Switch Theme:

just some quick math for IG anti-tank  [RSS] Share on facebook Share on Twitter Submit to Reddit
»
Author Message
Advert


Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you:
  • No adverts like this in the forums anymore.
  • Times and dates in your local timezone.
  • Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance.
  • Email notifications for threads you want to watch closely.
  • Being a part of the oldest wargaming community on the net.
If you are already a member then feel free to login now.




Made in us
Nigel Stillman





Austin, TX

 schadenfreude wrote:
The op said the chart ignores scatter for the melticide stormies. That's a pretty big deal.


It's a very big deal because it puts all of the data into question and reveals a pretty obvious bias. If the OP is going to assume that melticide stormies are always going to arrive safely within 6" then he might as well assume that all of the Manticores always hit.
   
Made in us
Regular Dakkanaut





OP said "Given X assumptions, and measuring Y results, I come to Z conclusions."

If your assumptions match or closely resemble X, and you consider Y relevant, then Z is valuable to you. Enjoy this thread, and use it to improve your list making/target priority.

If your assumptions do not match/resemble X, or you consider Y irrelevant/insufficiently broad, then Z is useless to you. Quit complaining, do your own math based on your particular preferences, and quit highjacking threads that others find useful.


I am a grammar Nazi only because grammar democracy is ineffective. 
   
Made in nl
Lone Wolf Sentinel Pilot





Serving with the 197th

Oh, well. I like stormtroopers, thus this thread just reassures me that it's a good decision.

Overall Record W-L-D = 22-24-15
Bataviran 197th/222nd Catachan "Iron Wolves", arrogant, dedicated and ruthless!
Captain Detlev Vordon, regimental commander.
Colonel Vladimir Russki, regimental commander 222nd Catachan. 
   
Made in us
Nigel Stillman





Austin, TX

 Tuagh wrote:
OP said "Given X assumptions, and measuring Y results, I come to Z conclusions."

If your assumptions match or closely resemble X, and you consider Y relevant, then Z is valuable to you. Enjoy this thread, and use it to improve your list making/target priority.

If your assumptions do not match/resemble X, or you consider Y irrelevant/insufficiently broad, then Z is useless to you. Quit complaining, do your own math based on your particular preferences, and quit highjacking threads that others find useful.



Except that "X" assumption is a massively in favor of the Melta stormtroopers. You're right, I could do my own calculations, except only include when the melta Stormtroopers end up within 13" of a vehicle, making them useless.

Apparently pointing out flaws is complaining now, did you wake up on the wrong side of the bed?
   
Made in us
Regular Dakkanaut





 Vladsimpaler wrote:

Except that "X" assumption is a massively in favor of the Melta stormtroopers. You're right, I could do my own calculations, except only include when the melta Stormtroopers end up within 13" of a vehicle, making them useless.

Apparently pointing out flaws is complaining now, did you wake up on the wrong side of the bed?


The point, however, is that this assumption is equivalent to the assumptions made for all other units in the chart- It is assumed that all units are in range and have clear line of sight to the target with all weapons. The chart should probably have a "Meltacide outside 1/2 range" entry for comparison, and pointing this flaw out would have been productive. Saying the lack of scatter consideration (which, like glancing hits, is arguably outside the scope of the chart) "calls all of the data into question" (even data about units for which deepstrike is irrelevant) is hyperbole, and comparing it to a chart for units that land out of range is, at best, a straw-man argument.

The fact is deep striking units land on target or within range almost 2/3 of the time. Basing their entry off of the assumption that they will do so is at least as fair an assumption as assuming that your opponent will not have cover for their valuable AV units.


I am a grammar Nazi only because grammar democracy is ineffective. 
   
Made in us
Decrepit Dakkanaut





Vallejo, CA

TheLionOfTheForest wrote:This data would be more useful for the 5th Ed ruleset since there was no glancing to death.

It's actually because of this attitude that I did this math.

For some reason, now that you can glance vehicles to death, I'm getting this feeling that people are ignoring the fact that you can still kill things outright. Efficiency is not always a good substitute for effectiveness - there are always going to be some things that you need killed, and need killed now, not three turns from now. Plus, kind of the whole point of alpha striking is that losing units early in the game has an impact that is felt in every subsequent turn of the game. Killing that barge lord turn 1 is going to have a much bigger impact on the game than killing that barge lord on turn 6.

schadenfreude wrote:The op said the chart ignores scatter for the melticide stormies.

