The Way of the Troll

[Hannibal]

OK, let's try the Giant War Elephant then.  Less than 3% more costly than the CG, no always-on-close issue, no restriction on command cards.  Leaving only the MC difference which is negligible since the CG can sprint for a CA.

Except you can't do that.  Saying Sprint = MC 5" simply isn't true.  Sprint soaks up valuable limited resources whereas MC 5" is always on, and so the GWE pays points for that MC 5".  IIRC, it's usually about a 10% increase to get a + or - MC from the 3.5" baseline.  So the GWE is functionally less than the CG if you just pair them up.

That being said, I will say that the CG vs the GWE is going to result in the GWE getting owned pretty hard.  The GWE is one of those big guy units (like trolls) and big guy units are usually a poorly optimized build in this game.  CG are notorious for being an uber-optimized build.  They are pricey, but every one of those 500 pts is spent on useful stuff that the game rewards.

I had a situation arise this weekend in a game against Brook where I had Centaurs that managed to get 2 Javelin throws before charging head on into Trolls.  There was additional pinching that occurred for both sides on following turns, but guess which unit was the sole survivor?  Would you believe it was the Trolls?  Well, it was.  "Anything" did not win that time around.

And two months ago, I rolled 20 dice with the Lord of Dusk and did 1 pt of damage.  But that doesn't mean the Lord of Dusk is crappy unit...  Anecdotal evidence is evidence only of an anecdote.

Put 400 pts of cavalry Trolls on 400 pts of cavalry Trolls.  Bet you its a coin-toss to see who wins.

Can we at least try to come up with a better rationale for something being balanced than that it has even odds to beat itself?

Okay, how about cavalry?  How about we pair up 400 pts of, say, Hawkshold Knights vs 400 pts of Trolls.  Oh gee, look Hawkshold Knights wipe the floor with them.  Hmm, usually don't big guys wear down cavalry after that initial hit? 

Or how about putting 400 pts of Troll to tank against, say, 350 pts of HE Elder Blades.  Oh gee, Trolls lose again.

Kevin, as someone who routinely says "Put X against Y.  X wins yet Y costs more, ergo X/Y needs tweaking." your argument comes off as silly, my friend.

Big guy units are not a good buy in this game, and Trolls are the paragons of suck.  They are not situationally useful: they suck.  They cannot breakthrough, they cannot tank, and they cannot even fight at their weight (or within 50 pts of it).  I'm done tilting at the Status Quo windmill here as far as changes go, but anyone who says they don't suck is just engaging in massive amounts of wishful thinking.

What Trolls have is staying power. Heavy cav has two or three green boxes whereas Trolls have seven. So once they blow their load on the charge they're pretty much done, the Trolls can keep going.

Gull, take out Hawkshold and Orcs.  Put Knights opposite the Trolls.  Roll the dice, assuming a charge.  Be really nice and take Bravery out of it.  Do it a couple of times.  See if a pattern develops.

[gornhorror]

Just remember the story of my Stag Cavalry taking out a unit of trolls in two turns.  Go Ravenwood!!!!!!!!!!  Gotta love 271 points bitchslappin' 406 points.  

Also, did anybody think about my idea of giving the trolls 6 hit dice?  Might be a good enough upgrade.  I'm in your corner on this one Hannibal.  I agree that the trolls need a boost.  Not sure if I agree with you about how to do it, however.  I would like for it to keep the green/red stat bar as this gives the unit character.  That being said, if this configuration makes the unit inherently weak, then some other part of the unit needs to be made strong to balance it out. 

So here are my suggestions:

1) Make the regeneration every turn
2) Give the unit 6 hit dice(my favorite)
3) Make two red health boxes green.

[gull2112]

Anecdotal evidence is evidence only of an anecdote.

That's golden. That's going to be my FB status today.

What Trolls have is staying power. Heavy cav has two or three green boxes whereas Trolls have seven. So once they blow their load on the charge they're pretty much done, the Trolls can keep going.

Gull, take out Hawkshold and Orcs.  Put Knights opposite the Trolls.  Roll the dice, assuming a charge.  Be really nice and take Bravery out of it.  Do it a couple of times.  See if a pattern develops.

