Total Research Time

[MikeTheRed]

Hi folks,

 

I am revisiting Total Research Time to correct the totals listed there. But it occurs to me that "total" is a little tricky when it comes to aliens...

 

The wiki list of all research topics includes a line that says "Live Aliens (except Terrorists and Sectoid Commander) (22 types)". But I think that this glosses over some details. Specifically, some aliens provide for more than one research topic.

 

To me, it seems like the minimum number of aliens to research in order to get research info on all topics that can be gotten from aliens, is 20, as follows: 8 Navigators (for alien missions), 8 Engineers (for UFO types), 2 Medics (for "live" sectopod and cyberdisk), 1 LDR or CDR, and 1 CDR.

 

This does not include the 11 corpses and 4 other live terrorists shown in the research list. (And you don't specifically have to do the Sectoid Commander.)

 

You'd have to do the medics carefully, namely, research them after doing all the other corpses and races except the two shown for them.

 

Otherwise, you can double-up and get research on alien races while doing the specialists, above.

 

Does this make sense? Am I missing anything?

 

Of course, in a real game, it's entirely likely that you will capture and/or research more live aliens than this minimum. But still... what's the minimum?

 

MTR

[Zombie]

I'd add a stipulation that one of the aliens must be psi-capable, otherwise you would miss out on the Psi Lab and Psi Amp topics.

 

And it's also a stars aligned thing too, if you want the 2 medics to just give the info on the live Sectopod & Cyberdisc. It's possible, as the medics give a random alien research topic, but highly unlikely. Sure, you can limit the possibilities by researching a unique race each time, but you'll still have 11 corpses and 4 live alien terrorists "cluttering" up the list once you get to the medics.

 

But why do you want the live Cyberdisc and Sectopod? They don't unlock anything of importance really besides the race type. Cut these two out and it's 18 total. If you mean just live aliens, then yes, 20 is the minimum

 

If you are after the minimum of everything possible, then it's 8 Nav + 8 Eng + 1 Ldr (or Com) + 1 Com + 11 corpses + 4 live alien terrorists + 2 medics (for the live Sectopod and Cyberdisc = 35 aliens total minimum for everything.

 

You might be able to skip ahead in the order by researching the 8 Navigators and 8 Engineers first, then the leader/commander and commander next, and then go balls to the wall with a boatload (17) of just alien medics. You might get lucky with one of them divulging the Sectopod and/or Cyberdisc interrogations early. Once you get those, then swap over to any missing corpses or terrorists.

 

Of course, to get everything done in the minimum amount of time, you really need to stick to the order outlined above. Just don't research a leader, focus only on 2 Sectoid commanders as they take 2 hours less. (It's a bug, but who cares? Let's exploit it)!

 

- Zombie

[MikeTheRed]

Hiya Z,

 

Good point about having one of them be psi capable... it is a little helpful in a game. Just a little.

 

There are some other (slight) constraints on how you'd have to research, to be most efficient. For example, 1 leader or commander would have to be ethereal (ignoring psi capability for the moment). Because we don't want soldiers, and that's all ethereals have besides soldiers. Et cetera. But it's not very complicated.

 

I did not realize that medics only divulge random results. You're sure about that? There's nothing on the wiki about it, that I can see.

 

That changes things... there isn't a fixed "total research time" any more; the medics invoke probability. But realistically, since those 2 aliens don't really matter, it becomes more like "all research except the two medic aliens [18 live aliens], plus probability distribution for medic aliens if you are interested". It looks like a favorite section title on a wiki page to me already! lol

 

Actually, I guess it would be close enough, and true enough, to quote the number with 20 aliens as being the "absolute minimum (if you get entirely lucky with medics)". And the "average minimum" is more like 35 (8 eng, 8 nav, 2 cdr = 18, plus 17 medics).

 

With random medic results, IF you want to get live sectopod and cyberdisk in minimum time, it's a good point to do them before doing any corpses or live terrorists... since the medics are sure to cough up some of that already. (That's an interesting wrinkle!) Now we have to compute how many terrorists and corpses are likely to have been done already, on average! (to subtract that from the total)

 

But are you saying there's some reason to do leaders and commanders after engineers and navigators?

 

You don't want to do 2 Sectoid CDRs. Because then, you haven't locked in an Ethereal. (But medics might reveal it already, depending on your approach.)

