TGs vs. Forts

[hydroxon]

That bad huh ... :(

[Danrelle]

LOL, Realest's comment reminds me of Soviet Propaganda.

[Zauper]

Orc: 9-point elite, 5-point spec, military_size = 100, Forts%+TG% = 30%, Building Efficiency 80%, off/def = 2/1.

Setup 1 (spam TG):
Forts% 0
TG% 30
Forts Effect 0.0000
TG Effect 0.2520

D-Specs 52.98
Elites 47.02
Total 100

Mod Def 264.91
Mod Off 529.81
Off/Def 2/1
Total Mod Def+Off 794.72

Setup 2 (Optimal):
Forts% 14
TG% 16
Forts Effect 0.1445
TG Effect 0.1613

D-Specs 47.73
Elites 52.27
Total 100

Mod Def 273.14
Mod Off 546.28
Off/Def 2/1
Total Mod Def+Off 819.42

FINALLY
Setup_2 / Setup_1 = 819.42 / 794.72 = 1.0311
I.e., Setup_2 is 3% stronger than Setup_1


Q.E.D.

Edit: I would normally include the Minor Protection self-spell into the calculation, but deliberately excluded it so that the noobs out there can easily verify my numbers :P

I'll try to make this short:
There are two phases of the game. The phase where defense matters, and the phase where defense doesn't matter. Defense only matters in war, in the two scenarios I outlined earlier. When I come oop, I'll come oop with 3-4000 elites and 200 defspecs. If you gave me 15% more DME from forts in exchange for 10% OME from TGs, and wanted a 'set' target defense of 1000, that would free up what, 30 troops? Obviously building TGs would give superior results. [FWIW, you also didn't include generals]

The reality is, the game doesn't exist in a static environment. Yes, sure, if I were setting a defense ratio, or potentially even a target defense, it is possible that forts would give me more total military power. The issue is that that isn't optimal, because it forces you to be static while you balance your tg/forts against eachother, and then your offense/defense afterwards. While you're trying to adjust things to give you the most military on your current landsize, your targets are growing and training defense.

The thing is, if I'm randoming, forts will never give me more than my current offense. Where your setups fall short is that there is nothing preventing the person in setup 1 from running 47.73 dspecs and 52.27 elites. In that scenario, they have 588.98 offense and 238.65 defense. That total -- 827.63 -- outperforms your setup 2. That is entirely my point. If I don't care about defense -- and I only care about defense where it either: makes me unhittable, or is the difference between causing me to be singled or doubled, doubled or tripled, tripled or quaded -- TGs provide me with better offense. This becomes particularly true if you're looking at more of a turtling oriented race and checking your total max offense (for doing things like randoming explorers in ghettos); because you overbuild elites, TGs will provide you with a higher max offense. And a higher total amount of power. QED.

The reason why you think forts are better is because you keep trying to force a ratio of offense to defense, or force a set amount of defense on it. There's really no reason to do that. I'll also point out that most people that set ratios that I know of, set it as offunits:defunits. If you look at the analysis above, you'd find that pure TGs are preferrable for that kind of ratio.

That's also beyond the simple point that an offense heavy strat will be more able to survive chains, and a forts heavy strat will find its defense dropping as it gets land in, which means if you rely on forts to hit your 'def target', you need to overbuild past it by at least 15%.

I'll also point out that my adjustment to your 'setup 1' causes it to go from being DT'd by setup 2 and only STing back to trading DTs -- which will favor setup 1 over setup 2 because setup 1 will lose less total ME due to having a more concentrated amount of TGs.

[Danrelle]

The reason why you think forts are better is because you keep trying to force a ratio of offense to defense, or force a set amount of defense on it. There's really no reason to do that.

So you understand that for everyone who wants to set a value for def and then look at increasing off, then adding some Forts to the build is better than having only TGs?

[Zauper]

So you understand that for everyone who wants to set a value for def and then look at increasing off, then adding some Forts to the build is better than having only TGs?

