jhrovii Posted August 31, 2011 Share Posted August 31, 2011 Hi! This thread is an attempt at mathhammer analysis. I am using the standard set by BoLS to quantify common units' potential and efficiency. I recommend reading their article. This thread is titled Planet Cueball because mathhammer is like Warhammer on a planet small enough to always be in range and never deal with cover. This world is confusing, because results that might correlate with list success might not reflect common perception. My goal is to figure out why this exists. Either the models are valuable in ways not quantifiable, or they are overvalued. I don't mean each metric to predict a unit's performance, though I do expect each metric to usefully compare units. Each units per-turn stat was multiplied by 5 to make the numbers more comprehensible, and give an idea of a unit's output throughout an entire game. Though this assumes a certain “best-case scenario,” I will say again the focus is on sheer output and efficiency- not predicted performance and strategy. Essentially, this is how I measured units: Dead MEQs by Shooting: Assuming no movement and within range. Dead MEQs by Close Combat: Assuming unit gets charge. Glanced Rhino (AV 11): Assuming no movement and within range. Dead Rhino (AV 11): “ “ Glanced Land Raider: “ “ Dead Land Raider: “ “ Dead Dreadknight by Shooting: “ “ I have chosen these units because I was considering them for my own list (I am missing many units). From red to violet is first to last place in offensive points efficiency. Point Efficiency = Cost of unit divided by models removed in 5 turns. The smaller the number, the less points are spent removing each enemy model. The numbers in each box is not the PE number: This number represents the quantity of models in question removed or damaged over 5 turns. Essentially what I'm comparing is similar units at similar jobs. Efficiency is useful for this: It means that some units will do their jobs better if allocated the same amount of points as another. For example- 270pts of Psyflemen (2 Dreads) versus 280pts of that 1 Purg squad with regards to glancing LR's; Purgs are more efficient even in a maxed squad. In fact, it's not even close: The dread PE is 45 while the Purgs is 31.11. Only the anti-tank Raven in Melting range comes close. http://img638.imageshack.us/img638/4939/mathammer.png I'll tell you what I find interesting: - The maxed out Purgation Squad with 4 Psycannons topped 5 of the 8 categories in point efficiency and 2nd in another. It is also the most expensive unit listed. - *sigh* And I just converted 3 AoBR Dreads... The Psyfleman performed disappointly poor in point efficiency. It performed above all others in penetrating rhinos, but came third in glancing them. 3rd from the bottom against the Dreadknight. Heavy tanks require heavy artillery, but frankly, Psyflemen can't even pop a LR, and are 5th in efficiency and 2nd-last in output at glancing it. Look there's no doubt what makes them special is their range and ability to move, but maybe there are better options for say, dropping that Russ that's pie-plating your troops from the board corner. -The 5-man Psycannon-wielding Purifiers you see nowadays are superbly efficient. Not only that, they're mobile, likely scoring, and good at everything. Dude, a Crowe list looks mighty appetizing with these guys. Overall, they're the second-best performing unit next to the Purg Squad. One thing to note is that because they have 2 base attacks, they get less of an efficiency bonus from the permanent-charge metric (that +1A is more an increase to 1 than it is to 2). Also, I forgot to account for cleansing flame. *edit*- I also forgot to factor in Hammerhand in relevant units. Ah well, I'm going to bed... -5-man Strike Psyback squads are terribly average. Footsloggers would be better I think. -The Raven looks bad in some areas, but I only care about the ability to get to heavy tanks camping in a corner and nuke them. Can it do that? ...To tell you the truth I don't know what this all means, I just had a day off today and time to kill. The BoLS guys seemed to like it and I had fun with it. Definitely the least surprising thing is the "math-hammeryness" of the Psyback Purifiers, and the most surprising thing is the performance of the Purgation squad. Of course, 10 purgs can't kill 4 rhinos a turn. 4 Pen hits is a good bet one will go though. But if nothing else, I'll have to rethink them. Not like the last codex, when I tore out their entry and burned it. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/ Share on other sites More sharing options...
grimz Posted August 31, 2011 Share Posted August 31, 2011 Nice analysis, What would be really good would be writing two similar lists and trying to test your theory regarding purgation and psynaughts. Like you have mentioned this is quite theoretical, and basically in a vaccuum; I don't see many lists or Bat Reps which use Purgation Squads so some empirical testing would be really awesome! Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2862811 Share on other sites More sharing options...
yperihitikos Posted August 31, 2011 Share Posted August 31, 2011 Nikephoros' mathhammer is made to compare lists at the same point level or equal units. If you want to compare similar or different units from point perspective you would get better comparable results if you calculate how many points you have to spend to achieve one kill. Have you ever tried any max psycannon foot list? Just try them. If you want mobility take only one full interceptor squad for support. You don't need other toys. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2862833 Share on other sites More sharing options...
jhrovii Posted August 31, 2011 Author Share Posted August 31, 2011 Nikephoros' mathhammer is made to compare lists at the same point level or equal units. If you want to compare similar or different units from point perspective you would get better comparable results if you calculate how many points you have to spend to achieve one kill. That's what I did. :) Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2862972 Share on other sites More sharing options...
