response 249 of 1578:

Jan 7 01:56 UTC 2003

(complain, complain, complain... someday, if I win again and
the contest lasts a while, I'll write a script to score things.)

awful

response 251 of 1578:

Jan 7 02:40 UTC 2003

freak

response 253 of 1578:

Jan 7 03:17 UTC 2003

blank

response 255 of 1578:

Jan 7 03:50 UTC 2003

skink

response 257 of 1578:

Jan 7 04:18 UTC 2003

servo 0 lastword
-----
awful 0 russ
crest 0 remmers
freak 1 jep
-----
blank 2 jep
-----
skink 0 gelinas
think 0 russ

(I think one way to get high numbers of guesses is to score
infrequently. I could have let people guess words ending in "k" a long
time.  As it is, this one is almost solved.)

response 259 of 1578:

Jan 7 04:57 UTC 2003

flare

russ, if you'd like a copy of my scoring script, let me know.

response 261 of 1578:

Jan 7 05:11 UTC 2003

Well, you gotta remember, I've been playing since 1994.  And at various
points during the last nine years, feathers have been ruffled because of
scoring mistakes, so it made sense to do scoring automatically.  kentn wrote
a system that several people have used, and I wrote a simple awk program
that does much the same thing.  I'll mail it.

response 263 of 1578:

Jan 7 05:23 UTC 2003

There is one ~gelinas/Public/do.mm, but it would need some tweaking.

response 265 of 1578:

Jan 7 05:38 UTC 2003

crumb

response 267 of 1578:

Jan 7 08:20 UTC 2003

The scoring script I use is in HyperCard, and includes the utility I 
wrote to keep track of what letters have been eliminated and crosscheck 
against a list of 5-letter words.  I keep considering writing it again in 
BASIC so I can compile a faster version, but I'd have to write a whole 
interface for it, and I just haven't felt motivated to do it.

response 269 of 1578:

Jan 7 14:23 UTC 2003

servo 0 lastword
-----
awful 0 russ
crest 0 remmers
freak 1 jep
-----
blank 2 jep
-----
skink 0 gelinas
think 0 russ
-----
flash 3 jep
flare 3 aruba
-----
crumb 0 gelinas
-----
scrip 0 other

(I sometimes try to figure out other player's strategies based on the
guesses they make.  Usually I can make some sense of it, and even draw
some conclusions about what kind of analysis is behind it, but those
last two guesses are a mystery to me.  "Scrip" for example, contradicts
six out of ten previous results.  It does eliminate one possible
solution word, but if you wanted to eliminate that, the guessing it
would be the better course, since it might be right.  "Crumb" is pretty
much the same, contradicting 4 of 9 previous guesses and eliminating
only one possible solution.  Probably I'm over-analysing.  Some people
are just guessing words as they pass through the conference, without
playing very seriously.)

response 271 of 1578:

Jan 7 15:50 UTC 2003

Generally I just look at the list of words and try to find a new word 
which might fit all the criteria.  My early guesses are just whatever 
word I think of first.

Once in a while I try to get more analytical, and list all the letters 
which are possibles, and the ones which are eliminated for each 
position, and use pattern searches in on-line dictionaries, and all 
that.  I'm not very good at doing that.

No, that's not true, either.  I'm pretty good at it.  It's just that 
several others here are much better.  Jan seems to be the most 
successful, but I suspect aruba is the real master.  He never makes bad 
guesses.