MLB In-Running Odds

[Backslider]

Assume that you want to build a tool that will tell you, given certain search criteria, how often a team wins in that situation and/or how often the final score of the game has more than N runs.

The more criteria you pick, the more specifically you can define the situation, but the less data you will have to create your percentages.

If you had all the following information at your disposal, which do you think would be the most useful for getting a good population for calculating the win percentage?

For each you could use an exact match, or a range. You could also combine (add/subtract) any of them.

Game Situation:
1. Away Score
2. Home Score
3. Inning (Half Inning)
4. Outs
5. Balls
6. Strikes
7. Man on First
8. Man on Second
9. Man on Third
Game Details:
10. Day of the Year (or Week or Month)
11. Day of the Week
12. Starting Time
13. Played at NL Ballpark
14. Indoors/Outdoors
15. Distance travelled by away team
Starting Pitchers:
16. ERA
17. Throwing Hand
18. K per 9 innings
19. GB/FB Ratio
20. WHIP
21. HR Allowed per game
22. Percentage caught stealing
23. OPS against
Odds:
24. Home or away favored
25. Home or away moneyline odds
26. Over/Under total
27. Over/Under price
28. Home or away runline price
Team Records (for either team):
29. Winning %
30. Winning % at home
31. Winning % on the Road
32. Winning % against LHP
33. Winning % against RHP
34. RPI
35. Winning % last 10 games
Team Offense (for either team):
65. Runs scored per game
36. Hits per game
37. HR per game
38. Total bases per game
39. BA
40. OBP
41. SLG
42. OPS
43. K per Game
44. BB per game
45. SB per game
46. SB %
47. GB/FB Ratio
48. GIDP per game
49. OPS vs Righty
50. OPS vs Lefty
51. OPS at Home
52. OPS on the Road
Team Pitching (either team):
66. Runs Allowed per game
53. ERA
54. Saves per game
55. BB allowed per game
56. BA allowed
57. OPS Against
58. SLG Against
59. SB Allowed per game
60. % CS
61. WHIP
62. K per 9
63. K per BB
Other:
64. Team Fielding %

[Backslider]

My first inclination..

For sure:
1-9 - Game situation
13. NL Ballpark
24, 25, 26 - odds to win and over/under
29. Win % (or RPI?)
65+66 for home team = Runs per game at this park
65&66 avg of runs scored by home team and runs allowed by away team
65&66 avg of runs scored by away team and runs allowed by home team

Maybe:
10. week of the year
16. SP ERA
35. Win % in last 10

[Jizay]

I don't know much from baseball, but I would be most interested in the game situation cluster. The other stuff, I'm not sure if I would use any of it. The idea is that I want a purely actuarial account based on a large sample of data. I then adjust for "this particular game" factors based on what I'm seeing/other scouting. The latter is hard to capture statistically. My guess is that when the price deviates too much from what the actuarial data suggest, it's a good play. Also, odds information is stuff you usually had, so absent a complicated theory about how it interacts with game situations, it doesn't do much but complicate the model.

[Backslider]

I don't have enough games in my database. I could add more games that I don't have other info for, but then I think I would just end up with the same odds as ESPN probably, and there is no point.

[Backslider]

Jizay pointed out that there may be a point for the over/under since espn doesn't have that. So I've continued game pbp into the system. I am up to 12,000 games, 3.2M pitches. I'm part way through 2003, working backwards. I'm not sure how far back I want to go, because it seems like the further you go back, the less relevant the data is.

[California]

For "real time" links to MLB games broadcast around the country, try HDTHE.net . . . the signal delay, as measured against local television, appears to be only about 1/2 to 2 seconds.

For price makers, this minimal delay works well, as orders on Matchbook have, in my experience, a 4 - 5 second delay . . . this usually gives price makers time to pull their offers if something dramatic (e.g., a home run) happens.

Recently, I have began to cross check MB Live Betting games with those listed links on HDTHE.net . . . if they are listed in both places, I can bet in-running games anywhere without being constrained to what is available on local TV . . . without paying a subscription fee.

For today, Saturday, July 11, 2009, all 15 MLB games are available thru HDTHE.net.

If anyone knows of other sites that provide quality MLB games link service, let us know . . . my experience with a few others has been that they either do not work or intermittently work . . . HDTHE.net links appear to work over 95% of the time.

[fotso]

My first inclination..

For sure:
1-9 - Game situation
13. NL Ballpark
24, 25, 26 - odds to win and over/under
29. Win % (or RPI?)
65+66 for home team = Runs per game at this park
65&66 avg of runs scored by home team and runs allowed by away team
65&66 avg of runs scored by away team and runs allowed by home team

Maybe:
10. week of the year
16. SP ERA
35. Win % in last 10