TomG's Robot Professional MLB Betting

[TomG]

[ QUOTE ]
A couple stupid questions if you don't mind. I also came up with a spreadsheet that closely resembles what you have. I have correction factors for the bullpens, starting pitchers, and each teams offense. You talk about offensive power ratings and pitching power ratings. What are these? Also. Where do you get your stats?

Thanks in advance.

[/ QUOTE ]
I get my current season stats from espn.com/mlb. Historical stuff from previous seasons are from Doug's Stats. The Baseball Archive has even more extensive data already formatted into MS Access Database format.

Like I said in my intro, offensive power rating is generated based upon the team's starting line-up's OB% and SLG% stats. Pitching power ratings are a weighted average between the starting pitcher's xERA and his bullpen. Michael Murray gives a very specific example between Houston and St. Louis from last year's playoffs that helps explain this. Buy the book if you don't have it already.

[TomG]

[ QUOTE ]
I've been trying to simply use average runs scored by game (for either left or right handed pitching as appropriate)

[/ QUOTE ]
Also, in my limited experience I think you should be careful about doing something like this. Separating out left-handed stats vs. right-handed stats may introduce too much bias. Sample size issues could arise depending upon whether the left-handed pitchers the batters faced were aces or duds. Is there *that* much of a difference between lefties and righties that a team's probability of winning will change by 10-20%? You might be better off 1st creating your own line using the blended data. Once you have a blended line, then you can adjust it a few points in either direction based upon your own handicapping. Doing the adjustment beforehand is may give you lines that are too far from the posted line. But I may be wrong. Try it both ways and see which method will give you more reliable lines.

[macavity]

Many thanks for your detailed replies, Tom. I'm going to be travelling for the next week or so and won't have time to do anything until I'm back but I'll keep an eye on this thread. Thanks again, and I'll keep in touch.

Cheers,

Mac

[TomG]

We go from 1 play on Monday to betting half the board on Tuesday. The model really disagrees with a lot of the posted lines for tomorrow. Here are the ones that I am most confident about...

Toronto at Baltimore
Toronto -114
Risk 1.14 units to win 1 unit

LA Angels at Detroit
Detroit +117
Risk 1 unit to win 1.17 units

Boston at NY Yankees
Boston +164
Risk 1 unit to win 1.64 units

Philadelphia at Florida
Florida +143
Risk 1 unit to win 1.43 units

Washington at Cincinatti (this is even after being generous to M. Chico expected ERA)
Cincinatti -154
Risk 1.54 units to win 1 unit

New York Mets at Atlanta
Atlanta +127
Risk 1 unit to win 1.27 units

Milwaukee at LA Dodgers
Milwaukee -119
Risk 1.19 units to win 1 unit

Houst at San Francisco
San Francisco +119
Risk 1 unit to win 1.19 units

[TomG]

Updated YTD:
22.51 units wagered
+1.34 units
+5.9% ROI

Tues will be a big day... we're wagering 8.87 units--over a third of the YTD units wagered so far! I'm pleased with the overnight line movements. It reinforces my idea that one of the biggest advantages about creating my own lines is the ability to bet early with confidence. Interestingly, the two games where my initial picks got crushed by line movements were two favorites (Toronto and Milwaukee)... all of the dogs moved quite sharply in our favor (especially Florida, wow). It looks like for today's games simply betting every dog on the opening line would have been +EV (assuming the lines really do get sharper over time).

[TomCowley]

Interesting system, and I think the strong majority of the lines going in your direction is a good sign. With the lines going nuts between last night and today, did anything else become playable?

[TomG]

Wednesday Baseball...

Toronto at Baltimore
Toronto +127
Risk 1 unit to win 1.27 units

Seattle at Tampa Bay
Seattle +107
Risk 1 unit to win 1.07

NY Mets at Atlanta
Atlants +101
Risk 1 unit to win 1.01 units

Pittsburgh at St. Louis
St. Louis +101
Risk 1 unit to win 1.01

Oakland at Chicago Sox
Oakland +120
Risk 1 unit to win 1.20 units

Colorado at Arizona
Arizona -122
Risk 1.22 units to win 1 unit

Milwaukee at LA Dodgers
Milwaukee +124
Risk 1 unit to win 1.24 units

[TomG]

[ QUOTE ]
Interesting system, and I think the strong majority of the lines going in your direction is a good sign. With the lines going nuts between last night and today, did anything else become playable?

[/ QUOTE ]
I waver back and forth on the significance of beating opening lines depending upon whether or not I was most recently able to do! [img]/images/graemlins/smile.gif[/img]

I like to handicap ASAP and then be done with it for the day. It is possible that a line will over-adjust and by game time have become a value play according to the model. But since the system seems to have no shortage of plays lately, I don't feel like spending the time to go back and check.

[TomG]

Updated YTD:
31.38 units wagered
+4.14 units
+13.2% ROI

[rjp]

Nice job Tom. Keep up the winning.