michthro
Level: Smitemaestro
Rank Points: 679
Registered: 05-01-2005
IP: Logged

Look, StuartK, I understand how you feel about existing scores. I'm in much the same position as you. The big thing to bear in mind is that currently there are really only a handful of players who play for #1 scores. So, what's been happening is that a lot of #1 scores are non-optimal first-run attempts. So far, you've been getting away with it, but this is bound to change as more players start playing for high scores. I classify rooms into three categories: Those I want to optimise, those I'm not interested in, and those I'll replay once for the sake of getting a reasonable score. Sometimes I happen to get a #1 score for rooms in the second category. These scores I write off as "going to be improved sooner or later", and move on. I fully expect that as more players join up, my number of #1's is only going to decrease, until only the optimal ones remain (i.e. very few). Without UU it's only a matter of time before there are enough players for this to happen anyway. UU may speed up this process, but five years from now it will be all the same (hopefully). What I'm trying to say is, the game is bound to change fundamentally for you anyway (assuming you continue to play for high scores), as more players join in.
Besides, nothing stops you from going back and improving your scores. You only have the advantage over newcomers of experience, the knowledge of which scores are already optimal etc. As for the amount of effort you already put in: Well, considering that admittedly they are mostly first-run scores, it isn't really such a hell of a lot, is it?
I think you overestimate the value of UU. You seem to think it will enable every rank newbie to improve your existing scores. Well, I don't think so at all. UU isn't some hugely powerful tool enabling anyone who is the first to play a room to optimise it. With your skill and experience you will only be using UU with according skill, obtaining better scores for the same amount of effort. What's more, your scores will more often be "secure", which is something you surely prefer.

robin
Level: Smitemaster

Rank Points: 842
Registered: 09-01-2004
IP: Logged

OK, I can live with how it's done.
Good job mrimer!!

(I'll add it to my list )

____________________________

---

[Last edited by robin at 02-20-2006 02:18 PM]

StuartK
Level: Smitemaster

Rank Points: 564
Registered: 06-10-2003
IP: Logged

KevG wrote:
You are seriously over-estimating how much UU is going to effect things.

Perhaps I am overestimating (or overstating my case) or is it you who's underestimating? We don't have any actual data, just lots of supposition and future predicting on both sides of the argument. The thing is, once the err... genie is out the bottle, it's going to be very difficult to force/persuade/trick it back in, if it turns out not to be quite as benign as many are assuming it will be.

Anyway, my issue isn't just with the highscores (though it is an issue) I think my single worst fear is it's going to have a negative impact on the long term playability of the game. It just seems/feels like a cheat mode to me, and cheat modes have always made games boring, after the initial fun of greater freedom wears off. And to continue competing for highscores, I'd be effectively forced to change my playstyle and 'cheat'. Or worse, I'll just use the feature because it's there, come to rely on it, and eventually find it has indeed sucked all the fun out of the game. Where's the tension, with the constant availability of UU? Are we only to have a sense of it the very first time we play a room? What if we want to replay a hold from the beginning?

2. Hold author has the option to disable it

I'm completely against this.

A fair point, for the reasons you state.

3. Instead of UU, have a single, portable checkpoint.

Actually, as far as I'm concerned this would be my first choice for a way of avoiding having to redo previous work.

I still have reservations, but if we have to have a change at all, I'd vastly prefer such a compromise

StuartK
Level: Smitemaster

Rank Points: 564
Registered: 06-10-2003
IP: Logged

Michthro - I agree with much of what you say, and, on reflection I don't think I'm quite so attached to my highscores as I might have led the board to believe. In trying to seperate selfish motivations from the overall picture, I'd go so far as to say (and I've said it before, even if my actions have not reflected it) I'd like to see the highscores more open to newer players. Again though, its questionable to me whether UU would help or hinder this. It's often after all newbies who start off by grabbing those #1s from non optimal solutions. Perhaps you can identify with this, being a relative newcomer? I don't know. We're still only a tiny proportion of the entire player base.

