Baseball's Quick Revenge System Bet only favorites that are playing at home and were defeated by three or more runs the previous game by the same team they are playing in game 2 of the series.
Killersports.com SQDL Query t:site=home and t:line<-105 and opo:runs+3<=op:runs and SG=2 and season = 2009
2009 SU: 83-54 (0.9 rpg) average line: -154 / +140 on / against: +$95 / -$835 ROI: +0.4% / -6.1% 2010 SU: 63-50 (0.5 rpg) average line: -155 / +140 on / against: -$956 / +$341 ROI: -5.5% / +3.0% 2011 SU: 82-53 (0.3 rpg) average line: -147 / +135 on / against: +$662 / -$1,283 ROI: +3.3% / -9.5% 2012 SU: 35-18 (1.5 rpg) average line: -145 / +133 on / against: +$930 / -$1,141 ROI: +12.1% / -21.5%
Results are nothing to write home about - plus took a loss in 2010 BUT.......if you add the filter that the series 1 game opponent was a dog in that game - so the dog came in and beat them at home - they fair better in game 2
Killersports.com SQDL Query t:site=home and t:line<-105 and op:line >-105 and opo:runs+3<=op:runs and SG=2
2009 SU: 67-35 (1.0 rpg) average line: -161 / +146 on / against: +$1,055 / -$1,585 ROI: +6.4% / -15.5% 2010 SU: 44-38 (0.5 rpg) average line: -163 / +147 on / against: -$1,296 / +$801 ROI: -9.7% / +9.8% 2011 SU: 70-39 (0.6 rpg) average line: -151 / +139 on / against: +$1,375 / -$1,815 ROI: +8.3% / -16.6% 2012 SU: 33-15 (1.8 rpg) average line: -147 / +135 on / against: +$1,103 / -$1,284 ROI: +15.6% / -26.7%
The results are better - except for the loss in 2010 still - I would still would want to find filters that increased the win rate
0
Baseball's Quick Revenge System Bet only favorites that are playing at home and were defeated by three or more runs the previous game by the same team they are playing in game 2 of the series.
Killersports.com SQDL Query t:site=home and t:line<-105 and opo:runs+3<=op:runs and SG=2 and season = 2009
2009 SU: 83-54 (0.9 rpg) average line: -154 / +140 on / against: +$95 / -$835 ROI: +0.4% / -6.1% 2010 SU: 63-50 (0.5 rpg) average line: -155 / +140 on / against: -$956 / +$341 ROI: -5.5% / +3.0% 2011 SU: 82-53 (0.3 rpg) average line: -147 / +135 on / against: +$662 / -$1,283 ROI: +3.3% / -9.5% 2012 SU: 35-18 (1.5 rpg) average line: -145 / +133 on / against: +$930 / -$1,141 ROI: +12.1% / -21.5%
Results are nothing to write home about - plus took a loss in 2010 BUT.......if you add the filter that the series 1 game opponent was a dog in that game - so the dog came in and beat them at home - they fair better in game 2
Killersports.com SQDL Query t:site=home and t:line<-105 and op:line >-105 and opo:runs+3<=op:runs and SG=2
2009 SU: 67-35 (1.0 rpg) average line: -161 / +146 on / against: +$1,055 / -$1,585 ROI: +6.4% / -15.5% 2010 SU: 44-38 (0.5 rpg) average line: -163 / +147 on / against: -$1,296 / +$801 ROI: -9.7% / +9.8% 2011 SU: 70-39 (0.6 rpg) average line: -151 / +139 on / against: +$1,375 / -$1,815 ROI: +8.3% / -16.6% 2012 SU: 33-15 (1.8 rpg) average line: -147 / +135 on / against: +$1,103 / -$1,284 ROI: +15.6% / -26.7%
The results are better - except for the loss in 2010 still - I would still would want to find filters that increased the win rate
Baseball's Departing Dogs Bet the underdog that won the previous game as an underdog, while allowing six hits or less, and the game being played is the final game in the series.
