HoMM III: Best level 3 creature

[Metathron]

While in practise no one goes with Elves alone

Who can speak with the kind of authority to say that no one does something? I may do exactly that sometimes, though my targets will be chosen with care.

[Bandobras Took]

My definition of "best" was:
Which one do I always use as a backbone when choosing that town type?

Orc Chieftain.

With the other towns, I'll vary, but Orc Chieftains are cheap and combine with a Barbarian's high attack for good damage relative to their cost.

[Muszka]

What version are you playing on? I've never seen neutrals split into 7 stacks.

Does it matter the version? If you play impossible and you attack 20 Tbirds or Minotaurs or Vamps. with 50 Gelves, They will surely split up in 7 stacks and they'll wait, that's also sure. (Let's say you split in a 22 and a 23 unit stack + 5*1 unit guards.) if in first round you kill two stacks, than Tbirds/MinK/Vamp move, and move again killing all the guards, and then you kill two more, that leaves three stack standing, (in the best 3,3,2) but they come first in the next round and I doubt that 23 Gelves can kill in melee 3 Birds/MinKs/Vamps.

Than I agree GC that nor 50 DragonF can win such fight alone, but I say that 50 RGriffins can. And we didn't calculated the Growt.

Anyway that's absurd, cos noone would attack such army, may they win or lust it would be stupidity.
And if that's only for the sake of the example, than I say that's just AN EXAMPLE, should we rate an unit only because of conclusion from that example, we are blind.

[Banedon]

Does it matter the version? If you play impossible and you attack 20 Tbirds or Minotaurs or Vamps. with 50 Gelves, They will surely split up in 7 stacks and they'll wait, that's also sure. (Let's say you split in a 22 and a 23 unit stack + 5*1 unit guards.) if in first round you kill two stacks, than Tbirds/MinK/Vamp move, and move again killing all the guards, and then you kill two more, that leaves three stack standing, (in the best 3,3,2) but they come first in the next round and I doubt that 23 Gelves can kill in melee 3 Birds/MinKs/Vamps.

How do you predict the amount of stacks the AI is going to split into?

Indeed, after a night's sleep I recall the AI splitting into 7 stacks, which would make it highly risky to attack 20 Thunderbirds with 50 Grand Elves. But even so, the Grand Elves stand a better chance than 50 Dragonflies As for Royal Griffins winning the fight, well maybe they can, but the losses would be heavy. Looking at the Minotaur calculations again, under the best case the Grand Elves can kill 15 Minotaurs with associated loss of 4 Elves, but Royal Griffins can't get anywhere close to that.

In the best possible scenario (without any special terrain), you have one Minotaur stack, and will get in 1 volley at half strength and 4 at full, killing some 14 Minotaurs, and have melee capabilities for another. (All this disregarding morale, which could screw up for the Elves).

Why are you getting only 5 shots off? If the Minotaurs go in a single stack, they'll take three turns before they can start hitting the Grand Elves, and then another four turns before they can start hitting the main stack.

[Muszka]

How do you predict the amount of stacks the AI is going to split into?

Though I don't know the exact numbers, I know that it's related with my armies strenght and his armies strength. I also assume that it has some randomness in it. The Weaker my army is(compared to his), the more stacks will be. If my army is much stronger than his, there is a random chance that neutral armies to come in 1, 2 or 3 stacks.

Indeed, after a night's sleep I recall the AI splitting into 7 stacks, which would make it highly risky to attack 20 Thunderbirds with 50 Grand Elves. But even so, the Grand Elves stand a better chance than 50 Dragonflies As for Royal Griffins winning the fight, well maybe they can, but the losses would be heavy. Looking at the Minotaur calculations again, under the best case the Grand Elves can kill 15 Minotaurs with associated loss of 4 Elves, but Royal Griffins can't get anywhere close to that.

I've mentioned that DF couldn't kill them. But I seriously doubt that 50 RGriffins won't kill 20 MinKings.

In the best possible scenario (without any special terrain), you have one Minotaur stack, and will get in 1 volley at half strength and 4 at full, killing some 14 Minotaurs, and have melee capabilities for another. (All this disregarding morale, which could screw up for the Elves).

Why are you getting only 5 shots off? If the Minotaurs go in a single stack, they'll take three turns before they can start hitting the Grand Elves, and then another four turns before they can start hitting the main stack.

Why would they attack in one stack? I don't think MinKs know the word 'pity'. I don't know if on other difficulties could happen, but on impossible they would come at least in 6 stacks. And that would be a situation much like with Tunderbirds, since even with 8 speed(*2 = 16) can reach a lot of areas.

PS: But IF we take that the MinKings come in 1 stack, even if the chance of that is 0%, then you are right, the GE would win.

