MLB Money Line Calculation

[Formulawiz]

Does anyone know how to calculate a MLB money line and how they arrived at it

[abev]

You mean where did "-140" come from? Simple terms: probability to win but many other factors. And you can read Here.

[Formulawiz]

You mean where did "-140" come from? Simple terms: probability to win but many other factors. And you can read Here.

Thats not what I am asking. What I would like to know does anyone have a good formula as how to calculate your own ML and then compare to oddsmakers line.

[turola]

Does anyone know how to calculate a MLB money line and how they arrived at it

http://forum.sbrforum.com/players-ta...ics-video.html that's a good start

[twister]

The bookmakers money line SHOULD be based purely on probability. In reality, the line at any given time is slightly more based upon the bookmaker achieving equilibrium in terms of money on both sides. Unless ofcourse, the line is opinionated and the bookmaker is willing to take a loss on one side because it thinks the other side has a better chance of hitting.

Now, as for creating your own line - as I said earlier, it is all about probability. Firstly, you need to decide what stats you want to compare. There are numerous stats you could compare, but the 4 main categories are Team Batting, Starting Pitcher, Bullpen, Fielding. You can also throw in total W/L, and/or Home/Road W/L. Note: There are loads of other categories you could throw in, and that is where things get complicated - such as weather, ballpark, umpire, streaks, etc etc

Then you need to establish how much you think each of those categories contributes to predicting the outcome of a result, out of 100%. i.e Team Batting 30%, Starting Pitcher 35%, Bullpen 25%, Fielding 5%, Previous results 5%.

Then you need to pick the stat you think is the most important from each category, turn them into % by comparing the two teams, then weight them by their category.

e.g Team A SP ERA 3.40, Team B SP ERA 7.20. 3.40/(3.40+7.20) = 32%.
Thus Team A is 68% and Team B 32%. Then multiply both by the weighting, which is 35% for a SP, and you get 23.8% for Team A, and 11.2% for Team B.

You do that for each stat, add up the stat x weighted percentages, and arrive at percentages for each team. e.g Team A 71%, Team B 29%.

Convert those into ML and compare. Now, the problem is finding which stats are useful, what weightings to use, and which stats to include. Anyone that has spent the required amount of time it takes to do such a thing to an acceptable level of accuracy, probably will never share that information freely. You'd also need a huge database of stats just to find out which ones are predictive, and which aren't.

All in all...it is hard work.

[Formulawiz]

http://forum.sbrforum.com/players-ta...ics-video.html that's a good start

with all due respect, this video basically tells us what everyone already knows. What I am looking for is examples and formulas. I thought the forumn format was to help each other so that systems developed by people when viewed by others has a chance to get improved so we can all benefit.
I guess I barked up the wrong tree.

[floridagolfer]

with all due respect, this video basically tells us what everyone already knows. What I am looking for is examples and formulas. I thought the forumn format was to help each other so that systems developed by people when viewed by others has a chance to get improved so we can all benefit.
I guess I barked up the wrong tree.

I guess I would tell you that there is no precise formula. It's an educated guess based on half a dozen primary factors, not something that you can necessarily try to create with a three-page spreadsheet.

[MonkeyF0cker]

There's more than one way to skin a cat. In this case, an infinite amount of ways. Outside of someone handing over their model to you, there is no way to hand feed you this information. And anyone with a profitable model is certainly not going to do this.

[turola]

with all due respect, this video basically tells us what everyone already knows. What I am looking for is examples and formulas. I thought the forumn format was to help each other so that systems developed by people when viewed by others has a chance to get improved so we can all benefit.
I guess I barked up the wrong tree.

When you read that thread you will find this link: http://armchairgm.wikia.com/Pythagorean_Formula then you probably want to know how you come up with those score estimations.
Simply Google "run estimator" and you will find this: http://www.hardballtimes.com/main/ar...un-estimation/
Is that what you are looking for?

I guess that's the most you will get from forums and all that stuff was in the Justin7's thread.

