Scratches and Spot Plays

[JohnB]

Question for the "Spot Play" guys:

What kind of change in your ROI have you seen in your spot plays when plays are made before late scratches are taken into consideration? I have been working on a couple of high volume spot plays, looking for a way to take advantage of rebates. Ideally looking to put a bunch of bets in early in the morning, and hopefully count the winnings in the evening :)

I have 2 plays that have tested well forward and backward, using "after-scratch" horses. Spot Play #1 gets around 20% winners with 1.03 ROI (before rebate), generating around 400 plays per month (across all tracks), and Spot Play #2 gets around 25% winners with 1.07 ROI (again before rebate), generating around 300 plays per month. I have been tracking the "before-scratch" results for the last 2 months, and the "before-scratch" Spot Play #1 has underperformed the "after-scratch" results for the same time period by about .08, while the "before-scratch" Spot Play #2 has outperformed the "after-scratch" results by about .06. I realize that 2 months of data is not sufficient to draw any significant conclusions - wondering if anyone else has done a similar study of their spot plays in terms of "before-scratch" and "after-scratch" results. I seem to recall an old post by someone (Fred? Hurrikane?) that mentioned that their ROI generally drops by around .05 when they put in their spot plays before scratches, but I can't seem to find that post. Anyone have any thoughts on the subject?

Thanks,

John

[fred4now]

Ahh, something I actually know something about. This is exactly what I do everyday. How I solved the problem is I set up my db for research without the scratches so it would mimic how I actually bet. Has held up really well going forward. If you want to know more about how I do the db let me know.

[JohnB]

Fred -

Thanks for your reply. I would like to know how you maintain your database (and any other words of wisdom you may have). I seem to recall that you have 2 databases set up - a pre-scratch database, and another post-scratch database?

John

[fred4now]

OK,
First I run the B files thru the exporter and create an all_hx4 without results and import this into a table that I call "noresults".
Then I delete the all_hx4 file and bring in the c and q files and run the exporter again. I import this all_hx4 file into a table I call "results".
Then I have a append query that has all the fields from noresults table except win, pla, sho, naodd, raodd and the win, pla, sho, naodd, raodd from the results table. Create the joins between the tables on date, race, track, program. I then append this data to a main table and delete the data from the noresults and results tables. Now I can query this main table all I want and it is the same as putting my bets in before scratches. No more guessing on if the ROI drops or not.

I have to think a little on how I set up my ROI functions, because I do have it refund my $2 if there is a scratch.

[JohnB]

Fred -

Thanks for your response.

Pretty nifty - I follow what you do. I'll have to try that. I always struggled with the idea of running queries on before-scratched fields, but if your experience shows that spot plays queried in that manner have held up going forward, then I am going to take a look into doing that. A couple of other questions, if I may:

1) Do you generally play high volume/low ROI type spot plays, or plays that are more selective that usually have a higher ROI? My favorite spot play is a low hit rate (around 14% win percentage), high ROI play. I need to monitor scratches on that play for it to be profitable. Miss a couple of 50-1 horses, and there goes the ROI. I think I'll be more successful with before scratch plays on my high volume type spot plays that have a much lower average mutuel.

2) When you did follow the ROI with plays before/after scratches, did you notice a general trend up/down with ROI with plays made before and after scratches?

John

[fred4now]

1) I am playing 4 spot plays right now, 3 are around 18-20%win and 1-5%ROI before rebate and one is 80%place and break even ROI before rebate. Today I have 7 plays, a little low for a Thursday.

2) Not sure I understand the question. If I have a play that has a decent amount of plays that has shown a profit every year for the last 5 years, I am pretty confident. I really like looking at the data in time frames like years or months to see if a few big winners have skewed the data. The other thing that has made a big difference for me is looking at a graph like Victor is doing rather than just the ROI. So far this year I have been showing just mediocre results until this last week and then everything came in line. Watching the graph helps me keep the confidence to keep betting or I would have missed the great week. Even with these good percentage plays I am always surprised on the swings.

Also I have been looking at ROI a little differently, instead of just looking at the number, I want to know how long it takes to double my bank. With my present plays it took 3 months betting 2% of bank to double.

[JohnB]

Thanks for your replies Fred - sounds like you and I have a couple of spot plays that get similar results. I appreciate your time.

John