Get off the Trolls v Cav wagon, we have no disagreement there. My point was that if you want to do the whole job of holing a line and then rolling it up from the inside...Trolls baby, you want Trolls.

[RushAss]

That being said, I will say that the CG vs the GWE is going to result in the GWE getting owned pretty hard

That's an entirely different discussion and I suppose I'm guilty of dragging this thread there.  Back to the core of the argument...

Big guy units are not a good buy in this game, and Trolls are the paragons of suck.  They are not situationally useful: they suck.  They cannot breakthrough, they cannot tank, and they cannot even fight at their weight (or within 50 pts of it).  I'm done tilting at the Status Quo windmill here as far as changes go, but anyone who says they don't suck is just engaging in massive amounts of wishful thinking.

Variance of opinion here.  I will say that before they could regenerate out of the red that Trolls sucked.  Now I just consider them to be lame, but not useless.  But even if I thought they still sucked it wouldn't bother me much because the Orcs really don't need them at least 90% of the time and that other ~10% is when there's some sort of special condition or terrain funkiness where a 5MC Large unit could be useful.

[Kevin]

How about we pair up 400 pts of, say, Hawkshold Knights vs 400 pts of Trolls.  Oh gee, look Hawkshold Knights wipe the floor with them.  Hmm, usually don't big guys wear down cavalry after that initial hit?  

Trolls (403 406) vs. Hawkshold Knights (413).  Assume the engagement happens on the Knights' turn with both charging.  No command actions or cards.

In each case I'm rouding damage off to the nearest tenth of a point.  I track tenths over multiple turns (relying on the "law of averages,"), but total damage is based on rounding to the nearest integer.



Turn 1 - Knights must pass fear check (13).  Assume they pass, but are still outreached.
Trolls take 5 dice + impact at 5s and 5s.   damage = 4.3
Knights take 5 dice at 2s and 5s.  damage = 1.4

Both are in the green.


Turn 2:  trolls regenerate 1 damage to 3.3
Trolls take 6 dice at 5s and 3s.  damage = 2.5 + 3.3 = 5.8
Knights take 5 dice at 3s and 4s.  damage = 1.7 + 1.4 = 3.1

Knights are in the yellow, Trolls are in the green.  Knights must pass a rout check (12).



Turn 3:  
Trolls take 5 dice at 5s and 3s.  damage =  2.1 + 5.8 = 6.9.  
Knights take 5 dice at 3s and 4s.  damage = 1.7 + 3.1 = 4.8

Knights are in the red; Trolls are in the red.  Both must pass rout checks (Knights need an 11; trolls need a 12 11.)


Turn 4:  Trolls regenerate 1 point, which puts them back into the green with 5.9 damage.
Trolls take 4 dice at 5s and 3s.  damage = 1.7 + 5.9 = 7.6
Knights take 5  dice at 3s and 4s.  damage =  1.7 + 4.8 = 6.5.  

Trolls are in the red, but have 5 boxes left.  Knights are even(ish) odds to be destroyed or to be in the red with one hit point left.

Assume the Trolls got unlucky and the Knights still have a hit point, and that both pass their rout checks (Knights need an 11; Trolls need a 12 11.)


Turn 5:
Trolls take 4 dice at 5s and 3s.  damage = 1.7 + 7.6 = 9.3
Knights take 3 dice at 3s and 4s.  damage = 1 + 6.5 = 7.5

Trolls are in the red with 5 boxes left; Knights have been destroyed.

--------------------------

Conclusion:

If both sides engage on the Knights' turn, the Trolls, as they currently stand, should win a majority of the time.

Courage:  Knights had to make a fear check (or they'd fall way behind on the damage curve), and had to make a rout check before the trolls did, and had to make a red rout check--at less good number, than the trolls.

If the engagement happent on the Trolls' turn, it's probably closer to even odds.
---------------

Yes the Trolls will probably get slaughtered if the Knights play a good red card on their charge turn and the Trolls don't "parry" it with a good blue card, but that's pretty universal.  (The relative dearth of Blue Cards in the Orc deck is a universal weakness, but if you have some and this fight is coming, save them for the Trolls!) If you give the Knights free Bravery then they may be slightly advantaged, but the same happens if you give the trolls a free card.