[Formulawiz]

When you read that thread you will find this link: http://armchairgm.wikia.com/Pythagorean_Formula then you probably want to know how you come up with those score estimations.
Simply Google "run estimator" and you will find this: http://www.hardballtimes.com/main/ar...un-estimation/
Is that what you are looking for?

I guess that's the most you will get from forums and all that stuff was in the Justin7's thread.

Bill James theorm might have worked 25 years ago but now reputation alone wont get you anywhere especially trying to win games on a daily basis. Try using James's system now and I can assure you will lose your pants. Anyway I tried.

[Justin7]

Bill James theorm might have worked 25 years ago but now reputation alone wont get you anywhere especially trying to win games on a daily basis. Try using James's system now and I can assure you will lose your pants. Anyway I tried.

You can beat the hell out of openers just using Bill James info.

[Formulawiz]

You can beat the hell out of openers just using Bill James info.

Justin can you give us an example on a game today using james system

[Bogart45]

Thanks twister, that was very helpful. For me, you gave a very good explanation of how to start.

[hajune]

Baseball line making is exceedingly difficult. When I lived in Vegas in the early 70's and talked to the "nerd" linemakers they used material I was not familiar with. ERA ratings are not reliable. Pitchers go on 3 games cycles on average and then there is the visitors / home problem. There are things that can be done to establish a line of your own but it is truly complicated. It is best to use how a teams momentum is going into todays game. Some very promising patterns can be established. After finding, what looks like valid information. You have to research it for about 100 games to see if it has merit. Baseball is unigue because its70% defence and of that 90% pitching. It is far to difficult to calulate the unknown , such as who will be reliefing etc.

[Formulawiz]

The bookmakers money line SHOULD be based purely on probability. In reality, the line at any given time is slightly more based upon the bookmaker achieving equilibrium in terms of money on both sides. Unless ofcourse, the line is opinionated and the bookmaker is willing to take a loss on one side because it thinks the other side has a better chance of hitting.

Now, as for creating your own line - as I said earlier, it is all about probability. Firstly, you need to decide what stats you want to compare. There are numerous stats you could compare, but the 4 main categories are Team Batting, Starting Pitcher, Bullpen, Fielding. You can also throw in total W/L, and/or Home/Road W/L. Note: There are loads of other categories you could throw in, and that is where things get complicated - such as weather, ballpark, umpire, streaks, etc etc

Then you need to establish how much you think each of those categories contributes to predicting the outcome of a result, out of 100%. i.e Team Batting 30%, Starting Pitcher 35%, Bullpen 25%, Fielding 5%, Previous results 5%.

Then you need to pick the stat you think is the most important from each category, turn them into % by comparing the two teams, then weight them by their category.

e.g Team A SP ERA 3.40, Team B SP ERA 7.20. 3.40/(3.40+7.20) = 32%.
Thus Team A is 68% and Team B 32%. Then multiply both by the weighting, which is 35% for a SP, and you get 23.8% for Team A, and 11.2% for Team B.

You do that for each stat, add up the stat x weighted percentages, and arrive at percentages for each team. e.g Team A 71%, Team B 29%.

Convert those into ML and compare. Now, the problem is finding which stats are useful, what weightings to use, and which stats to include. Anyone that has spent the required amount of time it takes to do such a thing to an acceptable level of accuracy, probably will never share that information freely. You'd also need a huge database of stats just to find out which ones are predictive, and which aren't.

All in all...it is hard work.

Everything is hard work, but how can anyone wager on a game without calculating their own money line. How can you tell if there is value on a game. So my question to all is, give us a formula to calculate a line. I am sorry to say all I hear is how difficult it is but no examples.

[Dark Horse]

I thought the forumn format was to help each other so that systems developed by people when viewed by others has a chance to get improved so we can all benefit.

Wouldn't that be nice?

If you had developed such a formula or system, and you knew that it worked, how much value would you place on it? In dollars.

Would you share it for free? Even if you knew that sharing it would undermine the very edge you attained through it?

If only there were a market for systems that do work, but that are no longer of use to its developer...