Killersports.com SQDL Query 0<tp:line and 0<t:line and tpo:runs<tp:runs and op:hits<=6 and SG = 3 and SGS = 3 and season = 2009
2009 SU: 35-30 (0.6 rpg) average line: +135 / -149 on / against: +$1,815 / -$2,275 ROI: +27.9% / -23.5% 2010 SU: 24-41 (-1.3 rpg) average line: +149 / -165 on / against: -$640 / +$310 ROI: -9.8% / +2.9% 2011 SU: 15-34 (-1.7 rpg) average line: +133 / -146 on / against: -$1,554 / +$1,404 ROI: -31.7% / +19.7% 2012 SU: 11-15 (-0.9 rpg) average line: +129 / -140 on / against: -$153 / +$43 ROI: -5.9% / +1.2%
Pretty inconsistant - not sure if my query is wrong or the website with this system - but the numbers are off - because there numbers show no losing seasons
0
Baseball's Departing Dogs Bet the underdog that won the previous game as an underdog, while allowing six hits or less, and the game being played is the final game in the series.
Killersports.com SQDL Query 0<tp:line and 0<t:line and tpo:runs<tp:runs and op:hits<=6 and SG = 3 and SGS = 3 and season = 2009
2009 SU: 35-30 (0.6 rpg) average line: +135 / -149 on / against: +$1,815 / -$2,275 ROI: +27.9% / -23.5% 2010 SU: 24-41 (-1.3 rpg) average line: +149 / -165 on / against: -$640 / +$310 ROI: -9.8% / +2.9% 2011 SU: 15-34 (-1.7 rpg) average line: +133 / -146 on / against: -$1,554 / +$1,404 ROI: -31.7% / +19.7% 2012 SU: 11-15 (-0.9 rpg) average line: +129 / -140 on / against: -$153 / +$43 ROI: -5.9% / +1.2%
Pretty inconsistant - not sure if my query is wrong or the website with this system - but the numbers are off - because there numbers show no losing seasons
Yea the site showed a positive long term return on all of the systems. I'd rather trust you than some random About.com writer though. I think there was around 11-15 systems in total but there were only a handful that seem to make any sense while the rest is just predicting randomness.
The thing I wanted to show you was this.
I was trying to estimate runs with some more advanced metrics like Runs Created and Batting Runs and so on. What I'm trying to do is take the average estimated runs of a few of those metrics, average it out and then multiply it by a pitcher factor.
A worse pitcher would have a higher multiplier while a great pitcher will have a lower multiplier. To find the pitcher multiplier I was thinking of using current ERA, component ERA, and Expected ERA ( I found this on the about.com site ). So you would average those ERA values out. So originally I wanted this to be the pitcher multiplier but if a team is estimated to have 4 runs a game, multiplying by any ERA value is retarded since itll give you like 10 or 15 or 20 runs . So what I was thinking is to use natural logarithm.
Ex. Say a pitcher has an average ERA of 3. the ln(3)= 1.1
then you multiply the estimated runs by 1.1 instead of the 3.
So from the graph its easy to see how the change in the y value (pitcher multiplier) increases exponentially as the x value increases during the [1,6] interval.
Let's do an example.
estimated runs = 5
estimated ERA = 10 (horrible pitcher, me in MLB 2k12)
5 x ln(10) = 5 x 2.3 = 12
So a team that usually is expected to score 5 runs will now score 12 batting against this pitcher
Another example
estimated runs = 5
estimated ERA = 2
5 x ln(2) = 5 x .69(giggity) = 3
This team will score 2 runs less than expected against this amazing pitcher.
I don't know how well I explained this and it might all be a load of crap but take a look at it and feel free to critizize it and rip it up. There is obviously a lot more room for improvement
0
Yea the site showed a positive long term return on all of the systems. I'd rather trust you than some random About.com writer though. I think there was around 11-15 systems in total but there were only a handful that seem to make any sense while the rest is just predicting randomness.
