First strikes

elisk

Mongol Warlord
Joined
Dec 26, 2008
Messages
64
I understand that first strikes allow your unit to attack before the enemy can, but how do they work and how can I use units with them, immune to them, and Drilled units effectively?



Second thread started today!
 
"First strike" is a misnomer. A unit with a first-strike advantage does not hit first, but instead is immune to damage from its opponent during the first combat-round (and the second and so on if it has more than one first-strike, until they have all been used).The result is that they give a higher chance of victory unless, of course, the opponent is immune. So their effective use is situational, whereas the combat line of promotions is always effective.
"First-strike chances" may or may not be converted to first-strikes; each "chance" has some likelihood of conversion to an actual "first-strike".
 
elisk, I'm going to be posting a thread with the details on first strikes soon. I'll link you to it when it's done.

Bushface's summary is very accurate too, of course.
 
Thanks, guys.
 
Here http://forums.civfanatics.com/showthread.php?t=137615 is a old article by Arathorn explaining first strikes, along with anything else you might want to know about combat.
It's old so some details (collateral damage) might be a bit off in numbers. All the basic principles still apply tho.
 
The little paradox is... FS is better promoted for units that are stronger in the battle. But the stronger unit is more likely to win each round anyhow.. making the effect of FS less useful.
 
The little paradox is... FS is better promoted for units that are stronger in the battle. But the stronger unit is more likely to win each round anyhow.. making the effect of FS less useful.

What? I don't what to say to counter that at the moment. But would that mean the opposite?
 
Well the idea is that first strikes are better employed when the first striker already has the edge. So people are correct when they say that first strikes will not usually tip the odds in the first striker's favour and this is the standard (and fair) criticism directed at first strike promotions. On high defense terrain for example, you could say the unit was likely to defend well anyway, but with FS it just makes them that much harder to dislodge.

Using them in attacking roles, drill units are not good as generalist city attack units for example, but once defenders are a bit damaged and the drill units start to get the ratios in their favour, they become almost as good as other units like CR units. The main advantage drill has in these sorts of attack roles is they earn XP faster and they also keep maximum HP a bit more often, meaning no time required to stop and heal. Using them in attack is definitely not the most straightforward way to use them, but it can be done if you pay attention to the details of combat.
 
The little paradox is... FS is better promoted for units that are stronger in the battle. But the stronger unit is more likely to win each round anyhow.. making the effect of FS less useful.

Basically:

1. Anti-zerg, can handle more weak attackers because more likely to emerge unscathed from each battle. So the advantage is not in the first battle odds, but in subsequent battle odds. Think being attacked when you have infantry and are protective, but he has a stack of 100 grenadiers (I have seen AI stacks that big, but not all grenadiers of course) versus your 5 defending infantry. You're going to kill more of them if your units have drill promos instead of combat (following CG of course).
2. Anti-siege (less collateral taken).
3. As experience gained is based upon A : D strength ratio (R), first strikes allow you to gain exp faster at better odds, as PoM alluded to.

The problem: Drill 1 SUCKS, so protective leaders are going to be the ones to take advantage of drill promos. The drill line needs to be improved, but especially Drill 1.
 
Well the idea is that first strikes are better employed when the first striker already has the edge.

What I was getting at is the actual die-roll for the combat round is in favour of the stronger unit (after promos calced). So, you are more likely to get your HITS in than the other guy, meaning first strikes sort of get demoted a bit as you already have a bias toward getting those hits.

Maybe it's a bit tricky to explain right. Hope that makes it more better.
 
What I was getting at is the actual die-roll for the combat round is in favour of the stronger unit (after promos calced). So, you are more likely to get your HITS in than the other guy, meaning first strikes sort of get demoted a bit as you already have a bias toward getting those hits.

Maybe it's a bit tricky to explain right. Hope that makes it more better.

Yes I know exactly what you mean. Basically you had a good chance of getting those hits in the first place, so having them as first strikes was not a big advantage. However, it can significantly increase the expected hitpoints after battle for the first striker, and this is one of the main points that shouldn't be forgotten (I'm not however suggesting you forget this).

But back to the point, the situations where first strikes can have a big influence are when the odds are in the first striker's favour but not ridiculously so. If you are killing warriors with infantry, for example, the damage per hit and the probability of hit are so high first strikes are worthless. Similarly, attacking the infantry with your warrior, giving it tons of first strikes also does very little for the warrior.

However there is somewhere in the middle (or maybe towards one side) where first strikes offer a huge advantage. I might look into first strikes a bit futher to try and decide what the best odds/damager_per_hits are to use them at. It might be something like opt for as many first strikes as it takes to kill the opponent. eg. if it takes 4 hits to kill that mace, then a Drill IV longbow will have good results, assuming it has the ratio edge already.

This is sort of a rushed reply so forgive me for not being entirely clear, but you seem to have a pretty good grasp of the nuances of first strikes to see what I'm saying.
 
First strikes also seem to raise the likelihood of doing damage to a tough defender.

Which has a higher chance of damaging a high str defender? Say you could have a drill II rifle or a C II rifle vs infantry. Which one has a better chance of causing damage? I'm not sold on this yet. The difference isn't enough for me to eyeball it, but my feeling is that at some point drill is better for damaging a top defender using less units...just don't know what point that is.
 
Say you have a strong military pump city that can get 3-4 promos on its own. I find my best bet with taking first strikes is when my base odds are somewhere in the 70-80% range unmodified. Then, I could take C3-4, and I'll probably have like 99.8% odds to win, but I'll probably get 1 XP from it. But if I take D3-4, then I might only have about 90% odds to win, but I'll usually get more XP from that battle.
 