What would happen, though, if us veterans decide to go through the existing holds, optimising everything in sight with the new tools? We have a headstart, our position would be even more unassailable, and the club we belong to more exclusive and difficult to join than it is now. It raises the bar for newcomers from very high, to err... very very high. Perhaps inevitable anyway, but UU as it stands (or indeed any additional ease of use features) will do nothing but accelerate it.

Rabscuttle
Level: Smitemaster

Rank Points: 2460
Registered: 09-10-2004
IP: Logged

agaricus5 wrote:
Want a challenge?

Oooh...

What we should do, is every week or so pick a room to bash our heads against. We'll get some well-optimised rooms out of it, both from the competition, and from the "two heads are better than one" principle. We can discuss strategies and so on, which might make for some interesting reading, as well as encouraging others.

I reckon we should start with KDD2:L8:2S2E, seeing as Stuart has now found time to fix my imperfection. Except I've just realised that cleaning out the middle section isn't necessary.

---

[Last edited by Rabscuttle at 02-21-2006 10:03 AM]

Rabscuttle
Level: Smitemaster

Rank Points: 2460
Registered: 09-10-2004
IP: Logged

And what would also be cool would be if we could see how the best score changed over time.

eg (for the above room)

25/7/2005 10:02pm: Rabscuttle: 389
21/2/2006 6:00pm: StuartK: 381
21/2/2006 8:54pm: Rabscuttle: 330
21/2/2006 9:03pm: Rabscuttle: 304

Definitely looks like there's still room for improvement, but I have to go out now. I sure hope no-one beats it

---

[Last edited by Rabscuttle at 02-21-2006 10:26 AM]

michthro
Level: Smitemaestro
Rank Points: 679
Registered: 05-01-2005
IP: Logged

StuartK wrote:
What would happen, though, if us veterans decide to go through the existing holds, optimising everything in sight with the new tools? We have a headstart, our position would be even more unassailable, and the club we belong to more exclusive and difficult to join than it is now. It raises the bar for newcomers from very high, to err... very very high. Perhaps inevitable anyway, but UU as it stands (or indeed any additional ease of use features) will do nothing but accelerate it.

You're still overestimating UU. Many rooms will still be very diffcult to optimise. In fact, there are many rooms that would take billions of years to optimise with all the supercomputers in the world. So what would happen is it would take ages, and that's only for those rooms that can be optimised. By the time you've done that there will be so many more holds again. Besides, newcomers could do the same, they'd only be slower than you at optimising, but they'd get their fair share nevertheless (fair as in allowing for experience and skill). Those with a headstart will always have a headstart, but in the long run it will come to mean less and less. If it bothers newcomers they could always make themselves feel better by saying things like: "StuartK is useless, it's only because he had a headstart..."

About high scores being more open to newcomers: It would help a lot if newcomers could get some advice and tips somewhere from the veterans. One mistake I made, and I think it's a common and natural thing to do, was to start with JtRH. I don't have to tell you why that's a bad idea, and why it could very well put people off playing for high scores. It isn't obvious to a newcomer that JtRH is much more optimised than anything else. I kind of assumed it would be the same everywhere, until I tried some other holds.
Another thing is the way veterans stomp on newcomers if they dare to snatch a #1 score. That's how it felt from my point of view anyway. So it didn't put me off, but I can see how some people might be. Maybe let newcomers keep a few high scores to encourage them. You can always reclaim those scores later on, when they're not so new anymore.

michthro
Level: Smitemaestro
Rank Points: 679
Registered: 05-01-2005
IP: Logged

Rabscuttle wrote:
What we should do, is every week or so pick a room to bash our heads against. We'll get some well-optimised rooms out of it, both from the competition, and from the "two heads are better than one" principle. We can discuss strategies and so on, which might make for some interesting reading, as well as encouraging others.

I reckon we should start with KDD2:L8:2S2E, seeing as Stuart has now found time to fix my imperfection. Except I've just realised that cleaning out the middle section isn't necessary.

Great idea! How about starting a thread for this?

KevG
Level: Smiter

Rank Points: 333
Registered: 08-16-2004
IP: Logged

StuartK wrote:
Perhaps I am overestimating (or overstating my case) or is it you who's underestimating?

