PDA

View Full Version : At The Races review Pt II


Dick Schmidt
07-04-2002, 05:33 AM
Continued from Part I of the review of the At The Races betting program.




Now let’s get down to boxing and wheeling our handicapped picks. Directly below the Low 2/3 etc. boxes is an exacta layout. This assumes a $2 exacta and in the top row displays the win payoffs for each horse. Down the left side are the three types of exactas you can use: Box, Wheel and Back-wheel. To box up to five horses, just click on their numbers in the top row and the program will dutch them for you. As with win, you can also hedge. In this case, hit the payoff in the exacta matrix above and it will turn blue and hedge (bet to break even) your bet on that combination. For the first time, exacta betting becomes a science, not a guess.

You wheel and back-wheel quite simply. Pick the number you want to key and the bets to all the other horses are displayed. If you want to both back and full wheel a horse, you need to do two separate calculations. Don’t forget to adjust your total bet size. If you choose to try the “Pitz” betting, where you use an exacta bet instead of a win bet, the Wheel command is perfect. Since this allows you to play many races with low priced favorites, this feature alone is worth the price of admission. And don’t forget the Back-Wheel instead of a place bet for those of you who bet win and place. Many times it is far more profitable to bet a horse to win and then back-wheel it in the exactas. More work, but also more money.

A new feature just added at my suggestion is the ability to part-wheel. I suggested it Tuesday night and had the new feature in place on Wednesday afternoon. Did I mention that these guys offer very good customer service? Well, they do. Anyway, click on the square labeled Wheel and the Wheel and Back-Wheel rows become Top and Bottom for an exacta part-wheel. You click on the horses you want to win, and then again on the horses you want to place. It is intelligent enough to allow the same horse to win and place without getting confused, and the dutch is perfect. If you want, say, the 1, 2 and 3 horses to win and the 1, 2, 3, 4, and 5 horses to place, just enter them in exactly that way and the program will take care of everything. This was not an easy thing to add, and I’m impressed by the speed and accuracy of the upgrade.

If the race offers them, you also have click buttons for Daily Double and Quinella bets. They work just the same as the exacta screens and I’ll leave you to figure them out on your own. The part-wheel option was just added for Doubles as well. The last click button is Options, where you can set up the program to suit yourself. For those who like database research, you can save the data you have collected in a comma-delimited file at the touch of a key. That way you would have a record of the minute-by-minute changes in the pools, together with exacta, double and quinella information. For the database keeper who believes the answers are writ large in the toteboard lights, here is the answer to your prayers.

It is traditional in a review of any kind for the reviewer to find some fault with the book or computer program under scrutiny just to demonstrate he is a virtuous and righteous reviewer and not just a shill. OK, so here goes. I do have a couple of tiny faults I have found. I wish that on several of the screens I had the option of displaying odds instead of the payoff. They are both the same thing, I know, but I’ve been looking at odds for a very long time indeed and I’m quite familiar with them. Small thing, but there you are. I also wanted to complain about the inability of the program to dutch a part-wheel exacta, but Gary fixed it before I could get this into print.

So now I’m left with only the lame complaint that it is hard to bet when you have so many combinations, as when you dutch you of course get a different amount for each combination. That’s what dutching is all about. Three over five is 12 combinations. Add that to your win bets and you better start putting in bets early. Try to do it for three tracks all going off within 4 minutes of each other and it can be chaos. Now this isn’t the program’s fault, it’s just a fact of life, but you do need to plan ahead. Now if the betting services would just allow us to click a button and place all our bets electronically . . . Oh well, I can dream. That service isn’t available outside of North Dakota, and only for a very select few (like one, the famous Dakota Whale).

Well, I guess that about covers it. I didn’t mean to write a mini-manual here, but I do so very much like this program that I want it to succeed and I thought that a detailed description of what’s going on here might get people excited. I know everyone betting on any kind of serious level NEEDS this program. I just hope that after reading this, they WANT this program. It’s easy to get, you download it from their web site. Of course, it is very expensive. Free for 30 days, then the outrageous sum of $25 for using it through the end of 2003! If you order before September, they guarantee that the program fee will never be more than $25 a year. This has got to be the bargain of this new century.

