Statistician's Digest: How to Determine Standard Deviations ?
... How do I determine where a player who takes 1000 FGA in 2000 Min in a season, stands in terms of standard deviation as compared to the league rate of FGA/Min ? ...
Make the statistical computation FGA/Min for every player. Find the standard deviation of all those values. And then the formula is:
[ (1000/2000) - (Avg FGA/Min) ] / (stddev)
Also, it is possible to use to use percentiles. Sort all players by FGA/Min, give everyone a rank. Divide the rank by the total number of players in your sample.
Performing this calculation we must not mistake precision with accuracy. Nonparametric elements like percentiles make no hypothesis that is taken for granted about the shape of the distribution: they are more cautious measures and much less likely to return errors at the the highest degrees.
If you're searching leaders in a category, I recommend you to always use nonparametric measures with a common plan with parametric measure, merely to be sure there are no outliers into your results.
The weighted standard deviation doesn't succeed in seizing the unfairness between two players with the same rates in different quantities of playing time. Same rate is as if: that rate and no more. What it would succeed in catching is a significant change between the width of the distributions, having an effect upon the standard scores. The differences between player rates would be involved, but not if they have the same rate.
When analyzing numbers with measures that aren't well fully apprehended, I give preference to study at the numbers from multiple angles. In this case which you propose at your question, I would perform the computations using
1. Percentiles
2. Standard deviation
3. Weighted Standard deviation by minutes
4. Weighted Standard deviation by square root (minutes per game)
and examine and note the similarities or differences of them. There shouldn't be much significant change, but it would be holding the attention if there was.
( Prof. Roberto Azar )
Source: eBA Basketball Statistics Analysis System
You must be logged in to read or post a comment. Register
tags: basketball ~ statistician's digest ~ basketball analysis ~ statistics
visit: eBA Portal ~ eBA Forums ~ eBA Encyclopedia ~ eBA Clinics
No Comments/Pingbacks for this post yet...
Comments are not allowed from anonymous visitors.
| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| << < | > >> | |||||
| 1 | 2 | 3 | 4 | |||
| 5 | 6 | 7 | 8 | 9 | 10 | 11 |
| 12 | 13 | 14 | 15 | 16 | 17 | 18 |
| 19 | 20 | 21 | 22 | 23 | 24 | 25 |
| 26 | 27 | 28 | 29 | 30 | 31 | |
BASKETBALL Statistics, News, Video Clips & Photos eBA BLOG.










2007 Febrero ~ February
2007 Enero ~ January
2006 Diciemb ~ Decemb
2006 Noviemb ~ Novemb
2006 Octubre ~ October
2006 Septiem ~ Septemb
2006 Agosto ~ August
2006 Julio ~ July
2006 Junio ~ June
2006 Mayo ~ May
2006 Abril ~ April
2006 Marzo ~ March
2006 Febrero ~ February
2006 Enero ~ January
2006 Diciemb ~ Decemb
2005 Noviemb ~ Novemb
2005 Octubre ~ October
2005 Agosto ~ August
eBA Stats Statistics Portal
http://www.eBAstats.com/
Portal de Estadísticas eBA Stats
http://www.eBAstats.com/
Clínicas eBA Clinics
http://www.eBAclinics.com/
Foros eBA Forums
http://www.eBAforums.com/
Glosario eBA Glossary
http://www.eBAwords.com/
eBA Statistics Blogs
http://www.eBAblogs.com/
Tienda eBA Store
http://astore.amazon.com/ebastatspa-20/
Roberto E. Azar
Physical Education
Professor
Profesor de
Educación Física
Basketball
National Coach
Entrenador Nacional de
Básquetbol
Basketball Game Analyst
Analista de Partidos de Básquetbol
eBA Stats Group CEO
CEO Grupo eBA Stats
eBA Basketball Statistics Clinics Director
Director de las Clínicas eBA de Estadísticas
eBA WIKI Encyclopedia Supervisor
Supervisor de la Enciclopedia WIKI eBA
Buenos Aires, Argentina
azarober AT eba-stats DOT com