Averaging Internal Fractions

[coachv30]

Has anyone played around with averaging out internal fractions? I was thinking about experimenting with what I will call vertical and horizontal averages.

I am going to record the IFs from the 3 most recent races at todays distance. For 6 furlongs, i will use 1st Q, 1st Q to 2nd q, and 2nd Q to finish. My excel program will average out all the calls to get an average fraction- horizontal analysis. It will also average out each call point of the 3 races to get an E, M, and L average- vertical analysis.

I will also have an overall average for all fractions recorded from the 3 races. How I will use this info is still to be determined.

Any thoughts?

[coachv30]

Gonna mess around with a few races....


GULFSTREAM 7

$10 WP 3

$1 EX BOX- 3 / ALL

$1 TRI- 3,6 / 2,3,6 / ALL

[Tom]

Maybe if you use the 3 most recent good races at the distance....if there is a bad race, ie, bad break, it would drag down the horse's ratings.

BRIS has a summary - Avg SPeed Figure - that I modeled at AQU last year and it was the most predictive of all the stuff on the summary page - something like 85% of the winners were in the top 3.

I do something similar in HTR - using 2,3,4 pacleines in turf races averaged out.
Very good results.

Good luck - keep us posted.

[coachv30]

Originally Posted by Tom

Maybe if you use the 3 most recent good races at the distance....if there is a bad race, ie, bad break, it would drag down the horse's ratings.

BRIS has a summary - Avg SPeed Figure - that I modeled at AQU last year and it was the most predictive of all the stuff on the summary page - something like 85% of the winners were in the top 3.

I do something similar in HTR - using 2,3,4 pacleines in turf races averaged out.
Very good results.

Good luck - keep us posted.

Thanks Tom...I am indeed looking at comment lines for any red flags.

[coachv30]

Originally Posted by coachv30

Gonna mess around with a few races....


GULFSTREAM 7

$10 WP 3

$1 EX BOX- 3 / ALL

$1 TRI- 3,6 / 2,3,6 / ALL

Way too early to get excited but early reassurance is always nice when a race plays out like you hope.

Finished

$16.20...$7.40 ($81.00....$37.00)

$40.50

$143.20

[coachv30]

OAKLAWN 8

$10 WP 6

$1 EX- 6,8 / ALL

$1 TRI- 6,8 / 2,4,6,7,8 / 2,4,6,7,8

I have to beat the here....

[Mulerider]

Originally Posted by Tom

BRIS has a summary - Avg SPeed Figure - that I modeled at AQU last year and it was the most predictive of all the stuff on the summary page - something like 85% of the winners were in the top 3.

Tom -- Not sure which one you're referring to. There's the "Avg. Dist/Surf PACE-SPEED" column, and then the one that's labeled "Rcg Spd Avg." The latter is just an average of the overall speed ratings of the last three races.

[Tom]

Originally Posted by Mulerider

Tom -- Not sure which one you're referring to. There's the "Avg. Dist/Surf PACE-SPEED" column, and then the one that's labeled "Rcg Spd Avg." The latter is just an average of the overall speed ratings of the last three races.

Rcg Spd Avg is the one.
Sorry - didn't have a copy in front of me.

Using the Avg Dist/Surf section AFTER eliminating all but the to 4 or 5 from the Rcg Spd section gave me a good Sartin-style array to use. I then used the race shapes E,EP, etc, to decide which to bet - early of other than early.

[Mulerider]

Originally Posted by Tom

Rcg Spd Avg is the one.
Sorry - didn't have a copy in front of me.

Using the Avg Dist/Surf section AFTER eliminating all but the to 4 or 5 from the Rcg Spd section gave me a good Sartin-style array to use. I then used the race shapes E,EP, etc, to decide which to bet - early of other than early.

Same here. And if I line out a race that's included in the Rcg Spd section (which is the last 3 race avg) then I'll recalculate it.