 

MTR

[Zombie]

There are some other (slight) constraints on how you'd have to research, to be most efficient. For example, 1 leader or commander would have to be ethereal (ignoring psi capability for the moment). Because we don't want soldiers, and that's all ethereals have besides soldiers. Et cetera. But it's not very complicated.

 

Right, forgot about that. You would need an Ethereal Leader or Commander first, then a Sectoid Commander.

 

I did not realize that medics only divulge random results. You're sure about that? There's nothing on the wiki about it, that I can see.

 

That changes things... there isn't a fixed "total research time" any more; the medics invoke probability. But realistically, since those 2 aliens don't really matter, it becomes more like "all research except the two medic aliens [18 live aliens], plus probability distribution for medic aliens if you are interested". It looks like a favorite section title on a wiki page to me already! lol

 

Medics do indeed give a random alien research topic (100% positive on that). Tested it again just to be sure in a brand new game. Edited a Sectoid Medic in AC and then researched it a bunch of times - always provided a random entry in alien research (depends on what is left on the list though).

 

With random medic results, IF you want to get live sectopod and cyberdisk in minimum time, it's a good point to do them before doing any corpses or live terrorists... since the medics are sure to cough up some of that already. (That's an interesting wrinkle!) Now we have to compute how many terrorists and corpses are likely to have been done already, on average! (to subtract that from the total)

 

The trouble is that we are also including time in this discussion as well as the research itself. That brings up a whole bunch of variables into the equation. My theory is that when it's early in the game, you will have few scientists on your payroll. Thus it's better to research topics which take as little time as possible. Medics and live alien races take the most amount of time at 192 hours, so it's better to focus on live alien terrorists first (170 hours each), then corpses (180 hours), and finally the rest (engineers, navigators, leaders, commanders, medics, etc). You probably are not going to have a whole bunch of live terrorists to research early on (maybe one or two from your first terror site - if you get lucky), so you have to supplement the rarity of them by diving into corpses (probably will have a couple to pick from). Once you run out of your initial load of corpses though, you'll end up going into live alien research to supplement that. So as you can see, it's not a clear-cut A -> B -> C relationship. You will be constantly switching between them to minimize research downtime when you run out of the quick topics.

 

Another issue is getting enough medics. It's pretty tough to get a bunch of them early in the game to make it worthwhile to research them exclusively. Not only that, but their research time is longer than live terrorists or corpses. Would you rather whack out 4 live terrorists or 3 medics (and possibly end up with a live terrorist you have in containment anyway)? Me, I'd pick the 4 live terrorists simply because they take less time and give you some points for your end-of-month summary. They also remove them from the list of available alien research topics, which increases the chances of getting the live Sectopod/Cyberdisc later on. Let's face it, the Sectopod & Cyberdisc are not game changing research topics to begin with, so you may as well prolong them a bit until you a) get more medics, and b) remove more garbage alien research topics from the list.

 

But are you saying there's some reason to do leaders and commanders after engineers and navigators?

 

You don't want to do 2 Sectoid CDRs. Because then, you haven't locked in an Ethereal. (But medics might reveal it already, depending on your approach.)

 

Well, there's no specific reason to go engineers+navigators then leaders and commanders other than the fact that you'll probably have a heck of a lot more engineers and navigators than leaders and commanders early on. I'm just being realistic. However, you will need to research either a navigator or an engineer first so as to start research on The Minimum Three.

 

Trying to get an Ethereal Leader to research first is also a big issue as that race doesn't show up for a few months of game time. The Sectoid Commander isn't impossible to get early on assuming you get a battleship or base to raid. So you may be better off going 2 Sectoid Commanders to get the Cydonia research line going, and then let another medic give you the research on the Ethereal. This is assuming you are after researching the line as fast as you can with all the resources available. Realistically speaking, if you have a Sectoid Commander, you'll probably already have a Sectoid Leader as well. Then you are better off researching the leader first and then the commander instead of waiting to nab another commander. (You will lose 2 hours of research time by researching the leader instead of the commander first, but that 2 hours is meaningless when the limiting factor is getting another Sectoid Commander which could take much, much longer than 2 hours).