Right, but how do you account for this?

We know, straight away, that with airbone assault, the stormies have a .55 chance of landing dead on target. However, if you do your original placement close to the enemy vehicle, there are a LOT of possible combinations of directions and distances that the stormies can scatter in and still be in melta range. I have no practical way of calculating this.

In any case, if we give the stormies a hit rate of, say, .70, and we thus decrease their efficiency accordingly, melta stormies still come off up towards the top.

Tuagh wrote:The point, however, is that this assumption is equivalent to the assumptions made for all other units in the chart- It is assumed that all units are in range and have clear line of sight to the target with all weapons.

And this is also a part of it too, I suppose. Yes, stormies can scatter, but they can't get shot at (usually) before they arrive to shoot their meltaguns. Anything that starts on the board needs to have its effectiveness reduced by the chance that it gets killed before it gets a chance to fire, or that turn 1 is night fighting or something.

It feels like things would pretty quickly spiral out of control without keeping most things as control variables. I guess as Tuagh says, I leave it up to the interpretation of the reader how likely they think they're going to be in these various situations.


Your one-stop website for batreps, articles, and assorted goodies about the men of Folera: Foleran First Imperial Archives. Read Dakka's favorite narrative battle report series The Hand of the King. Also, check out my commission work, and my terrain.

Abstract Principles of 40k: Why game imbalance and list tailoring is good, and why tournaments are an absurd farce.

Read "The Geomides Affair", now on sale! No bolter porn. Not another inquisitor story. A book written by a dakkanought for dakkanoughts!
 
   
Made in gb
Rough Rider with Boomstick






 Ailaros wrote:
Efficiency is not always a good substitute for effectiveness



Definition for effectiveness:
power to be effective;
the quality of being able to bring about an effect.

Definition for effective:
Successful in producing a desired or intended result.

Definition for efficiency
The state or quality of being efficient: "greater energy efficiency".
An action designed to achieve this.

Definition for efficient
(esp. of a system or machine) Achieving maximum productivity with minimum wasted effort or expense.
(of a person) Working in a well-organized and competent way.

The effect? Dead thing. The more efficient something is the more effective it is. They are nearly the same word except except for the context they are used in. "It's effective" just means "it works", "it's efficient" means "it works" but implying "it works better". If anything you could argue that "effective" is as never as good a descriptor as "efficient".
   
Made in us
Blood Angel Terminator with Lightning Claws





New Jersey

 Ailaros wrote:
TheLionOfTheForest wrote:This data would be more useful for the 5th Ed ruleset since there was no glancing to death.

It's actually because of this attitude that I did this math.



All the more reason to do a cost analysis of glancing to death. A lot of us take mass quantities of ACs. You argue agains them due to the inability to kill outright, however maybe sometimes it's more efficient to drop 5 melta vets on something an nuke it and sometimes it's more efficiently use 3 ACs to just strip off the 2-3 HP. To remain objective and scientifically true we need all the data sets before we can make sweeping board conclusions.

   
Made in gb
Rough Rider with Boomstick






 TheLionOfTheForest wrote:
 Ailaros wrote:
TheLionOfTheForest wrote:This data would be more useful for the 5th Ed ruleset since there was no glancing to death.

It's actually because of this attitude that I did this math.



All the more reason to do a cost analysis of glancing to death. A lot of us take mass quantities of ACs. You argue agains them due to the inability to kill outright, however maybe sometimes it's more efficient to drop 5 melta vets on something an nuke it and sometimes it's more efficiently use 3 ACs to just strip off the 2-3 HP. To remain objective and scientifically true we need all the data sets before we can make sweeping board conclusions.


I've done AC and LC's elsewhere. ACs (on upgrade cost only) are significantly more effective than LCs. If we take the cost of the carrier then the LC becomes more effective as the cost of the carrier increases and when the carrier cost per AC/LC reach >70 the LC becomes more effective than the AC. My maths accounted for glancing and exploding, but it didn't consider immobilizing and weapon destroyed results, because gak that.

In short, veterans with ACs are as effective as killing armour as veterans with a LC. When the carriers are cheaper (IS, PCS, CCS) it is more effective to take a AC.

It also does better against almost all infantry targets except multi-wound T4 models. Against multi-would T5+ models the AC becomes better again.
   
Made in us
Douglas Bader






 Ailaros wrote:
there are always going to be some things that you need killed, and need killed now, not three turns from now.


