I not only read your post, but also digested your post throughly along with others. I appreciate your inputs, but I like my conclusion better, especially now I've finished a game on Hard. :-P You can find my own analysis on Leadership vs. Attack dilemma on this board, which should explain my conclusion in some detail.

Dude, what he is saying, is that the 120 leadership you were offered at level 2 (which amounted to 10% army strength at level 2) only amounts to like .4% army strength when you are looking back at level 30. The +1 attack remains a 3.3% damage increase whether you are level 2 or level 30.

He is absolutely correct and when balancing your level-up picks over 30 levels, you should ABSOLUTELY favor leadership more strongly in the later levels. The benefits of +1 str/def/int diminish with every level as the amount of leadership offered increases proportionally.

Yes but the problem with that sort of approach is that during 28 levels your choice was weaker than another choice:
At level 30 you get 30*60 = 1680
At level 28 you get 28*60 = 1800
Sum = 3480

25000 - 3480 = 21520
At level 12 you get 12*60 = 720
That's is 3.35% ie already a better choice than attack +1.

So if we suppose all this math is correct, the solution you quote is less good during 28 levels than the one I quote (start choose leadership since level 12 instead of 14). And yes my solution would be less good during 26 levels than a solution starting using leadership since level 10... And so on.

You can argue that the end is harder but is it really? Additionally is being more efficient during 28 levels worthless, probably in this game. But for example if money was important in the game then saving more money during 28 levels would largely counter balance the weaker attack during the last two levels.

I won't enter in detail but if we look at the average efficiency in the whole game and not only at the end like you did, your approach will be a quite weaker choice.

One question, where is explained the math behind the 3.3%?

Moreover about this "no leadership" during 13 levels, it isn't only a weaker choice than another method during 28 levels but also it takes a quite huge approximation that only attack care.

If you don't take Leadership during 12 levels and starts only at 14 this will be
60*(2+4+6+8+10+12) = 60*14*3 = 2520 Leadership you don't have at the end.

So that's 10.1% less units if you use 25k as a final base, 10.1% lower life or 10.1% higher command cost for anything but attack.

Well that's a quite huge approximation. If a good player never get hit then I wonder why he couldn't manage quite well the alternative.

I don't say that the no leadership during 13 levels choice isn't ok, just that it has also negative points and that alternative like always Leadership certainly worth roughly the same.

EDIT: And even "worse" in fact the "take always Leadership" method offers 11.2% more of anything but attack than the "don't take Leadership during 13 levels" method. In fact with a minus of 2520, the Leadership base becomes 22480 giving in fact 11.2%. Yeah the same than 10.1% less but still it's in fact 11.2% more.

EDIT2: All of that also don't consider low level units use. When attack < defense the minus is quickly cap ie at -10. That means that -17 or -10 is the same and in this case the additional +7 in attack is worthless. Sure it's not the majority of the case but still quite significant if you enjoy low level units. One more point in favor of the "take always Leadership" method.

When attack < defense, it is not capped at -10. It is capped at -60, for a 2/3 reduction in damage, just like attack > defense is capped at +60.

+60 with 3.3% malus for each point and you get a 300% malus of damages with weaker units? Obviously that's not the right point.
Tipple damages (attack > defense) and Third damages (attack < defense) si clearly not symmetrical. Between 300% and 100% there's 200%. 200/3.3 = 60.

Between 100 and 33% there's 66%. 66/3.3 = 20. Lol ok 20 not 10. :grin:

Now the point if I remember well is that in fact it's more complicated when defense > attack, except that each point worth much less than 3.3% anyway. So it's still quite a big approximation to look only at attack and to forget the weak units point when you look at character level up management.

That's the wrong way to think about Attack and Defense. You can't simply subtract the bonuses. Between 100 and 33%, there is indeed a 66% difference. But you need Defense-Attack=60 in order to get that 66% reduction. Why? Because each point of defense gives a smaller % than the previous one. Having D-A=60 reduces damage to 33.3%, while D-A=59 gives 33.7%, only a 0.4% difference.

Now, you're probably thinking that means Defense is worth less the more you have. But that's completely the wrong way to think about it. In fact, this system is completely fair and values Defense and Attack equally. Like I said, you can't just take the DIFFERENCE between the percents. For example, with A-D=30, you get a 100% bonus to damage, doubling your damage. But if D-A=30 gave a 100% reduction, that means you get ZERO damage, making you invincible against that troop. Clearly, this is not equal and would make Defense much more useful than Attack.

In fact, both Defense and Attack are equally valued. Attack increases damage output by 3.3% per point, while Defense increase how long you live by 3.3%. Give a troop 30 more defense, it'll live 100% longer against all attacks. It's very simple. In fact, the whole business about having different equations if D<A and if D>A can be misleading. They both ALWAYS give 3.3% more damage or 3.3% more life, regardless of what the other troop's stats are.

If you look at it mathematically, Attack helps as follows: Damage = BaseDamage * (1+A/30). Now, if Defense equals Attack, then the troop should do BaseDamage, right? So you divide by (1 + D/30), and voila, you have Damage=BaseDamage. This also tells you what the formula for defense should be: Damage = BaseDamage / (1+D/30). That's why Defense gives less % difference the more you have: because it gives a factor of 1/(1+D/30). You can see for yourself that if you take the derivative, you get 1/(30*(1+D/30)^2). But the % difference doesn't matter, it's still giving you 3.3% more life with each point! In fact, if Defense modified the unit's life instead of the attacker's damage, you could have it simply increase hit points by 3.3% for the purposes of that attack rather than reduce damage and get the exact same overall effect.

I wonder from what you deduce I was comparing attack and defense but that wasn't the point. The point was to compare Leadership and Attack.

Anyway you just confirm what I was saying, that an attack point doesn't worth 3.3% damages but less because of the case of Defense > Attack. That's exactly what I was trying to explain.

That put in question quite a lot the reasoning arguing that until level 14 Leadership doesn't worth much to be choose. Not too mention some other points linked to a bigger stack like some special unit skills linked to stack size and not at all to damages.

I don't want say that some variation like not pick much Leadership during first levels isn't a good option but it's not clear at all that it's the best and the single best.

EDIT: And about attack vs defense that wasn't my point, but about it, yes it's obvious it's symmetrical from a math point of view as it's just a difference between the two. But people argue that attack is still better because with good tactics your attack is more important, you take care to attack first and even if enemy stack counter attack it is already reduced, you even take care to attack more than you get attacked.

well obviously i hardly get attacked myself by enemy stacks
since i use slow or w/e spells i have avaible to make SURE i attack first!

and with that
ATTACK>defense

the hit is worth a penny :P
it removes a portion of the stack and completely vaporsied it
so what use is defense?:)

I understand that your larger point was about Leadership. My comment wasn't related to that. I just wanted to explain thoroughly the math behind the two stats, since you were calculating Defense bonuses completely incorrectly (like the -66% = 20 Def). This is something that I see all the time whenever this mechanic is used, from KB to HOMM and even Warcraft 3. I wanted to explain it in detail for everyone's benefit, not just yours.

On the issue of Attack vs. Defense, I completely agree that Attack is better. Over the course of the game, you should be dealing a lot more damage than you're receiving, if you're playing it right. My point wasn't about which is BETTER, but rather what they actually DO in terms of the math of combat. People in these forums are always (correctly) saying attack is better than defense, but then some people start thinking that defense is mathematically less advantageous, because it gives a smaller percent in damage reduction than 3.3%, rather than it being due to game tactics. I see it all the time, so again, I'm just explaining for everyone's benefit.