I am curious as to what exact particular process you applied get the particular demonstrated example optimum? I ask, since I was heavily inspired by your analysis and borrowed a number of its insights (primarily those concerning constraints), though not your literal actual procedure or algorithm ( I have no direct experience with the specific technical optimization methods you employed ), to try to program my own direct optimization algorithm, and was able to generate a solution that used 104 less XP (425 XP instead of 529 XP) to complete all the rooms, and ultimately had 79 more XP in your sense of the 7th free roam to allocate for work (460 XP instead of 381 XP), with respect to the same simplified floor division calculation of XP from raw points that you also appear to have employed. I bring this up since you mentioned the range of solutions should all be within 20XP within each other. So I am curious as to how to interpret your 20XP range, to see if the solution I found is actually within that range or potentially poses as a genuine counter example of some kind?( I do not have the sufficient know how of your algorithm to make that determination, which is why I programmed my own lol)
I am still working on formalizing and generalizing my optimization algorithm attempt, and trying to also extend it to work from different initial conditions, for the possibility that one is not interested in re-rolling their game to start from a relatively early portion of Episode 6, but simply wants to figure out what the best strategy will be to employ from wherever they are in their game. If I decide to go through with that (I cannot make any definite guarantees), this will probably take me at least 1 more month, more likely 2 more months, though I suppose this is a good time killer until Episode 8 is available, lol (this has taken me about 3 weeks so far, with several failed attempts along the way to the current algorithm; I've spent more time on this than on playing the actual game lol). Should I end up going through with this (again, I cannot make a definite guarantee due to possible additional difficulties in generalizing the problem as well as possible pragmatic difficulties with respect to working on this), I may consider posting the final program used, along with a follow up post detailing the finalized optimized algorithm and how to use the program, though as of now it is far less technically composed and analyzed than the excellent analysis and solution spreadsheet of Hurensohn Heigl. I believe that, so far, of the people who have made mansion guides here, Hurenson Heigl has attained the highest understanding of this mini-game's constraints and principles needed to optimize it.
For those interested in using what the attempted optimization algorithm came up with, below is the program's output of that, using the simplified floor calculations ( in the actual game the final XP will actually be 475 XP; however I ended up with 460XP in my actual program since I chose to use simple floor division each turn for making the code simpler and to serve as a comparison ( it also takes Upper 3 Section as worked on in turn 3 before working on the Mansion in turn 3 at the end of Episode 6 since this was found to be the best option of the 3 rooms I have found you can work on (might be more I haven't found yet): Right Hallway, Upper 2, and Upper 3). I also ended up using an equalization distribution of XP for stats across characters for each room's needed amount of the stat, which appeared to produce the best result of three distribution types that I tried to implement.
Note: If you look at the attempted solution below, "Final New Characters" refers to their final stat allocation of the characters after actually completing the turn, and what they will be at the start of the next turn, and "Adjusted Character Sets With Given XP For Solution" refers to how the characters and their sets must be allocated with respect to the XP to actually complete rooms for the turn ( I assume that the player goes up to level 5 dumpster before completing their first real allocation turn (costs $4, which brings us from $21 to $17), which is where the solution starts). I also did not take any other upgrades besides full Dumpster since they are expensive and make it impossible to finish the mansion and buy the windows as fast as possible, and do not provide significant enough bonuses, in my opinion, to merit their costs.
I cannot claim that this is the optimal solution; the brute forcing required for determining that with complete certainty is computationally infeasible. If someone has found an even superior solution, which I expect to be possible, then it would be neat if it were posted. I have also not found any real nice algorithm for determining how to distribute XP among the characters with respect to the rooms they are put to work on; of three relatively simple methods I tried (equalization among the stats for each given room, all XP to strongest character relative to the room, and all XP given to the strongest overall character of each set applied to the room with only respect to the characters' stats themselves rather than anything to do with the room itself), I found equalization to yield the best result. The two strongest character methods appeared to yield the same result, which could just coincidentally happen to be the case, or I might need to reanalyze the methods again to disambiguate them if they really end up being distinctive (or it could be the math happens to work out such that is always necessarily the case and thus their distributive distinction has no consequential difference even if they follow different logical methods that happen to converge on a same result due to conditions of the rooms and the characters' stats themselves).
