Pacific Coast League 1920-30?

[Splitter24]

1946 American Association stat files scanned and ready to go. Just waiting on LGO to take care of something first...

[Le Grande Orange]

I did! Spritze said he was going to send some scans tomorrow, but it looks like you got the jump on him.

Anyway, it'd probably be a good idea if you two co-ordinated your scanning so you don't both end up doing the same league/stats categories. You two are the only ones with access to the very valuable original resources, so seeing your efforts be accidentally duplicated would be unfortunate.

Just a thought at any rate.

[Splitter24]

LGO: Resent the 1946 American Association stats to your first backup email address (4 emails, 7 total files).

Spritze: I believe you have the other leagues scanned. Once this first round of inputting is done, we need to coordinate a "scanning schedule" going forward like Le Grande Organizer suggests. I would guess you have a little more time than I do at the moment.

[Le Grande Orange]

Just got the files. No problems this time.

And thanks for taking the time to do the scans, I know I speak for everyone involved when I say it's much appreciated.

[seth70liz76]

Ready for action!

[Le Grande Orange]

Okay, having looked over the AA scans, it looks like there might be a bit more to this project than I anticipated. It appears there are several issues that it would be best to resolve before handing stuff out, to keep things as efficient as possible.

I think it would be good to discuss these items here with the team before proceeding.

But, I have to write these matters up so it'll be clear what I'm talking about. So, look for a post here on Saturday afternoon. Your input on the subjects raised will be important.

[Spritze]

Sent the Orange feller scans of 1930 PCL and templates for 1930 thru 1959.

I don't want a PCL
I just want to ride my motorcycle

Which I'm gonna do this am for the firstest time since last fall.

1993 Harley. Vroooom.................

[Le Grande Orange]

First, enjoy the motocycle ride! The only cycle I have is one that requires peddaling. Not that I mind that though.

Second, let's not get too far ahead of ourselves. There are a number of little issues to be straightened out first before plunging in. With that in mind, I'd prefer to just get the PCL and IL stats for 1946 so that year will have everything available first, and then see how things go getting that year all done before moving on to other seasons.

Third, I'll start covering these issues in my next post.

[Splitter24]

The suspense is killing me, LGO.

What issues? Tell me! Tell me!

[Le Grande Orange]

Okay, let's see if I can cover things without taking 20 pages to do it.

First, a bit of background. It seems to me the most efficient and speedy way to get the statistical data input is to make the Excel files as easy to use as possible. The simpler they are to use, the quicker one can enter in the info. But there are some questions which arise from this approach, namely is it possible to take these simple Excel files and recomposite them into something more complex later without a lot of trouble?

Now for some details.

BATTING STATS

The player stats for batting included in the Lahman Database are:
G AB R H 2B 3B HR RBI SB CS BB SO IBB HBP SH SF GDP

Naturally, the more of these statistical categories we can capture, the better, though of course it will be limited to what's published in the TSN Guides.

Here are the individual batting stats recorded for the 1946 AA, listed in the same order as they appear. Players who bat left or who are switch hitters are noted in the list.

10 or more games (ranked by batting average):
LastName FirstName Club G AB R H TB 2B 3B HR SH SB Avg.

Fewer than 10 games (ranked by batting average):
LastName FirstName Club G AB H Avg.

To me, the quickest and easiest way to enter the stats into an Excel file is for the file to match the same stats order as are in the Guides. This means that one simply types the data just as it appears in the scan, without having to do any searching or switching back and forth between columns (for our purposes we can skip entering TB and Avg., since these are calculated stats).

The question which immediately arises are the players with fewer than 10 games. They have very limited data, so should we include such players? I would say no, since they had so little playing time I don't think it's worth including them. What is the rest of the team's opinion?

A number of important stats that are captured in Lahman, such as BB and SO, aren't included in the main AA batting list but are instead listed separately. Below shows the categories and the order:

Additional Batting - 10 or more games (ranked by RBI):
LastName Club G BB HBP RBI SO GDP

I definitely think these stats should be entered, but the fact they are included in a separate list raises some issues. The players are only listed by their last name, which is going to complicate the data entry. There are two ways to handle this, it seems to me.

The first way is to enter these extra stat areas on the main list in the Excel file. But this means that to enter in the stats you'll have to scroll up and down the list of players to find the correct one before you can enter in the numbers. This will get time-consuming very quickly and would slow down data entry significantly. The second way is to enter in these extra batting stats on their own worksheet; this would make the data input very quick. But the drawback here is the question of how difficult will it be to recomposite these other batting stats with the main list. Is more time lost in recompositing together the batting stats or is more time lost in trying to enter all the batting data onto one single list?

I personally think it's the latter, but I want to hear everyone else's opinion as well. Particularly those who are knowledgeable with Excel and who can give some idea on how difficult it would be to combine the batting data from two different lists into one consolidated list.

***End of part I***

[Le Grande Orange]

FIELDING STATS

The player stats for batting included in the Lahman Database are:
Pos G GS InnOuts PO A E DP ZR

Here are the individual fielding stats recorded for the 1946 AA, listed in the same order as they appear. Players who throw left are noted in the list, and the list is separated out into the following positions: 1B, 2B, 3B, SS, OF, C, P.