Yes, if people are looking to play in a static fashion and are trying to set their defense (though there is no point in setting your defense unless it makes you safe, or prevents a DT/TT/QT), forts will be superior. I said that in my first post, if you read it. Here, I'll even quote it for you:

Forts are good when they're good, and they're not when they're not. Generally speaking, if the forts don't make me safe (or pevent DTs/TTs), I'd rather have GS.

Generally speaking, offense is more valuable than defense, unless you're safe or could be safe.

It's pretty explicit.

p.s. that last line? It's about more than just forts. It's also about how you set your military up, as reflected in my response to hydro.

So, in short:
Yes, if you are playing in a inferior manner and setting your defense and then increasing offense, forts are better.

However, if you care about optimizing your offense, TGs are better. You agree?

[Ethan]

I'll try to make this short:
The reason why you think forts are better is because you keep trying to force a ratio of offense to defense, or force a set amount of defense on it. There's really no reason to do that.

Assume you are big. Assume you scout your opposing kingdom that you want to war. You find their fairly big orc is throwing down 98 OPA, but he is only 80% your size. (Noone else has more offense) If you keep your target def at 40 dpa, you get to double him, he can't double back. But you decide, in your infinite wisdom, that target defense is stupid, and you can get "better than optimal" by increasing your elites with pure TGs. You drop to 35 DPA (the ratio you gave in your post, rounded up) so you can have "more military".
Congrats - you just got doubled and are on your way to an easier chain *because you ignored your target defense!* See that there... the 40 dpa target? Maybe it's a super suicidal orc, packing 118 OPA, and you gotta keep 50 dpa to be safe. Maybe you are smaller, and don't care about being doubled, only avoiding a triple, so you keep 30 dpa. In every single case, *you have a target defense*, and going below it is bad.

I'm starting to think the problem here is remembering our algebra. In particular, that (strange, abstract) thing called a "variable". Target defense is one of those "variable" objects, and for this discussion we don't care *how* it is picked, just that it is. That's the part where strategy, game knowledge, kingdom plan, etc. all comes into play - picking the target defense "variable". If you want to claim there is no minimum defense *for each given situation* that you should be using... ya, that's where I give up. But I'm pretty sure the fact that you need some defense is sufficiently evident even for you.

That's also beyond the simple point that an offense heavy strat will be more able to survive chains, and a forts heavy strat will find its defense dropping as it gets land in, which means if you rely on forts to hit your 'def target', you need to overbuild past it by at least 15%.

Offense helps survive chains, right? So, I get chained with TG/Forts, you get chained with TGs. I started with more offense, we both lose almost all our defense... I still have more offense. Funny that.

As to the "def drop" when land comes in... ok, we don't get chained, land comes in from first wave. I lose some offense, some defense. You lose offense, and more offense. Now I've got *even more* offense to help survive the potential future chain. As well as having *even more* offense *right now* I can use to make big hits or chain them better.

We've both got troops in training, right? (Cause we lost troops on the first attack.) I'm in danger of falling below my minimum defense, so I build mostly def specs. You have enough defense, but need some more offense for the second round of hits... so you build... elites? O-specs? Either you pay more, or you lose space efficiency sooner, neither of which is good.

If you seriously need 15% extra forts to "protect" your 14% forts... you really need to learn how to play. Do you seriously expect to grow 100% larger *without building a single acre?* Training should normally start right as war launches (yay, dynamic observation!) so you have at least a *small* amount of flux in everything as soon as you make your first hit. Even on a max gain hit, you've got a grand total loss of 2.5% forts, which is small enough to compensate by adjusting your other actions (training) - especially when the adjustment is in your favor for other reasons too.

If you gave me 15% more DME from forts in exchange for 10% OME from TGs, and wanted a 'set' target defense of 1000, that would free up what, 30 troops? Obviously building TGs would give superior results.

So here, you grant a target defense (what target defense is unclear, but you explicitly grant it in theory), and grant an improved total (sum) ME by adding forts. (Which makes sense because of DBE). You've even got the 1.5 vs. 1 ratio about right - that's what you get for going from 30%:0% to 15%:15%. Then... you make up numbers, and claim "30 troops" without a shred of actual math. Then state that TGs are "obvious". Now, at least, I know why my old high-school math teacher said "obvious" was insulting. It is used to disguise the fact that there is no actual reason being presented, and the poster is just assuming the readers are too dumb to know the difference.

