[Rejected] Research Progress based on actual project cost

[Delian]

Rejection reason: after long discussions, it was decided that this is an unwanted feature (not even as optional feature!)


Currently, the research progress of a research project can show 6 values:

NONE - no scientists are assigned to the project
UNKNOWN - the scientist-days spent on the research project is less or equal to 33.3% of base research project cost
POOR - the amount of assigned scientists is less or equal to 7% of the base research project cost
AVERAGE - the amount of assigned scientists is over 7% and less or equal to 13% of the base research project cost
GOOD: Shown if the amount of assigned scientists is over 13% and less or equal to 25% of the base research project cost
EXCELLENT: Shown if the amount of assigned scientists is over 25% of the base research project cost

Note that the "base research project cost" isn't the same as "actual research project cost", which is base cost modified by RNG (+/- 50%).

This is how it's always been, however, I have a reason to believe that this is actually a bug in the original game. I believe that the original developers intended for the progress to be based on the actual cost, but due to a bug (perhaps a late addition of the rng element), the progress uses base cost instead.
The reasoning is obvious: the progress shown is more or less useless because of how incorrect the information it provides is. A project with "poor" progress can finish in 1 day, and a project with "excellent" progress can take forever to finish. It's impossible to use it to predict how long the research will take, so any player trying to rely on it has only ever ended up confused. Players that try to be efficient are forced to split research into several projects.

Therefore, I suggest that a new option is added to the game that would change how research progress is shown:
"Actual Research Progress"
"Show progress of research projects based on actual (RNG-modified) research costs"

I've already made the required code changes, but I have a few questions:
1. Should I make a pull request for OXCE, or for OXC? (Option under Geoscape or Extended?)
2. What to do with localization?

[Meridian]

This was already discussed in the past and (overwhelmingly) rejected.

Neither OXC nor OXCE will accept such PR.

[Delian]

Discussed (and overwhelmingly rejected) where?

[Meridian]

on Discord

[Delian]

Ok. What were the findings and conclusions of that discussion? Pros and cons of such a user option? Who was for it and who was against it?

[Meridian]

The discussion was very long (2+ weeks) and I don't remember much, so just to recap the (for me) important results from the discussions:

1. show progress based on remaining effort (instead of number of assigned scientists), rejected because:
- not realistic (vast majority of real-world discoveries happen without being able to say when they happen)
- immersion-breaking (whatever that means)
- players would know exactly how much remains, which destroys the point of random (more about that in point 2.)

2. show progress based on actual project cost (instead of base project cost), rejected because:
- allows players to cheat RNG (not super effectively, but still way more than acceptable)

Example:
"Back to school" needs 1000 mandays.
Game rolls 1500 mandays.
The player invests 334 mandays.
At this point, the player would be able to extrapolate the real cost to 1500 md.
The player cancels the whole project and starts it again, because statistically he still has a better chance to roll low and spend less time in total (only 834 md in the best case = 334 + 500).

The feature was rejected by all devs, all major modders present on Discord at that time, and by many players alike.

[mutantlord]

I will add my rejection to this feature. Pointless.

[Juku121]

How exactly does the above example show 'cheating RNG'? The average cost for a 'cheater' is going to be ((500+1000)/2 + (334+1000))/2 = 1042 > 1000. Plus all the time wasted on the extra effort. There's a number of scenarios where this does pay off, but they're a definite minority and I don't see how said 'cheater' is going to distinguish between actual costs of 1335 (where it's likely to give a minuscule profit) and 1003 (where they lose out considerably).

Edit: Okay, by only rolling back from costs extrapolated to be 1335+? So average savings of about 5% at the cost of a lot of min-maxing micro? Doesn't sound game-breaking to me./edit

Not to mention that if one were to change the 'can see progress' threshold to 1/3 of actual progress, said trick would lose a lot of its usefulness.

Was that the only 'objective' argument against the "progress speed based on actual cost" version?

I mean, "nobody but Delian wants it" is a much stronger line of reasoning. I don't want it, because a lot of the issues raised in Meridian's post also apply to the vanilla formula to one degree or another, and getting that chucked is a no-go.