10 or more games - all positions except for catchers (ranked by fielding pct.):
LastName Club G PO A E DP Pct.

10 or more games - catchers (ranked by fielding pct.):
LastName Club G PO A E DP PB Pct.

Two issues arise here. First, players are only listed by their last names, so again we are confronted with the question of how to handle this. I feel entering in the fielding stats on their own and recompositing the data later is the best route, but everyone should have a say on this approach. Second, outfielders are lumped together as one entity, with no breaking down into LF, CF, or RF. I'm not knowledgeable enough on the Lahman Database and importing players to say how much affect this has, so anyone with insight here should add their views.

PITCHING STATS

The player stats for pitching included in the Lahman Database are:
W L G GS CG SHO SV IPOuts* H ER HR BB SO OBA ERA
* Outs pitched (innings pitched x 3)

Here are the individual pitching stats recorded for the 1946 AA, listed in the same order as they appear. Players who throw left are noted in the list, and balks are listed at the end of the table.

45 or more innings (ranked by ERA):
LastName Club G W L Pct. IP AB H R ER SH BB SO HB WP ERA

Less than 45 innings (ranked by winning percentage):
LastName FirstName Club G W L Pct. IP H R ER BB SO HB WP BK

Pitchers without a decision (ranked by games):
LastName FirstName Club G IP

Several items of note here.

First, yet again only last names are used (for players with 45+ IP), so once more the question of how to handle this arises. Given that pitching is its own speciality, it really ought to be entered on its own. But some way of getting the players' first names will be needed, preferrably at a compositing stage so it won't have to be done at the data entry stage.

Second, players with less than 45 IP, while having their full names listed, are missing a couple of statistical categories. So, similar to hitters with less than 10 games, should we include these lower pitching-time players or not?

Third, HRs given up are not tracked. While it may be a very important stat for DIPS purposes, there isn't much we can do in this area since the information was not recorded.

Additional Pitching (ranked by complete games):
LastName Club XIG SHO NoHit 1hit 2hit 3hit 4hit GS TakenOut GF IncompleteGames* CG
* Games not finished as a relief pitcher

As with the additional batting stats, several important and useful pitching stats, such as SHO, GS, and CG, are listed in a separate table. I feel these are worth capturing, but again we face the problem of last names only being listed and getting this information linked up with the main pitching stats. I think an approach similar to that used for the extra batting stats should be followed; what's everyone else's view?

***End of part II***

[Le Grande Orange]

MISCELLANEOUS ITEMS

Games Played

One thing that's nice in the 1946 AA stats is that players who spent time with two clubs have the number of games played for each club listed. While the stats are unified and total number of games played is listed, beside each player is a notation of how many games were spent with each club.

But the question is should be attempt to record this information? It's there, so I'm tempted to say let's grab it, but what would be the best way of doing it if we do decide to input it?

Team Abbreviations

Rather than having everyone type in the full city for each club, I've got my own list of abbreviations which can be used. In fact, I confronted the issue of league and team abbreviations very early on when I started working on the minor league schedules.

The whole question of a standardized set of league and team abbreviations is actually an important one, one which I've not seen SABR nor anyone else address, not even for the major league level. It makes no sense to me to have different sources using different abbreviations; why not have one master list which everyone uses? That way, no matter where you'd look, you'd always see the same abbreviations being used. Seems to me this would make things easier for everyone.

I have rethought somewhat the exact system I would want to see used, but in terms of this project it isn't a big problem. Abbreviations can be changed quite easily through a simple search & replace in Excel.

Scan Quality

The scans I've received so far look pretty good, and shouldn't pose too many problems in terms of reading the stats. The only issue where readability comes into play is on the right-hand pages of the Guides. This is because the page margins are quite close to the book's spine.

Left-hand pages aren't an issue, because whatever numbers that fall near the spine are not needed by us since they are calculated stats. But for right-hand book pages, the player names fall close to the spine meaning the names can become hard to make out since the book can't be flattened enough. I don't think much can be done about this, since we don't want to wreck the books in getting the scans done.

So, some occasional checking may be needed to confirm the name if someone is having a hard time making out a name.

Below is an example of what I'm talking about. The pic is part of a TSN page showing the spine issue. Though cropped, the pic is presented at full size so what you see below is what you'd be looking at when you'd be inputting the stats. Any questions or comments on the scan quality should be raised now if some folks find the example below would be hard to work with.

And I think that covers all the issues I wanted to raise. Now it's time for your replies.
.

[Eugene Church]

LGO, as far as I am concerned, I trust you to devise the best and most efficient way to do this. I think you and Spritze need to work things out for his convenience in creating the database. Consult him and you two make the decision.

As far as the format, I think you should not split up the batting into two groups...include the SO BB HBP RBI in with the main group...it would be up to the data entry person to look back and forth to get the info. This way you would have all the stats in one row. It might be slow to do it this way, but I would be willing to do it. I have already had to do this working with the Southern Association and Texas League scans Splitter was kind enough to share with me.