[Pol]

The MK have speed eight and high probability for additional attack (always good morale). So they will reach GE from first to the third round. (battleground is 15x11 hexes)

Considering only offensive (A,dmg, sp) attributes is obviously very limited view. In any tough battle you will not survive only with blitz attack strategy. For equal forces, or stronger opponent here inevitable come a defend phase to consider, sometime you may start with it. Here are numerous strategies how to destroy GE along with theirs guards, not only spells(dd, area damage, stone skin + air shield, blind) but also Dragons So regarding fact that GE are excellent shooters, often picked for that reason they haven't highest combat value. They are specilized units, they have theirs pros and cons and both are huge. From that GE are predetermined to be everyone choice to kill - dangerous with dangerously low hp. A primavera candiate for quick death. (Same goes for Marksmen though)

Finally comparing shooters and hand-to-hand combat units is somewhat cumbersome.

How do you predict the amount of stacks the AI is going to split into?

Easy. As a matter of fact. You should too

[Jolly Joker]

Heroes is a game of chance, so consequently you'd have to do the following.
1) Define the number of creatures to be tested. The relevant time for testing a unit against neutrals would be weeks 2 and 3, difficulty hard (average difficulty). This would be akin to test with about 2 full populations for week 2 and 3 for week 3, avarage 2 and 1/2 weeks = 35 Elves (40 Flies, 32 or 33 Cerberi, 42 or 43 Griffins and so on)
2) Define the number of creatures to fight against: I'd suggest 1 1/2 weekly populations.
3) Define the split for the unit to be tested. This is a difficult point. Shooters are at a disadvantage here because against fast units they can't form fast enough and slow ones will have difficulties forming a defensive formation. This is certainly open to discussion on how to do it.
4) Tests would have to be conducted with neutrals split into 1-7 stacks. Final result would be the average result. A couple of tests might bring the result that average result is always the x-stack result which would make things simple.

For Elves the principle result is pretty clear: they will have a sharply defined "breaking point" since they are so vulnerable and their damage output is quartered in melee. That means, ultimately they'll either beat a stack losing a few at most or lose the fight completely. (Others will have more battles with a result of losing half their numbers or so.

[Gaidal Cain]

In the best possible scenario (without any special terrain), you have one Minotaur stack, and will get in 1 volley at half strength and 4 at full, killing some 14 Minotaurs, and have melee capabilities for another. (All this disregarding morale, which could screw up for the Elves).

Why are you getting only 5 shots off? If the Minotaurs go in a single stack, they'll take three turns before they can start hitting the Grand Elves, and then another four turns before they can start hitting the main stack.

Well, I thought minotaurs were speedier than they actually are

And though I did the numbers here, I'd still like to point out that such calculations don't have much point to them, unless you do them for every possible opponent (and several times to rule out freak morale incidents and the like). Minotaurs are grunts, which makes them ideal enemies for Elves but dangerous enemies for the Dragonflies. The result would be very different against Liches or Zealots, were Dragonflies would be the ones to with the better result.

[Muszka]

Minotaurs are grunts, which makes them ideal enemies for Elves but dangerous enemies for the Dragonflies.

I wouldn't say that MK are so ideal enemy for GE, since they have greater speed and awesome damage. Of course if you have mass slow things change (but that's another story)

The result would be very different against Liches or Zealots, were Dragonflies would be the ones to with the better result.

Zealots? I'm sure they have no melee penalty so DF won't stand so high chance against them too.

But since when we rate units after 'who dies slower' criterias?

[Banedon]

But since when we rate units after 'who dies slower' criterias?

No, but it's meant to show JJ that his misconception that Grand Elves work only with bodyguards stems from the pre-conceived notion that it's not possible to split Grand Elves to guard themselves. I would've thought it so obvious that everyone does it when they have no other blockers, and yet...

And though I did the numbers here, I'd still like to point out that such calculations don't have much point to them, unless you do them for every possible opponent (and several times to rule out freak morale incidents and the like). Minotaurs are grunts, which makes them ideal enemies for Elves but dangerous enemies for the Dragonflies. The result would be very different against Liches or Zealots, were Dragonflies would be the ones to with the better result.

Another point is that there're a lot more grunts than non-grunts. Among the lower-tier creatures (levels 1-5), there are few fast fliers (off the top of my head, Dragonflies are the only ones that can cross the battlefield in a turn) and relatively few shooters (I think every level includes more walkers than shooters). This just makes Grand Elves a lot more effective against the map than Dragonflies and Royal Griffins, because they have such a wider range of viability.

[Muszka]

Another point is that there're a lot more grunts than non-grunts. Among the lower-tier creatures (levels 1-5), there are few fast fliers (off the top of my head, Dragonflies are the only ones that can cross the battlefield in a turn) and relatively few shooters (I think every level includes more walkers than shooters). This just makes Grand Elves a lot more effective against the map than Dragonflies and Royal Griffins, because they have such a wider range of viability.

It's true that there are more grunts than non grunts in the low lvl troops, and that makes the GE with bodyguards an awesome unit in the early game. And since the Griffin building is off the build-chain to PoG, you rarely use them in the begining(a fact that you've mentioned), but starting from midgame they are a base unit of the Castle, 'cos they got quite a good growt. I'm pretty sure that they can be used in many ways, even more than GE (who can only attack)
Also I think that it wasn't a random thing that NWC put Griffins as emblem of the Empire.