This is a complete misunderstanding. HP removal in 6th is NOT the same as glancing to death in 5th, where it was a slow and unreliable process that took many turns to finally kill the target. Death by HP removal is MUCH faster, and, in many cases, actually FASTER than trying to kill through the damage chart. If you have a priority target that you need dead ASAP the best way of killing it is to set up an average of {target's HP} glances or better (with some margin for random chance) and consider an early "explodes" result a nice bonus.

And of course, once again, this is why the melta CCS is good: 90% hit rate with BiD, and then if three of the four melta guns glance or better (pretty likely) the target is dead, period. Likewise for other units that can consistently deliver a point efficient volume of glances or better, as they inflict death by HP removal in a single round of shooting no matter how badly you roll on the damage table.

There is no such thing as a hobby without politics. "Leave politics at the door" is itself a political statement, an endorsement of the status quo and an attempt to silence dissenting voices. 
   
Made in us
Nigel Stillman





Austin, TX

 Tuagh wrote:
 Vladsimpaler wrote:

Except that "X" assumption is a massively in favor of the Melta stormtroopers. You're right, I could do my own calculations, except only include when the melta Stormtroopers end up within 13" of a vehicle, making them useless.

Apparently pointing out flaws is complaining now, did you wake up on the wrong side of the bed?


The point, however, is that this assumption is equivalent to the assumptions made for all other units in the chart- It is assumed that all units are in range and have clear line of sight to the target with all weapons. The chart should probably have a "Meltacide outside 1/2 range" entry for comparison, and pointing this flaw out would have been productive. Saying the lack of scatter consideration (which, like glancing hits, is arguably outside the scope of the chart) "calls all of the data into question" (even data about units for which deepstrike is irrelevant) is hyperbole, and comparing it to a chart for units that land out of range is, at best, a straw-man argument.

How is it hyperbole if it actually does call of the data into question? If a person is creating a chart of statistical probability of several different units and they want to prove that they're right or that their unit is the best one, they can skew the data to attempt to "prove" their point.


The fact is deep striking units land on target or within range almost 2/3 of the time. Basing their entry off of the assumption that they will do so is at least as fair an assumption as assuming that your opponent will not have cover for their valuable AV units.

That's interesting, I'm aware that 1/3 of the time you'll score a "direct" hit, but melta stormtroopers ride a fine line between being in melta range and being out of range. I'm curious as to where the 2/3rds came from.

This message was edited 1 time. Last update was at 2012/11/21 23:16:33


 
   
Made in gb
Junior Officer with Laspistol




Manchester, UK

Stormtroopers get a reroll if they pick the right ability.

1/3 + (2/3 x 1/3) = 5/9

So nearly two thirds, without factoring in scatter distances that land you in a suitable location.

The Tvashtan 422nd "Fire Leopards" - Updated 19/03/11

"Never attribute to malice that which is adequately explained by stupidity." - Hanlon's Razor 
   
Made in us
Battleship Captain





NYC

 Trickstick wrote:
Stormtroopers get a reroll if they pick the right ability.

1/3 + (2/3 x 1/3) = 5/9

So nearly two thirds, without factoring in scatter distances that land you in a suitable location.


Not to mention, if placed centered adjacent to the vehicle at a 3" distance, in the 4/9 chance that it has to scatter, a direct forward or backward scatter of 3" or less will keep it in melta range.

If it scatters in a vector to either horizontal axis in addition, or instead of forward/backward, then it can scatter even further than 3" and still stay in melta range.

So really, yes, the data is skewed by that assumption, but it would be too abstract to make properly accurate.

Counting probabilities when dealing with scatter is too hard. Too many variables. It'd involve "Size of the vehicle, X and Y axis distance scattered, terrain tests, if the board edge can be scattered off of" etc.


Dakka member since 2012/01/09 16:44:06

Rick's Cards&Games 1000pt Tourney: 2nd
Legion's Winter Showdown 1850: 2nd Place
Snake Eyes 1000pt Mixed Doubles: 3rd Place

Elysian 105th Skylance W:37-L:3-D:6 in 6th Edition

The Captain does HH:Imperial Fists! Tale of Four Gamers Plog (New Batrep posted!) 
   
Made in us
Regular Dakkanaut





 Vladsimpaler wrote:

How is it hyperbole if it actually does call of the data into question? If a person is creating a chart of statistical probability of several different units and they want to prove that they're right or that their unit is the best one, they can skew the data to attempt to "prove" their point.


