Baseball Run Line System

I took it upon myself to find the real value in 1 run games in baseball, and whether the reduced juice from the moneyline odds added up correctly when compared to my findings.

What I found is I think that run line odds are off a little bit, based off what I believe they should be in accordance to the moneyline odds of the same games.

I derived this syestem from regression analysis of past games, a linear predictive equation, and algebra. I am not trying to hide what I am doing here, it would just be really quite long to explain but I am willing to if anyone is interested. Just ask and Ill try my best to explain.

I only had the run line odds from tonight and last night, so those were the only two days that I could test the system, and the sample size is ridiculously small to infer anything yet, but it has been profitable.

The cool thing about this is that I can predict an expected return from each bet, which really only matters when the bets start to add up because of variance from a small sample size but I can mathematically show my expectations. Here are the plays from the last two nights..

8/26/09 - Wednesday
* (ER = average expected return per bet)

1. Pirates +1.5 ...  $100 to win $100 ... ER - $12
2. Mets +1.5 ... $100 to win $115 ... ER - $11
3. Nationals +1.5 .... $100 to win $125 .... ER $8
4. Blue Jays +1.5 .... $150 to win $100 .... ER $5
5. Angles -1.5 .... $100 to win $150 .... ER $4

Pirates -$100
Mets -$100
Nationals -$100
Blue Jays +$100
Angels +$155

Net: -$45.00



8/27/09 Thursday

1. Pirates +1.5 ... $100 to win $105 .... ER - $12
2. Astros +1.5 .... $100 to win $120 .... ER - $12
3. Mets +1.5 ... $100 to win $105 ... ER - $4
4. Nationals +1.5 ... $100 to win $100 ... ER - $3
5. Rangers +1.5 .... $100 to win 100 ... ER - $3

Pirates +$105
Astros +$120
Mets +$105
Nationals +$100
Rangers +$100

Net: +$530.00

Two day net: +$485.00

If you could explain that would be great.  I'm a little confused by this.  How do you determine which games to bet?

Yankees....

I take the ML odds of the game first and convert them into expected win % for each team.

For example the Red Sox at -230 on the ML equates to them expecting to win straight up roughly 69.7% of the time. Meanwhile the Blue Jays at +212 equates to them winning outright roughly 32.1% of the time.

Note: The expected win % adds up to over 100% because of the vig, so I then adjust thoes percentage in relation to each other to make it out of 100% for a true expected winning percentage.

For example: (69.7% + 32.1%) = 101.8%
69.7/101.8 = 68.5%
32.1/101.8 = 31.5%
Those are the adjusted win % against the money line.

Also for each game, I predict the chance that it will be a 1 run game based solely on the total runs line for the game. I developed an equation from past games that predicts it quite accurately. (The lower the run total for the game, the higher the chance of a 1 run game (although not too much greater of a chance))

For example in the Red Sox game to continue with the example.. 9 runs is the O/U. I predict that there is a 28.53 % chance of a 1 run game with an O/U of 9.

From there I do algebra based on the chance of a 1 run game, and the expected winning % I calculated earlier to find the chance of each team covering their respective run lines, and then convert those percentages into odds for the RL.

Example: Blue Jays, calculated 31.5% of winning outright, so they will cover the run line (1.5) by winning straight up 31.5% of the time. Then I derive their chances of only losing by 1 run (which would cover the run line) by multiplying their chances of losing (68.5%) with the chance of a 1 run game (28.53%). This comes to 19.5%, and you add this to their chance of winning outright (31.5%) to get 51.0%. This is their expected chance of covering the run line.

From there I take that percentage and convert it into betting odds. 51.0% converts into approximately -104.

And from there, if a game has odds greater than I have punched out for each game, it would qualify as a play.

Example, the Blue Jays (+1.5) are EVEN odds on the run line, while I have them calculated at -104, so it would qualify as a play, but a very very minor one at that.

Also you can calculate your expected return based on the assumption that yoru percentages are exactly what will happen, (although that's a great assumption, and an untrue one at that, but its how to go about this). So if the Blue Jays actually did cover 51.0% of the time while continuously getting EVEN odds, the expected return would be .02 units for each bet of 1 unit.

I only take the top 5 plays each night, based on expected returns and count them into the system... I just ran the numbers for tonight's games, based on Pinnacle Odds and here they are...

