RE: Rules etc.

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

Remove the + run time bit and apply the rule as it was supposed to be
applied. We must not penalise mice for running their top speed in search of
ultra reliability or everything gets very boring.

-----Original Message-----
From: owner-micromouse@cs.rhul.ac.uk [mailto:owner-micromouse@cs.rhul.ac.uk]
On Behalf Of Duncan Louttit
Sent: 19 October 2004 20:09
To: micromouse@cs.rhul.ac.uk
Subject: RE: Rules etc.

Hi Everyone,

         An easier way to score is to make the touch penalty infinite so 
that all untouched mice beat all touched mice.

         I don't like that suggestion much either!

         The current system does seem to give the desired result but it is 
so complicated that it is not followed at informal competitions. Any rule 
that needs an example to explain it may be considered on the borderline of 
complexity, however I don't have any better proposal.

         Regards,

         Duncan

At 22:32 18/10/04, you wrote:

>Hi all
>
>In reply to Ken - the rules allow it, but it does incur a penalty.
>
>"4. When the micromouse reaches the destination square, it may stop and 
>remain at the micromouse maze centre, or it may continue to explore other 
>parts of the micromouse maze, or make its own way back to the start. If 
>the micromouse stops at the centre, it may be lifted out, manually, and 
>restarted by the handler. Manually lifting it out shall be considered 
>touching the micromouse and will cause a touch penalty to be added on all 
>subsequent runs."
>
>
>This is one that had me confused for a while, as I thought I could see an 
>opportunity for a bit of unsportsman-like behaviour - perish the thought!
>
>However, the loss of 3+(runtime/10) seconds for a touch seems to heavily 
>outweigh any advantage to be gained by the mouse not retracing its steps. 
>If a mouse had found its way to the centre (run), and returned via the 
>same path, one can assume a similar return-time to the run. This would 
>amount to only (runtime/30) seconds additional search time - clearly much 
>less of a penalty than the 'touch' method of returning.
>
>
>The last sentence of the quoted section of the rules above states that a 
>touch penalty will be added on ALL subsequent runs - this surely means 
>that the touch penalty is cumulative - or to be more accurate, more 
>cumulative than I thought!
>
>If we ignore the (run-time/10) contribution for simplicity, 3 runs each 
>with touches would incur a penalty of 3+6 = 9 seconds on the final run.
>5 runs each with touches would incur a penalty of 3+6+9+12 = 30 seconds 
>penalty on the final run. This implies that the mouse would need to go 21 
>seconds faster on the 5th run to gain any advantage over the 3rd run, and 
>that in general it would need to go significantly faster than any other 
>mouse in the competition. The outcome would surely be that the mouse peaks 
>at run 2 or run 3, thereafter the handicap would rule out any further gain.
>
>This suggests that the penalty for repeated infringment is extreme .. 
>probably as it should be.
>
>Have I interpreted the rule correctly?... does anyone have an alternate 
>interpretation?
>
>Does anyone else think there must be an easier way to score?
>
>Regards
>
>Tony
>
>
>
>-----Original Message-----
>From:   owner-micromouse@cs.rhul.ac.uk on behalf of Ken Warwick
>Sent:   Mon 10/18/2004 7:02 PM
>To:     micromouse@cs.rhul.ac.uk
>Cc:
>Subject:        RE: Rules etc.
>Duncan
>
>Am I right in assuming that the rules disallow picking up a mouse when it
>has reached the centre and returning it to the start for another run?
>
>Ken
>
>-----Original Message-----
>From: owner-micromouse@cs.rhul.ac.uk
>[mailto:owner-micromouse@cs.rhul.ac.uk]On Behalf Of Duncan Louttit
>Sent: 18 October 2004 12:35
>To: micromouse@cs.rhul.ac.uk
>Subject: RE: Rules etc.
>
>
>Hi Everyone,
>
>         There seem to be no comments about the last couple of proposals 
> so they
>have been written up. The sensor system is left at 1cm above the floor for
>compatibility, but I hope that we can better specify the sensor system at a
>later stage.
>
>         On with the rest of it, rule 1 of the micromouse spec.:
>
>"1. Although the superstructure of a micromouse may 'bulge' above the top
>of the micromouse maze walls, a micromouse is subject to the following size
>constraints -width 25cm, length 25cm. There is no height limit. A
>micromouse must be completely self-contained and must receive no outside
>assistance. This rule will be relaxed for that part of the competition
>allowing entries from competitors or teams under 18 years old to compete in
>a sub-section for mice using external guidance systems, automatic or
>manual."
>
>         I seem to remember someone proposed that we remove the rule for
size
>limitation but I can't remember the reason. Whoever it was, please try
>again.
>
>         The specification of width and length has caused some controversy 
> in the
>past. I prefer something along the lines of "the micromouse must fit in a
>box 25cm square" to avoid arguments about mice like Michael Shumouser that
>has one horizontal distance of more than 25cm.
>
>         Comments please!
>
>         Regards
>
>         Duncan
>
>
>
>
>