Ok - you want to claim the TGs are *obviously* better, even when playing by "our" rules? (Since I've shown yours are bunk anyway.) Go ahead - prove it. Give us the situation (or situations), and set some numbers. Then there are 2 questions, one important, one less so. Less important is "what is optimal?" - I can answer this with simple approximation. More important, but a judgment call... is "does this situation actually happen in the game?" Cause ya, I grant pure TGs is optimal if you run only land defense anyway... but I don't care about modeling that, cause it's a bad way to try and play.

So come up with a reasonable, typical scenario. (Being able to 32x yourself oop doesn't count. BTW, want to give me your loc on PM? I'd love to take your land from you in a quad, elites or no.) Post it up here. I'll find optimal, I'm sure others will as well, and you are welcome to do so on your own too. But if you want to attack our math, you need math to fight back with. Otherwise, stick to your "dynamic" arguments, as refuted above.

[kalorful]

I'm thinking the easier way is to assume
1. Only elites and dspecs are built (human attacker)
2. A set Dspec/acre (5-6) and a set draft rate (65%) you can then figure out how many knights you have
3. Find ratio of TG/Fort that maximises (off points + def points)? according to the above assumptions

I'm no maths major, and I know left out a lot of assumptions but this sounds as if its finally solvable, using excel (what's PERL? Finance major here, I deal with simple maths) :P
Is this approach valid?

[Bishop]

Its easily solvable, the whole point of Zaupers argument is that its useless to solve for it statically.

Offense helps survive chains, right? So, I get chained with TG/Forts, you get chained with TGs. I started with more offense, we both lose almost all our defense... I still have more offense. Funny that.

More offensive troops, not more offense. You also have some useless land with forts. Accurate that.

[Zauper]

Assume you are big. Assume you scout your opposing kingdom that you want to war. You find their fairly big orc is throwing down 98 OPA, but he is only 80% your size. (Noone else has more offense) If you keep your target def at 40 dpa, you get to double him, he can't double back. But you decide, in your infinite wisdom, that target defense is stupid, and you can get "better than optimal" by increasing your elites with pure TGs. You drop to 35 DPA (the ratio you gave in your post, rounded up) so you can have "more military".

Hi, you just made yourself look like even more of an idiot.
Here's my post again:

there is no point in setting your defense unless it makes you safe, or prevents a DT/TT/QT

I have said that numerous times. It is surprising to me that you still haven't seen it. The scenario you just posited creates a 'you're safe' argument. Or so you say at least -- I'm 5k acres, he's 4k acres, he has 392k off, I have 200k def -- they're going to make me doubletapable. I'm going to need to leave more than 40 DPA to be truly safe.

Congrats - you just got doubled and are on your way to an easier chain *because you ignored your target defense!* See that there... the 40 dpa target? Maybe it's a super suicidal orc, packing 118 OPA, and you gotta keep 50 dpa to be safe. Maybe you are smaller, and don't care about being doubled, only avoiding a triple, so you keep 30 dpa. In every single case, *you have a target defense*, and going below it is bad.

I'm starting to think the problem here is remembering our algebra. In particular, that (strange, abstract) thing called a "variable". Target defense is one of those "variable" objects, and for this discussion we don't care *how* it is picked, just that it is. That's the part where strategy, game knowledge, kingdom plan, etc. all comes into play - picking the target defense "variable". If you want to claim there is no minimum defense *for each given situation* that you should be using... ya, that's where I give up. But I'm pretty sure the fact that you need some defense is sufficiently evident even for you.

The issue is, your formulas, your buddies formulas, etc -- don't set it as a variable. You set it as a flat target. The reality is, most of the time, you're not sitting on the edge of a target defense. You need 30 DPA to avoid being tripled, and you have maybe 6.5 or 7 DSPA, with maybe 115% DME because of mp/pay. You're way over the target. Why do you need forts? What do they give you? You're not going to be going 'okay, I need to prepare for war with this kingdom, let me raze to put in some forts and retrain some of this D to offense!'. You may however say "Crap, with 5 more DPA I'd be safe attacking -- ok, let me speedbuild 15% forts".