That's interesting, I'm aware that 1/3 of the time you'll score a "direct" hit, but melta stormtroopers ride a fine line between being in melta range and being out of range. I'm curious as to where the 2/3rds came from.


1- The absence of a valid data point does not invalidate other valid data points: The fact that information on "Meltacide outside 1/2" is absent does not invalidate comparisons of "Meltacide inside 1/2" to "X tank under ideal circumstances", much less invalidate comparisons of "X tank" to "Y tank".

2- 2/3, quick and dirty math: Assume 1/3 chance of direct hit (actually higher for storms, but this is for deepstrike in general) The other 2/3 of the time they will scatter. Assuming the initial model placed is 6" away from the target, approximately half of the potential scatter directions will result in moving either closer, or only moving laterally. High distances or mishaps are possible, but choosing how your meltas are positioned in your rings gives some range flexibility and also some large numbers will cause "hop-over". Offsetting factors approximately counter, therefore assume 50% of scatters leave the unit in range. Therefore: 1/3 direct hit, 1/3 scatter but stay in range, and the final 1/3 mishap or scatter out of range. Dirty math, but should be within a reasonable margin of error.

I am a grammar Nazi only because grammar democracy is ineffective. 
   
Made in us
Decrepit Dakkanaut





Vallejo, CA

Peregrine wrote:Death by HP removal is MUCH faster, and, in many cases, actually FASTER than trying to kill through the damage chart.

?

Death by causing a vehicle explodes result happens with one instance of the event. For most vehicles, the glance takes 3. 3 sounds like longer to me than 1, not shorter.

Certainly it is easier to glance than it is to pen or to wreck, but that doesn't make it faster over all. Watching the SW player use missile longfangs to glance our local necron AV13 player to death is a painful experience that, as far as I've seen, has ended only in his defeat. Blowing up lord barges with lascannons on the top of turn 1, on the other hand, produces a very, very different kind of game.

Furthermore, this always seems a strange argument to me, given that people very much prize instant death against targets. Dragging down AV13 vehicles with autocannons is like dragging down paladins with lasguns. Yes, it is perhaps more efficient to do things this way, but that doesn't mean that bringing weapons that kill them dead, now, aren't better.

Blaggard wrote:The effect? Dead thing. The more efficient something is the more effective it is. They are nearly the same word except except for the context they are used in. "It's effective" just means "it works", "it's efficient" means "it works" but implying "it works better". If anything you could argue that "effective" is as never as good a descriptor as "efficient".

In this case, what I mean by "effective" is "killing stuff", and by "efficient" I mean "killing stuff per point".

These definitions being rather important when time is one of the things being controlled for.


Your one-stop website for batreps, articles, and assorted goodies about the men of Folera: Foleran First Imperial Archives. Read Dakka's favorite narrative battle report series The Hand of the King. Also, check out my commission work, and my terrain.

Abstract Principles of 40k: Why game imbalance and list tailoring is good, and why tournaments are an absurd farce.

Read "The Geomides Affair", now on sale! No bolter porn. Not another inquisitor story. A book written by a dakkanought for dakkanoughts!
 
   
Made in gb
Rough Rider with Boomstick






You're shooting yourself in the foot and banking on random dice rolls to hopefully get a pen and explode?

If you spend 1k points on dudes with lascannons the only thing they'll kill more off than 1k points of dudes with AC's is AV14, but they aren't good at that. That's the only thing you are paying the premium for, another AV14 threat.
   
Made in us
Douglas Bader






 Ailaros wrote:
Death by causing a vehicle explodes result happens with one instance of the event. For most vehicles, the glance takes 3. 3 sounds like longer to me than 1, not shorter.


You're missing the point entirely. Death by the damage table only happens in one shot if you roll "explodes". The simple fact is that rolling "explodes" is far from guaranteed, so in the real world vehicles often die faster to HP removal than by the damage chart (where you can just roll 1s every time).

Anyway, the point of this is that by excluding death by HP removal you artificially inflate the value of AP 1/2 weapons that are more likely to roll "explodes", even if on average they kill slower than poor-AP weapons. For example:

AC HWS x2 vs. AV 12 gives you a 32% chance of killing AV 12 outright even ignoring the damage table, with an efficiency rating of 232. That is a better kill outright chance than everything but melta stormtroopers and Vendettas, and better efficiency than everything but the 5-man stormtroopers, in addition to the small chance of killing it even faster with an "explodes" result.