I would assume that we're both under/overstating our respective cases. I can see where unlimited undo would be a bad thing for new players, but the conquerors only restriction means that only players experienced enough to conquer the dungeon can use it. I just don't see players suddenly ignoring the skills they learned solving the dungeon to revert to using pure trial and error. I played around with UU for a while before jumping into the conversation and it primarily was useful as a time-saving technique. More importantly, it allowed me to stay focused on the tactics and strategy instead of being constantly interrupted to repeat moves.

I just don't see longevity as an issue; only a small handful of players are ever going to play all the available rooms even once. And let's face it, optimizing as a practical mode of play has existed for less than a year. DROD lasted for years without it, even now most players don't take part in it. Making it easier won't seriously affect DROD overall.

UU isn't a cheat mode; giving a player greater freedom as a reward for completing the game isn't cheating. Most of the thrill of playing a room is gone when you replay it anyway; UU would encourage more people to replay. Also, it would encourage people to stick to and actually conquer holds. I'd feel a lot less pressure to perfect each room as I do it and be encouraged to move onward. Traditionally, DROD has been about completing rooms not perfecting them. Conqueror's only UU would seem to reinforce this.

KevG
Level: Smiter

Rank Points: 333
Registered: 08-16-2004
IP: Logged

michthro wrote:

Rabscuttle wrote:
What we should do, is every week or so pick a room to bash our heads against. We'll get some well-optimised rooms out of it, both from the competition, and from the "two heads are better than one" principle. We can discuss strategies and so on, which might make for some interesting reading, as well as encouraging others.

I reckon we should start with KDD2:L8:2S2E, seeing as Stuart has now found time to fix my imperfection. Except I've just realised that cleaning out the middle section isn't necessary.

Great idea! How about starting a thread for this?

I'd like to request that people not play this room. I'd like to keep my demo from July in third place. Although, if StuartK and Rabscuttle would be polite enough to achieve the same score that would be appreciated.

Seriously though, this is a great idea. As several people have alluded to in this thread, win or lose, actively competing with people is where the real fun is.

Edit: started topic in Challenge board.

---

[Last edited by KevG at 02-21-2006 02:26 PM]

roach strangler
Level: Delver
Rank Points: 57
Registered: 01-07-2006
IP: Logged

man these conversations are long it could make me

Chaco
Level: Smitemaster
Rank Points: 3626
Registered: 10-06-2005
IP: Logged

You, sir, need to think before you post.

____________________________
Quick links to my stuff (in case you forgot where it was):

StuartK
Level: Smitemaster

Rank Points: 564
Registered: 06-10-2003
IP: Logged

My last post on the topic*. I think I've said everything I want to say, and would only be repeating myself (as I already to a degree)

KevG wrote:
I just don't see longevity as an issue; only a small handful of players are ever going to play all the available rooms even once.

Longevity isn't simply directly proportional to the number of rooms - it's also a function of the amount of gameplay in those rooms. Once the player has access to a game feature that makes that gameplay less interesting (for some) it's possible to lose interest in the game more quickly and irrespective of how many rooms are available.

Also, I'll never play all the available rooms, but I have, and will play some holds more than once - some many times - and not just because I want higher scores - more than anything else because I enjoyed them.

It's the wrong solution to the problem, because it's an excessive solution - it gives the player too much freedom, eliminates the element of tension from the game, and makes some proportion of rooms simplistic (maybe not all rooms, but definitely hack+slash rooms) And it does so to satisfy only a subset of the playerbase, which is less than the set of people who will be replaying holds. A subset of a subset.

And let's face it, optimizing as a practical mode of play has existed for less than a year. DROD lasted for years without it, even now most players don't take part in it. Making it easier won't seriously affect DROD overall.

I agree that the first run through a hold will remain unaffected, except when you're going for highscores on that first run, and other players have already completed the hold, have access to UU, and are now optimising rooms. So what to do? Rush through as quickly as possible?

UU isn't a cheat mode; giving a player greater freedom as a reward for completing the game isn't cheating.

You've said it will make the game easier yourself. And it has been used as a cheat mode already. That was it's original purpose for inclusion - for testing. Where's the tension, if there are no consequences to a misstep or mistake? Robbed of obvious negative consequences to your actions, DROD is no longer a game - it's an exercise.

