Angel ,'Chosen'
A thread to discuss naming threads, board policy, new thread suggestions, and anything else that has to do with board administration and maintenance. Guaranteed to include lively debate and polls. Natter discouraged, but not deleted.
Current Stompy Feet: ita, Jon B, DXMachina, P.M. Marcontell, Liese S., amych
Mathematical average seems to be easy to skew.
Not if we set a fixed range. Which we're doing.
In other words, we're saying choose between 10 and 100. If we said, "choose any number" then, yeah, someone could put in 10,000 and skew the results. With a max of 100? Not so much.
I'd prefer median to mathematical average. Mathematical average seems to be easy to skew.
Not really, I mean, median *IS* a mathematical average. It's a narrow enough range that it shouldn't be skewed to badly, and it's simple enough to run the numbers for all three measures for a sanity check. (Not that anyone would want to use mode, it's just that mode is fun.)
HEY! I want to use mode!!! ; )
If we said, "choose any number" then, yeah, someone could put in 10,000 and skew the results. With a max of 100? Not so much
Jon, would we be doing that adjustment for extremes, like in figure skating?
HEY! I want to use mode!!! ; )
Oh, sure, just because it's so stylish...
Jon, would we be doing that adjustment for extremes, like in figure skating?
Hee! If that's a serious question, I'd say no. In figure skating it's more important to drop the extremes because there are only, what?, 10 or 15 judges? So an extreme score can have a large impact. We'll probably get 100 votes so dropping one "10" vote and one "100" vote doesn't gain us much.
Hee! If that's a serious question, I'd say no. In figure skating it's more important to drop the extremes because there are only, what?, 10 or 15 judges? We'll probably get 100 votes votes so dropping one "10" vote and one "100" vote doesn't gain us much.
It actually was. *g*
I'm so used to doing it, because well, we did it all the time the year I took stats (it was important to do so with the data we were studying, to clarify), and I do it (or did it, as I'm no longer doing that) when adjusting data for horse racing information, that I tend to assume the correction.
First of all, I’m all for the new way of selecting MVT and seconds. Letting people enter their choice and taking the mean is cooler than preferential voting on a fixed list. (All about the stats, baby! I wanna see mean, median, mode, standard deviation, distribution curves, yeah!) And having limits on the range is obviously necessary.
Now we’ve just got to quibble about the limits. Typo Boy wants seconds to be 1-15 rather than 1-10. Fine with me.
Somebody’s got to quibble about the MVT limits, and I guess that’s me. I think both 10 and 100 are too high. I’d suggest that 100 is unreasonably high. The second ballot we did got only 100 votes, so I think 100 for a limit is way, way too high. If the upper limit is unreasonably high, the lower limit should be unreasonably low, like say 2. 10 is, well, reasonable. If the lower limit is reasonable, the the upper limit should be as well. What’s a reasonable number for the upper limit? I’m not sure we have enough data to say yet. Last week the number 65 was mentioned several times. I’d be willing to accept that as reasonable for now.
So either 2-100 or 10-65, please?
Why quibble over 2 vs. 10 for the lower limit? 32% of us voted against MVT on the first ballot, so consistency would dictate that we vote for the lower limit, which will skew the mean lower. I’m warning y’all, I’m ready to fight every step of the way – if not 2, then 3. If not 3, then 4.
I got a whole lot more to say on this, but it’ll take me a while to get my thoughts together.
Jon, would we be doing that adjustment for extremes, like in figure skating?
t pedant Gymnastics and diving drop the high and low score--figure skating uses the marks as placeholders to rank the skaters, so that the winner isn't the skater with the highest point total, but the one who gets the majority of first-place rankings from the judges. Until this year, all the marks counted, but now they've got the stupid anonymous system where a computer randomly selects which judges' marks will count. The I-D-O-T idiot in charge of the ISU, who comes from the speedskating side of the sport and doesn't know from figure skating really, wants to change the scoring system even further, rather than something logical like, oh, I dunno, CRACKING DOWN ON CORRUPT JUDGES AND FEDERATIONS. t /pedant
t natter Susan, if you're interested, Math Horizons (a math magazine for undergrads) had an article about figure skating judging this month. t /natter