User talk:Spike

Message Board

---

By any chance, have you taken another look at my talk page since your return?

Yes yes, I'm lazy and should really be able to interpret the data myself... But I can only really get as far as saying "that's not a linear distribution" and then my lack of understanding re what all the trig functions are for leaves me stranded. :/

- Bomb Bloke 09:20, 10 April 2009 (EDT)

Sorry what with starting a new job, moving, Easter, I've been a bit busy. Did you say you still had the original data? You sent me it in an Access database, is that right? It all seems a long time ago! Spike 16:28, 14 April 2009 (EDT)

I don't remember sending you the data specifically (I certainly haven't done so within the last year), but you can find it all linked on my talk page now. - Bomb Bloke 22:31, 14 April 2009 (EDT)

Apocalypse Blog QA

Uncle NKF to the rescue!

Don't ever unload troops at the mission site. To select which agents go into a mission, just highlight them before picking the attack or investigate button or click on the vehicle name to select them all. Get into this habit - always.

The reason why will become incredibly apparent if you play long enough to do an alien dimension building. The moment you complete the mission, the alien building will collapse and your ship will take off automatically (as per standard practice). However, the kicker is that you'll never be able to land at the building site again. This will result in forever stranding any troops that did not get killed by the building crashing down around them.

---

Alien's won't attack a dead body on purpose, though HE explosions caused by nearby combat will certainly destroy them. Watch where you are fighting. Also some enemies tend to pick up anything that's not bolted down, and may even carelessly wander towards one of the area exit pads. It happens - so be vigilant!

---

Dual-Wield does penalise your accuracy, and depending on some weapons, will slow you down considerably. The only reason to dual weild is as you say: to increase volume of fire. If you're like me and like close range combat (you'll find a lot of this in some buildings), then you want to get out as much firepower as you can because the enemy will have just as much of an advantage at such a close range as you do (and there's often more of them than you or they have one-hit big-bang weapons like the Poppers!). Out in the open, single weapons with aimed shots get better results - until you are so accurate that you don't even feel the penalty.

The game's super weapons are easy enough to be effective without dual-wielding as they're ramped up to be the best weapons available. In fact, most people make do with just one devestator or toxigun and cloak combo exclusively.

---

Apocalypse uses the volume of successful hits to determine accuracy improvements. The M4000 just gets lots of hits in, while the laser sniper rifle may hit a few times. By the way, aimed + dual M4000 garners more accuracy for a slightly slower fire rate than full-auto, allowing you to not waste as much ammo. Not as accurate as one M4000 on aimed or as fast as dual M4k's on full auto.

Plasma gun ammo is scarce at first, but you do have the option of attacking your enemies like the Cult of Sirius and win some off their fallen brethren. In fact, a lot of items that don't get released until a week or two in can be won off raiding enemy organizations. Better to get a good start than try to hobble along. I'd recommend stealing borrowing a power sword too.

The Autocannon with AP shells is one of the most powerful non-explosive ranged weapons you can buy. Stronger than the plasma gun and the ammo is easier to obtain. I'd actually recommend only carrying 1 HE and 1 Incendiary clip and carry AP as your primary clips. Beside, you don't want to do too much property damage with the HE rounds (companies don't mind the burning corpses of civilians - they could go hostile towards X-Com if you track too much dirt on their carpets). The HE and In clips are for special uses, like blasting brainsuckers off if you're on your own, or to wipe out hyperworm mobs. Incendiary can be break up tight mobs of enemies, etc. The AC will however be very slow in turn based - so if that's what you play then it may not suit you very well.

---

Psi is a long term investment in Apocalypse. You only earn 3 times your base Psi stats. Hence humans will only ever be useful for low level simple attacks like probes. Hybrids will be your core psi users. Even then they may not be able to do much at the start. They need plenty of uninterrupted practice time in the psi lab before their skills pay off (when I say uninterrupted, I mean: no health loss when the clock ticks over midnight).

Psi may not be immediately useful, but they come into their own once they've built up their stats and can easily strip enemy anthropods or human guards of their disrupter shields in a flash while setting all of their carried grenades to explode on impact with the ground and dropping them.

The only time I've ever got a neutral organization hostile towards me when doing psi practice was when one of my psi troopers mind controlled a gangster then made him jump off the building. When he hit the ground, the gangster was still under my psi trooper's control. The loss in health was attributed to X-COM, and they started shooting. However, if my psi trooper had broken the psi link before the gangster hit the ground, the gangster would've eventually died of critical wounds (or the impact), and the psi trooper would've been free to roam about the map as if nothing happened. A morbid way of killing neutrals and not getting blamed for it.

Stun attacks will have the same effect as you'd get with stun gas or a stun grapple (marvelous weapon by the way - bounces of shields but still great for conserving ammo).

Psi attacks require line of sight. Breaking line of sight breaks the link.