UU would encourage more people to replay. Also, it would encourage people to stick to and actually conquer holds. I'd feel a lot less pressure to perfect each room as I do it and be encouraged to move onward. Traditionally, DROD has been about completing rooms not perfecting them. Conqueror's only UU would seem to reinforce this.

The original run through the game will be the same as it is now. The game after that point will however be changed. Any other effects are down to the perception of the individual.

So, as you feel less pressure to compete for highscores, I'd feel more pressure to complete a hold quickly, in order to compete with others who are doing the same on a level playing field. Again, my playstyle throughout the game is fundamentally changed, and not IMO for the better.


* probably

KevG
Level: Smiter

Rank Points: 333
Registered: 08-16-2004
IP: Logged

StuartK wrote:
My last post on the topic*. I think I've said everything I want to say, and would only be repeating myself (as I already to a degree)

Then I'll make this brief. You're right that we're getting to (or past) the point where we don't have anything new to say.

You've said it will make the game easier yourself. And it has been used as a cheat mode already. That was it's original purpose for inclusion - for testing. Where's the tension, if there are no consequences to a misstep or mistake? Robbed of obvious negative consequences to your actions, DROD is no longer a game - it's an exercise.

I'm aware of the history of UU; it's one of the reasons I don't believe people are rushing to judgement. Regardless of how it was used in the past, using it as a reward for solving a hold doesn't qualify it as a cheat mode in my opinion.

Solving and optimizing are different modes of play. The focus of optimizing should be in concentrating on whether or not you can shave off moves, not on simply surviving. You already know you can survive the room, now it's how can you do it better. Having to repeat hundreds of moves to compare approaches doesn't add tension, it simply adds tedium. To me, that's when DROD becomes an exercise.

agaricus5
Level: Smitemaster
Rank Points: 1838
Registered: 02-04-2003
IP: Logged

This discussion was continued from here

Syntax wrote:
Well... I'm disappointed from the point of view that there are high scores at stake. As previous posts have pointed out, I'd like to think the playing field was even. To think that some of the #1 battles I've had could be against someone with an unfair advantage makes me very unhappy.

With UU implemented in TCB, everyone will be on the same level playing field

In fact, the day it gets introduced to DROD is the day I stop playing.

The main problem with UU is that it eliminates the need for instinct, and unusual solutions would become common-place simply because everything becomes *triable* without the backfire of having to start again.

What you argue here is that guesswork and experimentation should be penalised by having to replay masses of moves repeatedly, i.e. only people who can consider and calculate the entire move sequence they want to make in advance should have an advantage.

The problem is, as has been said many times before, this is very unfair towards people who do not have the capability to do this, mentally. The total number of possible pathways increases exponentially as the number of moves in the sequence increases (for small sequences), and unless you have a computer-like mind able to process and remember that many different outcomes all at once, I highly doubt you'll be able to find an optimal solution without an element of guesswork involved.

My optimisation method is a kind of complicated Markov algorithm/process thingy (minus the randomness). At each point where I no-longer know if I am optimal or not, I attempt to use "instinct" to choose the best of the 11 next moves by considering the effects of the next or next 2 moves. The best option prevails, it is likely to be chosen, and then the decision process is repeated for the next move, and so on. If the move I pick for a given turn seems to work, and allows me to continue moving, I give it a higher priority, and if I have to restart from the original point, I am more likely to make that move a second time when I reach that turn. Similarly, if a move leads to disaster, or a whole chain of unoptimal sequences, I will give it a lower priority, and not pick that move the next time I get to that turn. So, the way I look at moves changes as I play. The more a move seems to work, the more I use it, and the more it fails, or becomes unoptimal, the less I use it. That way, I don't have to remember the actual move sequences I have played as such, I only need to remember what the general overall effect of each move was (i.e. is it more likely to lead to partial success or failure?).

Basically, it's a kind of an intelligent search function - I can't read 50 moves in advance, so I'll split it into 2-move chunks.

When I play chess, I'll let my opponent move their piece back if they made a stupid move. I wouldn't, however, go further back than that in the game. Making mistakes is part of learning, and understanding movement order and why roaches gap for example would become obsolete. DROD is a hard game to play, but even harder to master... UU would make the latter meaningless.