Under normal circumstances, the Trolls, as they currently stand, are better-than-even odds to defeat Hawkshold Knights.

Once again I fail to see the problem.

[RushAss]

The Trolls first courage check would be at an 11, not a 12.  Straight to the red, no yellow boxes.

[gull2112]

And Trolls are 406, but lets not quibble with Kevin's thesis, Trolls wouldn't get totally pwned by the knights which is the critical point here. The only certainty would be seen if Marcus were playing the Knights, as they would break even if they had their courage circled, or if I were playing the Trolls as I always roll exactly what I need, except when its not important.

[Hannibal]

Trolls are in the red, but have 5 boxes left.  Knights are even(ish) odds to be destroyed or to be in the red with one hit point left.

Assume the Trolls got unlucky and the Knights still have a hit point, and that both pass their rout checks (Knights need an 11; Trolls need a 11.)

Whoa, there's where you lose me.  Assume that the Trolls make 2 Red checks?  Crazy.  That's like a 40% chance.  The Knights are at a 50/50 of passing the Red check.  I mean if you're going to assume both units always pass their Courage checks, functionally fearless, then that screws with the points costs and performance.

If you're assuming that, then it's equally likely to assume the Knights will do 6 pts of damage on turn 2 (since they do 5. bringing that first red check on Turn 2.  But it also lowers the damage inflicted to 1 pt by the Trolls on turn 3, which means no Red check for the Knights and another Red check for the Trolls which they will fail 60% of the time.

Really the fight comes down to that second turn of combat, since the Knights can't do 5.8 pts.  They'll do 5 or 6.  And it's fair to say that they'll do 6 points of damage 8 out of 10 times.  At which point, the Trolls will rout the next turn 6 out of 10 times.

It roughly works out that 72% of the time the Trolls rout on turn 3 or 4 with the Knights running them down.  About 21% of the time both run away, and about 7% of the time the Trolls don't run away.  Out of that, a little under half the time the Knights will do 6 pts of damage followed by the trolls running on turn 3 and being run down without putting the Knights into the Yellow.

I hardly call a 1-in-14 chance of winning against equal points being worth it.



PS:  I re-read my post and it came across as really snide and mean.  Sorry guys, it wasn't intentional at all.  Specifically, my part to kevin was meant to be lots of  and came out as lots of  .  My bad, sorry!

[Kevin]

No worries.  As you know, my ego has a 4 Toughness and a Fumble card.

Now regarding rout checks.   Let's try this OK again, slowly this time.

The  Knights  must  make  equally  bad  rout  checks  when-or-before  the  Trolls  do.  


To review the above example, looking at just the courage checks:

Turn 1 (No rout checks, but Knights need a 13 Fear check or their first attack will do about 2 less damage.)
Turn 2:  Knights need a 12.
Turn 3:  Knights need an 11; Trolls need an 11.
Turn 4:  Knights need an 11; Trolls need an 11.
Turn 5:  Knights are destroyed; Trolls need an 11.

So is there a good chance that Trolls will lose by blowing a rout check?  Of course!  But there's a greater chance that the Knights will lose by blowing a rout check.  (Unless you give the Knights Bravery, but then you should give the Trolls a command card and it should once again balance out.)  So factoring in rout checks, if anything, reinforces the basic thesis that Trolls are better-than-even odds to win a straight up fight.

As an aside, I don't think Trolls "pwn" Knights.  Do this battle 100 times, and Trolls will probably win 53.  But the question was whether Knights "wipe the floor" with Trolls.

This thread has shown that Trolls function like heavy cav. in disguise:  underpowered vs. heavy infantry & monsters, but good at taking down twerps.  Knights are better at the quick kill, but Trolls are more resistant to sniper fire (and slightly tougher for their cost)--and of course the Orcs don't have heavy cav, so it's Trolls or bust.

[Hannibal]

The  Knights  must  make  equally  bad  rout  checks  when-or-before  the  Trolls  do.

But that's the point:  they won't.  The Knights have a 46% chance of passing both the Yellow Check and the Red Check.  Very close to 50/50.

The Trolls by contrast have a 39% chance of passing two Red checks.  Well enough below 50/50 to assume that they'll be gone by the time they take that second Rout check.


