| |
BGonline.org Forums
Nacbracs
Posted By: Nack Ballard In Response To: Nacbracs (Matt Ryder)
Date: Monday, 28 December 2009, at 7:20 p.m.
Nack (previous): The problem with writing nacbracs in hundredths or quasi-hundredths is that it is in direct conflict with writing them in thousandths.
Matt: I considered this problem, which is why in my original post I suggested starting the ordinal scale at 0, as follows:
0.Tied: 0.00 – 0.01
1.Barely wrong or Very close: 0.01 – 0.02
2.Marginally wrong or Close: 0.02 – 0.03
3.Wrong, or Mistake: 0.03 - 0.06
4.Blunder: 0.06 - 0.10
5.Whopper: 0.10 - 0.20
6.Double whopper: >0.20Yes, I understood that. At the low end, this is exactly the same as rounded down hundredths (instead of thousandths).
With that scale, I'd notate your example [$=U S2 D3] as
[$0 U0 S2 D3]
If you write it either of those two ways, it is the ordinal scale's translation of our Gnu 51S-21 example [$ U3 S28 D34]. The $ and U are tied at "0" (the .00-.01 range) which is also what "=" means, S is at "2" (the .02-.03 range) and D is at "3" (the .03-.04 range).
The use of zeroes should clearly signal that the B/W scale is used here.
Not unless I create a special convention that has not yet been discussed. If someone currently writes [$0 U0 S2 D3], I interpret that as $ and U being exactly tied to the nearest thousandth (-.000), S being 2 thousandths back (-.002) back, and D being -.003. It is shorter to write [$0 U0 S2 D3] as either [$ U0 S2 D3] or [$=U S2 D3], but all three mean the same thing to me.
Assuming such a system caught on, I expect readers would become familiar with the convention that if the numbers are all in the range 0-6, they should ordinarily expect the B/W scale. In the very rare circumstance where you want to use the thousandths of a point convention and all the numbers are in the unlikely range 0 through 6, you could notate as follows:
[$ U1 S2 D3]
Without contextual reference, I cannot tell if this is being written as normal (in thousandths) or as ordinal (in hundredths). Perhaps I am missing your point.
Also, I don't find it at all rare for all plays listed to be in the .00 to .01 range (especially when there are only two). For example, our current Gnu 51S-21 example is often [$ U3] without mention of other plays, and Stick's XG rollout, which included only those two plays, produces a bot average of [U $9]. And, off the top of my head, 51S-62 at DMP (recently addressed by David Rockwell) is [$=H R5 S7]. In part, this is because Nactation caters to early game moves more than to later stages of the game.
This might be a tad confusing in the beginning, but I'm sure with some careful explanation initially it would be a snap for all the very smart folks who'd be inclined to use it.
Granted.
Assuming you have a list of close picks, it might be useful to employ your "=" convention as follows:
[$0=U=R Z2=S D3]
This would imply that the Slot, Up and Run plays are all in the range 0.00 – 0.01. The Reverse Split and Split plays are in the range 0.02 – 0.03. The Down play is in the range 0.03 - 0.06.
Note that it's possible to illustrate equality ("=") here for categories other than just 'Tied'.
The = sign is a redundancy that can be dispensed with altogether; e.g., [$UR ZS2 D3]. I just use = as a special alert that the top plays are exactly tied. I write the second set as Z2 S2 if I want to retain added clarity or as ZS2 for brevity.
I would tend to resist the term "quasi-hundredths" to describe the increments of such a scale, as only the first three members of the set really fit the description. The scale should be understood as ordinal, not cardinal.
The reason I included "quasi-" is because the scale describes wider ranges from 3 up, but I like your term "ordinal" better anyway. :)
As before, I have no enthusiasm for creating a break-away 'dialect' of nacbracs. If these ideas (or something similar) were "sanctioned" by yourself and added to the corpus, I would make use of the convention. My thoughts are offered to you as mere suggestions for your consideration, and you are quite welcome to ignore them if you don't see their merit.
Much appreciated. As I see it, there are three issues:
(1) Currently, it is only top players using nacbracs. Thousandths are important to them, especially for early game situations, where Nactation is most valuable.
(2) I believe the concept of a second set of numbers can be improved upon. I'd feel more comfortable proposing the idea of "tvcmbwd" (for tied, very close, close, etc.), letters that even stand for what they mean (much of the intent of the B-W scale).
(3) The use of nacbracs is not yet widespread enough to withstand an alternative system.
When (3) is far enough along, and players of different skill levels find nacbracs useful, some may prefer to think in B/W ranges rather than thousandths. At that point, perhaps we can gently introduce the idea.
Thanks very much for your input.
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.