But what about BiD? It's a fair question since the HWS are probably within range of the CCS: chance to kill outright from HP removal alone goes up to 60%, while the efficiency improves to 125. IOW, the AC HWS with BiD gives you the best chance of killing AV 12 outright, and the best efficiency as well. Better than melta stormtroopers, better than Vendettas, better than everything. And that's through HP removal alone, add in the damage table (even at AP 4) and the numbers are even more one-sided.

Or, since the AC HWS get to start the game on the table, let's see what happens if it gets to fire on turn 1 and 2 while the stormtroopers arrive on turn 2: kill outright chance goes up to 78.8%, and efficiency goes up to 95. Again, without considering the chance of "explodes" results.


End result: if you want to efficiently remove AV 12 in a single round of shooting, buy mass autocannons.

Furthermore, this always seems a strange argument to me, given that people very much prize instant death against targets. Dragging down AV13 vehicles with autocannons is like dragging down paladins with lasguns. Yes, it is perhaps more efficient to do things this way, but that doesn't mean that bringing weapons that kill them dead, now, aren't better.


The difference is that instant death happens 100% of the time. If your STR 8 weapon hits a T4 target it dies instantly, you don't have to roll on a special "bonus wounds" table where only one of six results is instant death.

This message was edited 1 time. Last update was at 2012/11/22 01:37:01


There is no such thing as a hobby without politics. "Leave politics at the door" is itself a political statement, an endorsement of the status quo and an attempt to silence dissenting voices. 
   
Made in us
Decrepit Dakkanaut





Vallejo, CA

Blaggard wrote:You're shooting yourself in the foot and banking on random dice rolls to hopefully get a pen and explode?

What? We're playing a dice game. You're banking on random dice rolls to get glances too.

Peregrine wrote:AC HWS x2 vs. AV 12 gives you a 32% chance of killing AV 12 outright even ignoring the damage table, with an efficiency rating of 232.

And if you want to include glancing, then you have to also add the fact that lascannons can glance AV12 to death as well.

Which creates a methodology problem. How do you blend the glances and the explosions together into a single metric? After all, once a vehicle explodes, then it doesn't matter how many hull points it has left to be glanced off.

In the case of a lascannon HWS, we're talking about an average of 1 HP AND a 1 in 4 chance to wreck said AV12 vehicle outright.

The problem is that this reintroduces time into the equation. If you bump the scale to try and get rid of this problem, then you still have a complicated problem. After all, if we put 4x AC HWSs against 3x LC HWSs, we're comparing 4 HP with a .33 to wreck outright against 3 HP with a .75 to wreck outright. In this case, given that they both have enough firepower to peel off HP, then the lascannons would be superior because they have a much better chance of not NEEDING to glance it to death, freeing up the extra shots after the explosion to handle other targets.

Peregrine wrote:The difference is that instant death happens 100% of the time. If your STR 8 weapon hits a T4 target it dies instantly, you don't have to roll on a special "bonus wounds" table where only one of six results is instant death.

So? It happens less often, but the result is no less important. Rather, it's more so. Causing instant death to a land raider is rather more important than causing instant death to a single biker, so the odds being worse of it happening can easily mean it's just as big a deal.



This message was edited 1 time. Last update was at 2012/11/22 02:49:44


Your one-stop website for batreps, articles, and assorted goodies about the men of Folera: Foleran First Imperial Archives. Read Dakka's favorite narrative battle report series The Hand of the King. Also, check out my commission work, and my terrain.

Abstract Principles of 40k: Why game imbalance and list tailoring is good, and why tournaments are an absurd farce.

Read "The Geomides Affair", now on sale! No bolter porn. Not another inquisitor story. A book written by a dakkanought for dakkanoughts!
 
   
Made in us
Douglas Bader






 Ailaros wrote:
Which creates a methodology problem. How do you blend the glances and the explosions together into a single metric? After all, once a vehicle explodes, then it doesn't matter how many hull points it has left to be glanced off.


Easily. For a given unit (or combination of units) calculate the combined probability of getting an "explodes" result or removing 3 HP, since dead is dead regardless of method. Like everything else involving dice the only "difficulty" is in taking the extra time to do the additional math.

The problem is that this reintroduces time into the equation.