8/28/09

Reds +1.5 .... $100 to win $102 .... ER $8
Giants -1.5 .... $100 to win $198 .... ER $7
Tigers +1.5 .... $145 to win $100 .... ER $5
Orioles +1.5 .... $148 to win $100 .... ER $5
Twins +1.5 .... $155 to win $100 .... ER $4

Total Expected Return : $29

sounds interested and I see where you are coming from now.  Will you post your plays everyday? I wouldn't mind following this for the rest of the season to see how it does

Yeah, sure as long as I have the time to put in the information each day

Hahah... No I don't have past results to back test mainly because I can't find any previous run line odds to back test against. I mean I would jump all over it if I could find such data, but I don't know of any site where I could find such a thing. I mean Covers has old MoneyLine odds and Over/Unders but not run lines. If anyone knows of such a site where I could find all three, that would be greatly appreciated.

Dufresne, I came up with that 28.53% this way..

I took every baseball game from the past few years and did this...

Broke them up into groups based on the Over/Under total, and calculated the percentage of 1 run games for each different Over/Under number group. I then ran a regression analysis on these numbers, to see if there was any negative correlation between the higher the O/U and the lower the % of 1 run games. I threw out the 6.5 data and anything higher than 11 because there just wasn't enough data for these games (to much chance for error to screw with the analysis). What I found was that there was a negative correlation and it was quite strong with a R-squared value of aprroximately .6.

I know there is some error here along the way, but I tried to minimize the error as much as I thought was possible.

Hahah wow, I never expected something like this but last night ended up 5-0 again! wow, that's 10-0 the last two days!

8/29/09 - Friday

Reds +$102
Giatns +$198
Tigers +$100
Orioles +$100
Twins +$100

Net Return: $600



Keeping Track (8/26-8/29)

Total Net Return: $1,185
Total Expected Return: $103

Keep it going fresh, not quite the involvement of the sabermetrics approach SN has been putting out there, but I believe this approach is actually much more straight forward and makes perfect sense all around, continued success to you and please continue to post your plays when you can.

Thanks jerm, I appreciate the support, obviously though this type of success won't be sustained over a long period of time, as it will definitely cool off but hopefully it is still profitable, which it has given signs of.

Here are today's plays  8/29/09 - Saturday

1. Reds +1.5 .... $100 to win $105 .... ER $13
2. Tigers +1.5 .... $126 to win $100 .... ER $7
3. Pirates +1.5 .... $100 to win $103 .... ER $6
4. D'backs -1.5 .... $100 to win $165 .... ER $6
5. Giants +1.5 .... $165 to win $100 .... ER $6

Total Expected Return: $38.00

Results from 8/29/09 - Satuirday

Reds -$100
Tigers -$126
Pirates -$100
Diamondbacks +$165
Giants +$100

Net Return: -$61.00


Keeping Track: (8/26 - 8/30)
*Betting Unit = $100*
Net Return: $1,124.00
Total Expected Return: $141.00

8/30/09 Sunday's Plays

1. Reds +1.5 .... $100 to win $100 ..... ER $13
2. Mariners +1.5 .... $130 to win $100 .... ER $12
3. Nationals +1.5 .... $100 to win $130 .... ER $11
4. Orioles +1.5 .... $155 to win $100 .... ER $5
5. Red Sox +1.5 .... $140 to win $100 .... ER 3

Total Expected Return: $44.00

Hey JBrent here is what I got..

Over/Under      % of 1 Run Games

7        - 35.62%
7.5     -  34.15%
8        -  25.74%
8.5     -  28.30%
9        -  30.07%
9.5     -  24.64%
10      -  23.86%
11      -  25.49%
11.5    -  25.00%

The percentage for all games was 28.05%

I didn't use games below 7 and above 11.5 because there wasn't enough data for it to be reliable.

 

Fresh, 

Why don't you add the underdog's chance to win to the "percentage of underdog losses by 1 run"?  In looking at the data for the 2008 AL, I notice that the underdog lost by 1 run 14.69% of the time for all games.  This is just 1 season and 1 league of data.

o/u      % Underdog loss by 1-run

7        0%  (15 games)

7.5      25.33% (75 games)

8        15.7%  (121 games)

8.5     13.22%  (242 games)

9        15.03%  (259 games)

9.5      15.15%  (198 games)

10        9.42%  (95 games)

10.5     13.51%  (74 games)

11        13.89%  (36 games)

11.5     13.33%  (15 games)

 

 

 

 

Today's Results 8/30  Sunday

Reds +$100
Mariners -$130
Nationals +$130
Orioles +$100
Red Sox $100

Net Result: $300.00


Keeping Track 8/26 - 8/30 : (5Days)
Net Return: $1,424.00         (+14.24 units)
Expected Return: $185.00     (+1.85 units)

Thanks IceAge for the catch, I realized I added an extra $100 dollars on the 29th. I will make sure to fix that. I am going to start switching everything over to units anyways from now on, just because that is the only thing that really matters. Anyone can adjust the unit on their own afterwards....