I was bored today, so I looked into fading the ESPN Streak For Cash MLB games, because realistically..you'd be fading the public. So I started by going back to April 6th where the season began essentially and I broke it down into 3 categories to begin with:
Less than 5% chance of winning
5%-10% chance of winning
10%-25% chance of winning
(I rounded to the nearest percent. )
The record I got for these are below
5%- 21-16
10%- 13-18
25%- 40-32
total record: 74-66
I then went a step further and filtered out the games labeled as cold and came up with these records. Which didn't have much of a significant change.
5%- 17-16
10%- 12-13
25%- 25-17
total record: 54-46
I also broke them down into the following:
Hot Games
5% 4-4
10% 2-2
25% 3-1
Warmer Games
5% 7-6
10% 5-4
25% 10-3
Warm Games
5% 6-6
10% 5-7
25% 12-13
So filtering out the cold and warm games seemed to help the record a bit
The records will update to the following:
5% 11-10
10% 7-6
25% 13-4
total: 31-20
Obviously its only from this year so its not a very large test sample. But I would imagine the majority of these teams are dogs. I have not actually went back and looked at the lines. But that will be my next step in this test. I will have more of a clearer vision to see if this is something that will work.
Any input would be awesome!
thanks
Less than 5% chance of winning
5%-10% chance of winning
10%-25% chance of winning
(I rounded to the nearest percent. )
The record I got for these are below
5%- 21-16
10%- 13-18
25%- 40-32
total record: 74-66
I then went a step further and filtered out the games labeled as cold and came up with these records. Which didn't have much of a significant change.
5%- 17-16
10%- 12-13
25%- 25-17
total record: 54-46
I also broke them down into the following:
Hot Games
5% 4-4
10% 2-2
25% 3-1
Warmer Games
5% 7-6
10% 5-4
25% 10-3
Warm Games
5% 6-6
10% 5-7
25% 12-13
So filtering out the cold and warm games seemed to help the record a bit
The records will update to the following:
5% 11-10
10% 7-6
25% 13-4
total: 31-20
Obviously its only from this year so its not a very large test sample. But I would imagine the majority of these teams are dogs. I have not actually went back and looked at the lines. But that will be my next step in this test. I will have more of a clearer vision to see if this is something that will work.
Any input would be awesome!
thanks