SABR Syllabus

[Tyrus4189Cobb]

You are asked by a university to teach three courses for a total of $12,000. You will teach three sections of sabermetrics (SABR 212) for math students who need a quantitative and/or reasoning elective. Class meets twice a week for 75 minutes each. There will be a total of 29 classes. I will use the dates for fall semester 2012.

What topics will you cover each class? What are the required reading materials? How are students graded? For my syllabus, parenthesis wouldn't appear on the actual syllabus but are there to explain what materials I would cover

Sep 5th- Intro to sabermetrics (what they are, what they do, a little background of baseball stats, no stat is perfect, important to remember humans can't be assigned to a number)
Sep 10th- Basic statistics (more history of stats, calculating basic stats, all-time leaders)
Sep 12th- Basic statistics pt 2
Sep 17- Measuring skill (applying basic stats to playing)
Sep 19- Movie: Major League (assignment: was team managed effectively? Could it be made better using stats?
Sep 24- Major League (cont)
Sep 26- Quiz One
Oct 1-Bill James (Bio. Rightfully credited? How did he and others change the scheme?)
Oct 3- Advanced Statistics (borderline sabermetrics; more involved stats like OPS+)
Oct 8- Sabermetrics (more in-depth history, use, sabermeticians, basic overview [linear weights])
Oct 10- Calculating Sabermetrics (math behind sabermetrics)
Oct 15- Calculating Sabermetrics (calculating some sabermetrics)
Oct 17- Calculating Sabermetrics pt 2
Oct 22- Quiz 2
Oct 24- SABR Flaws (opposition against sabermetrics. Assignment: what do you like better?)
Oct 29- Building a Team on Lunch Money (story of Billy Beane and 2002 A's
Oct 31- Movie: Moneyball (assignment: how was Beane unconventional? Where do sabermetrics come into play?)
Nov 5- Moneyball cont
Nov 7- Fielding: Harder to Measure (UZR, TZR, etc)
Nov 12- Fielding pt 2 (stats continued and differing needs for each position type)
Nov 14- Quiz 3
Nov 19- In-class assignment: Create your own statistic (doesn't have to be too involved. What are its perks and flaws?)
Nov 21- Going to WAR (precursors needed to calculate WAR like wOBA)
Nov 26- What is It Good For? (calculating WAR)
Nov 28- Statistical trends (how stats have different throughout years and why)
Dec 3- Quiz 4 (WAR calculations done using given info)
Dec 5- Where are stats headed?
Dec 10- Movie: A League of Their Own (assignment: can sabermetrics account for environment and personality?
Dec 12- LOTO continued
**Dec 18 FINAL

Assingments: 25%
Quizzes: 40%
Final: 20%
Homework/participation: 15%

Required reading: Beyond Batting Average by Lee Panas, The Book by Tom Tango and co., Whatever Happened to the Hall of Fame? by Bill James

[Iowanic]

If I had the money; I'd been the first student signed for that course...

[EdTarbusz]

I would take the class.

[brett]

I would include something on predicting future player productivity (based on age, positions, skill set etc.)
I'd like to see correlation coefficients of different team stats to winning.
Maybe rather than jumping into sabermetrics, "why do historical stats fail?" I think we might also call slugging percentage and OB% early saber stats and compare it to batting average.
Something about DIPS. Why do we look at actual runs for pitchers, rather than OPS+ allowed, but OPS+ for hitters rather than actual runs produced?
Park factors
Is clutch real?
It would be neat to see a table comparing gold glove winners to their actual rank in defensive metrics, ie perception versus reality. Why might a fielder that looks good not be so?
Do managers matter? Home field advantage?

I have designed curricula for science classes and there is not perfect formula.

[Tyrus4189Cobb]

I would include something on predicting future player productivity (based on age, positions, skill set etc.)
I'd like to see correlation coefficients of different team stats to winning.
Maybe rather than jumping into sabermetrics, "why do historical stats fail?" I think we might also call slugging percentage and OB% early saber stats and compare it to batting average.
Something about DIPS. Why do we look at actual runs for pitchers, rather than OPS+ allowed, but OPS+ for hitters rather than actual runs produced?
Park factors
Is clutch real?
It would be neat to see a table comparing gold glove winners to their actual rank in defensive metrics, ie perception versus reality. Why might a fielder that looks good not be so?
Do managers matter? Home field advantage?

I have designed curricula for science classes and there is not perfect formula.

Great points. I would include park factor in OPS/ERA+. Students wouldn't calculate it but what see how it works. Everything else could be its own topic or embedded in another except for the managers issue. That's tough to cover and I'd just leave students to make their own opinions based on the movies and what is learned in class.

If anyone wants to sneak me into a university as a professor, feel free to do so...

[Bothrops Atrox]

I'd have a 2-3 week study on DIPS and BABIP, etc. I'd end the third week of the study analyzing pitchers who defy DIPS and why/how.

You could probably make the defensive system comparison/analysis section 10 weeks if you wanted to.

[brett]

Essay question, explain 3 adjustments that can be made to a player's slugging or on-base percentage to make them better measures of value.

a) relativize to league rates (varies from 3.41 to 5.41 runs per game-as much as a 58% difference, or a little over +/-22% from mean)
b) adjust for the player's home park (parks vary from about 88% to 114% of average. Player's vary by about half this, or +/-6.5%park factor, except for Colorado)
c) properly weigh discrete events based on historically probabalistic value. (actually a small percent deviation from taking straight bases)

d) possibly weigh events based on probabalistic value IN a given run setting

Also a look at some player splits like Dimaggio (hurt PARTICULARLY by home park?) Klein, and Ott.

I like those 3 cases because Dimaggio hit much better on the road, Klein had both a boost due to park factor, AND probably a particular boost, and Ott probably was as good on the road despite particular differences.

I might be worried about having people with limited math skills, but most of it is ratios.

Oh, one last one, possibly pythagorianesque winning percentage prediction.

[Jackaroo Dave]

In addition to the formula calculation, I think there should be a math section on probability early on. It's impossible, in my opinion, to appreciate the power and limitation of advanced baseball statistics without grasping the huge roles of (a) random variation on the one hand and (b) the law of large numbers on the other.

Otherwise, you get anecdote-based resistance and "You can prove anything with statistics."

[Tyrus4189Cobb]

In addition to the formula calculation, I think there should be a math section on probability early on. It's impossible, in my opinion, to appreciate the power and limitation of advanced baseball statistics without grasping the huge roles of (a) random variation on the one hand and (b) the law of large numbers on the other.

Otherwise, you get anecdote-based resistance and "You can prove anything with statistics."

That's why I made it SABR 212. It required prerequisite courses like STATS 101

[Jackaroo Dave]

That's why I made it SABR 212. It required prerequisite courses like STATS 101

Very shrewd. In that case, I'd have something on regression, maybe even logistic regression. Just one lesson on each, showing the basis, how to run it, and the interpretations, since they'd already have the context and maybe already a taste of regression.

[Jackaroo Dave]

duplicate post

[honus14]

Something on Branch Rickey's attempts to advance statistical analysis should definitely be covered in the historical section.