Is combat hollow, or is the calculator wrong?

[korn469]

After reading the combat explained thread over at civ fanatics I started playing around with the combat calculator. According to it a knight (strength 10) attacking an unfortified axeman (strength 5) on open terrain (no attack or defend bonuses for either unit) has a 99% chance of victory. I haven't played the game nearly enough to know if observations back this up or not. If it holds true then that would mean infantry (strength 20) attacking a knight also has a 99% of victory, as does modern armor (strength 40) attacking infantry.

Here's the link for the combat calculator:

http://c4combat.narod.ru/c4c_v0_13.htm

Those instances don't worry me nearly as much as a Jaguar (strength 5) attacking a chariot (strength 4) and having 75% chance of victory. Or a knight (strength 10) attacking an unfortified musketman (strength 9) and having a 68% chance of victory. Or a knight with combat 1 (strength 11) attacking a knight without any combat modifiers, and the attacking knight also having a 68% chance of victory.

What have you observed, and is the combat calculator right? If it is, should combat function as it currently does?

[Simon Appleton]

I suspect the calculator is correct, but I don't see the problem you are alluding to. Do you think superior units have too much of a chance of victory? It may look like that, but bear in mind that they will typically be badly wounded afterwards.

Often, I find a tank attacks and ends up with its strength little higher than a redcoats afterwards. Plus, if a I understand the combat system correctly, a wounded unit with strength 4/8 is easier to kill than full health lower tech unit with strength 4/4. (It has only half the hit points).

If anything, from my gameplay experience, I think the game understates the power of historically superior units but it does that for a good gameplay reason - to give low tech combatants a chance and to allow quantity to have a chance against quality. [I came to appreciate this in my last game when I had to fight modern wars without oil - ouch!]

[korn469]

Simon Appleton

I guess the problem I have with it is that even a small bonus (less than 10% basically just one hit point difference) makes the odds of victory for evenly matched units go from having a 50% chance of victory to having about a 2/3's chance of winning. Then a unit with a 60% advantage has about a 90% chance of victory. I guess this is where your statement about modern units comes into play, because as soon as a unit suffers any damage at all it loses a significant amount of combat power. So while an undamaged tank might never lose to a spearman, a tank that has lost a few hitpoints certainly has a much lower chance because the combat engine exaggerates any changes in combat power.

[Arrian]

I've lost battles where I've had the battle odds advantage (if you right click & drag over the opponent, it gives you a strength vs. strength breakdown) quite a few times. I've also won a few that were against the odds. So there is definitely a chance that the weaker unit will win, but the general idea is that it shouldn't happen often. People tend to get upset when their Tanks lose to spearmen

-Arrian

[Simon Appleton]

I guess part of the point is that a 10% difference in strength is not a small difference - it's one promotion (combat 1). Going from a 50% to 2/3 chance of victory means that such a promotion really matters. I originally thought a combat 1 promotion was a bit lame, compared to the beefer situational promotions (25% for city raider, garrison etc). However, as I played the game, I came to see how important even such a apparently small edge can be. But as Arrian says, this is not to say that nasty (or happy) surprises won't happen from time to time.

A 10% difference may look small when you compare units with different technologies, but then arguably those kind of units should have very different power. Even then, as I said in my original post, numbers are the great leveller. The AI typically does not attack piecemeal and nor should we. As Stalin said, quantity has a quality all of its own.

[Hauptman]

If i use a veteran tank against an untrained tank yes it better have a big advantage in that fight. 10% increase should mean 75% win chance, and in this game, it does. However i may win that fight but now my tank is out of the war for a while as it is badly mauled. As now it is only that 10% thats remaining...

IMO they did very well with this system over the last. In civ 3 a double strength advantage gave you a 66% chance to win. That wasnt anywhere near enough.

[Arkay]

The combat calculator is correct. For an intuitive answer as to why, its because 'on the fly' hitpoints are calculated for each battle based on relative strenghts. Thus, a more powerful unit has more hitpoints as well. The weaker unit will have to win more times with lower odds each 'round' to win, which explains the hefty percentages.

As to your second question, I am happy that it works this way. In civ3, if you attacked a city with a unit and did no damage, you actually ran the risk of promoting the defender and making it more powerful. I didnt like that at all. Furthermore, a unit's defense (or offence) number never changed, so a 1 hit point pikeman was still as strong every round as a 5-hitpoint pikeman. Taking (lots of) cities was just about impossible unless you had the proper tech.

Now, it is very possible to overrun a better-teched opponent just on sheer numbers. The first guy chips the armor, the second guy wears him way down, and the third guy mops up. In civ 3, it would be the first guy makes the defender stronger, the second guy takes off that hit point, guys 3-10 take him down to 1 hit point, your last 3 guys fail and then you exit the game and play call to power 2. I like the idea of being able to commie-man-rush a city and take it over. "Sir! They're everywhere! We're outnumbered 10-1 and we just ran out of ammo!" Kinda like the movie Zulu.