Borinar Posted August 31, 2011 Share Posted August 31, 2011 Did you account for armor saves during the shooting phase. I dont see strikes taking down 18.5 marines per round it is more like 7 and thats with psybolts. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2862979 Share on other sites More sharing options...
jhrovii Posted August 31, 2011 Author Share Posted August 31, 2011 Did you account for armor saves during the shooting phase. I dont see strikes taking down 18.5 marines per round it is more like 7 and thats with psybolts. Per 5 turns. And yes. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2862989 Share on other sites More sharing options...
yperihitikos Posted August 31, 2011 Share Posted August 31, 2011 If I read your table correctly, 10 Purgators of 280pts can kill 26,1 marines. Instead of this, I suggest that you would have more comparable results if you had written that when you take these purgators every marine that you kill has cost you 10.73pts when 5 terminators cost 36.36pts to kill just one marine. Or you can calculate how many kills you can get for every point you spend on a unit. In this way you can plan better when you make a list and you can always compare units of different points at a glance. I prefer points per kill (in excel it was better) 12,43____ 9,06___ 17,36___ 26,44___ 51,11___ 77,97___ 25,84 11,76____ 9,84___ 11,36___ 28,74___ 55,56___ 84,75___ 25,77 13,18___ 14,05___ 15,45___ 25,76___ 75,56___ 114,86___ 31,48 11,68___ 12,66___ 11,08___ 17,81___ 43,33___ 66,10___ 23,49 27,84____16,27___ 11,54____ 7,63___ 45,00_______ - ___ 55,10 23,44___ 18,07___ 25,64___ 32,97___ 41,10_______ ?___ 52,63 28,08______ - ___ 24,12___ 33,06___ 32,54____ 69,49___ 42,71 10,73___ 14,32___ 10,61___ 16,09___ 31,11____ 47,06___ 19,44 36,36___ 9,41_______ -_______ -_______ -_______ -___ 210,53 24,18___ 6,27_______ -_______ -_______ -_______ -___ 140,00 Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863170 Share on other sites More sharing options...
jhrovii Posted August 31, 2011 Author Share Posted August 31, 2011 If I read your table correctly, 10 Purgators of 280pts can kill 26,1 marines. Instead of this, I suggest that you would have more comparable results if you had written that when you take these purgators every marine that you kill has cost you 10.73pts when 5 terminators cost 36.36pts to kill just one marine. Or you can calculate how many kills you can get for every point you spend on a unit. In this way you can plan better when you make a list and you can always compare units of different points at a glance. I prefer points per kill (in excel it was better) 12,43____ 9,06___ 17,36___ 26,44___ 51,11___ 77,97___ 25,84 11,76____ 9,84___ 11,36___ 28,74___ 55,56___ 84,75___ 25,77 13,18___ 14,05___ 15,45___ 25,76___ 75,56___ 114,86___ 31,48 11,68___ 12,66___ 11,08___ 17,81___ 43,33___ 66,10___ 23,49 27,84____16,27___ 11,54____ 7,63___ 45,00_______ - ___ 55,10 23,44___ 18,07___ 25,64___ 32,97___ 41,10_______ ?___ 52,63 28,08______ - ___ 24,12___ 33,06___ 32,54____ 69,49___ 42,71 10,73___ 14,32___ 10,61___ 16,09___ 31,11____ 47,06___ 19,44 36,36___ 9,41_______ -_______ -_______ -_______ -___ 210,53 24,18___ 6,27_______ -_______ -_______ -_______ -___ 140,00 This is my Points Efficiency number. Our data is exactly the same. For example, the first column: 10 Strikes (DH, 2 PC): PE of 12.43 10 Strikes (DH, 2 PC, PA): PE of 11.76 5 Strikes (DH, PC, Psyback): PE of 13.18 ... They're all the same. PE is the points cost of the unit divided by the digits in the box (models damaged/removed in 5 turns). You can do the math yourself from my table above. Again, for box 1A you can divide 230 by 18.5 and get 12.43. I liked seeing both data, so I represented the relative value of PE through color-coding. I should have posted those numbers because they are useful as well, you're right. *EDIT*- What I need to do is add a bunch more units and tests, factor in Hammerhand as a separate option to see the disparity (and Cleansing Flame where I missed it), and create tables like yours that list all the remaining data. Look, here's what I think we'll find. Once we flesh this out, we can then factor in movement. This should realign the order of efficiency back to favoring traditional anti-tank. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863182 Share on other sites More sharing options...
