HoMM III: Best level 1 creature

[Metathron]

You should add more wimfrits, Conflux is entirely capable of deploying Firebirds by the first week (second week latest), which means...

(and conflux' lack of a decent retaliation stealer )

Heehee!

Finally decided on the centaurs meself (Yes, I've been deciding all this time).

[Pol]

Pardon?

No worry, your objection is righty. The above example is calculated in one to one ratio. Why? Well, because question asked 'What is the best level 1 creature?'

It's none of plural.

Wait a second..

I think not, it's likely but it will not happen always, when other factors have something to add to or against. Two terms 'Tactics' and 'Speed Artifacts' crossed my mind.

My reason why I picked CC over Sprites is that they are better material to work with. To enchant in ordinary battles, CC can do more - theirs base number premise is better.

Don't know why other creatures are going to be mostly forgotten. Certainly Troglodytes ability can be critical. Gnolls are one of the best under theirs leader Drakon. They have a great use to the start and sometimes in the final too. Familiars ability can also decide some battles, by one more but lethal Magic Arrow.

[Banedon]

@JJ - I guess we just agree to disagree, there isn't much more to debate (other than Conflux imbalance, which I guess we all agree on)

[Muszka]

I've got curious. I made some tests today.

1) Firstly I tried the 'one on one' or/and 'many vs. many'.
I've only involved Centaur Captains, Halberdiers, and Sprites.
I worked with lvl 1 heroes, without seconday stats, with and without primary stats too, neutral-cursed ground.
I tried many variations and I got some big surprises.

1.1.1)Sprites vs. Centaur Captain with equal numbers.
Fight formula was wait with Sprites and than strike twice, than CCs strike, Sprites retaliate, Sprites strike, and so one.
The numbers were the follows: 1 Sprite vs. 1 CC, then 5, 10, 50, 100, 500, 1000.
Obviously the CCs won; with some unrelevant numbers CCs remained in 25 to 50%. No partial fight was won by Sprites.

1.1.2) Sprites vs. Halberdiers.
I haven't make the one with equal numbers, because the -3 speed doesn't count as Srites still strike twice, and so the +2 defense will make the losses even smaller.

1.1.3) Centaur Captains vs. Halberdiers

Here I tried two different ways.
(Same numbers 1,5,10,50,100,500,1000)
A) Fight formula: Wait with CCs, let Haldberdiers[Hs] come closer, than strike, (Hs retaliate), strike again (Hs retaliate), Hs strike, CCs retaliate, and so one.
Up to 30% of the CCs stacks survived.
I have to mention, that sometime in the 1 vs. 1 and 5 vs. 5 fight Hs won.

B) I simulated a rush-fight.(eg. with Haste)
Fight formula follows, CCs strike, Hs retal., Hs strike, CCs retal., and so on.
It seems this were the evenest fights of al. Mostly Hs won. Up to 20% of stacks survived, with an average of 10-15%.
Here I have to mention that in the 20% of fights CCs could win, but only in the cases of the stacks below of 100 creatures

1.2.1) Sprites vs. CCs, counting in the Growt.
(28/50 base growt means that if 28 is 100%, 50 is somewhere 178.5%), so resulted, 1 CC was fighting 2 Sprite, 5 vs. 9, 10 vs. 18, 50 vs. 90, 100 vs. 180, 500 vs. 900, 1000 vs. 1800.
I also tried two kind of fights.
A) Fight formula: the logical wait, wait, strike mode.
Between 60 and 80% of Sprite stacks survived.

B) The rush strike mode.
Fight formula: strike, enemy strike, and so.
The Sprite stacks survived in around 30%
Mention: the smaller partial fights (1 CC vs. 2 Sp and 5 CC vs. 9 Sp) may be winnable by CCs.

1.2.1) Sprite. vs Halberdies calculateing the growt in (same numbers as with CCs)
The results were much the same in both type of fights, only the losses of Sprites were greater with from 5 to 15%.


2) I played a map concentrating especially on Sprites usefulness and durability.
(I played WoG on Impossible difficulty. With Brissa, M sized map, Grail on)
I must say that (As Jolly Joker said) you can survive only with Sprites till the Firebirds/Phoenix/Sacred Phoenixes are ready, and not only.
I could flag all mines, and get to some other buildings. I hesitated only with some shooters , because I had no mass haste.
I survived the First two weeks with no losses, around 300 Sprites and 8 Firebirds (nothing else used).
I've won the map at the middle of second month with no problems, and with about 1200 Sprites in my army.

