In reply to tom_in_edinburgh:
> I don't think so The suggestion I was responding to in my original post was to increase 'accuracy' by calculating a mean of suggested grades in the normal 'interval' style system which results in a decimal answer and truncating the decimal bit to get a 'more accurate' grade in the original system. If you do that the result will be wrong.
Increasing precision does not increase accuracy. If I say a climb is E3, and you say it is E3.4, and it is actually E3.3, you have arguably provided a less accurate answer than me. I have just claimed a lower precision.
Incidentally, if you are going to use numbers with discrete steps which fall exactly on 0.5 (e.g. the 0.1 steps suggested here), you should consider rounding-to-even rather than rounding-up. The rule you learn in school to round 0.5 up is neither the only rounding rule nor the 'correct' one... if you round to even then you round every other 0.5 up or down, which avoids adding an overall upwards bias to your rounding.
Or maybe we are massively overthinking this? :P
As already pointed out by several people, we could define the 'E2' grade (which is currently at least a category, not a number) as either E2.0 to E2.999...9 (where the second number is actually a number while the first is a category), or as E1.5000...1 to E2.4999...9 (and specify a rounding rule to deal with E2.5). Both are equally valid approaches, both have advantages and disadvantages, and neither is 'wrong'.
Personally I think the whole thing is mad; by all means solicit higher precision numbers from people to get better grade estimates, but then average them down to a reasonable level of precision when you actually publish the guide - almost certainly not 10 distinct subgrades!
PPS as any good scientist knows, a number without error limits is meaningless. Decimal numbers are often assumed to have an implied precision from the number of decimal points e.g. E3 is +/- 1, E3.5 would be +/- 0.1, but really I would like to see E2.7 +/- 0.3 before I was happy...
Post edited at 13:43