Fred Johnson the 3rd jr Posted August 31, 2011 Share Posted August 31, 2011 Hi! YO. Thanks for doing this man. Don't have time to take a serious look at the numbers, but this caught my eye: -5-man Strike Psyback squads are terribly average. Footsloggers would be better I think. If all, and I mean *all* you are concerned with is offense, sure maybe footsloggers are better. The thing about Rhinos / Razorbacks that your number crunching doesn't capture is that they are a good way to keep troops safe. Some games those smaller number of Strikers inside a transport will actually offensively outperform the other option of a squad of a couple more Strikers footslogging... simply because the sloggers will have been shot to pieces before they got to do much. In other games, where camping on objectives is important, troops inside a transport are often far more desirable. Similarly, take Paladins, who in a 5 man squad can take 5 non-ID wounds before losing any offensive power. Would this show up in the offensive data? No, it wouldn't, even though it is a great benefit to the squad and the army as a whole. Paladins in effect reduce the enemy's offensive power which gives you a net gain in power relative to them. I am certainly not knocking your numbers, or all the work put into it.... just elaborating on the differences between math on Planet Cueball and an actual game. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863378 Share on other sites More sharing options...
thade Posted August 31, 2011 Share Posted August 31, 2011 Don't fret about stuff you made that doesn't seem like it fits the math hammer mold. ^_^ Math hammer is very good for the cases where you have a task that needs be done on the table and you have one or more units that can potentially do it: numbers help you figure out which of your options will weather that task the best and/or if that unit will need help from another unit. Good way to allocate your resources quickly. It's terrible and misleading for list design. The system will show when a unit will perform poorly (in the implicit vacuum) where it is lined up against some (static and predictably upgraded) MEQ unit in a way that uniformly allows both units to bring their full payload to bear (at a perfect average performance rating). It will not take into account: the presence of other units; whether any of the units are at full strength; all possible wargear options across all possible third-party units; whether the units are mounted or have their transport nearby (even empty it's still a moving wall...one you can explode or waste shots on); how your dice rolling is doing; many other factors. So while your (insert unit that mathhammer doesn't favor here) may not look good in that idealistic one-vs-one vacuum, it may be very effective given your play-style or typical opponents. TL;DR: Math hammer is a great tool. Don't base every decision you make on it though. <3 Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863398 Share on other sites More sharing options...
jhrovii Posted August 31, 2011 Author Share Posted August 31, 2011 10 Purifiers (4 PC) = 280 pts 10 Purgators (4 PC) = 280 pts Purifiers are strictly better! Damn you math-hammer... :lol: You all make good points. Still, I like making tables. :lol: Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863529 Share on other sites More sharing options...
boreas Posted August 31, 2011 Share Posted August 31, 2011 Nice work... What I find interesting is the seeming lack of additionnal killing power given by psybolt ammo on SSGKs. The only thing that's significantly increased is Glacing Rhinos. But glancing/killing rhinos is something that lots of other units do very well (and anyways, the 10 ssGKs w/ 2 PC can kill 1 rhino per turn...). I like running a "naked" 10 ssGK w/PA because I perceive them as a cheap way to hold objectives while harassing enemy footsloggers (and denying DS in my zone). But, as of now, I'll switch them to 10 ssGKs w/ 2 PC... Weird how perceptions go... Phil Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863555 Share on other sites More sharing options...
Something Wycked Posted August 31, 2011 Share Posted August 31, 2011 Next task: add a chart that compares everything performing in less-than-ideal situations; shooting units are moving, units in assault are not getting the charge, etc. Get the bare baseline minimums going instead of the ultra-maximums :P Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863575 Share on other sites More sharing options...