[Bandobras Took]

Another point is that there're a lot more grunts than non-grunts. Among the lower-tier creatures (levels 1-5), there are few fast fliers (off the top of my head, Dragonflies are the only ones that can cross the battlefield in a turn) and relatively few shooters (I think every level includes more walkers than shooters).

Side note . . . Silver Pegasi are 1 slower than Dragonflies, but are a 2-hex creature, and so can cross the battlefield in one turn as well.

[UndeadHalfOrc]

I notice that nobody so far voted for the Iron Golems (although they were much more popular in the equivalent Heroes2 poll).

I agree, with SoD 3.2 data they are inferior, but I think this is the case mostly because of their very, very poor growth rate (only 12 per week when fully built, lowest of all towns). I buffed it to 14 per week in my personal balance patch. Their basic stats and cost are already excellent.

[Muszka]

I think Steel Golems were popular in H2 mostly because of the smaller battlefield and because of the fewer creature types. I don't think that even with a growt of 20 I would buy them, especially in the expensivest town.

[Gaidal Cain]

Stone Golems are too slow to use, and when you get to Iron, you usually have Nagas and thus not much of a need for them (until the enemy comes knocking on your door).

[Banedon]

It's true that there are more grunts than non grunts in the low lvl troops, and that makes the GE with bodyguards an awesome unit in the early game. And since the Griffin building is off the build-chain to PoG, you rarely use them in the begining(a fact that you've mentioned), but starting from midgame they are a base unit of the Castle, 'cos they got quite a good growt. I'm pretty sure that they can be used in many ways, even more than GE (who can only attack)
Also I think that it wasn't a random thing that NWC put Griffins as emblem of the Empire.

I think you're underestimating Castle. Castle forces are multivaried and versatile, they simply do not need Griffins:

1. Griffins fly - but so do Archangels.
2. Griffins are fast - but so are Archangels and Champions.
3. Griffins deal decent damage - but then so do Marksmen, Crusaders, Zealots, Champions and Archangels.
4. Griffins are expendable - but they still can't cross the map in a turn to block shooters.
5. Griffins are decent blocker units - but then why not use Halberdiers?

All in all I think being out of the build chain hurts the Griffins' cause. Castle never really requires Griffins; they can always improvise and replace Griffins with Crusaders / Pikemen, later upgrading to Champions and Archangels. There's no time to build Griffins and no need to. So though stats-speaking they're great units for level 3, there's simply no time to build them.

[Jolly Joker]

It's true that there are more grunts than non grunts in the low lvl troops, and that makes the GE with bodyguards an awesome unit in the early game. And since the Griffin building is off the build-chain to PoG, you rarely use them in the begining(a fact that you've mentioned), but starting from midgame they are a base unit of the Castle, 'cos they got quite a good growt. I'm pretty sure that they can be used in many ways, even more than GE (who can only attack)
Also I think that it wasn't a random thing that NWC put Griffins as emblem of the Empire.

I think you're underestimating Castle. Castle forces are multivaried and versatile, they simply do not need Griffins:

1. Griffins fly - but so do Archangels.
2. Griffins are fast - but so are Archangels and Champions.
3. Griffins deal decent damage - but then so do Marksmen, Crusaders, Zealots, Champions and Archangels.
4. Griffins are expendable - but they still can't cross the map in a turn to block shooters.
5. Griffins are decent blocker units - but then why not use Halberdiers?

All in all I think being out of the build chain hurts the Griffins' cause. Castle never really requires Griffins; they can always improvise and replace Griffins with Crusaders / Pikemen, later upgrading to Champions and Archangels. There's no time to build Griffins and no need to. So though stats-speaking they're great units for level 3, there's simply no time to build them.

I hate it to come in here again, but this is simply wrong. Your points 1-4 involve "Archangels", directly or indirectly; if you have the time to build Archangels you have the time to build Griffins. 1000 Gold + 5 Ore isn't much - and they are more than worth it.

[Banedon]

You're just wrong JollyJoker. You don't have time to build Royal Griffins because you're building every turn, aiming at the Portal of Glory. Of course you don't have to build the Royal Griffins and of course you don't have the time to build them. After you get the Portal of Glory you won't have any need for the Royal Griffins.

In the meantime, what do you mean by they're worth it? They're not worth it because you don't need them. What can you do with Royal Griffins you couldn't already do with your other units anyway?

[Pol]

You're just wrong JollyJoker.

Hehe, lets that not become your daily quote - it's not the best way how to start a comment, and looks like that would apply for any comment.

I expect that your experience is going out from difficulties to Hard. However on Impossible you may usually forget on Portal of Glory in short time outlook. Instead you will need to focus your attention toward griffins (and sometimes it can be only griffins) to be able to explore fast and hunt effectively a treasuries. Griffins are backbone (or better say additional backbone) of Castle Army. You never played campaings? - You would learned it here.

[ByteBandit]

I like Royal Griffins. I'll take them every time over anything listed here. I'll even take them over some Level 4 creatures.