[RTTY] JARTS RBN stats

Wed Oct 21 16:39:37 EDT 2015

Well Dai san.
You don't need to ask for excuses.
Mine rant was really as slicing a air in four ... callsign rules and how 
they apply over time are a nightmare.
BTW you have 4 errors over 62 right catches.
 From your data Dai san ... (quick and dirty)
rank   dx                number of record
233    IW1QN        62
232    IW1Q           1
234    IW1QNN      2
235    IW1UN         1

That error number over the right catches number, BER, is more than 
acceptable. It's a 6.45%, isn't?
Over the more big chunk of the data that Tomi showed us, the error rate  
is quite negligible 0.68% .
 From Tomi's data ... (quick and dirty)
Day of date    dx       Number of Records
17th           IW1QN     2.157
18th           W1QN      2.560
xx                                #4.717
18th           IW1Q         8
17th           IW1Q         8
xx                                 #16
18th           IW1QM      9
17th           IW1QM      7
xx                                  #16
18th            IW1QNN    6
17th            IW1QNN    6
xx                                  #12

(BTW the 0.68 contain also UR 4/62 Dai san, but negligible is still 
negligible, isn't?)

(May I consider that IW1UN would exist ... say hello to him ... 
mumbling, but somewhere that record was pruned, isn't?)

Also the same simple operations done on raw data files for real 
worldwide data (raw files on RBN) gives (even more quick and dirty and 
with strings samples catted out - believe me):
17th    43/2157  1.99%  error rate
18th    53/2560  2,07%  error rate
A raw attempt to sum the two days gives:.
96/4717  a mere 1.97% error rate, isn't?
Some strings/data_recognize errors may apply, as I sorted RAW csv for 
IW1* and then choose by hands all what may seemed to  me applicable, 
i.e. leaving out  IW1* in CW mode (easy) and other different calls 
IW1C/F and more ...
Not all mismatched call all shown everywhere, Dai's, Tomi's and RAW RBN 
csv, but percentages remain low to negligible, IMHO. I mean that not all 
the mismatched strings call had seen in every place. This leave space 
for any further speculation. As VE7CC wrote here, as example.

So, my rant went quite illegitimate as numbers, 32/4717 or 96/4717, 
speak more favorably than what I'd think-tanked alone. Did you seen, 
4717 record for Tomi's data and RAW RBN ... to have further conclusions 
is up to anyone.

I believe that I have had more AGN AGN or AGE? AGE? than those show by 
the ratio of right and wrong catches of the call I am eviscerating. This 
looking tree data sources, one included in the others, plus of course 
the memory of those wonderful 48 hours of the last JARTS that all, well 
we 3, have had operation the call IW1QN @IQ1RY.

I could stigmatize even more than what I have done how IW1Q is a 
remarkable error.
But IW1QNN and IW1UN aren't errors by rules. So theirs owners could pop 
up anytime from the dark.
That's is pretty right for IW1QNN, as this call is in the same series of 
IW1QN. Not so for IW1UN, that
to my knowledge was never assigned, but still it's a viable call to be 
seen on air.

That's will not as easy as it it is to discriminate as the case as a 
friend like IW1AU will come in RTTY ...

Anyway just looking,  I'd not deeply analyzed the whole set of calls 
furnished here by anyone, it seems that there are more higher error rate 
for other calls. But then my knowledge of other nations call creation is 
really limited.

Then we have to remember that propagation and other setup & working 
conditions plays theirs respective roles.

Anyway as Dai san data and more the worldwide data presented by Tomi 
shows very low error rate I have to ask my excuses to anyone as to have 
used the bandwidth here for an improper rant. Sorry.
May I say that I was catches by the easiest profiling of numbers don't 
looking at the whole thing. That's maybe my fault.

Well, I lose another good silence moment. :-(

             73 de iw1ayd Salvo

PS More, this time, differently than in January, February and the VOLTA, 
I haven't seen the call I was operating mismatched on the bandmap of my 
beloved N1MM+. This would confirm, at least to me, that things are going 
to be better and better since then. I was on VE7CC cluster by a common 
operators decision for the whole JARTS. :-)

On 21/10/2015 18:00, rtty-request at contesting.com wrote:
> Message: 4
> Date: Wed, 21 Oct 2015 13:30:27 +0900
> From: Dai NAGAKURA<ndai at cameo.plala.or.jp>
> To:rtty at contesting.com
> Subject: Re: [RTTY] JARTS RBN stats (Tomi kallio)
> Message-ID:<20151021130954.17AC.272C9EF at cameo.plala.or.jp>
> Content-Type: text/plain; charset="US-ASCII"
>
> de JF2IWL Good morning
>
> Here is analysis of my RTTY skimer server on JATRS by SH5 analyzer.
> http://59925.org/sh5/jf2iwl/2015/2015_jarts_ww_rtty_jf2iwl_skimmer/
> 8783 stations in 48h may be little fewer than Tomi's result.
>
> In 1,189 unique callsigns,
> http://59925.org/sh5/jf2iwl/2015/2015_jarts_ww_rtty_jf2iwl_skimmer/allcallsigns.htm
> Over half stations should be busted or "Caller's Callsign"
> , on White and Blue belt (once or twice captured) may be these.
>
> Of course I undesrtand RTTY SkimServ cannot avoid capture caller's call.
> and Busted
>
> Pse excuse me Salvo,
> I captured you as IW1UN ,IW1QNN etc.
> I did not use Master.dta to catch all station.
>
> Dai JF2IWL