Notes and Mentions:
a)Sprites double-no retalialiaton strike is devastating because of their number.
b)mass haste, can reach almost anything even without brissa.
c) IF taken care of them, they can have a mediocre->excelent killing potential from begining till end.
=> their usefulness is maybe the best among lvl1 units.

d)from the mid game, when the mediocre and big fights begin to occur, the Sprites will be disapper without Resurrection. In all mid->big fights I would have lost half->all my Sprites if not for the Resurrection.
e)They may survive longer if you don't use them.
f)They don't like sieges.
=>their durability is the lowest of all troops without resurrection.


At end I must admit, that I'm surprized, I never thought that they can kill Centaur Captains and/or Halberdiers face to face even with their growt. So their growt trully makes them strong, but without ressurection, they cannot be used effectively after mid game.
So my conclusion is that Jolly Joker was right in some more points.

[UndeadHalfOrc]

Summary of changes to Level 1 units in my personal patch
viewtopic.php?t=7079

Increased gold cost of Pikeman from 60 to 65
Decreased gold cost of Centaur Captain from 90 to 85
Increased basic growth of Gremlin/Master Gremlin from 16 to 18 (total
Increased basic growth of Imp/Familiar from 15 to 19 (total 38+8=46)
Decreased gold cost of Imp from 50 to 30
Decreased gold cost of Familiar from 60 to 40
Increased basic growth of Skeleton/Skeleton Warrior from 12 to 13 (total 26+6=32)
Increased basic growth of Troglodyte/Infernal Troglodyte from 14 to 15 (total 30+7=37)
Increased basic growth of Goblin/Hobgoblin from 15 to 17 (total 34+8=42)
Increased basic growth of Gnoll/Gnoll Marauder from 12 to 14 (total 28+6=34)
Increased gold cost of Gnoll from 50 to 55
Decreased wood cost of Upgraded Gnoll Hut from 10 to 5
Increased adventure map stack size of Peasant from 20-50 to 40-70
Decreased basic growth of Pixie/Sprite from 20 to 18
Increased gold cost of Sprite from 30 to 35

[Muszka]

Can your patch co-work with WoG?

[Jolly Joker]

@ Muszka:
Thanks
@ Banedon:
You can disagree with me as much as you like; it doesn't change a thing in the game.
@ Pol

1 on 1 without making the simple speed adjustment (faster unit waits and starts with 2 attacks in a row) except for units having at least speed 11 (both are big), 12 (one is big) and 13 (both are small) because they never come 1 on 1 - they have a different growth, at least in the lower ranks. Additionally, virtually 1 on 1 makes the Wraith better than a lot of higher level creatures and the Troll a monster of a unit which doesn't make sense. So fights against each other must include a certain number.

[Pol]

@JJ
Yes, that why my example was played 100:100.

As long as in question you are asked about 'creature', you should rate them individually. (Even if I think that plural was only omitted, but maybe not - considering one to one is more fair in some circumstances. )

Wraith, Troll...

You forget a Vampire Lord!

=>their durability is the lowest of all troops without resurrection.

You are looking on that from the wrong side, as I expect that you were fighting against AI. If you would play against human, from the exactly same reason, you will found yourself very uneasy in maintaining sprites in sheer numbers.

[Jolly Joker]

@ Pol
Pol, 1 Wraith beats 1 Vampire Lord with ease.
Pinning creatures 100:100 makes no sense when they come in different quantities. You can either measure them 1 on 1 or production against production. 100:100 (which is a mass comparison) is unlikely in game terms. You simply won't ever see 100 Sprites fighting 100 Centaur Caps. Moreover it devaluates speed and so on.
And, no, I'm not talking about the AI here. And I don't need to maintain Sprites in high quantities if they allow me to expand rapidly in week 1 on EVERY difficulty level, setting resources free to equip another hero on not-impossible levels.

[Pol]

You, hmmm, missing the point.

In the question in this topic you are asked about one creature, you did not contributed with your answer on it here. Instead your reformulated it (and not only you) and answered your new question. That could be a dangerous habit, mostly in some other pools.

I wasn't comparing VL against Wr. That wasn't my point. Better word would be along..

Of course, that here are maps when you will not met/use 'full week' production. It may be less or more, this is depending on the purpose of such scenario.

[Jolly Joker]

Pol, if this poll would be about making 100:100 comparisons there would be no need for a poll; you could simply post the results.

The poll simply says "best level 1 creature"; no explanation. My first post stated that "best" could be seen in different ways und was obviously seen in different ways, because it depends on what you define as "best".