I would include G GS CG SHO with the main group of stats...again the data entry person would have to look back and forth to get the info.

Same thing for first names...you would just have to look in the batting portion to get the first names of pitchers and first names for fielders.

Don't include the players who played in less than ten games.

I noticed from using Splitter's scans he did for me that in 1946 the Texas League did not break up the stats in batting and pitching...however the Southern Association did...the TL listed LF, CF, and RF in fielding stats, but the SA only listed outfielders in fielding stats.

As far as scan quality, I found that if I zoomed in I could see the edges of the pages better...but this could be a problem if Splitter and Spritze cannot get the pages flat enough to scan the entire page.

[plannine]

Originally Posted by Le Grande Orange

But there are some questions which arise from this approach, namely is it possible to take these simple Excel files and recomposite them into something more complex later without a lot of trouble?

If you output the excel files to txt (or csv) you can then copy the txt files into one large file (using dos commands) [i.e. copy pcl.txt + pcl19*.txt]


or am i missing what you are asking......

[seth70liz76]

I feel it will be easier to input all the battinginfo on one worksheets--seperate ones will just be more hassle.

I agree with EC, the info on players with less than 10 games (especially since we lack thier fielding info) is not worth it either. They can always be added at some point in the future if the information does become available. But the pitchers under 45 IP should be included--most of the info we need is there.

As for the HR/OF issues--we're just going to have to live without it, again, unless it turns up in some form. However, with Pitcher HR, would it be worth it to figure what would create a mid-point "HR allowed" rating? Not the most accurate, but it would prevent every P from being imported as really tough to homer against. Just a thought.

[gohorns98]

Batting

I think we should drop the 10 games or fewer. If we kept them then we would have the additional stats for some but not others. This could cause problems while importing players into OOTP. If everyone is missing a catagory that is one thing, but if some have ratings and other do not then that is another.

Fielding

I believe that assigning OF with only OF that OOTP will randomly give them ratings in 1,2, or 3 OF positions. I am not sure of this, but worse case we could create the FieldingOF sheet using random assignments. Or even we could give all the OF ratings at all three positions.

Pitching

I think we should include the pitchers with less then 45 innings pitched, but not the pitchers without decisions. We would simply not have enough info for the no decision players (only G and IP). Also, we should have the additional information included even if the OOTP engine does not use this information to get ratings. You never know it might in the future.

Combining Data

If I were given each of the data sets in individual worksheets (i.e. Batting stats and Additional Batting Stats in 2 worksheets). I could then make a templete that would get them into the right order. It might take some manual work if there are two players with the same last name, but it shouldn't be too hard.

Another issue is general information. Have you thought about Unique Player IDs? I just wanted to throw that out there for any thoughts.

Thanks for all your work LGO. As a side note, I agree that a uniform system of team abbreviations is needed. Do you know if OOTP would except 4 letter team IDs? If so, that would be a good way to differentiate between minor league teams and major league teams.

[Splitter24]

Random bullet points from a thoroughly exhausted guy:

1) I like having every player. But the data in certain seasons just isn't good enough. Like our fringe players in the 1946 AA for instance. You're going to run into the same data for minor leaguers during the Dead Ball Era, even for the biggest minor leagues. So we may just have to jettison the short-timers.

2) One thing to think about for Phase Two of this project (which could be a couple of years off) is to create a Minor League player database instead of a season database. That way you could insert a fringe player if he has other seasons with which to form some sort of composite rating from. For example, if Joe Ballplayer only played in 8 games in the AA in 1946, but had other minor league seasons where he played full time, then you might be able to go back and plug him into the the 1946 AA database with composite ratings for other years. Just a thought.

3) If you haven't already downloaded Irfanview (it's free), do it. I always check to see if I can use Irfanview to view the scans that I make. I can always zoom in and sharpen the image enough to make it easily readable using Irfanview.

4) I think all the batting stats should be kept together. If it takes a little longer to go back and fill in all of the stats together, then so be it. I don't know exactly when, but somewhere along the line all of the leagues had complete stat lines printed without the miscellaneous stat section. I think the IL and PCL don't have this silly issue. All my guides are packed up for the move, so I can't verify that; just going on memory here.

5) OF fielding positions are pretty much a non-issue. You'll find major league data for some years doesn't differentiate between LF, CF and RF, either. I think gohorns98 is right about the random OF assignments.

6) Let's use LGO's team abbreviations.

[Tiger Fan]

Quick note on Of fielding positions. If I am not mistaken OOTP will import every outfielder as a CF if there are no positions designated.

One other issue that if it has already been mentioned I apologize.

What about the minor leaguers who are already in the Lahman database. Do we ignore them? Being AAA I bet a huge percentage saw at least some MLB action.

[changewand]

BTW - another volunteer here! I own the Snelling book as I was thinking of tryign this project on myself! I'd be happy to take any date/era from the PCL.

[seth70liz76]

Originally Posted by changewand

BTW - another volunteer here! I own the Snelling book as I was thinking of tryign this project on myself! I'd be happy to take any date/era from the PCL.

Could this cover the "less than 10 game" players or fill in the missing information for the PCL players? That would be a huge help!