Monday, December 27, 2010

Draft proposal for SI units for bitching...

... as proposed by Jeri Ellsworth:  the "Jones", after EEVblog's Dave Jones.

As with all SI units, we must define it in terms of base units, so I propose this:

1 Jones = (# of complaints about a topic) * (sum of levels of annoyance caused by said complaints weighted by exposure frequency to causative events, in seconds between exposures) * (# of people likely to be exposed to your ranting)

Obviously, the level of annoyance varies from person to person; I imagine a level 1 annoyance ("mild eye rolling followed by repetition of action until desired result is achieved") for a more patient colleague might be a level 5 annoyance to me ("loud sigh with posture change and muttered profanity") and could rank to a level 100 for someone else ("widely distributed internet video rant with attendant behavioral change and increased probability of property damage to exact retribution on an inanimate object").

As a practical example I'll float my issues with OrCAD- there are five things which cause me pain every day with OrCAD, three level two issues ("eye rolling with quiet sigh") and one level 6 issue ("profanity audible but not comprehensible to workers in adjacent cubes") and one level 50 issue ("loud cursing and banging of mouse on desk").  Weighted, I'd say I get one level two event every 60 seconds of OrCAD usage (.033), one level 6 issue every 300 seconds (.02), and one level 50 issue every 3600 seconds (.014), for a total weighted annoyance of .067.  At best, normally, about four people are exposed to my ranting; thus my irritation with OrCAD is about 5*.067*4 = 1.34 Jones.

The nature of this is such that it is similar to the Schmidt Sting Pain Index- individual preferences and tolerances will affect the outcome heavily, as does personal audience size (clearly, Dave has a much bigger audience than I do).


  1. This is far more practical than the old fashioned unit, the "Cadence", which is equal to about 2.7*10^33 Jones.

  2. Would it be appropriate to suggest an attenuating factor that is proportional to the proximity to the end of the working week? Experiencing some noticable level of Jonesian excitation at 4:55pm on Friday has got to limit the annoyance factor by some empirically measurable quantity.