Would you want to play a 50-move sequence fifty times over (i.e. 2500 moves) just to see which of your possible move sequences were not mistakes? How would you know, without experimentation, if a move sequence was correct?

I agree that making mistakes helps you to learn, but the current system penalises mistakes. How am I supposed to learn anything if I can't actually make a mistake, because I'll have to play out 200 moves all over again? In fact, I assume that chess players and players of similar games, like Go, are in fact performing an intelligent search like the one I just described when playing in situations they do not recognise. The major difference is that while I may only be able to compute 2 moves in advance, a chess player could probably do about 20.

Unlimited undo will put people like me on a more level ground with people who can compute 20 moves in advance. I don't see why this makes DROD harder to master - is the goal of optimising in DROD to train you to see long-range (e.g. 20 move) sequences in advance, or to train you to perform a short-range intelligent search more efficiently? I think your concern is that the latter requires less effort and time, but I disagree - the further you can read, the less you'll undo, and the less time you'll use experimenting, while the more you rely on it, the more time you'll spend optimising and guessing. The differences stem from the type of mind you have, and how good you are at performing your random searches, given the tools you have.

Long-range reading will probably give more consistent results, but trial rates will be lower, while short-term experimentation may allow for faster trial rates, but will drastically increase the probability of finding really unoptimal results. Either way, they should end up the same; I think the reason why there is such a disparity between good players on the #1 table at present is the fact that people using the former require fewer trials, and so do not restart as much.

Having UU as an option doesn't even work in theory. High scores would be affected since once a room had been completed with it on, it becomes easy enough to mimic it with it off (that's assuming high scores would be disabled if it was on in the first place). Besides, demos could be hacked to spoof it being off.

That doesn't really make much sense. You can mimic the #1 leader at any time you like by downloading a demo, whether UU is on or not, and you could probably "hack" demos in JtRH anyway. The main idea of highscores and optimisation is not to copy someone else's solution, but to be the first to find a new one.

Finally, KevG's post can summarise mine for me.

KevG wrote:
Solving and optimizing are different modes of play. The focus of optimizing should be in concentrating on whether or not you can shave off moves, not on simply surviving. You already know you can survive the room, now it's how can you do it better. Having to repeat hundreds of moves to compare approaches doesn't add tension, it simply adds tedium. To me, that's when DROD becomes an exercise.

____________________________
Resident Medic/Mycologist

Maurog
Level: Smitemaster

Rank Points: 1501
Registered: 09-16-2004
IP: Logged

Hmm, can we turn the Unlimited Undo into an optional thing, even for hold conquerors? Some sort of a checkbox you need to check to enable it.

And then, and then maybe count any time anyone goes beyond the first undo and sort of put it in their highscore. With no negative implications of course. Just the number of the times you used that feature.



____________________________
Slay the living! Raise the dead!
Paint the sky in crimson red!

NiroZ
Level: Smitemaster
Rank Points: 1302
Registered: 02-12-2006
IP: Logged

Maurog wrote:
Hmm, can we turn the Unlimited Undo into an optional thing, even for hold conquerors? Some sort of a checkbox you need to check to enable it.

And then, and then maybe count any time anyone goes beyond the first undo and sort of put it in their highscore. With no negative implications of course. Just the number of the times you used that feature.

It might be interesting to have DROD tell caravelnet whether or not a particular demo used UU or not, but I don't think that sort of information should be available to everyone(well, perhaps anonymously), because then it becomes a shame issue, which defeats the purpose of implementing it.

Syntax
Level: Smitemaster
Rank Points: 1218
Registered: 05-12-2005
IP: Logged

Ok, agaricus5...

I can think several moves ahead. In fact, that's why I got 3 of the top 5 battled #1s. I may not spot the best solution straight off, but I can grab a #1 off almost anyone.

The reason for this is hours of trial and error. I can spot roaches gapping whilst they're still behind a wall. I can spot how many moves it will take to kill that damned roach queen. I can cut tar beautifully.