As far as making money goes, if you do not take the Pink Rose Pass and worked in the library, then you can allocate the remaining XP such that, with hopefully not too many re-rolls for jobs, you can make exactly enough additional money to buy the windows on the turn just before you sit down to eat Breakfast towards the end of Episode 7. You essentially specialize the allocation so that you have three sets of three characters that are each good at some of $3 jobs, so that you can earn $27, which, added to $38 you now have ( $17 Start Derived From Maximizing Dumpster + $10 For Giving Up Pass + $3 Story Work At Elderly Home + $2 Library + $6 Story Work At Elderly Home Or Pink Rose For MC, From All Characters' Combined Efforts = $38) gives you $65, exactly needed to repair windows on the very last turn of Episode 7, so that you do not have to do any more mansion work or jobs. I do not think that there is any way to generate enough cash to buy the windows and complete all the rooms prior to sitting down for Breakfast such that on your very last turn, you do not have to do any work or repair the mansion at all on that last turn, if you forfeit the Pink Rose pass, though correct me if I am wrong.
As far as determining exactly how to allocate the XP for those jobs, I have not made an allocation algorithm for that in particular yet; I simply eyeballed the list of $3 jobs and the characters' stats, and came up with the following, which seemed to work well (though I have not particularly rigorously assessed this yet), which depends on following the mansion repair plan from above.
First and foremost, thanks for the props. I have to return them thankfully.
To answer your question (as good as I can): The range of 20XP was related to the total XP generation process and not the remainder in turn 7. I see that you have generated a total of 825XP with your allocation plan while my (probably just local) "optimum" generated 840. As I mentioned in the corresponding section, I myself, know how to increase this 840 up to 853.2 (+13.2) as I have not allocated 91XP when I could have. This was for the same reason you are raising this question: I wanted to balance between maximizing XP generation and maximizing the left-over XP in turn 7. Neglecting the last objective, one could even achieve more - but to inferior conditions. Therefore my eductaed guess was 20 which is basically 50% more of what I could have done and am aware of.
In regards to the applied process:
1st subordinate minimization goal: Try to use all of the available XP in one round or get very close to it for the first three rounds. I.e. spend all 60 XP in turn 1. Furthermore, I couldn't further reduce the XP left over in turn 2 without violating the first assumption or not completing one room.
Reasoning: same as above
Limitations: Again, one could argue that spending all XP and leave one room unfinished would be superior, but this would not only be influenced by past decisions but also have to be accounted for in future decisions which is something I can totally not model (and this bothers me........)
2nd subordinate maximization goal: As mentioned in the post, there was this interesting finding that we are able to maximize the total score in turn 6 by clearing 9 rooms in one go. Therefore I artificially limited myself to doing that as it was superior to all previous scenarios.
Reasoning: see fun fact in constraint #2 of my second analysis.
Limitations: Since this assumption affects the previous decision, this is the most fragile one. Even though it maximizes the XP gain in the current turn, one has to account for the XP loss from not maximizing the previous turn with most attention. The solution I presented is maximizing the 5th turn subsequently and the XP loss is lower than the gain, but this only proofs that it is a superior reaction to the decisions made previously. If we change the initial approach (like loosing the restrictions of turn 3), this might look completely different.
However, I have tested multiple scenarios and identified the approach presented in the main post as the most optimal. But comparing ~20 scenarios out of multiple billions of possibilities is certainly not representative by any means.
I hope I could help and I am looking forward to reviving this discussion - to the pain of all other forum members, haha.
Its kind of amazing you don't realize how dumb a statement this is. BADIK isn't some science fiction reality bubble wackadoo present changing the past changing the future changing the present nonsense.
The wait for ep8 is really destroying people's minds.
The dev said that he likes to annoy his feminist wife by arguing things that actually he doesn´t think. And, of course, the character that "lambast the shit out of feminist" is the clown of the game, the one that nobody takes seriously. The dev is a woke.
I wouldn´t mind if, during second season, when it´s clear that Maya and Josy are lesbians, the player would be able to say "I´m out of here". But no, the player is forced to enter into the trap, and after that, is even forced to, at least, "remain friends" of them. It´s not even possible to forget about them and their far fetched, stupid, postmodern cliché soap opera drama with the evil religious father and all that jazz. And it seems will be even worse with Chad and the other guy.
I agree with you there. The past remains a constant when that past is confirmed.
But here lies the problem. There is much in the past that is not confirmed. For example, Lana.
We all suspect that she is the sister of Jill, who died. But where is that confirmed. It is only confirmed by Jill that her older sister died. Was it Lana? I don't know. Jill never said the name of the deceased older sister. You don't need to bring up the argument that she has only one sister, because that is not confirmed either. We just don't know.
And thus an alleged past constant becomes a varibale.
And this is also true of Partick, because Maya herself turns the father from a constant into a variable.
I agree with you there. The past remains a constant when that past is confirmed.
But here lies the problem. There is much in the past that is not confirmed. For example, Lana.