[coachv30]

EVANGELINE 2

$5 W/P- 3,7

$1 EX- 3,7 / 2,3,5,7

$1 TRI- 5,7 / 2,3,5,7 / ALL

[coachv30]

Originally Posted by coachv30

EVANGELINE 2

$5 W/P- 3,7

$1 EX- 3,7 / 2,3,5,7

$1 TRI- 5,7 / 2,3,5,7 / ALL

FINISHED

$29.60...$10.60 ($74.00....$26.50)

$59.20

$176.30

[iwearpurple]

coachv30

Through the years, I have been fascinated by the many ideas that people (who clearly are better handicappers than me) have posted on this board. As a numbers type guy, I like to try various ideas brought up here, and I was looking at your idea.

I quickly found what I believe is a flaw in your formulas. I'll use your original race that you "messed around" with. When you average the internal fractions, you are assuming that 22.1 is 22 and 1/10 second, in reality it should be calculated as 22.2 (22 and 1/5 second) when averaging.

So your Score (Avg Tot) for the winning horse 3 should be 23.48889 (22.8 + 22.6 + 25.6 / 3 = 23.48889) You used (22.4 + 22.6 + 25.6 /3 = 23.188889). While this horse (#3) would still be the pick, you would not have had the trifecta.

The 7 horse would have been the 3rd choice rather than the 6 who finished 3rd.

Scores using what I believe to be the correct internal fractions are:

1 - 23.82222
2 - 23.54444 - 2nd
3 - 23.48889 - 1st
4 - 24.2
6 - 23.7333
7 - 23.6667 - 3rd

I intend to do some more playing around with your idea, as this certainly hit a nice payoff for the first race you "messed around" with.

[coachv30]

PENN 5

$10 WIN 5

$10 EX 5-4

$5 TRI- 5 / 4 / 2,3

[coachv30]

Originally Posted by iwearpurple

coachv30

Through the years, I have been fascinated by the many ideas that people (who clearly are better handicappers than me) have posted on this board. As a numbers type guy, I like to try various ideas brought up here, and I was looking at your idea.

I quickly found what I believe is a flaw in your formulas. I'll use your original race that you "messed around" with. When you average the internal fractions, you are assuming that 22.1 is 22 and 1/10 second, in reality it should be calculated as 22.2 (22 and 1/5 second) when averaging.

So your Score (Avg Tot) for the winning horse 3 should be 23.48889 (22.8 + 22.6 + 25.6 / 3 = 23.48889) You used (22.4 + 22.6 + 25.6 /3 = 23.188889). While this horse (#3) would still be the pick, you would not have had the trifecta.

The 7 horse would have been the 3rd choice rather than the 6 who finished 3rd.

Scores using what I believe to be the correct internal fractions are:

1 - 23.82222
2 - 23.54444 - 2nd
3 - 23.48889 - 1st
4 - 24.2
6 - 23.7333
7 - 23.6667 - 3rd

I intend to do some more playing around with your idea, as this certainly hit a nice payoff for the first race you "messed around" with.

Funny thing is...I thought about that as well. I am indeed using 1/5 seconds but Excel doesn't recognize it. When I manually input, I will convert a 23.7 to a 24.2. However, when Excel does the averaging, it will not do so.

Therefore, when evaluating, I have to keep in mind that a 26.1 is actually better than a 25.8.

In the Penn 5 above, I did not make those adjustments. Therefore, a 24.6 stayed as is, instead of making it a 25.1. I believe that this here is the skewed way of doing it.

[iwearpurple]

Originally Posted by coachv30

Funny thing is...I thought about that as well. I am indeed using 1/5 seconds but Excel doesn't recognize it. When I manually input, I will convert a 23.7 to a 24.2. However, when Excel does the averaging, it will not do so.

Therefore, when evaluating, I have to keep in mind that a 26.1 is actually better than a 25.8.

In the Penn 5 above, I did not make those adjustments. Therefore, a 24.6 stayed as is, instead of making it a 25.1. I believe that this here is the skewed way of doing it.

Here is a formula that I have used to address this issue. There may be simpler formulas, but this is what I have used for many years

Lets call the cell with your IF1 "C2"

=((C2-INT(C2))*2+INT(C2)

So if cell C2 is 22.1, this formula (placed in another cell) will return 22.2.