RE: The Rules - What I was Looking For

["Derek HAll" <derek@micromouse.me.uk>]

I believe the wobbly wall suggestion was for judgment of the non contact
wall followers and can't be applied to the other events. The only walls I
have seen put out of place by any mouse are ones that are already damaged
and should have been fixed before the contest. No mouse I can remember has
ever actually damaged the walls. Mice like Thumper were not given that name
for nothing; it would be a crime to exclude them.

The UEL maze is nearly 10 years old and has been carted the length and
breadth of the country, including the wonderful treatment it received at
Tecnogames, I think it stood up very well. Time for a new maze.

Although I think the non post maze design is probably better most mazes that
are currently under construction are based on posts so I think we need to
work with what is available.

Has a maze been sorted out for the event in November at the observatory?

Derek

-----Original Message-----
From: owner-micromouse@cs.rhul.ac.uk [mailto:owner-micromouse@cs.rhul.ac.uk]
On Behalf Of Martin J.Barratt
Sent: 27 October 2004 10:31
To: micromouse@cs.rhul.ac.uk
Subject: Re: The Rules - What I was Looking For

All

Nick wrote:
>>       The first rule should be that the Judges will encourage all mice 
>>       to be displayed and run on the maze at least once regardless of 
>>       their size or ability.
I absolutely agree - no question.

>>       The maze should be as strong as possible so that errant mice 
>>       have some chance of showing their ability.  Mice should only be 
>>       removed from the maze if the judges think they will actually 
>>       damage the maze to the detriment of other competitors.
A 'strong' maze seems to be a good idea - the UEL maze, although no
longer perfect, seems to provide adequate strength for 99.9% of mice.
Wall Bangers obviously require a minimum level of maze strength and I've
seen solvers make good use of solidly fixed walls on occasions!

Duncan replied:
>            The problem with wall contact is that some mice, currently 
>    wall-followers but possibly maze solvers in the future, use the 
>    walls for braking purposes. The apparent limit on performance is 
>    how hard you hit the wall. This then comes down to the judges 
>    deciding, on the day, whether the fastest mouse might cause damage 
>    to the maze. The only realistic test on the current rules is to run 
>    it on the particular maze and see what happens. Smashing a maze to 
>    determine if it is strong enough on the day of the competition 
>    doesn't seem to be the best way of resolving the issue. I am 
>    looking for an objective test that can be carried out by the 
>    competitor before the competition. As long as this test allows 
>    moderate wall contact, such as Sterling Mouse or my own DOT, there 
>    is no problem and mouse builders can use wall braking up to the 
>    known limit.
I agree that an objective test is a nice idea, but at what level of
'force' do we set the limit? The UEL maze is less robust than one made
from the plastic walls-and-posts and so will withstand less abuse...so
different thresholds?
Two potential methods have been suggested:-
A 'wobbly wall' (perhaps one with the interlocking mechanism removed)
will, if touched, be too easily knocked over or displaced by any mouse,
whatever its type.
A weighted wall _might_ improve the situation but will, I think, still
be easily displaced by the current crop of Wall Bangers - even when
travelling down a straight!
My feeling is that it is obvious when a mouse is 'abusing' the maze and
making use of maze-strength to gain a speed advantage and so, as this is
not an easy problem to define, is there not a case for not trying to
hard to do so?

>            I agree that the rule set should encourage entry rather 
>    than discourage it. By having a well-specified problem, mouse 
>    builders are more likely to succeed. I have seen mice turn up for 
>    competition incapable of dealing with the real maze for many 
>    different reasons. Usually they do not come back the next year. I 
>    would prefer it if the problem was not caused by the rules not 
>    specifying the problem well.
I remember one particular mouse was so large that it didn't fit between
the walls, let alone be able to make a turn...and I haven't seen them
since.
My feeling is that this particular entrant did not need finely specified
rules in order to design a suitable mouse - he just needed to read the
most basic aspects of what there was!

                        Simple is often beautiful
                        -------------------------

Regards
-- 
Martin J. Barratt