Also, frankly -- when you're randoming, your defense is pretty much irrelevant. Most people I know run 15-20 DPA tops randoming until at least year 5/6.

Offense helps survive chains, right? So, I get chained with TG/Forts, you get chained with TGs. I started with more offense, we both lose almost all our defense... I still have more offense. Funny that.

No, you don't. Look at my response to hydro's scenario 1. I had more offense than he did. Not only that, but now even more of our land is barren! I guess we're lucky I had more offensive troops.

As to the "def drop" when land comes in... ok, we don't get chained, land comes in from first wave. I lose some offense, some defense. You lose offense, and more offense. Now I've got *even more* offense to help survive the potential future chain. As well as having *even more* offense *right now* I can use to make big hits or chain them better.

The drop from 30% TGs to 27% TGs and 15% TGs to 13.5% TGs and 13.5% forts -- which has a larger impact on your total military?

We've both got troops in training, right? (Cause we lost troops on the first attack.) I'm in danger of falling below my minimum defense, so I build mostly def specs. You have enough defense, but need some more offense for the second round of hits... so you build... elites? O-specs? Either you pay more, or you lose space efficiency sooner, neither of which is good.

Except I had more offense, so I don't have less offense? Additionally, I don't really need to train offense or defense to continue attacking, so I can afford to use my spec creds on D, train elites, stock gold for a dragon, whatever.

If you seriously need 15% extra forts to "protect" your 14% forts... you really need to learn how to play. Do you seriously expect to grow 100% larger *without building a single acre?* Training should normally start right as war launches (yay, dynamic observation!) so you have at least a *small* amount of flux in everything as soon as you make your first hit. Even on a max gain hit, you've got a grand total loss of 2.5% forts, which is small enough to compensate by adjusting your other actions (training) - especially when the adjustment is in your favor for other reasons too.

Sorry I didn't make this clear enough. When I say overbuild by 15%, if your target is 14% forts, you built 14*1.15 = 16.1% forts so you're still at ~14% after a hit.

So here, you grant a target defense (what target defense is unclear, but you explicitly grant it in theory), and grant an improved total (sum) ME by adding forts. (Which makes sense because of DBE). You've even got the 1.5 vs. 1 ratio about right - that's what you get for going from 30%:0% to 15%:15%. Then... you make up numbers, and claim "30 troops" without a shred of actual math. Then state that TGs are "obvious". Now, at least, I know why my old high-school math teacher said "obvious" was insulting. It is used to disguise the fact that there is no actual reason being presented, and the poster is just assuming the readers are too dumb to know the difference.

Really? Okay: here's some basic math for you. Target defense of 1000. 200 troops. You're right -- the forts saved me 24 troops, not 30. Silly me. On the other hand, I have 4000 elites. Do you really think that having pure TGs is a question here? 4000 elites, 200 defspecs. What produces more total military: retraining defspecs to elites and using forts, or pure TGs? I figured you were capable of basic math, but guess not.

Ok - you want to claim the TGs are *obviously* better, even when playing by "our" rules? (Since I've shown yours are bunk anyway.) Go ahead - prove it. Give us the situation (or situations), and set some numbers. Then there are 2 questions, one important, one less so. Less important is "what is optimal?" - I can answer this with simple approximation. More important, but a judgment call... is "does this situation actually happen in the game?" Cause ya, I grant pure TGs is optimal if you run only land defense anyway... but I don't care about modeling that, cause it's a bad way to try and play.

I already gave you two scenarios. My response to hydro's scenario one, and the oop scenario. And then I guess the scenario earlier in this post, where I pointed to a situation in which my naturally defense was already beyond my target point. Do you really spend days after scouting before you start hits to go into a conflict? Because that's a bad way to play.

So come up with a reasonable, typical scenario. (Being able to 32x yourself oop doesn't count. BTW, want to give me your loc on PM? I'd love to take your land from you in a quad, elites or no.) Post it up here. I'll find optimal, I'm sure others will as well, and you are welcome to do so on your own too. But if you want to attack our math, you need math to fight back with. Otherwise, stick to your "dynamic" arguments, as refuted above.