The reason is effort. UU will kill that and I think it's a shame. The skills of Rabscuttle, michthro and I will be wasted by it. We've spent hours memorising 2K+ rooms... and now it won't matter. Imagine telling Karpov he could replace his pieces. Now *that* would be silly...

[EDIT]

Typo

---

[Last edited by Syntax at 11-12-2006 08:43 AM]

NiroZ
Level: Smitemaster
Rank Points: 1302
Registered: 02-12-2006
IP: Logged

What we really need is a prototype to test this out on. At the moment, the best argument is but speculation, well thought out speculation, but still speculation. (anyone want to hack DROD:AE?)

But in reference to syntax, I must point out that you will still need to be able to count ahead. Its just now if you realise that you couldn't have killed the roaches in 9 moves instead of 10 after you have cleared the tar, you can just rewind.

zex20913
Level: Smitemaster

Rank Points: 1723
Registered: 02-04-2003
IP: Logged

Preamble: This is essentially a note to Syntax.

I think another bonus to UU would be for long rooms, as michthro sort of mentioned already. If there's a room that has an optimal solution of approximately 450 moves, and the #1 is currently 475. You know this. You also know the optimal solution for the first 300 moves. However, the last 150 moves are very tricky, and you die often. It turns out, the 300 move sequence is difficult to replicate. Wouldn't you rather spend less time on repetition and more on optimization?

I realize that this sort of "evens the playing field" in that it takes less to say "Hey! Why don't I try this instead!", but it is already possible, though convoluted (write down/type every move until a decision point, then go through a sort of Markov process, as ag described.)

Personally, I'm going to use it. It ends a lot of the frustration of repetition, for me, and I still need to think of the optimal solution. As you know, I'm at approximately "Optimal plus one-to-ten" for a lot of rooms. UU may get me to optimal. And yes, I may be taking advantage of the system, and not be "hardcore" enough of an optimizer, but I think I'm almost there already, even if I don't spend hours on a room.

Effort is still necessary. How much effort is up to the player. Some people don't want to memorize rooms. I don't. I sort of just intue, and if I feel I'm close to claiming a 1st, I try to improve.

By the way--if you get a #1 on a difficult room in minutes, and you're pretty sure it's optimal, is that any different than spending hours on it? It seems like one of the claims you're making is that "We've put hours into this room." Well, what if somebody comes up with a better optimal solution in minutes? I recently beat a Rabscuttle #1 by about 160 moves. You read that right. I was dillydallying, saw the #1, said "Quoi", and went back to improve upon my #1 score. Now, I'm not going to say which room, because I want to have the score for at least until Rabscuttle gets back, but is what I've done wrong? If so, then I see that your perspective is against UU. I see UU as a correction of typos moreso than an optimization finder.

Ramble over.

____________________________

michthro
Level: Smitemaestro
Rank Points: 679
Registered: 05-01-2005
IP: Logged

Syntax, UU won't enable you to try literally everything (11^1000 possibilities?), or even anywhere near every reasonable approach. You make it sound as if everyone will be able to optimise rooms by sheer brute force. A computer, working 100 000 000 times faster than any human, even with an algorithm more advanced than basic brute force search, would still take billions of years to optimise a typical 500-move room.

However, there is a lot a computer *could* do, and that's what UU is about to me. I'm generally not interested in solving problems a computer can solve, such as sudoku puzzles, orb puzzles, or the quickest way to kill five roaches. It's much more interesting then to actually write a program that does it. (I think Stefan has written a program that can solve small DROD problems.) Partly on principle, as in there's little point in doing something a computer can do much faster, but mostly coincidentally, as in the kind of puzzle computers can solve I tend not to like, especially if the most basic algorithm already works well. So, it's all about what a computer could and couldn't do. A computer would make very short work of a handful of roaches, but it would be one heck of a program that actually optimises whole rooms. I want UU as a substitute (a rather poor one, come to think of it) for something that automatically does the boring, tedious, a-computer-could-do-this job of optimally dispatching of a handful of monsters, so I can concentrate on ideas and strategy. Anyway, as little as it is that computers could do, UU will at most enable you to do that at snail's pace.