The thing I wanted to show you was this.
I was trying to estimate runs with some more advanced metrics like Runs Created and Batting Runs and so on. What I'm trying to do is take the average estimated runs of a few of those metrics, average it out and then multiply it by a pitcher factor.
A worse pitcher would have a higher multiplier while a great pitcher will have a lower multiplier. To find the pitcher multiplier I was thinking of using current ERA, component ERA, and Expected ERA ( I found this on the about.com site ). So you would average those ERA values out. So originally I wanted this to be the pitcher multiplier but if a team is estimated to have 4 runs a game, multiplying by any ERA value is retarded since itll give you like 10 or 15 or 20 runs . So what I was thinking is to use natural logarithm.
Ex. Say a pitcher has an average ERA of 3. the ln(3)= 1.1
then you multiply the estimated runs by 1.1 instead of the 3.
So from the graph its easy to see how the change in the y value (pitcher multiplier) increases exponentially as the x value increases during the [1,6] interval.
Let's do an example.
estimated runs = 5
estimated ERA = 10 (horrible pitcher, me in MLB 2k12)
5 x ln(10) = 5 x 2.3 = 12
So a team that usually is expected to score 5 runs will now score 12 batting against this pitcher
Another example
estimated runs = 5
estimated ERA = 2
5 x ln(2) = 5 x .69(giggity) = 3
This team will score 2 runs less than expected against this amazing pitcher.
I don't know how well I explained this and it might all be a load of crap but take a look at it and feel free to critizize it and rip it up. There is obviously a lot more room for improvement
Marko - looks interesting - but to be honest with you the math is a bit over my head - but if it can done done in excel I can probably do it.
And that wiki article on Natural Logarithm was defintely over my head.
But that being said - i actually have done some work on Base Runs and Runs Created. And I have automated it into an excel spreadsheet. The problem I see is more times than not - a pticher has not faced most players in a given lineup, OR they will have so little AB's against them that the calculation is skewed - for example if a batter is 1-2 against a pitcher and batting .500 but only faced him once - you can't expect that to happen every time. The 2nd issue I ran into is at it's core - Base Runs and RUns created are run estimators given the raw statistics of a player. The result being if that player batted every position for all 9 innings. Now again you could figure out what that estimate would be per AB and then multiple by how many AB's that player would have in a game - but you still run into the same problem - that player will not have that averaged liftime performace at every AB.
Now I could be interpreting Base Runs and Runs created incorrectly - who knows. But the one thing it does do - is give you a lean on which team COULD perform better on a given matchup IF they perform the way they did historically. Then you could layer other handicappign methods to exclude it as a play or not.
Again - this is just from what I gather from my research. I could be way off base.
But even saying that I still think that statistics are the key. I just don't know what statistics to look at. And I agree with you that you need a way to determine how the Starting Pitcher will perform against the opponents starting lineup. How to do that is the question....
0
Marko - looks interesting - but to be honest with you the math is a bit over my head - but if it can done done in excel I can probably do it.
And that wiki article on Natural Logarithm was defintely over my head.
But that being said - i actually have done some work on Base Runs and Runs Created. And I have automated it into an excel spreadsheet. The problem I see is more times than not - a pticher has not faced most players in a given lineup, OR they will have so little AB's against them that the calculation is skewed - for example if a batter is 1-2 against a pitcher and batting .500 but only faced him once - you can't expect that to happen every time. The 2nd issue I ran into is at it's core - Base Runs and RUns created are run estimators given the raw statistics of a player. The result being if that player batted every position for all 9 innings. Now again you could figure out what that estimate would be per AB and then multiple by how many AB's that player would have in a game - but you still run into the same problem - that player will not have that averaged liftime performace at every AB.
