Talk:Weapon Effectiveness (TFTD)

To the layman this appears nothing more than a brick wall of data; most of it is not relevant to user play in any meaningful way. Even though some wounded aliens have reduced accuracy, too many are just as dangerous at 1HP and don't care about average damage in the slightest. A player is going to care about two things- The odds of a weapon dealing ZERO damage, and the cumulative odds of continuous fire killing a target. For example null 20%, OHKO 20%, 2-hit 50%, 3-hit 90%. On the extreme end maybe knowing how many hits for a 98% kill rate can be useful. Those values are FAR more useful as it informs the player of how much firepower he needs to have ready against any particular threat. Oh, and if a target is totally immune, drop the numbers. It's IMMUNE. -- Bobucles Apr 28 2015

My thoughts exactly: it's nice to have all the numbers for those who enjoy the joys of data crunching, but also useful to have some sort of summary of the findings for quick reference. Hobbes (talk) 13:10, 28 April 2015 (EDT)

I'd also like to mention that "Armor fatigue" plays an important role for some weapons, especially explosives against 2x2 units. The hydro-jet can deal 12 potential armor scratching attacks in one burst! That is no negligible increase in damage, but the chart does not seem to model this potential at all. -- Bobucles 28 Apr 2015

Don't everyone thank me at once...

The point is to aid weapon selection, not to tell you "how much firepower you need" - the answer to "how much firepower should I bring" is literally always "as much as you can, taking into account encumbrance and 80-item limit". What this tells you is what form "as much as you can" takes.

Say, for instance, that I'm sending my Triton out to assault a touched-down Gillman Dreadnought. I know that I'm going to be facing Gillmen and Xarquid. Let's assume for the sake of argument that I have Gauss Rifles and the full suite of Sonics available. I go through the Xarquid "average damage per turn" column and note to my horror that Gauss Rifles are almost completely ineffective, and that Sonics usually take 3+ full turns of fire. But I notice that the Gas Cannon and Hydro-Jet Cannon HE rounds are superbly effective due to Xarquid's terrible under armour and large size, and so I pack several of those for my snipers to pick them off with. The remaining question is what I'm going to arm the rest of my troops with - since I'm going to be sniping the Xarquid with HE anyway, and I don't particularly want to spam HE against the Gillmen because I'll destroy loot, I'll be using energy weapons. Comparing the Gauss Rifle and Sonic Pistol numbers against Gillmen, I find that the Gauss Rifle's rate of fire wins out over the Sonic Pistol's damage (due to Gillmen having crappy health/armour and not resisting Gauss) so I arm my assault troops with that (because it's more effective and also because Gauss Rifle ammo is cheaper). Alternately, I could note that Xarquid take massive damage from drills, and therefore arm each of my assault troopers with a Thermic Lance (the most effective of the drills against Xarquid, according to the table) and a Sonic Pistol in their other hand (the best one-handed weapon against Gillmen). Not sure what I'd give my snipers in that case.

If you want something that isn't in these calcs, calc it yourself or ask me more nicely. Magic9mushroom (talk) 23:00, 28 April 2015 (EDT)

Am now adding N-hit kills, since I now have a CAS available. Unfortunately, without either a much faster computer or a more optimised algorithm, I cannot go beyond 4-hit kills. 5-hit kills would take days per calculation the way I'm doing it at the moment. Magic9mushroom (talk) 02:03, 30 December 2019 (CET)

Feedback requested

I'm considering switching the average-damage-per-turn to an average-TUs-per-kill. The advantage of this is that it's a better measure. The disadvantage is that I can only actually do it for alien/weapon combinations where all the N-hit-kill numbers are known; this means I'd be limiting it to those aliens where I can actually get a number (or a "negligible damage") for all weapons. I should be able to do at least some - but not all - 5-hit-kill calculations on my current computer, which helps, but it would still be only some aliens. So: is it better to have the bad measure on all aliens but have a standardised format, or to have the good measure on some aliens but not others and thus be less tidy? Magic9mushroom (talk) 11:45, 10 November 2023 (CET)

I'd prefer the average-TUs-per-kill, despite the mentioned limitation. Darkpast (talk) 18:47, 19 November 2023 (CET)

Okay, so the ones I can definitely do are the Aquatoid, Gill Man and Deep One (the Deep One I'd have to approximate the Dart Gun but who cares). That's why I did the 6-hit kills for the latter two (which are the only ones I can do; even the Jet Harpoon for Gill Men was only tractable because it was so close to 100%). The perennial problems for the rest are the Gauss Rifle (since it's fast-firing so 6-hit and 7-hit kills are relevant to its utility vs. many aliens) and in some cases the Sonic Pistol. The Triscene, uniquely, has a serious problem with the Heavy Thermic Lance; the 5-hit kill on that is completely and utterly intractable for me (it'd take a month, and I have Windows 10 so I can't actually get a month of uptime without cutting my Internet connection not to mention that running calculations heats up my keyboard and blocks me from gaming), although the rest of it quite neatly divides into "all kill chances known" and "immune or near-immune". If you have any fantastic ideas for things to do about the rest, feel free to mention them. Oh, and I'm kind of umming and ahhing regarding how to account for unusable TU (e.g. Gauss Rifle Auto leaves 20% TU unspendable, Sonic Cannon Aimed leaves 30% unspendable); just ignore it? Magic9mushroom (talk) 13:48, 4 December 2023 (CET)