Even with UU, you're still not going to try those "surprise" moves. Remember that one room in RR? We both thought I had it in the bag, and then you discovered that crucial wait. How would UU have helped me think of that? It's only one more thing to try, how much trouble would it have been to try it? I simply didn't think of it. UU would enable me to try more things (not *nearly* *everything*), but I'd still first have to think of what to try.

The way I see it is that UU will help a lot with details. You still have to think of the right approach, and come up with the good ideas yourself. It will enable you to find the quickest way to get rid of the five roaches here and the two goblins there along the way, but it's not going to help you realise that if you wait a turn right at the start, then by the time you get to the roaches they line up nicely (and again, you have to *realise* something like that - you can't simply try waiting once at the start, go to the roaches and see what they do, try waiting twice, see what happens when you get there, wait on the second turn, wait in the entrance and the next square....). UU is not going to tell you that if you take a little extra trouble from the start so as to end up facing S instead of N, you can save a move at the end because you don't have to backswipe that last roach that ends up right E of you, you can use the wall to get it SE of you.

I don't think UU is going to make much difference to me on relatively short rooms. It will save time, yes, but it's not going to affect the scores I get. On longer rooms, especially those with bad checkpoint placement, it will help a lot. Note, it will *help*. It's by no means suddenly going to be easy to optimise longer rooms.

eytanz
Level: Smitemaster

Rank Points: 2708
Registered: 02-05-2003
IP: Logged

I don't have much to contribute to this discussion, especially not compared to some of the well thought-out arguments seen above (on both sides). I'd just like to point out that I find my opinion has shifted in the past few months, and I'd now support UU if it is included.

____________________________
I got my avatar back! Yay!

agaricus5
Level: Smitemaster
Rank Points: 1838
Registered: 02-04-2003
IP: Logged

Syntax wrote:
The reason for this is hours of trial and error. I can spot roaches gapping whilst they're still behind a wall. I can spot how many moves it will take to kill that damned roach queen. I can cut tar beautifully.

The reason is effort. UU will kill that and I think it's a shame. The skills of Rabscuttle, michthro and I will be wasted by it. We've spent hours memorising 2K+ rooms... and now it won't matter. Imagine telling Karpov he could replace his pieces. Now *that* would be silly...

This is exactly the counter argument. You're using trial and error, and yet you're still trying to memorise room sequences (both are at odds with each other). I imagine that few people here have the capability or time to do this, and memorisation of 2000 move room sequences doesn't actually work in the context of DROD's rooms (while maybe 50-move memorisations in chess might). Each room is different, and requires a set of concepts to solve. While you might recognise a sort of move sequence to adopt in a local situation, such as with clearing 3*x width tar blobs, you can't simply cut and paste solutions or move sequences from different rooms and expect to put them together to make an optimal solution. A room needs to be considered as a whole; memorising a solution doesn't necessarily help you understand the room's mechanics, nor does it teach you about the room and its elements.

You and others may have put a lot of effort into memorising solutions, but is that the point of DROD and optimisation? Why not just copy a solution instead of memorising it? Optimisation, as you pointed out, requires trial-and-error to help find novel solutions. What you are saying is that only people who have the time to repeat and memorise move sequences over and over (which I most certainly do not) should be advantaged to optimise. Anyone else hasn't put in enough "effort", and so doesn't "deserve" it.

Is that really fair, given the definition of optimisation? I have a current DROD playing time allowance of 3-4 hours per week - are you saying that to qualify for having spent enough "effort", I should spend 80% of my precious time memorising and replaying a solution set (for every 10 minutes of optimising in longish rooms, I estimate I spend up to 50 replaying and memorising moves)? Add to that the fact that the chance I find a #1 solution is maybe less than 1 in 4, so perhaps in a week, I may only find half an optimal solution on average (i.e. to get just one solution, I have to waste 160% of my total weekly playing time). Compare that to people who may be able to devote up to that amount per day, and I end up more than 7 times disadvantaged, time-wise and competition-wise.

I might as well become an architect.