First strikes also seem to raise the likelihood of doing damage to a tough defender.

Which has a higher chance of damaging a high str defender? Say you could have a drill II rifle or a C II rifle vs infantry. Which one has a better chance of causing damage? I'm not sold on this yet. The difference isn't enough for me to eyeball it, but my feeling is that at some point drill is better for damaging a top defender using less units...just don't know what point that is.

I'll assume no defense modifiers are involved. Also, unless you want me to do otherwise, I'll include the infantry's +25% bonus vs. gunpowder units.


C2 rifle vs. infantry.
14*(1 + 0.2) : 20*(1 + 0.25)
16.8 : 25
Victory odds = 10.2%
Expected damage to defender = 50.45HP


D2 rifle vs. infantry
14 : 20*(1 + 0.25)
14 : 25
Victory odds = 4.8%
Expected damage to defender = 40.76HP




This is a situation where D2 is clearly inferior to C2 but it does not surprise me. Using first strikes at a ratio disadvantage I don't think is usually going to give a better expected damage to the opponent. Especially here since we are looking at the weak D2 promotion. A comparison of C4 vs. D4 would be interesting, but I wouldn't be wasting a D4 unit at such bad odds. I'm not sure how you got that suspicion TMIT but I can't see any evidence for drill being better at low odds like that. Actually the only thing going for the D2 rifle is he'd get 7xp if he won; the C2 rifle would only get 5xp. :D
 
Say you have a strong military pump city that can get 3-4 promos on its own. I find my best bet with taking first strikes is when my base odds are somewhere in the 70-80% range unmodified. Then, I could take C3-4, and I'll probably have like 99.8% odds to win, but I'll probably get 1 XP from it. But if I take D3-4, then I might only have about 90% odds to win, but I'll usually get more XP from that battle.

You have experienced first hand the xp-milking ability of drill units. In particular drill IV units enjoy very easy xp earning at the sort of odds you are mentioning.

Assuming one knows how to calculate the ratio R = attacker to defender ratio... This ratio takes into account the defense modifiers and unit damage.... ie. R = Amodified * HPA / (Dmodified * HPD )

XP earned is based on R. When attacking, it's just int(4/R).

The table below is for a Drill IV unit with 4-7 first strikes attacking another unit. It is assumed the defender has no first strikes. It is also assumed both combatants have 100HP.

Code:
R	       XP	P(win)	E[XP]
(2,4]		1	1	1
(1.33,2]	2	0.9996	1.999
(1,1.33]	3	0.9602	2.881
(0.8,1]		4	0.8364	3.346
(0,67,0.8]	5	0.5839	2.919
(0.57,0.67]	6	0.3182	1.909
(0.5,0.57]	7	0.1538	1.076
(0.44,0.5]	8	0.0538	0.431
(0.4,0.44]	9	0.0345	0.311
(0,0.4]		10	0.0094	0.094

The values of P(win) and E[XP] are for the best case for each bracket i.e. the far right of each bracket. For example, attacking at R = 1 would give 83.64% odds of victory and 4XP if victorious. Taking into account the number of times the unit will die, on average the unit will earn 3.346XP.

Basically this table shows you there is a 3XP "sweet spot" for drill IV units attacking at odds near 96% (over 96% is usually 2XP). This is for values of R greater than 1 but less than or equal to 4/3.

I posted a similar look at Guerilla 3 and its withdraw odds here.
 
If you can arrange a fight in which your unit has a strength of twice that of your opponent or very slightly less, you'll get about 99% winning odds and gain 2XP from the win. If your odds go to or above 99.18% then the XP gain will be only 1. Odds between 77% and 50.01% mean 3XP, 50% down to about 32% bring 4XP if you win.

EDIT: PoM has given much more detail: use his figures, NOT mine.
 
If you can arrange a fight in which your unit has a strength of twice that of your opponent or very slightly less, you'll get about 99% winning odds and gain 2XP from the win. If your odds go to or above 99.18% then the XP gain will be only 1. Odds between 77% and 50.01% mean 3XP, 50% down to about 32% bring 4XP if you win.

Just clarifying, this is with 0 first strikes I assume...

And I think the 32% should be about 25% if I'm not mistaken.
 
I'll assume no defense modifiers are involved. Also, unless you want me to do otherwise, I'll include the infantry's +25% bonus vs. gunpowder units.


C2 rifle vs. infantry.
14*(1 + 0.2) : 20*(1 + 0.25)
16.8 : 25
Victory odds = 10.2%
Expected damage to defender = 50.45HP


D2 rifle vs. infantry
14 : 20*(1 + 0.25)
14 : 25
Victory odds = 4.8%
Expected damage to defender = 40.76HP




This is a situation where D2 is clearly inferior to C2 but it does not surprise me. Using first strikes at a ratio disadvantage I don't think is usually going to give a better expected damage to the opponent. Especially here since we are looking at the weak D2 promotion. A comparison of C4 vs. D4 would be interesting, but I wouldn't be wasting a D4 unit at such bad odds. I'm not sure how you got that suspicion TMIT but I can't see any evidence for drill being better at low odds like that. Actually the only thing going for the D2 rifle is he'd get 7xp if he won; the C2 rifle would only get 5xp. :D

Well it's true I don't run too much rifle vs infantry, but I figured it as valid a consideration as any. I think I asked the question wrong however, as what I'm looking for is consistency of damage too. Admittedly that late in the game consistency is attained by numbers anyway.

The time where I actually noticed this was with archers attacking cities defended by archers when screwing around with early rushes and seeing what that looked like :p. My body of evidence is far too low, but it looks like even C I drill I may be > than Drill II for this (of course cover would probably be best, but barring that...).
 
Top Bottom