As I said before several times, I have found both authors to be very helpful and responsive to any questions or suggestions. Several anonymous friends report the same experience. I know it’s not “sexy” like a new method for pickin’ horses, but if you’re serious about making money at the track you owe it to yourself to give this a try. The first taste is free, and after you’ve tried it, you’ll beg to be allowed to buy it. Here is their address for downloads and up-to-date information:

http://www.homebased2.com/atr/at_the_races.htm

To contact the authors:

For technical support contact Gary at: garypitz@hotmail.com

For questions on how to use AT THE RACES contact Jack at: amazing3@hotmail.com

Jake
07-04-2002, 03:55 PM
Dick,

For a complicated program, that's an amazing review. The clarity of your explanations for the different screen functions and program features is outstanding. Hope they hire you to do the official documentation for the program! Thanks for your efforts.

Rpd
07-04-2002, 06:26 PM
Dick,

This was a very thoughtful and complete review. Thanks for sharing it with us.

Brought back memories of what I miss from the 'Follow-Up'

MikeDee
07-05-2002, 09:10 AM
I'ld like to add my thanks as well Dick, very nice of you to provide the information. The authors of the program should hire you to write their documenation for them.

rrbauer
07-08-2002, 04:46 PM
Review of At The Races (ATR) Part 2 - Functionality.

This is a nice programming job by Michael Pizzagati. The product documentation however is weak to non-existent. A playing “strategies” section with some examples would be helpful. I had some problems with the original download. Michael helped me initially and then turned me over to “Rick” who handles the web-site hosting for ATR. Rick was able to help me around the problem. (The problem was a missing Active-X plug-in for my IE5 browser.)

Since Dick Schmidt’s review addressed much of the “how to” of ATR, I looked more at actual usage in a real-time betting development. I concentrated my use of ATR on two tracks: Belmont and Hollywood. These tracks have big pools that may be less affected by the end-game volatility mentioned in Part 1. I will mention the capture-time interval control. Since the simulcast network data is updated on 60-second cycles, I’m not sure what is gained by being able to control the interval in one-second increments. I have been setting it for 45-second intervals. Also, I’m glad that Schmidt pointed out what the Internet “R” and “S” tote board symbols mean, because it isn’t mentioned anywhere else that I could find.

The WPS screen is a good tool in that it provides not only the payoff price for the win, but also the range of possible payoff prices for place and show. It would be better if the prices were displayed without applying the breakage formula. This would allow a player to track the prices towards post-time with an eye towards the breakpoint. If the amount displayed is before breakage, then place-price displays’ of 4.13, 4.07 and 4.02, for example, on the three (3) captures just prior to post-time, would alert you to not only the price trend, but that the price was poised to break through the 4.00 breakpoint. When displays’ are 4.00, 4.00 and 4.00, you lose that information.

At the same time, the breakage formula is not applied to the prices in the Pitz Exacta and Double columns. Consistency would be nice. Also, I noticed that the breakage formula being used for Belmont is not correct for payoffs of $10 and higher.

The win dutch is easy to accomplish with a couple clicks. I would prefer that once the dutch bet is entered that there was a display of the composite odds that you would receive on the bet. The profit for the individual bets is shown, but the profit is a function of both the amount bet and the odds. If I know what odds I’m getting overall on the dutch bet then I can figure out my profit pretty easily. Having the odds also allows you to make comparisons of different betting opportunities.

Enabling the option to “save” .csv files automatically over the last 2 minutes of betting will give you some data that can be imported into a spreadsheet for later review. This data will also give you some idea of the volatility that I’ve mentioned. For example, here is a race at Belmont and the WPS pool data as it is updated in the last minute of betting and after the race is off.

MTP.................Win.....................Place. ................Show
1................$ 92,388.............. $30,560............. $14,427
0.................116,607.................37,053.. ..............14,427
0.................142,416.................45,480.. ..............21,461
Off...............171,994 ............... 52,288............... 25,121
Off...............221,075................ 65,991................25,121


And, during these update cycles the win price on the already low-priced favorite dropped from $3.40 to $3.30. (At a track using dime-breakage that drop would’ve been to $3.20.)

Even with the data volatility, for the purposes of win-bet dutching, I think that ATR is useful if you’re willing to accept that your bets may be slightly out of proportion based on the final prices. The other features of the WPS screen, win betting via the exacta and double, have limited opportunities so long as there is a large difference between the straight takeout and the exotic takeout. Also, using the win-bet percents to derive place and show probabilities (Place and Show Outlook %) for evaluating the respective payoff estimates is going to be limited by the end-game volatility.