We all suspect that she is the sister of Jill, who died. But where is that confirmed. It is only confirmed by Jill that her older sister died. Was it Lana? I don't know. Jill never said the name of the deceased older sister. You don't need to bring up the argument that she has only one sister, because that is not confirmed either. We just don't know.
And thus an alleged past constant becomes a varibale.
And this is also true of Partick, because Maya herself turns the father from a constant into a variable.
Ok, this is long overdue, I've meant to make this post ages ago but just couldn't find the time to go through the game to extract all that Lily goodness.
So I know, you all loved my previous posts on Quinn, Bella, Heather and Josy, and while I'd like to break down all the characters, they don't all have interesting progression arcs, but I've always wanted to do one for Lily, so here goes:
The dev said that he likes to annoy his feminist wife by arguing things that actually he doesn´t think. And, of course, the character that "lambast the shit out of feminist" is the clown of the game, the one that nobody takes seriously. The dev is a woke.
I wouldn´t mind if, during second season, when it´s clear that Maya and Josy are lesbians, the player would be able to say "I´m out of here". But no, the player is forced to enter into the trap, and after that, is even forced to, at least, "remain friends" of them. It´s not even possible to forget about them and their far fetched, stupid, postmodern cliché soap opera drama with the evil religious father and all that jazz. And it seems will be even worse with Chad and the other guy.
Interesting catch. She looks a bit older than I’d expect, but… We know MCs mom is young (18?) when she meets Neil, but we don’t know how old she is when she gives birth. Neil looks 40+ and was early 20’s when they meet, so it’s possible he and Lynette could have been together years before she dies.
This is true, except what it could do is allow a 3rd party to reveal some home truths about Neil, Lynette, and her family.
Yeah, it could just be a shared Daz asset, but I think it's more likely that it's the same person in this case just because it'd be strange to introduce the character out of nowhere in episode 7 and devote a decent amount of dialogue to her with Jade if she has no relevance.
Ok, this is long overdue, I've meant to make this post ages ago but just couldn't find the time to go through the game to extract all that Lily goodness.
So I know, you all loved my previous posts on Quinn, Bella, Heather and Josy, and while I'd like to break down all the characters, they don't all have interesting progression arcs, but I've always wanted to do one for Lily, so here goes:
Huh. Now that necklace looks pretty familiar... I'm assuming someone already mentioned this at some point, but damn! It'd be pretty interesting if Cathy's replacement was an old friend of MC's mom.
Ok, this is long overdue, I've meant to make this post ages ago but just couldn't find the time to go through the game to extract all that Lily goodness.
So I know, you all loved my previous posts on Quinn, Bella, Heather and Josy, and while I'd like to break down all the characters, they don't all have interesting progression arcs, but I've always wanted to do one for Lily, so here goes:
Lily looks like a girl who wants to get the most out of student life. Even on the way to chick, she considers Tremolo a dangerous guy and she likes it. Moreover, she is the only one we don't pay for sex! She likes to show her body, fuck and love danger (drugs, booze, prostitution). I wonder if a potential threesome will be blocked if we choose Sage at the Hots party? As you remember, she says that she is ready to participate in group sex with main girls of MC. Of all the famous girls, she is closest to Josy the most and it would be a superdouet. I assume that sex with Josy and Lily is only possible on Josy's solo path.
Exactly i'll never play a dik route since i'm forced to remain friends with them. Atleast on the chick path you get something out of tolerating them. And its so ridiculous that he tries to paint the father as evil. Not everyone likes homosexuals. Get over yourself dev.
Yup, and it's high time these devs stopped forcing content we don't like down our throats. Who do they think they are? Writers? Artists? People who write the stories they want to write or make the characters they want to create? Hell no! They need to recognize what slave worms they are and only write content that we approve of!
Maybe we could make some kind of ministry that regulates their games and bans content we don't approve of. Like in China, how they've been banning male characters that aren't manly enough and are too feminised. And homosexual content too. Man, they do it right. Libtards may call us authoritarian, but we know the truth. We're just right and everyone must conform to our desires, otherwise they're infringing on our freedoms to rule them with an iron thumb.
Edit: Here be the obligatory sarcasm tag for the truly hopelessly.
Yup, and it's high time these devs stopped forcing content we don't like down our throats. Who do they think they are? Writers? Artists? People who write the stories they want to write or make the characters they want to create? Hell no! They need to recognize what slave worms they are and only write content that we approve of!
Maybe we could make some kind of ministry that regulates their games and bans content we don't approve of. Like in China, how they've been banning male characters that aren't manly enough and are too feminised. And homosexual content too. Man, they do it right. Libtards may call us authoritarian, but we know the truth. We're just right and everyone must conform to our desires, otherwise they're infringing on our freedoms to rule them with an iron thumb.