Ringer errors

Shows how the band as a whole rang when ringing in a given position or following other bells. When ringing in a given position shows for example whether the band is weak at lead or whether they are habitually late at backstroke in last place. For following other bells, shows how the band as a whole is affected by bells other than the one each is ringing. Might indicate "fear of the bells", or perhaps that the band is distracted by a ringer's handling style.

Bell Error Tables

Shows how each bell rang. There are eight tables covering all combinations of:

• Handstroke and backstroke
• Mean (i.e. average error) and standard deviation (i.e. amount of randomness)
• Bell-after-bell and Bell-in-position

Perfect ringing would have zeros in all squares of the tables. Each square is the mean or standard deviation of multiple individual striking errors and where there are fewer than five occurrences the square is marked grey as the value may not be a statistically significant indicator of striking faults. Otherwise pink highlights negative values and orange positive values, and the darker the colour the worse the error.

If the touch contains rows with errors above the threshold the rows containing the errors are ignored for the purpose of these tables. This improves the likelihood that the dark coloured cells reflect striking faults even when there are method errors which would otherwise make the numbers meaningless. The row(s) of this touch that were ignored were:   .

Temporal Analysis

This set of graphs shows how the touch evolves over time.

• Peal Speed - how the peal speed varies throughout the touch
• Handstroke Gap - how the leading evolves throughout the touch (typically varies with peal speed)
• “Distance from Rounds” - 0 = rounds, 1 = backgrounds. A measure of how mixed the bells are.
• Faults over time - holistically, then split by hand and back

Error Histogram - not a hugely useful graph

This graph shows the number of errors(*) of different sizes. The peak or "median" value represents the "everything going right" capability of the band.

The shape of the histogram should be roughly that of a bell with its mouth down but method or ropesight errors cause the curve to be lop-sided on the right. If this happens the reported Standard Deviation for the touch will be higher than the median and the difference between the two shows the disruptive effect of such errors.

When working on improving the overall standard of striking in the band, recording the median standard deviation of each touch is a good way of monitoring overall progress.

----------
(*) measured as the standard deviation from the ideal according to the Contest model

Downloadable Resources

• List of Changerows
  
• List of Changerows after correcting for method errors
  
• Strike data in Lowndes format, e.g. "H 1 0X0000"
  
• Strike data and ideal times in CSV format
  
• Bell Summary
  
• Recording of ringing (if available)
  

Touch Id = TOUCHID
Training Data Recorded = TRAINDATARECORDEDON
Training Data levels = TDATALEV
Trained on = TRAINEDON
Source filename = SOURCEFILENAME
Recording level (dB) = TOUCHDB
Transcription options = TOPTIONS
Analysis options = AOPTIONS
Code Versions = VERSIONS

 
Association = ASSOCIATION
Calling = CALLING
CompositionName = COMPOSITIONNAME
Footnotes = FOOTNOTES
Ringers = RINGERS

 
QUALITY