Now I could be interpreting Base Runs and Runs created incorrectly - who knows. But the one thing it does do - is give you a lean on which team COULD perform better on a given matchup IF they perform the way they did historically. Then you could layer other handicappign methods to exclude it as a play or not.
Again - this is just from what I gather from my research. I could be way off base.
But even saying that I still think that statistics are the key. I just don't know what statistics to look at. And I agree with you that you need a way to determine how the Starting Pitcher will perform against the opponents starting lineup. How to do that is the question....
Ideally I want to automate how to handicap a typical matchup in excel:
So I would think you need to do the following:
- Figure out how the starting pitcher would fair against the opponent lineup. Determine which team has the edge - Figure out how the bullpen would fair against the opponent lineup. Determine which team has the edge
- Now figure out if the pitching or batting of either team is streaking. This is huge - we all know that there are peaks and valleys in basball. Some time bats are hot and sometimes not - sometimes pitching is hot or not
I read in the one of the articles you sent - that a way to predict future pitcher performance is to look at a starting pitcher and their recent form was to take the pitcher's season ERA and double it. Then subtract the pitcher's ERA from the last three games to get an estimate of their expected performance. it's called - Baseball Pitcher Form Reversal
Not sure how to do this for a linup - but looking at their statistics in their last few games and has to be part of it. Specifically if they are performing the way they should against the team they were playing - or playing over or under their potential
That's all I have for now - and in theory what I have so far makes sense - but I could be missing something
0
Ideally I want to automate how to handicap a typical matchup in excel:
So I would think you need to do the following:
- Figure out how the starting pitcher would fair against the opponent lineup. Determine which team has the edge - Figure out how the bullpen would fair against the opponent lineup. Determine which team has the edge
- Now figure out if the pitching or batting of either team is streaking. This is huge - we all know that there are peaks and valleys in basball. Some time bats are hot and sometimes not - sometimes pitching is hot or not
I read in the one of the articles you sent - that a way to predict future pitcher performance is to look at a starting pitcher and their recent form was to take the pitcher's season ERA and double it. Then subtract the pitcher's ERA from the last three games to get an estimate of their expected performance. it's called - Baseball Pitcher Form Reversal
Not sure how to do this for a linup - but looking at their statistics in their last few games and has to be part of it. Specifically if they are performing the way they should against the team they were playing - or playing over or under their potential
That's all I have for now - and in theory what I have so far makes sense - but I could be missing something
Wow. I think you really have laid the foundations of something incredible here.
So essentially what you are trying to do is compare the SP and bullpen to each individual batter. Like you said, there will be a lack of data.
A solution to this I think would be to set up a simple rating system.If you have ever been on Fangraphs.com's glossary which explains many metrics, it gives what the league average of the metric is than it has ranges from bad, below average, average, above average, etc.
If you were to set the baseline to average = 0, then you could maybe give 1 point for above average, 2 points for great and the opposite with bad players. So -1 for below average and so on.
I think this point system can be given a maximum value and each player can have a percentage value. So like if a player has 5 of a possible 10 points from a few different metrics, his rating would be 50%.
I'm still not sure how all of this could apply to a pitcher but maybe this could work:
Lets say,
Batter 1: 50%
Batter 2: 80%
SP: 60%(still not sure how to get this number)
the difference between batter 1 and SP is -10% and the difference between batter 2 and SP is 20%.
So this could mean the lineup is 10% better than the SP which can somehow be converted into an actual number and not some arbitrary percentage.
This is really raw theory and I'm not sure about it but I think it should at least determine value in lineups and pitchers.
0
Wow. I think you really have laid the foundations of something incredible here.
So essentially what you are trying to do is compare the SP and bullpen to each individual batter. Like you said, there will be a lack of data.
A solution to this I think would be to set up a simple rating system.If you have ever been on Fangraphs.com's glossary which explains many metrics, it gives what the league average of the metric is than it has ranges from bad, below average, average, above average, etc.
If you were to set the baseline to average = 0, then you could maybe give 1 point for above average, 2 points for great and the opposite with bad players. So -1 for below average and so on.
