StuartK wrote:
It's not just tedium that will be lost with this change, however.
DROD for me is very much akin to a war game - the techniques for optimising it are all about strategy and tactics. Strategy (spotting the tree) will overall remain unaffected, and tactics (the nitty gritty of optimisation) will. The tactical skills some have developed will be rendered largely irrelevant with unlimited undo - others will be able to do the same, not by skill, but by excessive reliance on undoing.
But like I said, do you actually
see those tactics before you go off and play a room or a section of it? You may be able to say what sort of thing needs to be done, like for example, you know that you need to kill the two goblins as fast as possible because the wraithwings get themselves trapped, or that the roach in column 25 is the limiting factor, so you need to get it out
fast, but can you see immediately
exactly which moves are required? I personally cannot, for the reasons mentioned earlier, so I need to experiment and if something does seem to work, then I can try to understand why it happened. For example...
In KDD 2.0, L2 1E, Rabscuttle previously had a winning demo that looked pretty optimal. In such a situation, I couldn't see any tactics that could possibly increase my score, since his were already so good, and I could see why they worked. Therefore, I accepted defeat, and played the room myself, as he did it, just to see if I could also match his best. However, unknown to me, I made a logical, but nevertheless accidental typo that made all the difference in this room (you'll see it at move 111, I think). So, I was pleasantly surprised when I realised my move count was one lower than his, and that I would now hold the new #1 score, although I had no idea why at the time.
Unless you're a computer, experimentation as an iterative process is the key to finding new solutions, and the "
tactic"
is a test of whether you can use what you discover in your experiments to help you find an optimal solution, as well as what you can analyse theoretically.
To take another example, the optimal solution in KDD 2.0, L7 2N is highly dependent on monster movement order. If you never tested the room, how would you find out what the order was? You could work it out theoretically, admittedly, since this room is simple, but what about in a horde room like JtRH L1 2N 3W? I severely doubt even Ricky could have got his solution
intentionally, on his first try. Indeed, it probably took him several experiments in which he must have realised what the movement order was, and then used it to work out how to get those roaches to line up nicely. In fact, I don't see there is much difference between your definitions of "
strategy"
and "
tactics"
. Basically, in optimisation, you need to search for what seems to work, and go off and find out if it does, refining your technique if you notice or predict that what you're doing is making you more efficient. Optimisation is about iteration by definition. I just don't want to reiterate moves that are already optimal.
____________________________
Resident Medic/Mycologist
[Last edited by agaricus5 at 02-19-2006 03:42 PM]
)
Only noticed now you've caught up again in record time. Talk about not being able to keep up..
)! Also, did you do that without replaying the room at all, or did you use the restart key?