More to the point, since units can't inflict fractional damage, this situation:

Turn 1 (No rout checks, but Knights need a 13 Fear check or their first attack will do about 2 less damage.)
Turn 2:  Knights need a 12.
Turn 3:  Knights need an 11; Trolls need an 11.
Turn 4:  Knights need an 11; Trolls need an 11.
Turn 5:  Knights are destroyed; Trolls need an 11.

Won't happen very often at all.  What will happen is the following:


Turn 1:  Fear check.  A 1 in 6 chance that the Knights fail, at which point the Trolls will probably win the overall fight.

Turn 2:  Knights take a Yellow check.  Trolls make a Red check 8 times out of 10.  It's fair to assume both will pass.

Turn 3a:  If Trolls are in the Red, the Knights will not have to take a red check but the Trolls will.  They will fail 60% of the time.

Turn 3b:  If Trolls are not in the Red, both units will take a check this turn.

Turn 4:  If the Trolls are still there, both units will take a Rout check this turn, with a decent chance both will run.


80% of the time, the Knights will get that 2nd pt on turn 2 which sets in motion a cascade of events where the best the Trolls can hope for is a draw, but the odds don't favor it.  When you factor in damage inflicted and odds of passing consecutive Rout checks, the Knights will win 72% of the time.  The Trolls will get a draw 21% of the time and win 7% of the time. 

Winning 1 time in 14 does not equal balanced in any way.  When you lose 7 times out of 10 against a unit of equal points and comparable role, then your unit is underpowered/overcosted.  These numbers support my original point:  the Trolls are neither good enough at inflicting damage nor taking a beating.  It tries to do both and fails at both.


I agree the Trolls are the Orc version of heavy cavalry, however, if you take heavy cavalry and put them against similarly costed heavy cavalry, the battle becomes close to a coin flip.  I remember awhile back folks were comparing the Dusk Lances to Hawk Knights to HE Knights.  When the Dusk Lances matched up with the Hawk Knights, the Dusk Lances won, (as would be expected for a unit that costs 10% more) but it was a very close fight.  Similarly, the HE Knights lost to the Hawk Knights, but it was a very close fight.

However, as shown here, the fight isn't even close at all.  The Knights win 70% of the time.  If it was in the 60/40 of 55/45 range, you can say that's just a situationally bad matchup.  But its not even close to that.

Sure, obviously, cavalry wants to get engaged with cheaper units and blow through them, but that's the case with all heavy damage units.  That's pretty much your desire in every aspect of the game:  get a favorable matchup in terms of expensive units fighting cheaper units and cheap units pinning down more expensive ones.  But to say that something is only balanced when it fights cheaper units says that its not balanced.  Because an expensive unit can usually fight a similarly costed unit of the same type to a draw.

Trolls can't.  They can't fight cavalry.  They can't fight heavy infantry.  They can't fight monsters.  They can't tank.  The only thing they can do is spend 400 pts to grind out a fight against a 200 pt unit.  That's not a situationally useful, that's sucking.

[elgin_j]

I ran some numbers through the combat simulator to see how they matched up.  I didn't include the regeneration (you can't), so make of it what you will but that would doubtless improve the stats in favour of the Trolls.  Unit 1 is Knights; unit 2 is Trolls - the figures below are the averages:

Total Trials 1000
Turn End   3.2
Knights Win 47.2%   
Trolls Win 38.1%
Both Rout/Die 14.7%
Knights Die 36.9%
Trolls Die 29.6%
Knights Rout 30.6%
Trolls Rout 52.4%   

Factoring in the regeneration I think the Trolls won't fare too badly.  They aren't perhaps competing at their cost expectation but this is not a simple win for the Knights.

[Hannibal]

I say this all respect to Zelc, but my results seldom matchup with his combat simulator.  When I test out units, I get out the pen and paper and annotate the results to see where the 'break points' are, then I compare them with some actual dice rolling and compare them to my game experience.  Its seldom that those results match up with the simulator. 

So maybe I'm way off base here, but my feeling is that the mathhammer of his simulator is missing something that is crucial in context.  Given my working with the formula and how it can produce stuff that doesn't accurately capture the context of the table top, I'm thinking the same thing is happening there.