[howareyou32ny]

Made sense- randomness is fun in this game

[Meridian]

I didn't say anything about average costs.
I gave a specific example of cheating RNG by 666 mandays (1500-834).
(Average costs are btw. irrelevant, because the players are not going to restart every time, only when it's actually beneficial.)

And don't start on me with "time wasted on extra effort"... half the piratez player base will sacrifice a leg and an arm for 1 manday saved by micromanagement.

This topic is closed and dead for me.
I'm not going to waste another 2+ weeks going through everything again.

Feel free to continue the discussion without me though.

[Delian]

The player invests 334 mandays.

I was thinking it would have to be 500 mandays, since if the game rolled 1500, then 500 would be needed to pass the unknown and make an estimate, and then cheating wouldn't be possible.

Well, in either case, the players would be able to make a lot better estimation of how long a project takes. Which may or may not be realistic. I think that all the necessary realism is already included in that RNG roll that happens. And that people being able to give you good estimates on how long research takes isn't exactly unrealistic, but ok.

Hmm. As a player, I don't particularly care about whether I'm able to make a good estimate of a project or not. It's more about not wasting any scientist-days.
So here's a question. Do you remember anything said about a... research rollover feature? If a project rolls 40 and I put 50 on it, could those 10 wasted roll over to the next project?

[Juku121]

@Meridian: And your example was cherry-picked to be the most advantageous possible.

I was wrong in my initial assessment, but averages still matter because no matter what the player does, the second roll is still random and not a guaranteed -50% roll.

If Piratez wants do disable such a relatively meager advantage (even 834/1000 ~= 17% is not a gigantic advantage for that amount of micromanagement hassle), Piratez can always lock that option to disabled.

Edit2: I think the actual problem here is not the ability to get some relatively mild research savings (~5% on average without reloading, and once you're savescumming the whole issue goes out of the window anyway), but the fact that a player will be able to intuit the actual cost and remove any and all uncertainty from the process at will (and nominally without cheating). /edit


@Delian: Genuine, ground-breaking research doesn't really work that way. People can give you estimates, but unless they've already done the research (a practice of quite a few scientists, sadly), that is going to be pretty useless for anything other than project management ("You said you'd be done two years ago! You're fired!" ).

The current formula is a compromise between having a certain measure of uncertainty while being very simple and easy to grasp. Obviously, most people here, myself included, seem to be of the opinion that giving the player any more control over this state of affairs would tip it too far away from a 'desirable' amount of uncertainty.

I don't see much harm in this being an option, but then again I'm not the person who has to maintain this whole thing.

Edit: As to scientist roll-over, I imagine that discussion was done to death when the original bug was squashed (or not re-implemented) when moving forward with OpenXcom. That is certainly going to raise some immersion issues - so 499 scientists collectively managed to come up with the bright idea of blasting alien alloys with EM radiation, and then the other 335 went back in time, saw that they weren't needed to finish the project and decided to redo their day and study laser pistols instead?

[Vakrug]

I have an alternative proposal:
remove that research progress (NONE, UNKNOWN, ...) completely!
This "research progress" has no meaning since player already knows project's base cost in the upper left corner.
This values now and always was totally misleading. When I played original X-COM many years ago, I was confident that this "POOR" means "too little scientists are assigned, so an output from each individual scientist is reduced". Because what else could that mean realistically?

I think this value should be replaced with number how many scientist-days already spent on this project.

[krautbernd]

I have an alternative proposal:
remove that research progress (NONE, UNKNOWN, ...) completely!

I have a somewhat different proposal:

Don't do this.

[yergnoor]

I have an alternative proposal:
remove that research progress (NONE, UNKNOWN, ...) completely!

The authors of mods can easily remove these evaluation inscriptions, for example, by replacing ALL of the relevant lines with just "Exploring". Since the inscription will always be the same, the player will not be able to assess the progress of the study.
Except that the number of man-days will not be shown in this way.
Such a mod is very easy to make and for personal use, if someone does not like these inscriptions.
True, I'm not sure that these same strings are not used in some other places in the game.
Forcing everyone to use just this weird research option is a departure from the UFO game behavior that OXCE developers probably won't go for. Who needs it - let them use mods, that's what OpenXcom Extended was created for.