No it doesn't. The numbers I gave are for a single turn. It's still about killing outright with a given amount of shooting, but counting the fact that killing outright includes both "explodes" results and removing all of the target's HP.

If you bump the scale to try and get rid of this problem, then you still have a complicated problem. After all, if we put 4x AC HWSs against 3x LC HWSs, we're comparing 4 HP with a .33 to wreck outright against 3 HP with a .75 to wreck outright. In this case, given that they both have enough firepower to peel off HP, then the lascannons would be superior because they have a much better chance of not NEEDING to glance it to death, freeing up the extra shots after the explosion to handle other targets.


You're just asserting this based on intuition. Do the math, for both chance to kill and point efficiency.

So? It happens less often, but the result is no less important. Rather, it's more so. Causing instant death to a land raider is rather more important than causing instant death to a single biker, so the odds being worse of it happening can easily mean it's just as big a deal.


Of course it's a big deal, when it happens. It's very simple:

Instant death weapons are valuable because instant death is consistent, and this consistency makes them the fastest way to kill multi-wound targets.

Weapons that rely on the damage table are less valuable because "explodes" results are NOT consistent, and this inconsistency means that often the fastest and most efficient way of killing vehicles is to just remove all of their HP.

There isn't some inherent value to either, all that matters is how effective they are at getting the desired end result: a dead model.

There is no such thing as a hobby without politics. "Leave politics at the door" is itself a political statement, an endorsement of the status quo and an attempt to silence dissenting voices. 
   
Made in us
Regular Dakkanaut





The probability of a given unit destroying a target by any means is much more complicated to compute than you seem to be implying Peregrine.

The rough equation for destroying a 3 HP vehicle:

(%chance of by achieving at least 3 HP) + (%chance of achieving any 2 results of "Immobilized" without achieving at least 3 HP) + (%chance of achieving at least 1 result of "Explodes" without achieving either 3 HP or achieving 2 "Immobilized" results) = %chance of destroying 3 HP vehicle


This is entirely possible to compute, but is much more of a pain in the butt than I am personally willing to invest. I certainly don't blame Ailaros for generating slightly less accurate data by taking this shortcut.

I am a grammar Nazi only because grammar democracy is ineffective. 
   
Made in us
Douglas Bader






 Tuagh wrote:
This is entirely possible to compute, but is much more of a pain in the butt than I am personally willing to invest. I certainly don't blame Ailaros for generating slightly less accurate data by taking this shortcut.


Sure, use approximations when they're useful. There are two problems here, however:

1) Death by HP removal is a major factor and ignoring it skews the results significantly in favor of the desired conclusion (AP 1 is powerful, my melta stormtroopers are awesome). A good approximation can ignore minor details like death by double immobilized result (since, at best, it's a 1/36 chance from any two damage results), but once you start ignoring major factors (like death by HP removal, which out-performs most of the units against AV 12) the value of your numbers decreases significantly and the final result has more to do with your choice of approximation than any meaningful differences in the various options.

2) He stubbornly refuses to admit that death by HP removal is a major factor, and seems stuck in a 5th edition mindset where glancing to death can be safely ignored. It would be fine if he just compiled a chart of "explodes" probability and admitted clearly that it's only half of vehicle killing and therefore of limited value, but instead the whole thing is treated as "proof" that certain units are best.

There is no such thing as a hobby without politics. "Leave politics at the door" is itself a political statement, an endorsement of the status quo and an attempt to silence dissenting voices. 
   
Made in us
Calculating Commissar






 Peregrine wrote:
 Tuagh wrote:
This is entirely possible to compute, but is much more of a pain in the butt than I am personally willing to invest. I certainly don't blame Ailaros for generating slightly less accurate data by taking this shortcut.


Sure, use approximations when they're useful. There are two problems here, however:

1) Death by HP removal is a major factor and ignoring it skews the results significantly in favor of the desired conclusion (AP 1 is powerful, my melta stormtroopers are awesome). A good approximation can ignore minor details like death by double immobilized result (since, at best, it's a 1/36 chance from any two damage results), but once you start ignoring major factors (like death by HP removal, which out-performs most of the units against AV 12) the value of your numbers decreases significantly and the final result has more to do with your choice of approximation than any meaningful differences in the various options.

2) He stubbornly refuses to admit that death by HP removal is a major factor, and seems stuck in a 5th edition mindset where glancing to death can be safely ignored. It would be fine if he just compiled a chart of "explodes" probability and admitted clearly that it's only half of vehicle killing and therefore of limited value, but instead the whole thing is treated as "proof" that certain units are best.