 

- Zombie

[MikeTheRed]

Hmm, yes, the in-game timing issue can add a lot of complexity. But one way to keep it simple is to call it a minimum estimate of total research time (TRT). Seriously though, I'll have to think about it. There may be two total research times... the absolute minimum, and then a much more complex "realistic TRT" which may include some probabilities.

 

But answer me this: When I said "are you sure medics research randomly", I meant, not dependent on anything. You replied that you're sure of it, then said a couple of things like "Edited a Sectoid Medic in AC and then researched it a bunch of times - always provided a random entry in alien research (depends on what is left on the list though" and "They also remove them from the list of available alien research topics, which increases the chances of getting the live Sectopod/Cyberdisc later on".

 

Perhaps we're crossing wires over terms here, but the latter is not what I'd call random. Or at least, it's not what I meant... I assumed that that was the point you were making - that it's entirely random, and not dependent on what's been researched.

 

If the available pool is dependent on what's not researched, then doesn't my OP calling for 2 medics (after all the other alien research) stand?

 

Thanks - MTR

[Zombie]

But answer me this: When I said "are you sure medics research randomly", I meant, not dependent on anything. You replied that you're sure of it, then said a couple of things like *snip*

 

Perhaps we're crossing wires over terms here, but the latter is not what I'd call random. Or at least, it's not what I meant... I assumed that that was the point you were making - that it's entirely random, and not dependent on what's been researched.

 

As you do alien research, the total number of topics steadily decrease. Thus, when you research a medic the set of topics is smaller leading to increased probabilities of getting a specific one. Here's a hypothetical example of a total alien research set at the start of a game:

 

Reaper
Chryssalid
Silacoid
Celatid
Sectopod
Cyberdisc

 

As you research live aliens, the total size of the set shrinks until it is just two topics:

 

Sectopod
Cyberdisc

 

So pretend you researched an alien medic in the first set. You would have a 2 in 6 chance of getting a Sectopod or Cyberdisc interrogation (or a 33% chance). If you researched an alien medic in the second set, you have a 2 in 2 (or 100%) chance of getting a Sectopod or Cyberdisc interrogation. However, rolls within each set are still random: to get just the Sectopod in the first set you have a 1 in 6 (17%) chance. In the second set you have a 1 in 2 (or 50%) chance of getting the Sectopod. Your welcome to believe whatever you feel is random, but the fact still remains that the rolls are random within a set no matter how big it is. I consider that random even though the set size is decreasing leading to increased probabilities of getting a certain topic.

 

The point I was trying to make was that within a brand new game with an unedited, unresearched set, you have a totally random roll of the available topics if you research a medic.

 

If the available pool is dependent on what's not researched, then doesn't my OP calling for 2 medics (after all the other alien research) stand?

 

Of course it still does. I don't know of another way to naturally get the Sectopod and Cyberdisc interrogations, do you?

 

- Zombie

[MikeTheRed]

Cool... we're on the same page here. Those are good points about the order and approach.

 

As for "total research time", I guess I will rewrite the wiki citing the 20 aliens as said in my OP. And there isn't really any need to talk about more aliens than that, don't you think? A.k.a., a person is either paying close attention (and hoping to only use 2 medics at end of alien research), or they are probably not paying serious attention and just winging it. In which case, who cares.

 

But before I re-do the Total Research Time (TRT) numbers: there was an integer truncation issue whose write-up seems to have been lost on the current TRT page, but the Efficiency results from that are still shown. Do you (or anyone else here) still think integer truncation is happening?

 

I have a hard time visualizing it. Consider that the roll for Laser Weapons (average 50) comes up as 51. You assign your 10 scientists to it. The truncation principle says it will need 5 days. But what if, on the 5th day (40 hours done so far), you reduce the number of scientists to 1?

 

See, I don't see how the tracking for projects can get around the idea of actually comparing the number assigned (every midnight) to how many are needed. Which is to say, I don't see how truncation can work. Essentially, I think it has to be rounded up, not rounded down... you have to, in fact, actually do at least the number that was rolled. And you will probably also do more than that, because you can't see exactly how many hours were rolled (and therefore, probably over-estimated. unless you only have 1 person assigned. )

 

What are your thoughts?

 

MTR

[MikeTheRed]

Ok, I just added Minimum Alien Research, reply here or revise it if you want. I'll do Total Research Time shortly.