Also, how would I retrieve batted ball x,y locations prior to the Pitch F/X GameDay XML data ~2006?

"FROM (SELECT * from fielding
WHERE IF(yearID>1995 AND POS = "OF",1,0) != 1 ORDER BY G Desc) f"

I can't really figure out whats going on there.

primary pos data from: http://tangotiger.net/bdb/

for example

 select playerid, yearid, CASE WHEN pos = 'C' THEN (innings/162) * 12.5 WHEN pos = 'SS' THEN (innings/162) * 7.5 -- etc ELSE 1 -- use 'ELSE' as the catch-all, to be used on anything that wasn't listed above END as POS_ADJUSTMENT from fielding; 

there might be a more elegant way to do it with MySQL variables, but using CASE is a little more generic

If pos = 'c'
then tot = 12.5
else
if pos = 'ss'
then tot = 7.5
…….

It will be pretty long, but should work. If is the only method I know, but there might be another method.

http://basql.wikidot.com/full-year-statistics

Then, do something like:

SELECT p1.playerID, p1.yearID, p1.ERA, p2.ERA
FROM pitch_year_total_avg p1, pitch_year_total_avg p2
WHERE p1.playerID = p2.playerID
AND p1.yearID = p2.yearID + 1;

Select
p.playerID,
from pitching p
where (p.yearID > "1985"
and (p.IPouts/3) > "160")
group by p.playerID;

And then join that on your main query.

Select
sp.nameFirst,
sp.nameLast,
p.playerID,
sp.playerID,
sp.college,
sp.birthYear,
c.yearMin,
c.yearMax,
year (sp.debut),
year (sp.finalGame),
Count(p.YearID) as YS,
SUM(p.G) as G,
sum(p.GS) as GS,
sum(p.IPouts/3) as IP
from pitching p
join master sp on p.playerID=sp.playerID
left join schoolsplayers c on c.playerID=sp.playerID
where (p.yearID > "1985"
and (p.IPouts/3) > "160"
and birthCountry = 'USA'
and finalGame is not null)
group by p.playerID
;

I'd advise checking out the RetroSQL mailing group - there's a lot of good info there about setting up a play-by-play database. I think you have to be a member to search the archives, though. One of the things on my to-do list is to take that material and write it up as a comprehensive tutorial.

If anyone else has subject topics to request for the forum, let me know.

For instance in 2006 for RperOut instead of .182 that he gets, im getting .185. I'm wondering if the difference between Colin's table and Tango's for primpos is why there's a difference?

1. To provide a place for people to ask questions
2. To provide code samples for people to use and modify.

The pages tagged "snippets" are my current focus - I'm taking a lot of my old code and posting it basically as-is. I'm not saying improvements couldn't be made. Porting some of it to Access wouldn't be the worst idea.

I can host data files, so long as they aren't too large - I think we get 100 MB hosting total, I'll have to check. There are a few things that I've adapted from other sources that I'll see about getting the permission to share here.

And then the big one is to do tutorials for setting up play-by-play data. That's on my to-do list, somewhere…

I've seen Tango's mdb file which was great, but all that gave me is more questions.

I was wondering if people wanted to chime in with their ideas on what should be included in BaSQL, and what kind of content and features we hope to see. I know Colin said, in his post on statspeak, that he hopes it will be the "central clearinghouse for information on how to commit sabermetrics using a relational database." There's a lot of room for what can fit under that, though, and I figured it might be good to list some of those thoughts.

Are there thoughts on: what topics need to be addressed first (mysql vs. access vs others? implementing Win Shares/other advanced stats? connecting bdb data with retrosheet data? many others?) ? What features might be nice to implement, if possible (spotlight articles? beginning tutorials? downloadable files? data hosting?) ? How active people might want to communicate with each other (this forum? mailing list? private messages?) ? I'm sure there are other questions I'm forgetting here…