# .666

Discussion in 'Nuttin' but Net' started by gatordavisl, Dec 30, 2013.

1. Offline

### gatorjeff20017Active Member

Joined:
Dec 2, 2013
Messages:
290
37
Trophy Points:
28
+207
Your second paragraph is beyond incomprehensible. A 5 percentage point difference (much different than a 5 percent change, but you inexplicably changed terminology in mid-sentence) relative to 90 percent is not a 100 percent change. Was eager to read your paper, but wasn't clear how to access it after bringing up the link you provided.
Last edited: Jan 19, 2014
• Agree x 1
2. Offline

### bayou_gatorActive Member

Joined:
Sep 8, 2013
Messages:
448
46
Trophy Points:
28
Location:
Among the savages
+347
I was hoping Themi would clear this up.
3. Offline

### themistoclesWell-Known Member

Joined:
Apr 8, 2007
Messages:
12,530
73
Trophy Points:
48
+360
Variability is expected. I was merely making the point that percentage differences are different than simply arithmetic differences due to the loge nature of percentage distributions.

Statistical significance is really one of the least useful things to use, although the current meta-paradigm is utterly trapped within R.A. Fisher's 1925-30 Analysis of Variance model, which was adopted almost in total (original source of the probabilities [.01, .05, .10] was actually Fisher's perceived enemy Karl Pearson). The thing about significance in testing is simply that as the sample size increases everything, even absolutely meaningless differences become significant because the denominator is based on sample size. The question is meaningfulness. I am not sure whether that difference is in fact meaningful, but it may possibly be.

If your statistics professor taught you only about significance and not about relative magnitude and meaningfulness, then this clearly shows that this person, like most people who conduct statistical analysis throughout most of the world, have little understanding of what and why they are doing this.
4. Offline

### themistoclesWell-Known Member

Joined:
Apr 8, 2007
Messages:
12,530
73
Trophy Points:
48
+360
A 90 to 95% change is not in fact 100%, but rather 50%. I should instead have used 1% to explicate what I meant to say.

A difference from 50% to 51% is a 2% change. A change from 1% to 2% is obviously twice as big and is a 100% change, so is a change from 98% to 99% because you are looking at the difference from 100% and it doubles between 98% and 99%. This is the way the loge distribution moves.

Sorry about the error, I should have used 1% rather than 5% it's just that most people understand 5% better than 1%, thus my foolish error.
5. Offline

### gatordavislWell-Known Member

Joined:
Apr 8, 2007
Messages:
10,696
348
Trophy Points:
83
+1,205
Never took a course in statistics. My degrees are in the arts. We learned just enough about statistics to apply analyses in our dissertation/research. As you know, though, deriving conclusions from statistics that were applied to human behaviors (perhaps even more so in the arts) is tricky, if not altogether problematic.
Last edited: Jan 20, 2014
6. Offline

### gatordavislWell-Known Member

Joined:
Apr 8, 2007
Messages:
10,696
348
Trophy Points:
83
+1,205
Only 13 attempts today, but the Gators shot 76.9% on 10/13. Keep up the improvement Gators!
7. Offline

### 08gatorbaitWell-Known Member

Joined:
Nov 27, 2013
Messages:
1,631
191
Trophy Points:
63
Location:
Tampa
+965
When it matters we've done well lately. 8-10 at bama in final minutes. Im more concerned about the scoring droughts we hit sometimes and especially three point defense. Three point shooting mid majors pull upsets by going nuts from there...can't let that happen. Our offense vszone needs work too.
8. Offline

### gatordavislWell-Known Member

Joined:
Apr 8, 2007
Messages:
10,696
348
Trophy Points:
83
+1,205
Thought I'd provide a FT shooting update to reflect the first half of the conference season. Results indicate that the Gators have not improved at the stripe in SEC play. Wilbekin has upped his percentage, but DFS has sagged. The numbers:

Mizzou 24/33 72.7
Texas AM 18/29 62.1
Miss St 8/19 42.1
Tenn 10/13 76.9
Bama 12/18 66.7
Aub 23/33 69.7
Georgia 11/20 55.0
Ark 25/34 73.5
USCe 17/24 70.8
Totals 148/223 66.4
• Like x 1
9. Offline

Joined:
Apr 8, 2007
Messages:
5,156