[Barchan]

Originally posted by Arkay
Now, it is very possible to overrun a better-teched opponent just on sheer numbers. The first guy chips the armor, the second guy wears him way down, and the third guy mops up. In civ 3, it would be the first guy makes the defender stronger, the second guy takes off that hit point, guys 3-10 take him down to 1 hit point, your last 3 guys fail and then you exit the game and play call to power 2. I like the idea of being able to commie-man-rush a city and take it over. "Sir! They're everywhere! We're outnumbered 10-1 and we just ran out of ammo!" Kinda like the movie Zulu.

Or, if you prefer, the US strategy for how to have the wholly inferior M4 Sherman tank defeat the much more powerful German Tiger panzers....

[Saurus]

Originally posted by Arrian
I've lost battles where I've had the battle odds advantage

In a game, After toying around with the world editor,
I in fact somwhow managed to lose a Tank to the defending archers of a city.
My Tank (Ok - it's strenght was down to some 90% due to the fact that a pack of animals, panthers I think, had attacked it).
Then my Tank aproached Montezumas city. The first defending archer perished whitout inflicting any damage to my tank.
The second archer was also destroyed but damaged my Tank down to some 14.9 in strengt.
The third defending archer destroyed my tank.
Sure I attacked from behind a river but the Tank should have had a definiate advantage combat-odd wise.
So yes, most defineately you still can lose even with ods in your favor.

[Diadem]

Just today I attacked a musketeer with a high level cavalry unit. Relative strengths were 24.4 for me and 14.4 for him. My guy died without getting a single hit in. The musketeer made 7 hits in a row. Really amazing.

But in general the cIV combat system works pretty good. Stronger units have a distinct advantage, but it's not decisive. And you have to be careful to avoid having your injured units massacred.

By the way, if you read the combat logs, the game gives the chances of victory for each combat. I assume those are correct. Easy way to check your combat calculator...

[swat-spas2]

Originally posted by Barchan


Or, if you prefer, the US strategy for how to have the wholly inferior M4 Sherman tank defeat the much more powerful German Tiger panzers....

Sherman was in production as a result of the much smaller and manueverable (read: lighter armed) Panzer III and IV. It was assumed based on the initial blitzs over Poland and France (and later Russia) that a war of manuever was what tanks were for, and for this design the Sherman was wholely adequate, even superior to many European designs (see Patton's drive across France). Its not part of the strategy to toss equipment around recklessly in war.
But if you outnumber the other guy 50000 to 1500.. you have some room for error.
It was still a piece of junk compared to the later war tanks it was pitted against without the battlefield modifications crews would make, particularily to the armor plating.

In game terms.. giving the German's the bonus vs tanks is about where the Tiger was compared to western allied tanks. Russia had some awesome stuff in the armor department though, much of what into to the T-34 design went into the German Panther tanks and the Tiger (sloped armor for example)

[Ellestar]

Re: Is combat hollow, or is the calculator wrong?

Originally posted by korn469
Here's the link for the combat calculator:

http://c4combat.narod.ru/c4c_v0_13.htm

Those instances don't worry me nearly as much as a Jaguar (strength 5) attacking a chariot (strength 4) and having 75% chance of victory. Or a knight (strength 10) attacking an unfortified musketman (strength 9) and having a 68% chance of victory. Or a knight with combat 1 (strength 11) attacking a knight without any combat modifiers, and the attacking knight also having a 68% chance of victory.

What have you observed, and is the combat calculator right? If it is, should combat function as it currently does?

Newer version v0.15 is here http://c4combat.narod.ru/c4c.htm
AFAIK calculator is fine with the exception of that noone knows how things like "1-2 first strikes" are calculated.

Even if your unit is just barely stronger, enemy needs one more hit to finish your unit (5 vs 6 hits). That's why there is a big jump in probability of success.

[korn469]

Even if your unit is just barely stronger, enemy needs one more hit to finish your unit (5 vs 6 hits). That's why there is a big jump in probability of success.

Yes I realize this, and that's exactly the problem I have with the combat engine. If two units of the same type faced off, one with 100 hit points and one with 99, then the unit with 100 hitpoints would have like around a 62% chance of victory. That's because once you move off of 50% odd you hit a jump point. This isn't the way the combat odds should work.

[Diadem]

Originally posted by Diadem
By the way, if you read the combat logs, the game gives the chances of victory for each combat. I assume those are correct. Easy way to check your combat calculator...

Disregard that remark.

The 'defender odds' shown in the combat logs are actually the odds for each single round, not the total odds of combat.

[Diadem]

I did some testing on the calculator. I haven't tested first strikes yet (the math involved is complex. I can do it but I'm too lazy right now :P), but it's correct for all non-first strike battles.

I spotted some minor rounding errors in the average hp left, but nothing serious.

Yes, combat odds can go up very steep. That is because stronger units do more damage, recieve less damage, and have a higher chance of doing damage each round of combat.

If your strength is lower because of being damaged, you will also have less hitpoints, making a bad situation worse.