I think this point system can be given a maximum value and each player can have a percentage value. So like if a player has 5 of a possible 10 points from a few different metrics, his rating would be 50%.
I'm still not sure how all of this could apply to a pitcher but maybe this could work:
Lets say,
Batter 1: 50%
Batter 2: 80%
SP: 60%(still not sure how to get this number)
the difference between batter 1 and SP is -10% and the difference between batter 2 and SP is 20%.
So this could mean the lineup is 10% better than the SP which can somehow be converted into an actual number and not some arbitrary percentage.
This is really raw theory and I'm not sure about it but I think it should at least determine value in lineups and pitchers.
Wow. I think you really have laid the foundations of something incredible here.
So essentially what you are trying to do is compare the SP and bullpen to each individual batter. Like you said, there will be a lack of data.
A solution to this I think would be to set up a simple rating system.If you have ever been on Fangraphs.com's glossary which explains many metrics, it gives what the league average of the metric is than it has ranges from bad, below average, average, above average, etc.
If you were to set the baseline to average = 0, then you could maybe give 1 point for above average, 2 points for great and the opposite with bad players. So -1 for below average and so on.
I think this point system can be given a maximum value and each player can have a percentage value. So like if a player has 5 of a possible 10 points from a few different metrics, his rating would be 50%.
I'm still not sure how all of this could apply to a pitcher but maybe this could work:
Lets say,
Batter 1: 50%
Batter 2: 80%
SP: 60%(still not sure how to get this number)
the difference between batter 1 and SP is -10% and the difference between batter 2 and SP is 20%.
So this could mean the lineup is 10% better than the SP which can somehow be converted into an actual number and not some arbitrary percentage.
This is really raw theory and I'm not sure about it but I think it should at least determine value in lineups and pitchers.
interesting angle Marko - I think the only problem with that is that is players delta from the average based on who they have played so far in the season. A player may be ranked high if they have faced sub par pitching
0
Quote Originally Posted by marko123:
Wow. I think you really have laid the foundations of something incredible here.
So essentially what you are trying to do is compare the SP and bullpen to each individual batter. Like you said, there will be a lack of data.
A solution to this I think would be to set up a simple rating system.If you have ever been on Fangraphs.com's glossary which explains many metrics, it gives what the league average of the metric is than it has ranges from bad, below average, average, above average, etc.
If you were to set the baseline to average = 0, then you could maybe give 1 point for above average, 2 points for great and the opposite with bad players. So -1 for below average and so on.
I think this point system can be given a maximum value and each player can have a percentage value. So like if a player has 5 of a possible 10 points from a few different metrics, his rating would be 50%.
I'm still not sure how all of this could apply to a pitcher but maybe this could work:
Lets say,
Batter 1: 50%
Batter 2: 80%
SP: 60%(still not sure how to get this number)
the difference between batter 1 and SP is -10% and the difference between batter 2 and SP is 20%.
So this could mean the lineup is 10% better than the SP which can somehow be converted into an actual number and not some arbitrary percentage.
This is really raw theory and I'm not sure about it but I think it should at least determine value in lineups and pitchers.
interesting angle Marko - I think the only problem with that is that is players delta from the average based on who they have played so far in the season. A player may be ranked high if they have faced sub par pitching
If you choose to make use of any information on this website including online sports betting services from any websites that may be featured on
this website, we strongly recommend that you carefully check your local laws before doing so.It is your sole responsibility to understand your local laws and observe them strictly.Covers does not provide
any advice or guidance as to the legality of online sports betting or other online gambling activities within your jurisdiction and you are responsible for complying with laws that are applicable to you in
your relevant locality.Covers disclaims all liability associated with your use of this website and use of any information contained on it.As a condition of using this website, you agree to hold the owner
of this website harmless from any claims arising from your use of any services on any third party website that may be featured by Covers.
Hot Forum Topics