____________________________
Resident Medic/Mycologist

BDR
Level: Master Delver

Rank Points: 106
Registered: 10-03-2006
IP: Logged

zex20913 wrote:
By the way--if you get a #1 on a difficult room in minutes, and you're pretty sure it's optimal, is that any different than spending hours on it? It seems like one of the claims you're making is that "We've put hours into this room." Well, what if somebody comes up with a better optimal solution in minutes? I recently beat a Rabscuttle #1 by about 160 moves. You read that right. I was dillydallying, saw the #1, said "Quoi", and went back to improve upon my #1 score. Now, I'm not going to say which room, because I want to have the score for at least until Rabscuttle gets back, but is what I've done wrong? If so, then I see that your perspective is against UU. I see UU as a correction of typos moreso than an optimization finder.

I have done this myself; probably not nearly as dramatically but it made a real impression on me. Unlike zex, I won't mind much if I lose my #1, so take a look at this room in MetDroid Prime (Tallon Overworld 2S3E if you don't recall or have not played it). There, I managed to get the #1 by simply going through the secret entrance instead of the normal entrance (I don't quite recall, but I think now that in passing through the first time my logic was "too many trapdoors, let's go somewhere else" [because I'm not very fond of red doors, trapdoor mazes or large areas of trapdoor tiles in general]). I said "What the.." when 1st place showed up without 'tied' attached, restored back, downloaded the other demos and put the relevant improvements in. UU wouldn't have shown a player like me that coming in through the secret entrance would save 20 moves; the only thing it might have done would have made me less apprehensive about trying the room the normal way.

Schik
Level: Legendary Smitemaster

Rank Points: 5383
Registered: 02-04-2003
IP: Logged

Yay, I get to make an official notice. Unfortunately, it's one that's more likely to start some heated arguments than some fun new spoiler.

Unlimited Undo is now out of TCB. Since we're still in development, this kind of thing happens. Features come, features go. The reasons why we came to this decision are scattered throughout this thread, posted by various people, so I'm not going to repeat them here.

If you are a supporter of UU, sorry. Feel free to take your frustrations out on me. Mod me down, call me names (please use PM if they're naughty), throw tomatoes at me.

____________________________
The greatness of a nation and its moral progress can be judged by the way it treats its animals.
--Mahatma Gandhi

zex20913
Level: Smitemaster

Rank Points: 1723
Registered: 02-04-2003
IP: Logged

Wow...well, I don't think I'll mind too much. I haven't experienced it anyway.

That check mark in front of this post though...that's got to go.

____________________________

Jatopian
Level: Smitemaster
Rank Points: 1842
Registered: 07-31-2005
IP: Logged

*hopes for binary mods*

____________________________
DROD has some really great music.
Make your pressure plates 3.0 style!
DROD architecture idea generator

Schik
Level: Legendary Smitemaster

Rank Points: 5383
Registered: 02-04-2003
IP: Logged

Jatopian wrote:
*hopes for binary mods*

So you're hoping that you are able to gain an unfair advantage over everyone else for high scores?

As Caravel has said, binary mods will not be allowed on the forum. And if people are caught using them in high score demos, I would not be surprised to see those demos disappear.

____________________________
The greatness of a nation and its moral progress can be judged by the way it treats its animals.
--Mahatma Gandhi

Jatopian
Level: Smitemaster
Rank Points: 1842
Registered: 07-31-2005
IP: Logged

Schik wrote:

Jatopian wrote:
*hopes for binary mods*

So you're hoping that you are able to gain an unfair advantage over everyone else for high scores?

No, I just like having unlimited undo. I've never tried for high scores; I think I acquired one by accident once, but that's it.

____________________________
DROD has some really great music.
Make your pressure plates 3.0 style!
DROD architecture idea generator

Schik
Level: Legendary Smitemaster

Rank Points: 5383
Registered: 02-04-2003
IP: Logged

Jatopian wrote:

Schik wrote:

Jatopian wrote:*hopes for binary mods*

So you're hoping that you are able to gain an unfair advantage over everyone else for high scores?

No, I just like having unlimited undo. I've never tried for high scores; I think I acquired one by accident once, but that's it.

Ah. Well, all that was ever *in* TCB was unlimited undo after you've already finished the hold. So this change won't affect you.

____________________________
The greatness of a nation and its moral progress can be judged by the way it treats its animals.
--Mahatma Gandhi