I had some problems with the Exacta bets. First, there is a row in the bet selection area on the lower left that contains percentages. I don’t have a clue what those mean. But, the problem occurs when you click the low-2 box. It lists 6 bets, as if it were a 3-horse box.
I dunno? If the two low exactas are 7-1 and 7-5, then I would expect the low-2 box to be those two combinations and the 1/5 over the 7. Or, four combinations in total. ATR also boxes the 1 with 5. Schmidt mentioned having the part-wheel feature added and that seemed to work OK. (BTW. To modify an algorithm to process a list selectively instead of completely, requires one “If” statement! No big deal.) It would be good to be able to dutch mixed-bet types. For example play a horse to win and then bet it under in exactas from some other horses. And, dutch the whole enchilada.

The Exacta Stats screen analyzes the exacta probable payoffs vis-à-vis the implicit win odds contained in the win pool (i.e. If a horse has 20% of the money bet in the win pool then it is assumed to have a 20% chance of winning.) Using the same probabilities along with the condition that the horse didn’t win allows you to calculate a theoretical probability that the horse will run 2nd. Once you have those two probabilities for each horse you can calculate theoretical exacta payoffs. I’ve only ever seen this analysis in one other program, Super Dutch, and if you track a few races you will see that the “+ value”, often lies in the low-paying combinations; and, you will see the impact from the volatile payoff data as the close of betting nears. Referring to the .csv files and analyzing the exacta stats for specific tracks may give you some insights as to how to use that data best. I think that its use will likely be “track specific”.

The double and quinella functions are similar to the exacta functions, minus the “stats” screen. I don’t routinely use those bets so I leave the evaluation to those that do.

If you’re into dutching, then for sure you should take the free test drive of this program. Once you do that you will have a good feel for how valuable the program and its approach is to your playing style. The author refers to requiring updates for the takeout percentages as a reason to “lease” the program. I think that the takeout percentages should be in a file and not embedded in the program. Relying on the “free” tote-board data from either the Racing Channel or BRIS may pose some risk in the future if they choose to modify the way their data is deployed. But for now, if it works, don’t fix it!

superfecta
07-10-2002, 12:33 AM
have they ever tried to use the info they get (odds,probable xacta payoffs,pool size and avg. payoffs) to "project" what tris and supers would pay?If they know about what the exacta should pay,and do it quickly enough,it might be worthwhile to give possible horses for the third and fourth spot.But it would also need to factor if there is enough money to be won on tris and supers.Factors such as field size,odds on remaining horses ,pool amounts and average payoffs would be taken into consideration to "project" a profitable bet.
The handicapper (depending on his picks)could then decide if hammering the exacta or spreading on a tri/super would be more efficient use of his money.The speed factor is probably the key,being up to date on payoffs(as close as possible anyway)so it would save the bettor time trying to form a betting "matrix".Having to guess how much to bet on each exotic at the last minute and risk being terribly wrong.Just a thought.

Dick Schmidt
07-10-2002, 02:13 AM
Super,

Though they have issued 4 updates since I wrote my review (correcting every nit I picked), they have not added anything about tri's or superfectas. To do this, they would need to have access to the pools and amounts bet on each horse in each position. This would be really cool stuff if you could get in all on the screen, but as far as I know, no one outside of North Dakota has access to this information. And then you have the problem of betting. What upset so many people was that the guy in ND could bet over 100 tri's with just 4 or 5 seconds to go. Without that direct connection, this would be difficult to implement.

Dick

rrbauer
07-10-2002, 01:02 PM
Super, Dick S.

It's possible to compute a theoretical trifecta payoff using the pct of money on each horse in the win pool as a win probability estimate as your starting point (similar to what ATR does for the exacta stats screen) and then develop the conditional probability estimates for running 2nd and 3rd. Once you have each horse's probability estimate for running 1st, 2nd and 3rd you can compute the Tri payoff estimate.

Having said all of that, let me add that the resultant payoff price using this approach is flawed for Trifectas and I'm not sure why that is. (Exactas via this approach are much closer to reality.) I found this out when I was trying to come up with a valuation method for some software that builds plays for Tri bets. Adding the M/L odds into the mix gave me payoff estimates that were closer to reality. When I run my estimates against actual payoffs the correlation comes in at about .85.