(Actually, I think its a Schrodinger's thing.  A simulator will capture partial damage and average it out.  Real life doesn't do that though.  Something is or isn't.  Meaning that the matchup between the Troll and Knight is swinger than the math is capturing and that swing is going to go against the Trolls more often).

[lazyj]

How much money would it cost to put Kevin and Corey in a room full of Battleground cards for 4 hours and film the result? I'd be willing to contribute $5.   

What if we messed with some of the control variables that you are both using? Start with Courage: if I am a Hawk General who lets his Knights fight Trolls without the +3 Courage boost I deserve to watch them blow it. If I am the Orc General I can try very hard to find one of the Kill you Meself cards, but... How many draws do I realistically have before I really really need it?

Walking through it from the beginning with Trolls getting maximum amount of card draws, with both units piling towards each other in clear terrain and no other units available to engage or shoot:
Knights start in shooting phase
Trolls move 5" and draw 2
Hawk circle courage, move 5" and draw 3
Trolls charge 5" and draw 4

So now it's Knights with guaranteed Courage 16 vs Trolls with Courage 13 and, what a 33% chance of having the required check-passing card?

This is where I need to go more with Corey. Outside of the math exercise, in the game you are much more likely to have Bravery on your Knights than one of two courage cards for the Trolls. That's what really screws up the poor Trolls, who were not weighted to come out ahead in the rout checks even before courage modification enters the picture favoring the Hawks. So I think in gameplay terms it's even worse than Corey's analysis of how often Trolls will win their fight against Hawk Knights.

Which doesn't really contribute anything to this discussion on whether/what/how to "fix" Trolls. 

[Kevin]

I ran some numbers through the combat simulator to see how they matched up.  I didn't include the regeneration (you can't), so make of it what you will but that would doubtless improve the stats in favour of the Trolls.  Unit 1 is Knights; unit 2 is Trolls - the figures below are the averages:

Thanks for running the numbers!  I'd say Regeneration could be simulated by running the sim with the Trolls getting (a) 1 extra green and 2 extra red boxes. and (b) 2 green and 2 red.  Then do a 75/25 weighted average of the two.

So now it's Knights with guaranteed Courage 16 vs Trolls with Courage 13 and, what a 33% chance of having the required check-passing card?

46+% 36+% actually; 46+% 56+% if you start the trolls slightly farther back or oblique so that they get another draw before the 2nd turn of engagment.  But the key point here is that the Bravery costs Hawkshold a Command Action.  If you give the Trolls a single unanswered red or blue card instead of the CA that'll more-or-less either do or prevent a point of damage.  So if you're going to boost up the Knights' courage then either give the Trolls an extra green box or strip one off the Knights.

Now, it's true that Knights can mark Bravery and play a card every turn, while Trolls can just do the card.  And with Bravery marked the Knights are probably at least even odds to win the fight.  But in that case some other Orc-faction unit on the line is going to be having a very happy day as the Hawkshold player will be shy on cards.  It's just like how a Umenzi unit will outfight a 50-points pricier unit if you mark Faith Armor and throw a Hex every single turn in addition to playing cards.

Which doesn't really contribute anything to this discussion on whether/what/how to "fix" Trolls

The Trolls-vs.-Knights exercise is plenty useful!  If Trolls are even-ish odds to beat the same-MC-and-slightly-more-expensive Knights, it suggests that Trolls are priced more-or-less correctly, and therefore throwing any additional sugar cube at them will be counter-productive to game balance.  On the other hand, if the Trolls-vs.-Knights match does indeed heavily favor the Knights, then it suggests that Trolls are substantially underpowered and could perhaps use a little something.

Sadly, it seems like this discussion is slowly devolving into a debate between those whose conclusions are based on the evidence, and those whose evidence is based on their conclusion.

[GoIndy]

If I post a comment in this thread that in no way is germane to the discussion, is it trolling?

I will say this...I play orcs alot, and used to play them and dwarves exclusively.  I field dragons, giants, and hydras with impunity.  I've had double dragon builds, and double bombchucker builds.  I've never once, in any game, (even the 4,000 megagames we play), fielded a troll.  So, while I am willing to agree I have no evidence other than me looking at the card and going, "screw that", they seem lame to me.