confused (it doesn't take much)

[bugboy]

I only have 41 races for this certain angle. I know it is a very very small sample.
But. let me give this info.
bet win only.
11 wins out of 41
bet 82.00,ret. 128.10 a +of 46.10 a r.o.i. of 1.56
HOWEVER one paid 53.60/
lowest pay out was 4.40.
do i disregard the 4.40 and the 53.60 to get a realistic r.o.i for this angle?

thanks in advance to anyone who can help me.

[bugboy]

should add,only played races listed fast or firm.
looked at over 700 races found 41 with angle.

so it looks like this system goes into trash bin.

[jay68802]

Originally Posted by bugboy

should add,only played races listed fast or firm.
looked at over 700 races found 41 with angle.

so it looks like this system goes into trash bin.

Always good to have angles like this. I hate it when I see a angle that I occasionally play and they win without $ on it.

[bugboy]

agreed jay

[castaway01]

Originally Posted by bugboy

should add,only played races listed fast or firm.
looked at over 700 races found 41 with angle.

so it looks like this system goes into trash bin.

It's too early to say it goes in the trash bin. You'd have to see if the next 10,000 races go better. There might be more longshot winners in there. I don't know if you're saying you think 41 out of 700 isn't enough, but it would give you a handful of plays on an average Saturday. If it's profitable (and as you said, too soon to know) then that would be enough plays for me.

[Tom]

10,000 races!?

Just in time for the 2145 Derby!

You do realize you DIE in the long run, right?

[Jeff P]

Originally Posted by bugboy

11 wins out of 41
bet 82.00,ret. 128.10

$128.10 collected from 11 winners --

Avg Win Payoff = Amt Collected / Winners

-or:

Avg Win Payoff = 128.10 / 11

-or:

Avg Win Payoff = 11.65


Questions:

If you replace your larger (outlier) win payoffs with (say) the avg win payoff, are the numbers still interesting enough that you might want to track the concept the model is based on?

If so, what do the numbers look like going forward in time after the date you created the model or the date you first started tracking whatever the model is based on?

Are the win pct and avg win payoff (approximately) the same once you start forward testing the model?


-jp

.

[Tom]

If you look at all 11 winners, what is the 80% marks for them?
That is, what price are 80% (7) of the winner under? Say that number is $7. What happens if you put in an odds minimum of 5-2?
You will lose X number of plays, but will it improve your ROI?

Do you do this method by hand or by computer?
Do you include tracks you will not normally be playing?
I had a HTR spot play that was dynamite at Aqu, but threw severe losses at Prx, GP, Tam, and SA.

I would start playing now and keep track of the results.
At the very least, bet $2 on every selection and reevaluate every 20 play.

Data points you actual bet are worth more than those that come from mind bets.

[bugboy]

WOW. thanks guys.ya gave me a lot to digest,i really appreciate all the input.
I knew pace advantage members would come thru.
thanks again.

[bugboy]

Thanks for your input Tom.
Im using a computer program.
As far as trks i don't usually play,
don"t have enough data per trk as yet.
ex.apx 2/5
ctx 0/1
dmr.elp 0/2
flx 2/2
gpx 2/10
prm 1/1
prx 1/2
tdn 2/11
pid 1/1

[Light]

Originally Posted by bugboy

do i disregard the 4.40 and the 53.60 to get a realistic r.o.i for this angle?

I would disregard the 53.6 for now until you get more data. Here's how I know.

I have a system I am presently working on with 900 races. $1800 invested. $2300 returned. 28% ROI. But its 3 largest payouts are $73, $51 and $47. Then there are quite a few $31 and less.

The ones in the low $30's are not an anomaly. There are several. The 3 high payouts are an anomaly because only 1 occurrence of each has happened after 900 races. And unless I played those days I wouldn't have caught them. Whereas I could have not have played on the days when a $31 horse hit, but i would have caught them another day. So without those 3 high payouts my ROI is 18%. This is my true ROI.

[Robert Fischer]

Go back to the $53.60 winner

run a post-mortem

keep an open mind

[how did the favorites run in that race? Did they run their race? Did they all misfire leaving your $53.60 winner a 'default' win, or did your long shot do some running? Etc.. etc...]

What other significant models/factors were at play that day?

You may have accidentally found a long shot angle...

[castaway01]

Originally Posted by Tom

10,000 races!?

Just in time for the 2145 Derby!

You do realize you DIE in the long run, right?

I think the difference is in the interpretation. The only qualifications he gave were tracks fast and firm. Since 36,000 races were run in the U.S. last year, I have no idea why it would take until 2145 to get data on 10,000 races.

[bugboy]

hey cast.
i dont know about 10.000 races
I'm 78. but im gona try.

[thaskalos]

Originally Posted by bugboy

hey cast.
i dont know about 10.000 races
I'm 78. but im gona try.

78...and still a "bugboy"?