One of the gems that came out of my Tri payoff research was how to use a hard "bet down" in the Tri. (And I make the use decision with 1 MTP.) My rule is that if the horse is bet down to less than 1/2 of its ML odds (adjusting for scratches, etc., if necessary), for example, ML 15, Current 6, then I use it on top only. These are huge overlays in the Trifecta pool and I can only conclude that the reason why is that most players will use the bet down in the under slots, and as an after thought, but won't put it on top.

superfecta
07-10-2002, 10:48 PM
Originally posted by Dick Schmidt
Super,

Though they have issued 4 updates since I wrote my review (correcting every nit I picked), they have not added anything about tri's or superfectas. To do this, they would need to have access to the pools and amounts bet on each horse in each position. This would be really cool stuff if you could get in all on the screen, but as far as I know, no one outside of North Dakota has access to this information. And then you have the problem of betting. What upset so many people was that the guy in ND could bet over 100 tri's with just 4 or 5 seconds to go. Without that direct connection, this would be difficult to implement.

Dick So maybe the "whale" does have a similar program?Maybe...interesting food for thought.Thanks Dick for the reply.

superfecta
07-10-2002, 11:01 PM
Originally posted by rrbauer
Super, Dick S.

It's possible to compute a theoretical trifecta payoff using the pct of money on each horse in the win pool as a win probability estimate as your starting point (similar to what ATR does for the exacta stats screen) and then develop the conditional probability estimates for running 2nd and 3rd. Once you have each horse's probability estimate for running 1st, 2nd and 3rd you can compute the Tri payoff estimate.

Having said all of that, let me add that the resultant payoff price using this approach is flawed for Trifectas and I'm not sure why that is. (Exactas via this approach are much closer to reality.) I found this out when I was trying to come up with a valuation method for some software that builds plays for Tri bets. Adding the M/L odds into the mix gave me payoff estimates that were closer to reality. When I run my estimates against actual payoffs the correlation comes in at about .85.

One of the gems that came out of my Tri payoff research was how to use a hard "bet down" in the Tri. (And I make the use decision with 1 MTP.) My rule is that if the horse is bet down to less than 1/2 of its ML odds (adjusting for scratches, etc., if necessary), for example, ML 15, Current 6, then I use it on top only. These are huge overlays in the Trifecta pool and I can only conclude that the reason why is that most players will use the bet down in the under slots, and as an after thought, but won't put it on top. Might the higher "take" account for the 15% difference?Since the straight pools are around 17%,and exotics are 25-30%,maybe a fudge factor has to be added to the "projected" payoffs for the tri and super,since we don't know how many combos are bet?Good angle on the bet down ML ,rrbauer.Makes good sense.
I'm sure another factor has to be knowing whether the exotic bettor has a clue,or is just shooting for high payoffs at the given track.Seems like I see low paying tris and supers more on the east coast than in the midwest or west coast tracks.When compared to like odds horses of course.

rrbauer
07-11-2002, 12:18 AM
Super

Nah. The takeout is reflected in the exotic payoff calculation whatever it may be. And, the win pool takeout isn't a factor when determining the probability estimate because the money pools reflect the amount bet before takeout.

Not sure what your 15 percent reference is to.

I'm guessing that some of the difference between the theoretical payoff price and the actual payoff price reflects the randomness with the way some players throw everything but the kitchen sink into the 3rd slot of their trifecta bets. In other words, there is less handicapping involved when filling the 3rd slot of a trifecta than there is when filling the first two slots.

superfecta
07-11-2002, 09:59 PM
Originally posted by rrbauer
Super

Nah. The takeout is reflected in the exotic payoff calculation whatever it may be. And, the win pool takeout isn't a factor when determining the probability estimate because the money pools reflect the amount bet before takeout.

Not sure what your 15 percent reference is to.

What I mean is if you are using the actual odds of the horses to project payoffs,maybe adjusting these odds by the difference in takeout would get your projected tris more in line with what they actually pay.(example-figuring a horse with odds of 10/1,should be more like 8.5/1 in the tri because of the difference in take)A diffence of 15% in odds. You may be close enough now ,there will always be a laspe between the time you bet and the pools close.

rrbauer
07-12-2002, 11:02 AM
Super:
Not using "actual odds". Using the percent of money on each horse in the win pool. Those percents sum to 100%.

The Trifecta takeout percent is applied in the course of computing the estimated payoff.

Dick Schmidt
08-12-2002, 05:25 PM
I'm gonna "bumb" this half, too.

Dick