge bonus value, and then perhaps test more charge bonuses so we can try to find a trend if there is one.

The results of the various tests people have been doing are unexpected, but I wouldn't really be so sure that cavalry charges are bugged *just* yet. What I would like to know, is how charge bonuses SHOULD work, so we could compare how they are supposed to work with how they actually do instead of comparing the results to how we expected them to work. Armor bonuses are far more effective in deflecting charges than defensive skill bonuses, and as we all know, defensive skill has no effect against arrows. However, just looking at the unit card without that specific knowledge, you'd think that, say, a 9 chevron archer auxilia unit is on par with heavier infantry units in terms of defense. Perhaps something similar is happening here.

That said, even if there IS a problem with the charge bonus it hopefully shouldn't ruin the game for you. Cavalry are intended to outflank infantry and chase off skirmishers, archers, and routing units, and as of now they do it quite nicely.

posted 28 March 2005 06:44 EDT (US)     14 / 81  

Quoted from Myself:

(Oh yeah, the point of this thread is for argument on the charge bonus issue, especially with data from tests you've done. We're trying to actually get to the bottom of this...)

Quoted from Kaptain_Brad:

So this is another error. Isnt it still true the units in question are still Good units? They may not be incredible, but arent Companions and Preatorians still damn good compared to the unit below them? I've never been interested in the Romans or the Macedonians, so i cant say i care to much

Praetorian cavalry are still good units (especially for never existing in history). Companions, however, are not. They are the worst cavalry (and possibly just the plain out worst) unit to come at huge city. If their charge bonus makes no apparent difference, then Macedonian Cavalry, with 1 less attack, should probably perform just as well as they do. The only way to use Companions now to at least SOME effect is to hold down the alt button the whole game (so they at least get 12 attack instead of 10). The point I was trying to make was that, despite having the SAME mass and a LOWER charge bonus, with a higher attack, Praetorian Cavalry deliver a stronger charge than Companions, who SHOULD have the second strongest in the game. If we're going for historical accuracy here: Companion Cavalry should be the best cavalry in the game. Period.

I'm going to do some time tests with Companion Cav. vs Cataphracts, to see how fast the two of them move... Might be hard to do, though.


We are all in the gutter, but some of us are looking at the stars.

[This message has been edited by DoitzelKaiserIII (edited 04-14-2005 @ 09:45 PM).]

posted 28 March 2005 06:52 EDT (US)     15 / 81  

Quote:

If their charge bonus makes no apparent difference, then Macedonian Cavalry, with 1 less attack, should probably perform just as well as they do.


What about the +2 morale and +3 melee attack of the Companion?

Quote:

so they at least get 12 attack instead of 4


...you sure you feeling alright?
posted 28 March 2005 08:08 EDT (US)     16 / 81  

Quote:

I'm going to do some time tests with Companion Cav. vs Cataphracts, to see how fast the two of them move... Might be hard to do, though.


Take the Seleucids:

SL := silver legion
ca := cataphract
co := companion

....SL+ca+co ............................ SL
make sure the distance between the silver legions is far enough.
Let the first group (catas for example) move to the left legion, and then let them run to the other unit. Measure the time. Then do the same with the companions.

Another test you could do is let them both run a long distance and then compare their status (warmed up, tired, ...)


SUPERDROIDEKA
posted 28 March 2005 08:26 EDT (US)     17 / 81  
I just had a thought. What if there is a chane that the charge fails does not deiliver the charge bonous but only a standered attack or the equivilent of a lower charge bonus. Mabe companions are bugged so that thier charge fails too often?

May your victories be grand and your rule be long and wise

-Centurion Triarii

posted 28 March 2005 11:02 EDT (US)     18 / 81  
Ok fellas, don't tell my girlfriend but I'm no longer convinced that i am 100% right on this one, after all I am not infallible just perfect. And having done (long) tests I cannot disprove them outright, but just to be a smartass I might be able to cast (some) doubt over them.

What I realised is that when testing one on one, you really have to use a frontal charge, you cannot manauvre cleverly enough and still be far enough away. I assume the tests mentioned above were all done using frontal charges. This is important because once the enemy attacks you, the charge finishes. This essentially means that a frontal charge will never last more than a second. Not only does this mean that there is minimal time for any bonus to be in effect, but it also means there are no more than a handful of soldiers actually in the battle before the charge is over. Now, if you attack from the side or behind, the charge can last up to 6 seconds give or take a few, and this means all units are in the charge attacking. When this happens there is phenonemal damage inflicted and normally a route as well.

When I first did my first tests (see Which charge bonus? This charge bonus!) I used two of each cavalry to simulate a real cavalry manauvre (I would NEVER attack anything other than missile troops with even the best cavalry from the front alone) so I avoided this problem and my results seemed to support the idea that there was a unit specific charge bonus charge bonus (lancers did well, Companions did much better, early cohorts, which have similar stats to the companion, had no real charge effect). But I wanted definitative results so I tried baiting the enemy unit (this time an Iberian) with one unit while attacking from behind with another in order to see how a long charge from a single unit effects thing. Unfortunately the AI wasn't playing my game and rarely stayed still or went in a convenient direction for me to learn anything. The few times i got a high level (think Companion) unit to succesfully charge into the rear of an Iberian the enemy routed amd lost half its men by the end of the charge. The only way you could reliably test this is by using multiplayer.