Yes, psi attacks expend Psi Energy, a rechargeable resource. In turn based, psi attacks are generally a one time expense. In real-time, you are charged by the attack attempt. If successful, you'll start paying maintenance costs for as long as you maintain the psi link. You can break the link by going out of line of sight or readying another psi attack. Check the mind bender's information panel for the psi energy bar. Stun and panic for example require you to maintain the link long enough for stun increase or morale loss to take place. Probe lets you access the enemy information any time you want as long as the probe is in effect. Mind control lets you control the unit for as long as you want, but is also the most expensive in costs and maintenance. To make the best use of it, try not to spend any maintenance costs at all. Pause the game, control the enemy, go to their inventory and wreak as much a havoc as you can (for example: arm grenade. Right click grenade, then 'drop' the grenade). Once done, break the MC link and let time run and watch the fireworks.

-NKF 00:47, 16 October 2008 (CDT)

Dual Wield is useless in Turn Based. And when it comes to the M4000, I play turn based! Switch to autofire, and you only spend 1 TU per shot! Wildly inaccurate? Then sneak in ninja style and shoot everyone at point blank! OWNAGE! Note: Might need grenades for when you round the corner thinking there's only 1 guy and find there's half the Cult of Sirius hiding in the room. One of my favourite Dual Wields in Apocalypse is Autocannon with HE ammo, autofire, and Marsec Flying Armor. Only to be used when collateral damage is not an issue. Or in the case of COS raids, collateral damage is preferred! Of course, in real life, dual wielding large weapons like that will lead to such a loss in aiming control that you are very likely to shoot yourself in the hands.

Realistic UFO Economics an Oxymoron?

... Trying to form "realistic" economics in UFO defense is just ridiculous. Capturing one intact Medium Scout should realistically give enough cash to outfit an entire platoon of elite soldiers with all the best equipment money can buy. The power source+Elerium should rightfully be worth enough to buy a fleet of Interceptors. Instead, it gives $250k+$250k... not even enough to rent 1 interceptor for 1 month! Jasonred 08:01, 28 February 2009 (CST)

Yeah you're right about realistic UFO economics. :) But we do try. It's still good to try to smooth out the more flagrantly illogical aspects, and to try and level out the game balance. Removing Exploits goes a long way to making the economic aspects of the game better. For improving realism, I think one of the best things we came up with - working with Hobbes - is the idea that X-Com has a fixed-price "tarrif" for all the alien loot it hands over to the Council, established in the X-Com Constitution. This explains why there is no "free market" in the prices of alien items, because as you say, the prices of certain items ought to be astronomical. The other 'hard problem' in UFO economics is one that you highlighted in a recent post - why aren't the resources of the whole world harnessed, budgets in the billions etc. This is harder to explain away, you need a mixture of reasons/excuses. Some of the more useful excuses are - need for secrecy keeps the operation small scale (and as you also recently pointed out, the aliens seem to 'conspire' to keep the conflict small-scale); Earth governments are playing a 'double game' and don't want to risk offending the aliens by funding X-Com heavily, in case X-Com loses and the aliens punish them for supporting X-Com; Earth governments are not convinced that the Alien threat is real. (Obviously they can easily be convinced the Aliens are real, and are violent, but are they really a 'threat' to governments, or can deals be done - in a way this is like the 'double game' argument.) Spike 08:51, 28 February 2009 (CST)

Weapon Analysis

While I'm sure I was wondering about something. When you do Weapon Analysis for HE, I'm guessing you figure a GZ attack against the Under Armor. However, do you also factor in that Large units take nearly 4x damage from an HE attack, since they get hit on every square they occupy? Arrow Quivershaft 19:40, 1 March 2009 (CST)

Indeed I do. I calculate the other 3 squares of GZ+1 effect in some detail, including getting the averages right (which as you know is slightly tricky). One thing I don't calculate is the effects of any near misses. So the figures I give should be taken as a minimum. In practice, HE effectiveness will be somewhat higher due to the near misses - but by how much? It depends on a lot of factors, some I don't know and some that are highly variable (terrain density etc). Spike 03:55, 2 March 2009 (CST)

Hi Spike, to quote you from your user page:

One more problem with modelling Near Misses: the benefit of lucky near misses - whatever its value - is inversely proportional to overall accuracy. A perfect shooter gets exactly zero benefit from near misses. Which raises an interesting possibility. Maybe the "near miss benefit" could be estimated using a repeated experiments with a logger and a shooter with Firing Accuracy=0. Any damage done to the target would be due to a lucky near miss. This would be the other limiting case - the opposite of the perfect shooter. Then extrapolate between those two extremes to find a "near miss benefit" for that weapon as some function of {TA, range].

