My twenty race cycle

[hansend]

Before Ted posted about Dick Schmidt's "Beginner play" I had heard about it numerous times in listening to audio and reading of the followups.

So after a while of a little bit of learning RDSS and doing a few things here and there I set out to use this as my beginning strategy. I did a slightly modified version of 60/40 with $5 (3/2) instead of $10 as he suggest.

Unfortunately I did not get to "Step 3 - Take a $100 profit...". As a matter of fact I can't even say I made a profit. On top of that I only was able to hit 30% of those races.

I'm not unmotivated by this, but anything I can learn by doing this can only be a positive for me.

In the attachment I have included my 20 race cycle from these cards:
MNR 5/24
MNR 5/25
PEN 5/25
PEN 6/02
PEN 6/03
CT 6/03

In there you will find my models I downloaded and my bets (a few comments on some of the bets). I have been keeping the full models of all contenders in case one of them wins way down on the BL/BL but does have a factor to include modeling which I have started to do. In most cases when I lowered to five for betting purposes it did not change many of them, but if you see a bit of a discrepency between what I bet and model that may be why.

Again if by chance you did any portion of these cards and can assist in my learning I would appreciate it. I will continue to be doing these for my next twenty race cycle and will probably continue to post them.

Thanks,
Derek

[barb craven]

Derek, I'm bit confused. Why do you keep your models if you don't include your winners in all your races. I put the winner in even if I don't win so I can see the trends for that particular track.

What are you going to use your model for? I see very little results put in those races. And yes you will see things move around on your BL when you hide down to 5.

You asked bout CT race 5 and why you would use #2 so far down. Well my model tells me that at this time TT is winning a great deal of the time there.


I hope this helps a bit.

Barb

[Ted Craven]

Derek,

Thank you very much for sharing this with us, it takes a little bravery. I'll take a look at it over the weekend and see if I have anything. I'll go over the cards you worked, not just the races you worked and check the odds available at 2 MTP (via the HPI website odds intervals archive).

I suggest you not include the cost of data when calculating your ROI, just to keep your returns from betting operations distinct from your operating costs. You also don't include annual handicapping software costs, costs of maintaining an internet connection, phone line, computer amortization, other software licenses (e.g. Excel), or if you bet on track: costs of travel, admittance, food, etc, etc. These are all costs for sure of a betting business, but your income statement should separate gross margin (wagers won less wagers made) from operational costs.

Thanks again for putting this out in public - you are serious about this.

Ted

[hansend]

Barb,

Thanks for the information about TT and CT. I was (until you said something) very trepid in relying on only one or two indicators especially something like TT where in my opinion it seemed like an indicator that needed an additional supporting factor.

About your other comment, if I am interpreting it correctly. I am modeling the winner for every single race and keeping track of it all. I only included the models as to which I actually made the bets on and included only my bets. Mainly as a guide to let me know if I chose a wrong pace line and/or made a bad betting decision. The races not included I elected to pass or were outside my comfort level.

If desired I could attach another upload with all my models for each race covering those days. I will not have some of the maidens because until I get a nice comfort level I did not keep a model that only had 2-3 contenders in an 8 horse field (just an example) where the top 2-3 were the top betting odds and won. Correct me if I am wrong, but a model seems best when you can get a full list of contenders that may "spread out" some of the indicators on what is needed to win a race.

[barb craven]

Derek, I'm still not clear on what you do with your models. My understanding of a track model is to find out who the winners are no matter what they pay. I put in only the winners, no matter whether I win, lose or pass in my models. If I work a race, it goes into a model. I break down my models by distance ie. 4.5,5,5.5,6 etc.. that way if a trend shows up at a particular distance, I pay attention to it, of course along with other factors.

All the other things you talk about I record on my 20 race cycle sheets. That tells me exactly how I'm doing and where my winners are coming from. I find the biggest ones are usually coming from the third or fourth tier. I lose more than I win, but my wins are usually in double digits.

The modeling is part of my handicapping decision. When I've chosen my contenders, that is over and done.

The wagering part is a whole other thing. That's when I have to decide whether it's a bet or pass depending on the odds.

Of course this is only my way of doing things, everyone has their own way.

If you find modeling all that other information is helpful, could you explain how you use it please?

I think Ted is going to go over your races and will probably come up with more good advice than I can give you. Good luck. You're doing the right thing by showing what you do and asking for advice. That's the only way you learn. Ask me how I know. lol I'm still pretty new to this myself.

Barb

[Ted Craven]

Derek,

The Excel files you posts are not (yet) Decision Models, they are only raw output data from the horses you included in your Analysis set, race by race. This data turns into a model when you copy the line for the winner into a Model file for a given track, separated by surface and distance structure.

For example, I took your MNR races from May 24 and 25, looked up the Results, marked the Win, Place and Show finishers in Column I (labeled Result) of the output files and added the Win and Place mutuels in Columns BX and BY respectively (labeled $Win and $Place). Then I took the Master Template Excel file found here: \rdss\Models\Templates\MasterTemplate.xls and made a copy of it which I renamed MNR.xls. That template comes with one initial Tab, which I duplicated into a second Tab, then labeled the first one Sprints and the second one Routes.