While not the most accurate data in the world, that chart dose give a good idea of how some common anti-tank weapons stack up against each other.

40k: IG "The Poli-Aima 1st" ~3500pts (and various allies)
KHADOR
X-Wing (Empire Strong)
 Ouze wrote:
I can't wait to buy one of these, open the box, peek at the sprues, and then put it back in the box and store it unpainted for years.
 
   
Made in us
Decrepit Dakkanaut





Vallejo, CA

no, taugh is right. It really is a lot more complicated. For example, as mentioned, if a hit causes an explosion, then any glances or other penetrating hits are meaningless. The problem is that you have variables that are contingent on other variables, and not in a nice, clean way.

Furthermore, if a set of rolls produces some glances but doesn't kill the vehicle, then the squad in question has infinite inefficiency for killing something in a single round, which is what the scope of this study is looking at.

Also, I'd note that you have to be careful when you look at units assuming an infinite number of them. Two autocannon HWSs on average peel off 2 hull points in a single turn of shooting. The difference between a unit with 1 HP and a unit that is dead is pretty important here. Now, yes, you can talk about adding in more autocannons, but then you have to add in more AV12 vehicles as well. This quickly escapes the bounds of reality, wherein you have a finite number of autocannon teams and a finite number of AV12 vehicles, precisely because there are these breaking points in the metric, and what side of the breaks you fall on is really important.



This message was edited 1 time. Last update was at 2012/11/22 05:18:33


Your one-stop website for batreps, articles, and assorted goodies about the men of Folera: Foleran First Imperial Archives. Read Dakka's favorite narrative battle report series The Hand of the King. Also, check out my commission work, and my terrain.

Abstract Principles of 40k: Why game imbalance and list tailoring is good, and why tournaments are an absurd farce.

Read "The Geomides Affair", now on sale! No bolter porn. Not another inquisitor story. A book written by a dakkanought for dakkanoughts!
 
   
Made in us
Douglas Bader






 Ailaros wrote:
no, taugh is right. It really is a lot more complicated. For example, as mentioned, if a hit causes an explosion, then any glances or other penetrating hits are meaningless. The problem is that you have variables that are contingent on other variables, and not in a nice, clean way.


"Complicated" is not the same thing as "impossible". If you don't want to do the complete math, concede that it's an approximation at best and death by HP removal is a major factor that you're ignoring. There's nothing wrong with compiling a table of "explodes" probabilities, the problem is when you treat it as a complete answer to the question and assume that you can draw any meaningful conclusions about which units are best from it. And it just makes it worse when you repeat the same old 5th edition mindset about how the damage table is the only practical solution and insist that death by HP removal doesn't get the job done faster and more efficiently in many cases.

Furthermore, if a set of rolls produces some glances but doesn't kill the vehicle, then the squad in question has infinite inefficiency for killing something in a single round, which is what the scope of this study is looking at.


Just like if a set of rolls produces some penetrating hits but no "explodes" results the squad in question has infinite inefficiency. I really don't see what your point here is.

This message was edited 1 time. Last update was at 2012/11/22 05:20:11


There is no such thing as a hobby without politics. "Leave politics at the door" is itself a political statement, an endorsement of the status quo and an attempt to silence dissenting voices. 
   
Made in us
Decrepit Dakkanaut





Vallejo, CA

I don't have to concede anything. I made it plain what I was trying to accomplish in the OP. If you can create a chart that properly takes into effect interacting contingent variables, then feel free. I hope you're good with differential equations, as that's what it's going to take with this kind of a problem.

Basically, it's the same problem we were talking about with what makes it difficult to do the math for stormtroopers, except it's much more complex, as you're talking about a lot more die rolls.


Your one-stop website for batreps, articles, and assorted goodies about the men of Folera: Foleran First Imperial Archives. Read Dakka's favorite narrative battle report series The Hand of the King. Also, check out my commission work, and my terrain.

Abstract Principles of 40k: Why game imbalance and list tailoring is good, and why tournaments are an absurd farce.

Read "The Geomides Affair", now on sale! No bolter porn. Not another inquisitor story. A book written by a dakkanought for dakkanoughts!
 
   
Made in us
Douglas Bader






 Ailaros wrote:
I don't have to concede anything. I made it plain what I was trying to accomplish in the OP.