If you've considered this and its irelevent fine, I just have a real hard time believing that Lancers and milita cavalry are the same unit except one can fire missiles. And I can't explain how changing the bonus in the data file has no little effect unless that number is simply a representation of a game mechanic like some people have suggested.


*Coming soon to this space*
something clever and topical
posted 28 March 2005 11:51 EDT (US)     19 / 81  
I ran a few tests after hearing this, I think I found something else: charges lose effectiveness if they're distant! Charging from the other side of the battlefield and charging five feet away give great differences! I don't think that it's only fatigue, because I used the companion/iberian setup, and I'm pretty sure they were still "fresh" by the time they got to the Iberians. The results (all one unit of Macedonians versus ine unit of Iberians on grassy flatland with neither of them having any upgrades):
I got the number by subtracting the number of companions lost in the charge by the number of Iberians lost in the charge
(surprisingly enough, I lost no Companions in the charge)

1: 22 (charged about halfway to the Iberians)
2: 13 (charged aftfer a few steps of walking)
3: 12 (charged first thing)
4: 15 (charged about halfway)
5: 12 (charged first thing)

I tried to walk them once, but they charged on their own, and that had an effectiveness of 15. My guess on the first one is I may have taken a flank by accident, as opposed to a frontal charge.


__/\--*~N64~*--/\__
The Tale of Sir Gusson-3%
Sole competitor and winner of the Green Map contest.
So cool!
posted 28 March 2005 12:10 EDT (US)     20 / 81  
I don't mean to disrespect the work you did, but I think you are jumping to a conclusion with the evidence you have provided.

There are many variables to consider. The tests would need to be repeated at various difficulty levels, with the charges coming from different directions, and the charge beginning at different ranges from the enemy. You would need to test on various enemy units too, like light infantry, heavy infantry, archers, etc. There is also terrain, level ground, uphill and downhill to consider. It surely would make a difference if the opposing units were already engaged when charged vs seeing it coming. That's a large amount of testing!

I would say that while your evidence is interesting, it does not support a defacto conclusion that charge bonuses are not functioning as designed.

posted 28 March 2005 12:20 EDT (US)     21 / 81  
Old Celt: Every solid test done, by every person, has shown little to no impact from changing the charge bonus. Different units, different maps, different charge bonuses, etc etc. I think that says plenty. I myself tried it with equites, round shields, and companions. All on level ground. The receiving units were cataphracts, iberian infantry, and urban cohorts. The charge bonus didn't make any kind of noticeable difference in ANY of the tests. I can't see how it is possibly working.

[This message has been edited by Themistocles472 (edited 03-28-2005 @ 12:22 PM).]

posted 28 March 2005 12:38 EDT (US)     22 / 81  
A user named Red Harvest at Totalwar.org has posted his findings in batteries of tests which indicate the opposite conclusion: that in fact, those bonuses ARE working. The actual mechanics of how the bonus is applied are unknown to us. Only CA staff know the exact specifications for the performance of the battle engine. Since there is evidence from both sides of the issue readily available, the objective viewpoint is to say conclusions for either side are speculation only.
posted 28 March 2005 14:57 EDT (US)     23 / 81  

Quote:

...you sure you feeling alright?

I haven't been up before 7 AM in a week, and wasn't in bed until 2... I was thinking 12 and 4 (attack and charge) and typed the 4 instead of 10 I guess. I fixed it

Quote:

You would need to test on various enemy units too, like light infantry, heavy infantry, archers, etc. There is also terrain, level ground, uphill and downhill to consider. It surely would make a difference if the opposing units were already engaged when charged vs seeing it coming.

I'm only one man...

Quote:

I would say that while your evidence is interesting, it does not support a defacto conclusion that charge bonuses are not functioning as designed.

As I said, I can't draw a definite conclusion yet. Much more testing to be done. MUCH more. However, based on what I've already done the evidence is in favour of the charge bonus bug's existence.


We are all in the gutter, but some of us are looking at the stars.
posted 28 March 2005 16:53 EDT (US)     24 / 81  

Quote:

As I said, I can't draw a definite conclusion yet. Much more testing to be done. MUCH more.


no, CA should explain how the charge works and then we will have results.

SUPERDROIDEKA
posted 28 March 2005 22:49 EDT (US)     25 / 81  
Perhaps one of the reasons we do not see the charge bonus is that higher charge bonuses are overkill. A single cavalryman would probably only be able to attack one enemy soldier before the charge runs out. A Companion with 63 charge bonus would just as easily kill an Iberian Infantry as a Companion with a charge bouns of 1. Higher bonuses might not equal higher kills.
« Previous Page  1 2 3 ··· 4  Next Page »
Total War Heaven » Forums » Rome: Total War Discussion » Collective Data on Charge Bonuses
Top
You must be logged in to post messages.
Please login or register
Hop to:    
Total War Heaven | HeavenGames