Then I copied the line for the Winner (Column I = W) into the worksheet tab for the appropriate distance structure in my newly created MNR.xls file. You had 7 sprints ranging from 5 - 6 furlongs, and one route.

The resulting Decision Model for MNR, consisting of your chosen races and pacelines, is attached here:

MNR.xls

The raw output file for MNR May 25, Race 5 (for example) which I added Result info to, is here:

MNR20100525-05.xls

Take a look at those factors where the Summary shows in red, representing factors where the top 2 ranks exceeds 63%. The usual suspects, V/DC and Supplementary show 71% in the Top 2 (though V//DC has ties), but interestingly LPR and L/ep also show 71% Top 2, including your $117 horse on May 25, race 7.

If you are connected to the toteboard when you export the working files, the mutuel results and tote data are also exported, so you don't have to add in that data by hand.

You should redo this yourself, or start with the MNR Decision Model attached, and update the rest of the tracks you worked to see if you can get any further guidance as to which horses make good contenders for the races you're working.

Now, this has nothing to do so far with Wagering decisions, as you still need to determine if, in any given race, knowing now a little about what factors indicate a win contender, whether there is a wager available. Some contenders indicated by your models will be obvious to everyone, thus no value, at least in the Win pool. There may be value in other pools. For example, even with a Win model, you may have found that $117 horse inscrutable, but at the odds, given some Late factor intelligence (it was also the MOST Late on the E/L graph) you may have found a place for it in an Exacta or to Place and Show. Etc.

I would advise you to have the same number of horses in each race output set, so the ranks are comparable (i.e. rank 2 of 10 is less comparable than rank 2 of 5). I recommend 5 horses. If you simply left the winner out of the mix, add it back in as the 6th horse, perhaps flagging it with a note to yourself in a column that it had been omitted. I bet as time goes on, you find this happening very infrequently!

Again, these are Decision Models based on your before race analysis and a consistent paceline selection strategy. It is not a Wager Decision model, which I would keep separately, and which we can discuss separately.

I wills till go over your output files further and see if I have any advice about Analysis or Wager Decisions, both int he races you worked and reported, and in races you chose not to work.

Now, on to today's Belmont Stakes card!

cheers,

Ted

[hansend]

Thanks Ted and Barb,

It finally clicked what you were meaning Barb further with Ted's explanation.

I am creating those decision models like you had mentioned and Ted has attached.

I was giving my BL/BL downloads since I did not do screen shots of my top 5. The main purpose for those sheets was to give anyone who looked at it a chance to see what my final pace lines were per horse to see if I made any glaring mistakes.

I will go over your MNR model Ted and compare it with the one I have and see if we have agreements.

Lots of work still for me to do!

Not sure how much I'll be getting in next week, but I might just bet by top two odds based on M/L and/or what model says are winners as I may have a busy week, but going to keep it going.

Thanks for the assistance.

Derek

[JimG]

Great thread! Hansend, it takes alot of guts to go public like that. I admire you for that.

I would like to know of those here that apply the 2 horses to win way of betting and keep records, what is your hit rate? I ask because Pizzolla has indicated that when he was in Pirco, he had a high hit rate (65%-70%) and was winning many races, but not money. I have not done any formal studies, but it seems to me that very short win prices and long win prices are more commonplace now than in the past.

If I ever write a book, I am going to entitle it...whatever happened to the $12 win horse?

My guess is if success is to be had betting two horses to win you must be very selective in your races wagered and have a very high hit rate or bet more races searching for value, expecting a lower winning percentage than what was talked about in the follow-ups. I look forward to any comments.

Jim

[For The Lead]

Originally Posted by JimG

Great thread! Hansend, it takes alot of guts to go public like that. I admire you for that.

I would like to know of those here that apply the 2 horses to win way of betting and keep records, what is your hit rate? I ask because Pizzolla has indicated that when he was in Pirco, he had a high hit rate (65%-70%) and was winning many races, but not money. I have not done any formal studies, but it seems to me that very short win prices and long win prices are more commonplace now than in the past.

If I ever write a book, I am going to entitle it...whatever happened to the $12 win horse?

My guess is if success is to be had betting two horses to win you must be very selective in your races wagered and have a very high hit rate or bet more races searching for value, expecting a lower winning percentage than what was talked about in the follow-ups. I look forward to any comments.

Jim

The $12 horse is as alive and well today as it was 14 years ago. In databases that have more than 40,000 winners in each of those years, the average winning price is over $12 in each and every year and, I suspect, in the many years before that as well, although I do not have the information to support that belief. That comes to more than 500,000 winners, so the result is quite stable.

I can also tell you that over these same races, winners paying 5/2 or less make up 58% to 60% of all the winners, so "selectivity" is a must. It would behoove all handicappers, regardless of methodology, to have an understanding of basic handicapping principles, including class levels and race conditions, in order to help them reach a determination as to which of these short priced winners are legitimate and potentially the winners of their races and which ones are not so they can be bet against. It is from those races that value can be had.

[For The Lead]

Originally Posted by JimG

If I ever write a book, I am going to entitle it...whatever happened to the $12 win horse?

Jim

Just thought I would give you one more stat on the $12 horses you talked about.

Each year, give or take 1/100ths of a percent, $12 horses win 3% of all the races.