TomG's Robot Professional MLB Betting - Page 20

basementproject · #**191** 06-12-2007, 02:32 AM

Ah. Carry on then.

[img]/images/graemlins/grin.gif[/img]

TomG · #**192** 06-12-2007, 04:48 PM

Wednesday MLB...

Texas at Pittsburgh
Texas +158
Risk 1 unit to win 1.58 units

St. Louis at Kansas City
St. Louis -105
Risk 1.05 units to win 1 unit

Atlanta at Minnesota
Atlanta +107
Risk 1 unit to win 1.07 units

AllinDan · #**193** 06-13-2007, 12:40 AM

for anyone in the SF Bay Area-

CSU East Bay (in hayward) is hosting a "symposium on statistics and operations research in baseball" on July 11.

link

the speakers include Nate Silver (the mind behind BP's PECOTA system) and Jeff Ma, the creator of Protrade and the subject of "Bringing Down the House" about the MIT blackjack team, the director of statistics for mlb.com, along with many others.

My dad is also on the panel (he teaches a course on baseball history). I don't know exactly what to expect, but it should be interesting. the panel will also be recorded for broadcast on mlb.com at a later date.

TomG · #**194** 06-13-2007, 12:57 AM

Updated YTD:
118.34 units wagered
+17.57 units
+14.8% ROI

It was the wrong day to go against Detroit today... And so adding one more pick for Wednesday...

Milwaukee at Detroit
Milwaukee +115
Risk 1 unit to win 1.15 units

TomG · #**195** 06-13-2007, 01:53 AM

O'Reilly has a book that walks through techniques in building an MLB database and the types of analyses you can do with it. I plan on purchasing it and thought others might also be interested.

Baseball Hacks: Tips & Tools for Analyzing and Winning with Statistics

eternalnewbie · #**196** 06-13-2007, 11:11 AM

Don't post here a lot, but lurk around a lot and have really been enjoying this thread.

I have this book and its very good. Even more than building a database, it even goes a little into programming. Very useful stuff. The only thing that is a little annoying is that as you actually work through the book there are a number of errors in the code - I believe there is an errata page somewhere, though, where most of these have been identified.

eternalnewbie · #**197** 06-13-2007, 11:12 AM

[ QUOTE ]
Don't post here a lot, but lurk around a lot and have really been enjoying this thread.

I have this book and its very good. Even more than building a database, it even goes a little into programming. Very useful stuff. The only thing that is a little annoying is that as you actually work through the book there are a number of errors in the code - I believe there is an errata page somewhere, though, where most of these have been identified.

[/ QUOTE ]

Should have mentioned that obviously the book I'm referring to is the Baseball Hacks book from the previous post.

fun160 · #**198** 06-13-2007, 03:46 PM

TomG, love this thread, thanks. Do you ever compare your lines to the lines Murray posts?

alwardc4 · #**199** 06-13-2007, 05:58 PM

Tom,

Great thread. I have noticed latlely you are maybe leaning a bit too much toward the NL. Are you weighting the interleague matchups a bit? The numbers you crunch for xERA and ops are mainly against the respective leagues and it seems quite clear that the AL is better. So for example, if a player has an ops of 750 in the NL, it may only be worth 745 or 740 in the AL. Just a thought.

TomG · #**200** 06-13-2007, 06:56 PM

[ QUOTE ]
Tom,

Great thread. I have noticed latlely you are maybe leaning a bit too much toward the NL. Are you weighting the interleague matchups a bit? The numbers you crunch for xERA and ops are mainly against the respective leagues and it seems quite clear that the AL is better. So for example, if a player has an ops of 750 in the NL, it may only be worth 745 or 740 in the AL. Just a thought.

[/ QUOTE ]
I do adjust for league difference. Offensive power ratings are OB% * SLG%. I add that up for each player in the line-up to get a sum for the team. I then compare that against the league average statistic. For example, right now here are the league stats...

American League Avg Offensive Power Rating: 5.32
American League Avg Runs Scored: 5.43

National League Avg Offensive Power Rating: 4.87
National League Avg Runs Scored: 4.41

You'll notice the American league is quite a bit stronger than the National league. I'm guessing this is a mixture of the DH rule and that the AL has slightly better hitters in general. I don't see a wide range of ability in pitching between the two leagues.

Anyway, to run through a calculation... for example here is the Yankees average lineup using hitters stats from 2006-2007 YTD...

Hideki Matsui 0.702
Alex Rodriguez 0.874
Robinson Cano 0.662
Bobby Abreu 0.718
Johnny Damon 0.632
Melky Cabrera 0.522
Derek Jeter 0.765
Miguel Cairo 0.337
Jorge Posada 0.775

Summing that up gives us an offensive rating of 5.99. 5.99 / 5.32 = 1.13. That is, the Yankees are 113% of the league offensive power. 1.13 * 5.43 = 6.14. Meaning the Yankees on average score 6.14 runs against an American league average pitching staff (in reality Yankees have scored 346 runs over 62 games for an avg of 5.58. The difference is due to Yankees poor performance this year vs. last year when we include last year's hitting stats). Since I don't see much difference in the pitching between the two league's I believe a team's expected runs scored in their individual league should approximately translate to opposing league's during interleague play.