Question For Ganchrow or anybody who knows.

[ManOfValue]

How much , or how strong of a correlation is it if I'm able to bet parlays with team A 1st 5 innings on the moneyline to "team A's pitcher won't allow a home run in the game" .

It surely seems like you have an edge in doing that I would just like to know what the numbers say...

[Ganchrow]

How much , or how strong of a correlation is it if I'm able to bet parlays with team A 1st 5 innings on the moneyline to "team A's pitcher won't allow a home run in the game" .

It surely seems like you have an edge in doing that I would just like to know what the numbers say...

I don't personally keep data on first 5 inning MLs, but what did the stats indicate when you looked at them?

[ManOfValue]

Thats the thing , I don't keep 1st five inning stats either..I know a guy that is convinced that there is very little correlation and would not mind taking those plays..
And I have this other guy who thinks its a HUGE correlation, I'm kinda in the middle..

If you think about it , your pitcher not allowing a HR in the 1st 5 does help you to win that team's moneyline 1st 5..However I thought that since, a single ,double,triple, pass ball ,sac fly can all score runs then the correlation was not much..

[Ganchrow]

Thats the thing , I don't keep 1st five inning stats either..I know a guy that is convinced that there is very little correlation and would not mind taking those plays..
And I have this other guy who thinks its a HUGE correlation, I'm kinda in the middle..

If you think about it , your pitcher not allowing a HR in the 1st 5 does help you to win that team's moneyline 1st 5..However I thought that since, a single ,double,triple, pass ball ,sac fly can all score runs then the correlation was not much..

So perhaps you should initially look at the correlation between winning the 1st 5 and not allowing a HR irrespective of price?

[smitch124]

Does anyone have the correlation between the pitcher not giving up a home run and a team winning the full game?

I would imagine if the pitcher is the starting pitcher the correlation would be at least as high for 5 innings as it would be for the entire game.

[Ganchrow]

Does anyone have the correlation between the pitcher not giving up a home run and a team winning the full game?

From the 1990 - 2007 seasons, over 41,015 decided games, the winning team did not give up a defensive home run 49.24% of the time with a standard error of 0.25%.

Data source: retrosheet.org game logs.

[Ganchrow]

From the 1990 - 2007 seasons, over 41,015 decided games, the winning team did not give up a defensive home run 49.24% of the time with a standard error of 0.25%.

Data source: retrosheet.org game logs.

Broken down per year:

Year	N	Coinc.	Std. Err
1990	2,105	42.33%	1.08%
1991	2,104	43.35%	1.08%
1992	2,106	38.08%	1.06%
1993	2,268	46.96%	1.05%
1994	1,599	52.78%	1.25%
1995	2,016	50.79%	1.11%
1996	2,266	54.99%	1.05%
1997	2,266	49.96%	1.05%
1998	2,430	51.73%	1.01%
1999	2,427	55.50%	1.01%
90s Tot	21,587	48.73%	0.34%
2000	2,428	56.71%	1.01%
2001	2,428	55.07%	1.01%
2002	2,425	51.75%	1.01%
2003	2,429	52.16%	1.01%
2004	2,428	53.62%	1.01%
2005	2,430	49.88%	1.01%
2006	2,429	53.48%	1.01%
2007	2,431	51.38%	1.01%
00s Tot	19,428	53.01%	0.36%
Grd Tot	41,015	50.76%	0.25%

[smitch124]

Certainly no correlation there, at least without knowing the price, especially in the 2000s.

[ManOfValue]

OK , so for there to be a correlation what does the percentage have to be ? above 60 ?

[ICE-BLOOD]

i would say you could find a real lot of good plays over the course of a full mlb season if you could get that parlay booked on a daily basis for all games

[ManOfValue]

I think you're right Ice-Blood...I'm looking at those numbers and 1994 stands out..Anybody know why so few games were played that year ?

[ICE-BLOOD]

1994, 2000whatever, if you're not allowing a hr within the 1st five your more than likely leading
apply the other handicapping factors and odds and that guy wont be letting you make that bet within a couple weeks

[smitch124]

I think you're right Ice-Blood...I'm looking at those numbers and 1994 stands out..Anybody know why so few games were played that year ?

They went on Strike that year, early August if I remember right...

[RickySteve]

I think you're right Ice-Blood...I'm looking at those numbers and 1994 stands out..Anybody know why so few games were played that year ?

You're kidding, right?

[Ganchrow]

OK , so for there to be a correlation what does the percentage have to be ? above 60 ?

You're missing the point.

We have that for all else being equal (which in general isn't a valid assumption, btw) the probability of the winning pitcher not giving up a home run is 49.24%.

So at 4.545% juice on the ML we'd need, all else being equal, a decimal line of about 1/(49.24%*(1-4.545%)) ≈ on the ML to breakeven on the correlated parlay.

This should give you a good starting point from which to begin to consider individual matchup specifics (i.e., where the "all else equal" assumption breaks down).

I think you're right Ice-Blood...I'm looking at those numbers and 1994 stands out..Anybody know why so few games were played that year ?

Strike.

[Data]

I looked at my spreadsheet of roughly 20,000 games and I see this: the probability of a team not allowing a home run is 35% while the probability of a team not allowing a home run and winning after 5 inning is 44%. This translates to a large advantage.

[ICE-BLOOD]

the wording of this get rich quick parlay is tricky
i think team A's pitcher means starting pitcher no hr allowed, not team's pitchers allows hr

[Ganchrow]

the wording of this get rich quick parlay is tricky
i think team A's pitcher means starting pitcher no hr allowed, not team's pitchers allows hr

I wouldn't read too deeply into this. The reason why this may well be a get-rich-quick parlay without blatantly violating market efficiency is because while one might find many locals and D-rated books that'd take it, no serious book ever would (at least without upping the parlay vig).

[ICE-BLOOD]

I wouldn't read too deeply into this. The reason why this may well be a get-rich-quick parlay without blatantly violating market efficiency is because while one might find many locals and D-rated books that'd take it, no serious book ever would (at least without upping the parlay vig).

completely agree

probably was 1 isolated local book at the corner bar who was winning and thought he could attract more action by offering outrageous prop parlays

[pico]

I looked at my spreadsheet of roughly 20,000 games and I see this: the probability of a team not allowing a home run is 35% while the probability of a team not allowing a home run and winning after 5 inning is 44%. This translates to a large advantage.

where can you buy/get these type of data, data?

[Ganchrow]

where can you buy/get these type of data, data?

Try retrosheet.org.

[pico]

Try retrosheet.org.

wow, thanks. now i can run simulations.

[BDiddy]

How come when I open the files from retrosheet, they all come out scrambled in notepad?

[JVP3122]

How come when I open the files from retrosheet, they all come out scrambled in notepad?

Because they're .csv files. If you're working with Excel you might need to import the .txt file. If you're using Libreoffice or Openoffice you can just open the .txt file with Calc. You might be able to do the same thing with Excel but I'm not sure about it.

[uva3021]

i highly suggest organizing into an SQL database, search tangotiger.net for a good howto, or basql.wikidot.com

[pedro803]

Wow! those look like some good sources of learning UVA -- Thanks a Bunnch!

[BDiddy]

JVP, UVA.... Thank you both!