BLACK BLΠFLY Posted August 31, 2011 Share Posted August 31, 2011 Thanks for sharing your results. The information is very useful. My comment on psyflemen is that my experience matches your results - they perform best when used to destroy transports. That is very good if you are a competitive tournament player since the meta is dominated by mechanized lists. If you can halt the rush of chimeras and rhinos that can go a long way towards winning a game. They are also a nightmare for mechanized dark eldar lists. Obviously they are not well suited for killing Marines due to the 3+ save. G :P Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863580 Share on other sites More sharing options...
thade Posted September 1, 2011 Share Posted September 1, 2011 10 Purifiers (4 PC) = 280 pts10 Purgators (4 PC) = 280 pts Purifiers are strictly better! Damn you math-hammer... You all make good points. Still, I like making tables. :D Strictly better at shooting through walls. Purgation Squads have more expensive Halberds, one less attack a piece, no Cleansing Flame, and can potentially break and run from shooting (whereas Purifiers are Fearless). Purifiers can be made into Troop units; Purgation squads have no such ability. Good? Yes. Strictly better? No. Math-hammer in a vacuum is only your friend if you understand it's in a vacuum. It's useful when deciding whether to charge a unit with your Purgation Squad and/or Purifier squad. It's not nearly as useful in army design. You don't know what your opponent will bring in all cases and - even if you do - you don't know what the table set up will be every time, or how your opponent will choose to deploy or move, or a million other factors. <3 Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863676 Share on other sites More sharing options...
Algesan Posted September 1, 2011 Share Posted September 1, 2011 Next task: add a chart that compares everything performing in less-than-ideal situations; shooting units are moving, units in assault are not getting the charge, etc. Get the bare baseline minimums going instead of the ultra-maximums :D This is not the point to the ratings. The idea is to not play "what if a butterfly flew by your die as it left your hand?" type of game, but to get a baseline measure of the raw power of the units. If you read the build up and examination of various 0-4 and 4-0 lists you will find that certain bits of advice are confirmed. If all other things are equal: MSU lists are the most killy, the ability to kill light mech is virtually required (to the point where, IIRC, being unable to potentially glance/penetrate less than 50 Rhinos per game was a losing list) and other common bits of advice on list building. It does not account for the fact that you are the uber general of 40k generals and I'm the stupidest of 40k geeks. A great player using a "sub-standard" list will probably beat a doofus using an "uber" list. It also doesn't cover dice, it does you no good at all to be able to kill X MEQ a turn when your opponent rolls 17 of 18 armor/cover saves in one turn. It also doesn't do you any good to be able to nail X Rhinos per game when your opponent spends three turns without missing a cover save. It also doesn't deal with range and a number of other factors. Pretty much all this does at this point is let you know if your list is too weak in the ability to kill MEQ and AV11 to be viable with average luck vs. an opponent of comparable skill on a board that does not have terrain biased against you. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863686 Share on other sites More sharing options...
thade Posted September 1, 2011 Share Posted September 1, 2011 ...pretty much the smartest post I've read this week... Algesan should guest lecture at that hypothetical B&C college they're chatting about over in Amicus. Read that post and commit it to memory. Essay quiz tomorrow. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863690 Share on other sites More sharing options...
yperihitikos Posted September 1, 2011 Share Posted September 1, 2011 If I read your table correctly, 10 Purgators of 280pts can kill 26,1 marines. Instead of this, I suggest that you would have more comparable results if you had written that when you take these purgators every marine that you kill has cost you 10.73pts when 5 terminators cost 36.36pts to kill just one marine. Or you can calculate how many kills you can get for every point you spend on a unit. In this way you can plan better when you make a list and you can always compare units of different points at a glance. I prefer points per kill (in excel it was better) 12,43____ 9,06___ 17,36___ 26,44___ 51,11___ 77,97___ 25,84 11,76____ 9,84___ 11,36___ 28,74___ 55,56___ 84,75___ 25,77 13,18___ 14,05___ 15,45___ 25,76___ 75,56___ 114,86___ 31,48 11,68___ 12,66___ 11,08___ 17,81___ 43,33___ 66,10___ 23,49 27,84____16,27___ 11,54____ 7,63___ 45,00_______ - ___ 55,10 23,44___ 18,07___ 25,64___ 32,97___ 41,10_______ ?