Give me your loc in PM, we'll be happy to war you in about 2-3 weeks.

[Tightpants]

Hmmm so 4000 elites 200 dspecs.

I'm all for balls out offense, but when the faery quads you to start a chain you might have problems. For that build tho you'd probably want a different ratio of TG/forts (i.e. 0% forts) so you're right! :)

Also, "utopia isn't static!" is just an excuse to be bad at math.

[Zauper]

Hmmm so 4000 elites 200 dspecs.

I'm all for balls out offense, but when the faery quads you to start a chain you might have problems. For that build tho you'd probably want a different ratio of TG/forts (i.e. 0% forts) so you're right! :)

Also, "utopia isn't static!" is just an excuse to be bad at math.

OOP? You don't need defense OOP. No one attacks active players OOP. (There are also better targets than someone with 200 dspecs at oop) That scenario is entirely posited at OOP.

Let me ask you this then: Why haven't the best kingdoms in uto history been run by statisticians if "utopia isn't static" is just an excuse to be bad at math? I can tell you that the monarch of the best KD in uto history agrees with me.

[Ethan]

More offensive troops, not more offense.

More offensive troops. More offense. Unless you claim BE shoots up so high your TGs are suddenly far better than before.

Lets simplify by saying BE won't change in 12 hours, and that no land is razed or otherwise "unbuilt" (just trads). Let's approximate with 30% TGs with 1,000 offense, vs. 15% TG 15% Forts with 1,030 offense. Assume all troops used on attacks, and so suffer no losses. Let's also simplify by ignoring any GBP ME effects, since they are the same on each province.
We both get chanied...
Since no land is unbuilt from the trads, we have the same built %s, just on much smaller land. Since BE is the same, we have the same OME mods from build as before. Since base ME is ignored, we have the same total OME as before. Since troops on attack don't die, we have the same offensive troops as before, as well.

Thus, you have 1,000 offense, and I have 1,030 offense. *Nothing changes on offense because of the chain!* My "useless" forts have nothing to do with this... all it means is I can stop building them now, and skew towards TGs (or GS, or farms), while you can't because you aren't building any already. So sure, 28 hours into the war my "useless" forts start providing a secondary benefit *beyond the 3% extra offense they let me keep when chained.*

[Zauper]

More offensive troops. More offense. Unless you claim BE shoots up so high your TGs are suddenly far better than before.

Lets simplify by saying BE won't change in 12 hours, and that no land is razed or otherwise "unbuilt" (just trads). Let's approximate with 30% TGs with 1,000 offense, vs. 15% TG 15% Forts with 1,030 offense. Assume all troops used on attacks, and so suffer no losses. Let's also simplify by ignoring any GBP ME effects, since they are the same on each province.
We both get chanied...
Since no land is unbuilt from the trads, we have the same built %s, just on much smaller land. Since BE is the same, we have the same OME mods from build as before. Since base ME is ignored, we have the same total OME as before. Since troops on attack don't die, we have the same offensive troops as before, as well.

Thus, you have 1,000 offense, and I have 1,030 offense. *Nothing changes on offense because of the chain!* My "useless" forts have nothing to do with this... all it means is I can stop building them now, and skew towards TGs (or GS, or farms), while you can't because you aren't building any already. So sure, 28 hours into the war my "useless" forts start providing a secondary benefit *beyond the 3% extra offense they let me keep when chained.*

Let's start with the assumption that you have more offense because you had forts. You don't.

Now, about the rest of it:

You're not accounting for % drop (which is substantially higher) due to land coming in post-chain.

[Tightpants]

Let's start with the assumption that you have more offense because you had forts. You don't.

Now, about the rest of it:

You're not accounting for % drop (which is substantially higher) due to land coming in post-chain.

You do have more offense with forts unless you are chained. Forts make it harder to chain you, and all your land is pretty much gonna be unbuilt after you're chained anyway. Saying "but after you're chained they're worthless) is a bad argument because you can rebuilt your land after people destroy it for you, if you want.

[kalorful]

Is Zauper saying 0% forts or x% of forts which is not practical to solve?