And whether there are maps or not where you won't see no full population or whatever is completely beside the point: I made a map myself where you are supposed to beat everything with Walking Dead (not even Zombies), from low levels up to Titans and a full-blown Necro army within 3 months. If you do that, Walking Dead don't become strongest creature just because a mapmaker decided to give them special treatment.

Again, the question is open what best means. I'm rather sure that a battle 100:100 for level 1 creatures is rather inconclusive and misleading.

[Pol]

I'm very sure that is very inspiring. As it will leads anyone back to the roots.

Sorry, I didn't noticed your first 'best' statement. So, it's beside the point now.

Anyway I'm for stats best unit!

[Jolly Joker]

That would be the Centaur Captain. For a SINGLE unit.
If you factor the different growth in - which IS a factor quite obviously in the game (not for level 6, though) - it will be either Sprites or Skeletons depending on what you prefare.

Of course you can consider ALL factors and pick then - which seems to make a lot of sense, if you don't want to pick Wraith as best level 3.

It would be easier if the darn Conflux wasn't included because it's so far off with everything.

[Gaidal Cain]

@ Pol
Pol, 1 Wraith beats 1 Vampire Lord with ease.
Pinning creatures 100:100 makes no sense when they come in different quantities. You can either measure them 1 on 1 or production against production.

I don't think comparing creature vs creature that way has any merit to determine what the best creature is in a normal game. Apart from bias in the numbers (since at some points money, not availability, might be the limiting factor, and you also build different creatures at different stages, etc.), it also gives results like Lizard Warriors being better than Marksmen - and yet no one would ever pick the warriors if given the choice between them.

[Pitsu]

it also gives results like Lizard Warriors being better than Marksmen - and yet no one would ever pick the warriors if given the choice between them.

Nemo my name for evermore...

[UndeadHalfOrc]

it also gives results like Lizard Warriors being better than Marksmen - and yet no one would ever pick the warriors if given the choice between them.

Given 2.2 SOD data, I certainly would pick Lizard Warriors ANYDAY over Marksmen. Marksmen simply DON'T survive in any fight other than vs neutral stacks. Lizard Warrior's 2-5 damage is VERY, VERY good for a level 2 unit. And let's not mention the fact that they will do roughly 3 times more damage than Marksmen in melee.

With my current personal patch's data however, it could go either way.

[Muszka]

=>their durability is the lowest of all troops without resurrection.

You are looking on that from the wrong side, as I expect that you were fighting against AI. If you would play against human, from the exactly same reason, you will found yourself very uneasy in maintaining sprites in sheer numbers.

@Pol
I don't exactly understand.
How does that matter, with who I play, in maintaining Sprites numbers?
It does if I have only one enemy, and he's human, and of course, we have only one battle in the end.

And, as about the 1 vs. 1 and many vs. many you've read my earlier post. Altough it's not with exact numbers, but the result are from many battles and they pretty cover the one on one versions. And they also reflect the best creature based on stats and most abilities with and without growt.
But I agree JJ in that, there will be no such situation on a map, that 100 CCs to fight with 100 or 180 Sprites, without spells, on a cursed ground, especially with lvl 1 heroes.
And one more thing JJ was right in, is the fact that Sprites can give up to +3 speed on the adv.map over the other lvl 1 creatures. And that can mean +1 day walk within 5 days, but beacause theory is rarely equal with practice let's say that you can move +1 day further in a week. And a day can make big differences, as we all know.
And furthermore they have greater killing potencial than CCs.


@JJ
you're most welcome.
The truth is I made those tests, to have cover for my big mouth, but I failed, since I had only right in the three most unprobaby things.
The equal number fights, the fight between CCs and Halbs when CCs don't wait (irrelevant anyway), the Sprites are fragile and are missing in late game without resurrection.

[Pol]

..I don't exactly understand.

Matter by a poke.
(if that term doesn't exists I'm hereby proudly introducing it)
AI is not able to using 'Sprite Sweep' strategy(suffers losses for nothing, with excp. of cheat combat). From its point, clearly best are CC.

[Muszka]

..I don't exactly understand.

Matter by a poke.
(if that term doesn't exists I'm hereby proudly introducing it)
AI is not able to using 'Sprite Sweep' strategy(suffers losses for nothing, with excp. of cheat combat). From its point, clearly best are CC.

I've never said they are not.
I like them best.
I've voted for them.
I've only said, that Sprites are excelent, but they as fragile as good.

[Pol]

@Muska
Sure k, just you'd asked.

@JJ
You wouldn't believed how much I know, that is a true. And still it seems that this is the best AI among all Heroes series. (Thanks for extra point, stored for next time. )