___ 52,63 28,08______ - ___ 24,12___ 33,06___ 32,54____ 69,49___ 42,71 10,73___ 14,32___ 10,61___ 16,09___ 31,11____ 47,06___ 19,44 36,36___ 9,41_______ -_______ -_______ -_______ -___ 210,53 24,18___ 6,27_______ -_______ -_______ -_______ -___ 140,00 This is my Points Efficiency number. Our data is exactly the same. For example, the first column: 10 Strikes (DH, 2 PC): PE of 12.43 10 Strikes (DH, 2 PC, PA): PE of 11.76 5 Strikes (DH, PC, Psyback): PE of 13.18 ... They're all the same. PE is the points cost of the unit divided by the digits in the box (models damaged/removed in 5 turns). You can do the math yourself from my table above. Again, for box 1A you can divide 230 by 18.5 and get 12.43. I liked seeing both data, so I represented the relative value of PE through color-coding. I should have posted those numbers because they are useful as well, you're right. *EDIT*- What I need to do is add a bunch more units and tests, factor in Hammerhand as a separate option to see the disparity (and Cleansing Flame where I missed it), and create tables like yours that list all the remaining data. Look, here's what I think we'll find. Once we flesh this out, we can then factor in movement. This should realign the order of efficiency back to favoring traditional anti-tank. My calculations were made using your table. I'll try to explain my thoughts again. Using your table, you see what kill/damage a unit can make throughout the battle against a specific target. What doesn't clearly show is how many more points you devoted to manage every kill/damage instead of using other units for the same task. I'll give an extreme example. A unit that kills 45 marines and costs 800pts in your table would be coloured dark red, because it would seem better than anything else in killing marines. Of course, you give next to them how many points the unit cost. I don't say that your table has no use. I say that your table can mislead because you haven't embodied the points you paid to get that unit. A table is useful when you don't have to carry a calculator. There are three ways to build a list. The first is to take what you like, the second is to take what seems effective from experience. The third is to take what is mathhammerd to be the most effective. Effective isn't the most brutal in shooting or assaulting, it is the most brutal in a hypothetical table and opponent. You begin using your experience. You recall by memory what the other players have used, how was the terrain and the lines of sight. Then you ask yourself how much your list will have to depend on cover, how far you will shoot, on what targets etc. Then, you select a group of units that fulfil the requirements you have set. Finally, you look to your table and see which of these units are the most cheap to achieve the expected result and in which tasks there are better units. That is how you make your local balanced list. If you expand your table to more toughness and armour values for your personal use, you will have a complete guide to which is the best unit for your expected threats, specially when you lack experience against your opponent's race. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2863898 Share on other sites More sharing options...
jhrovii Posted September 1, 2011 Author Share Posted September 1, 2011 I absolutely agree with you guys. Personally, I highly prefer venerable dreads to regular ones. This has an offensive benefit, but it's primarily a stamina increase. Let's look at this example of what we've been talking about: Link Here a GK player has 6 Purifier Psyback squads like what you saw in my table, along with a full set of Psyflemen. Against a mech list you'd think they do great; after all, each squad has 2 psycannons and a psyback. http://i1219.photobucket.com/albums/dd425/Woodbok2/Grey%20knights%20vs%20guard/IGD.png The problem is twofold. First, I would have general'd differently. I don't know the layout of the physical table but chances are a banzai-charge straight into the maw of a shooty army from hell is exactly what the IG player wanted. The main problem though is failure of mathhammer to account for the bum rush to the opponents side in the first few turns. This neuters the Pure-Psyback squads. How would the game progressed differently with Venerable dreads and deepstriking doods? Hell, even an orbital barrage? The redundant "pwns-all-things" units simply weren't able to pwn in this scenario. It wasn't because they lacked killing power per se, but of course because they weren't playing on planet cueball. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2864081 Share on other sites More sharing options...
yperihitikos Posted September 1, 2011 Share Posted September 1, 2011 To begin with, the IG player has more long range weapons and superior. The GK has less and more vulnerable front armour to make a frontal assault. The GK player should have placed as many los blocking terrain in the middle he could get to be able to approach. On this table the GK should have deployed on the other deployment zone, the dreadnoughts behind the atmosperic processor and a bit left and the rest army behind the building and to the right. Manticores are expected to be deployed on the opposite building. The main flaw of the GK list is that it has too many KPs. GK are not SW. SW are cheap and they can take land speeder typhoons and predators or long fangs to support their las/plas razorbacks. In this list the razorbacks and their infantry squads have the same survivability, but the 3 Dreadnoughts can't offer the same support. To make it worse, these razorbacks are not a serious threat to the IG vehicles as you can only glance them. The GK list seems to have been made to face only MEQs. Even against Eldar and possibly Tau it would be probably doomed. Link to comment https://bolterandchainsword.com/topic/237407-the-confusing-world-of-planet-cueball/#findComment-2864302 Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.