No you didn't. You said "kill outright", and killing outright includes death by HP removal. And you have continued to insist that death by HP removal isn't a relevant factor, death by HP removal isn't immediate, death by HP removal can't be faster, etc. This entire thread is one giant example of how to lie with statistics.

If you can create a chart that properly takes into effect interacting contingent variables, then feel free. I hope you're good with differential equations, as that's what it's going to take with this kind of a problem.


Why? I've made it very clear what my judge of effectiveness is: count HP removed, and treat 'explodes' as a nice bonus.

Basically, it's the same problem we were talking about with what makes it difficult to do the math for stormtroopers, except it's much more complex, as you're talking about a lot more die rolls.


It's not at all the same.

Accounting for HP removal is just extra math. Everything about the situation is known, and no approximations are needed.

Accounting for deep strike error depends heavily on the table setup (mishap danger, space for models to arrive, exact direction of scatter vs. size of vehicle, scattering near alternate targets you're happy to kill, etc) and is impossible to model as a pure theoretical question. You have to create a table map and pick an exact arrival point, and then your numbers will still only cover that exact setup.

There is no such thing as a hobby without politics. "Leave politics at the door" is itself a political statement, an endorsement of the status quo and an attempt to silence dissenting voices. 
   
Made in us
Calculating Commissar






 Peregrine wrote:

If you can create a chart that properly takes into effect interacting contingent variables, then feel free. I hope you're good with differential equations, as that's what it's going to take with this kind of a problem.


Why? I've made it very clear what my judge of effectiveness is: count HP removed, and treat 'explodes' as a nice bonus.



If that is how you judge effectiveness, why not make your own chart instead of attacking Ailaros for his?

40k: IG "The Poli-Aima 1st" ~3500pts (and various allies)
KHADOR
X-Wing (Empire Strong)
 Ouze wrote:
I can't wait to buy one of these, open the box, peek at the sprues, and then put it back in the box and store it unpainted for years.
 
   
Made in us
Decrepit Dakkanaut





Vallejo, CA

Peregrine wrote:I've made it very clear what my judge of effectiveness is: count HP removed, and treat 'explodes' as a nice bonus.

I think you said it best:

Peregrine wrote:If you don't want to do the complete math, concede that it's an approximation at best




Your one-stop website for batreps, articles, and assorted goodies about the men of Folera: Foleran First Imperial Archives. Read Dakka's favorite narrative battle report series The Hand of the King. Also, check out my commission work, and my terrain.

Abstract Principles of 40k: Why game imbalance and list tailoring is good, and why tournaments are an absurd farce.

Read "The Geomides Affair", now on sale! No bolter porn. Not another inquisitor story. A book written by a dakkanought for dakkanoughts!
 
   
Made in us
Battleship Captain





NYC

 Happygrunt wrote:
 Peregrine wrote:

If you can create a chart that properly takes into effect interacting contingent variables, then feel free. I hope you're good with differential equations, as that's what it's going to take with this kind of a problem.


Why? I've made it very clear what my judge of effectiveness is: count HP removed, and treat 'explodes' as a nice bonus.



If that is how you judge effectiveness, why not make your own chart instead of attacking Ailaros for his?


I agree.

If your math is better; prove it.

Ailaros balled out and put together a very comprehensive table. Not specific enough for you?

Then...too bad?

Dakka member since 2012/01/09 16:44:06

Rick's Cards&Games 1000pt Tourney: 2nd
Legion's Winter Showdown 1850: 2nd Place
Snake Eyes 1000pt Mixed Doubles: 3rd Place

Elysian 105th Skylance W:37-L:3-D:6 in 6th Edition

The Captain does HH:Imperial Fists! Tale of Four Gamers Plog (New Batrep posted!) 
   
Made in us
Regular Dakkanaut





Just as a side point, there is intrinsic value in achieving "explodes" results over wrecking a vehicle by HP removal- the explosion killing nearby and/or transported infantry is value-added, as is the removal of LOS blocking wreckage. I would rather fire at a unit in a crater rather than not fire at a unit huddled behind a wreck.


Personally (admittedly from a primarily Xenos standpoint) I prefer to go for an explosion. If my explosion-producing unit falls short of its goal it will still (usually) remove several HPs, allowing me to use minor glance-producing units to bat clean-up and knock that last HP off.

I am a grammar Nazi only because grammar democracy is ineffective. 
   
 
Forum Index » 40K General Discussion
Go to: