Friday, May 4, 2012

Negro League Testing

Since the Carnie's are in a death spiral, most likely bound for a first round playoff exit, I have spent much time this week preparing the new negro league player set for our next ATB season.  Data entry is in its infancy stage, but before I spent time on entering scores of players I needed to prove that diamond mind would handle the negro league raw data in a manner we could all live with.

So far it has.

There are gaps in some of the data of course, but DMB  has built in estimators that bridge any omissions.  As the negro league seasons are generally very short I still have to figure out innings pitched and plate appearance minimums, but one idea I am toying with is allowing negro league players with fewer at bats into the game as "injury prone" to naturally cut down their playing time.

Anyway, here are a few results after simming 9 seasons in our "79 sim environment" - 60 teams, neutral parks, computer generated lineups, and multiple instances of each player.

Turkey Stearnes, a Vg ranging right fielder with an about average error rate, will end up being a first rounder barring any changes.  His season average was .317 / .415 / .543 with a boatload of doubles and triples.  A fair comparison might be Mickey Mantle with a bit more power.

Oscar Charleston now has  legit seasonal data and seems to have improved.  While losing the ability to play left field, the Ex/55 center fielder batted .305 / .390 / .452 with 30+ doubles a year.

Heavy Johnson can't field a lick but he carries a big stick, large enough to be mentioned with the likes of perhaps Ted Williams.  The soon to be DH (though rated in the outfield) batted .338 / .442 / .570, averaging around 20 HR, 10 3B, and 30 doubles.

A new catcher has entered the fray, the Excellent armed Biz Mackey batted .274 / .345 / .376 in limited duty.  Also look for Hall of Famer Willie Wells to make an impact, the shortstop batted .281 / .335 / .516 and with Ex/79 fielding may be in the Top 5.

Pitching wasn't nearly as dominant, but that may be because of my limited selection.  In this initial test I required the basic 20 games started rule which whittled the player pool tremendously.  Bill Tyler was likely the best, compiling a 3.69 ERA and giving up about a hit per inning, walking 42 against 70 strikeouts.

Nip Winters had 4 eligible seasons and the AI didn't give him many starts, but he holds hope to be a #2 type of starter.  At his best, and holding a terrible resim season aside, he managed a 3.78 ERA and a hit per inning as well.

Coming out of the pen, Red Ryan, Luther Farrell, and Norman Yokely each held their own with ERA's in the low 4.00s.

That's all for now.  Please feel free to comment, would love some opinions on the data.  I added roughly 60 player seasons and these ten were the best of them.  Most others would be late round picks or even un-draftable.

6 comments:

  1. Lou how did Leon Day do? He was considered the top pitcher in that league...

    -Sean

    ReplyDelete
  2. The only valid data on him shows only 1 decent year. I suspect it is not that he is bad, but no one has been able to compile official data on him.

    http://www.baseball-reference.com/nlb/player.cgi?id=day---001leo

    ReplyDelete
    Replies
    1. Its funny, because the 1937 13-0 season he had is well known (google it and you will see multiple resources), yet baseball reference only shows 5 of those games.

      Delete
    2. Is baseball reference owned by FOX News, perchance?

      Delete
    3. Ha:)

      The real issue is that folks have been saying Leon Day was 13-0 and what a great season, but there really is no proof. I am sure Day was 13-0 with a 3.02 ERA like has been reported, but DMB needs more than that. I need a full pitching line, and it just doesn't exist.

      Delete