You bring up some interesting ideas on near misses. But wouldn't the benefits of 0% accuracy (leading to unintended splash damage) be negated by the probability of the soldier shooting himself? I would think that would be a major consideration when taking stuff like this into account. I should also mention that a soldier with max firing accuracy (125%) can pinpoint where shots should hit, thus the shooter intentionally doesn't aim for one target but the center of a group of targets thus creating a "near miss benefit" scenario. What do you think about this? --Zombie 12:45, 2 March 2009 (CST)

---

Great points Zombie. On the soldier shooting himself, I was wondering if we could just assume, as a simplication, that 'bad misses' (friendly fire casualties, including self-inflicted damage) even out with 'good misses' (lucky near misses). That would be much simpler! I have no idea if it's true though. Certainly if doing this logger exercise, you would want to distinguish damage done to self vs damage done to target. Actually, it's not really adequate to just 'net out' friendly damage vs target damage. After all, we expect weapons to do more damage to the enemy than to ourselves! So you might want to subtract 2, 3, 4... 10 times the friendly damage from the 'target' damage when determining the 'net benefit' from inaccurate fire.

It's a good point about the marksman firing at a midpoint, and of course that's a tactic we probably all try to use from time to time. In fact it's often frustrating when the game won't let you target a certain optimum point just because there's no object or creature there. These situations are slightly easier to model mathematically - though still cumbersome - because with a super-marksman you know exactly where the shell is going to land. The bit I really can't get my head around in the normal case (average accuracty) is the probability of all the different squares where the round might end up. I know there is some experimentatal data around with error angles but... I guess I'm trying to talk myself into the view that the terrain features are more important than the error angles, therefore it's not worth doing all the heavy math to figure out the effect of the error angles. :)

For the time being I'm probably just going to add a Note to my weapon rankings tables saying "Does not consider the beneficial or harmful effects of misses, near-misses, etc".

Spike 14:51, 2 March 2009 (CST)

Stuff like this is always fun to ponder and debate I think. It brings up a lot of good ideas/tactics too. I was just thinking about the second part of your reply a little. For the sake of simplification, I think we should just ignore terrain for the time being. Yes, it plays a role (sometimes huge), but it eliminates quite a few variables to arrive at some sort of conclusion. What would you think about a testing scenario where a soldier with 0% FA (standing on flat level ground) fires at a static object (such as a soldier who can't be hurt)? I'm thinking about using Bomb Bloke's numerical tileset as the substituted desert terrain and then firing about 100 or so rounds and looking at the overall damage done to the terrain by inspecting the numbers. Then you could see any potential "hot zones" where shots may hit more often for example, or even just concentrate on the normal damage area of the weapon and look for hot zones in there. Might be an interesting trial to run. Your thoughts? --Zombie 15:22, 2 March 2009 (CST)

I think that sounds great! It would be great to build up a 2D (or even 3D) histogram of where the shots landed. Tweak an autocannon so it has 250 rounds for ease. Work it from a standard range of say 10 or 20. It would also be interesting to see the results at an accuracy of 50%. From data like that you could definitely get some kind of rule of thumb to estimate the good and bad effects of misses. Cool! Spike 15:39, 2 March 2009 (CST)

Rocket Launcher Firepower Incorrect

Hi Spike, on the Firepower Tables it seems the Rocket Launcher is incorrectly computed. This might have been fixed in the alien specific firepower tables, point blank, etc, I don't yet know (and I dont have an xls viewer at the moment, either). However, at skirmish range, when the Launcher is 45%, 75% TU for snap and aimed, respectively, and 55% and 115% on accuracy, how can aimed fire have more than twice the firepower of snap? I ran the numbers and both aimed and snap were run at 75% TU usage. Furthermore, damage on aimed, small and large rockets, is exactly 75% of the actual average. The damages should read 46 and 58 for small rockets and 61 and 77 for large. --Talon81 04:29, 15 May 2009 (EDT)

I bet I know why, talon. It's cause the Rocket Launcher can only fire once per turn, on snap. And once per turn, on aimed. Tadaa... Jasonred 14:00, 15 May 2009 (EDT)

That cannot be the cause because it takes 45% of TUs to fire snap (therefore can fire twice). Secondly, it SHOULD not be the cause because even if it could only fire once, that is something to apply to the Sustained Rates chart, and not the Instantaneous Rates chart. On the Instantaneous Rates chart, it is figured as if you could use every TU. --Talon81 14:17, 16 May 2009 (EDT)

You *DO* know that the Rocket Launcher has a max ammo of 1, and therefore it CANNOT fire twice in 1 round? ... to fire twice per round requires 149 TU minimum. 67 to fire snap, 15 to reload, 67 to fire again. ... also, corrected your link. Maybe you can view it now. ... I think this belongs to the firepower table discussion page, not here, anyhow. BTW, after careul consideration, I think you are right... depending on definition of Instantaneous Rate. Jasonred 14:51, 20 May 2009 (EDT)

Large Units and Fire

If I'd remembered earlier I would've piped up about this, but I recall years ago when I set up a savegame for Zombie to test damage done to sectopod quarters, I physically split a sectopod into its various quarters by moving its individual unitpos segment coordinates to different locations around the map. Blasting one quarterpod with incendiary rounds set fire to all the others, even though they weren't in the same locality. This referring to the unit sticky fire, not the ground fire. Can't remember if it uses the same "apply this effect to the next four segments" rule that messed up the mind control of large units in TFTD though. -NKF 02:20, 16 March 2009 (EDT)