Yonkers Morning Lines

[mrroyboy]

That is a very good article Sea. Thanks for posting it.

[zico20]

Originally Posted by pandy

I saw the statistics, and the TrackMaster/USTA computerized line was clearly, and, quite frankly, remarkably more consistent at predicting the actual final odds at almost every track that runs a regular harness racing meet. Cal Expo and the Meadowlands were the ONLY tracks that race a regular meet that beat the computer.

Yes, there will be some bad lines on horses, probably due to information that the computer can't analyze, but the bottom line is that the computerized line was consistently more accurate than the live line makers.

http://handicapping.ustrotting.com/pandycapping.cfm

Don't be fooled that the tracks don't take the ML seriously. They do. The tracks goal is to maximize the amount of money bet. And they do this by manipulating the ML. I should actually say by purposely being dishonest.

At Pocono on Wednesday, not one ML horse was under 5-2 yet 8 horses went off at 6-5 or under, including a 1-9 and a 1-5 shot.

As far as the computer generated odds for the ML goes, this is an absolute joke as well. Are you trying to tell me that at Yonkers the computer hasn't identified one horse this past month that was going to go off at 50-1 or higher. The other day three horses were over 99-1 in one race. Yet the ML had them no higher than 20-1. And if the computer had them at 99-1 and listed them at 20-1 then it throws off all the other odds for that race.

The only way to get an accurate ML is to give an experienced and knowledgeable handicapper free reign to make whatever odds he so desires. If that means 1-9,10-1, 15-1, 40-1, and 5 horses listed at 99-1 then so be it.

If the computer is restricted to 20-1 at the high end, then what is the point of using a computer based ML. Like I said at the start, the ML is all about bringing in the most money, period.

[pandy]

I can't argue with that, but the point is, the computer line is proven statistically to be more accurate.

This has been discussed on this forum before, yes, tracks don't like to make horses 1-9 because they're afraid no one will bet the race. On my Meadowlands line, I do make horses 1-9, but I don't work for the track.

[zico20]

Originally Posted by pandy

I can't argue with that, but the point is, the computer line is proven statistically to be more accurate.

This has been discussed on this forum before, yes, tracks don't like to make horses 1-9 because they're afraid no one will bet the race. On my Meadowlands line, I do make horses 1-9, but I don't work for the track.

With the way the current rules are set up, yes the computer model is proven to be better. Not going to argue that. But we will never know which way is the most accurate until all restrictions on a race is lifted. And that is never going to happen. Sorry if I came across a little harsh, that was not my intention. I just don't like the dishonesty that goes with the ML. I enjoy reading your posts, like so many others on here.

[pandy]

Thank you.

[The Inside Scoop]

The latest faviurite number for the morning lines is 9-1. In my many years around this business, I have never saw a horse listed at 9-1 morning line, yet now that R2D2 has taken over, there are horses morning lines listed at 9-1 everywhere.

Heres yet another example of a rediculous computer generated morning line Sunday in race 8 at Yonkers. Artomatic Pilot is listed at 20-1 from the 8 hole with Daniel Tuccillo Jr. up. I would venture to say that I have more chance of living on Mars some day than this horse has at going off at 20-1 or less.

[David Siegel]

All:

As I have mentioned, this is the sole forum I post on as most of the questions and criticisms are reasonable. I do ask though that if you have questions about something, ask. It is very rare I do not address a legitimate issue. Hopefully, this post can clear up a few things.

1) A previous comment was made about things that the computer does not consider. It said "A computer generated system fails to take into effect trainer changes, driver changes, troubled trips, etc etc." This is partially incorrect and partially correct. Trainer changes and driver changes ARE directly reflected in the automated morning line. Troubled trips identified as an individual race are not reflected however for the most part, they are ignored. For example, use Speed Ratings are part of a few of about 45-50 variables that are used in the creation of the line. However, one of the many Speed Rating variables, say the average of the best 5 of the last 10 within 90 days, would likely exclude the troubled line, so indirectly, they are taken into account.

2) About certain "bad" lines being too low. I could not agree more that this happens. We have given the tracks the option to put a ceiling and floor on lines, as long as they are not totally ridiculous. Many have opted for a 15/1, 20/1 or 25/1 cap. The floors actually rarely come into play. Every day we have horses where the "fully" automated line would put a horse at 40/1 and it gets published at 20/1. We don't just move that horse, but redistribute odds pro-rata so they are always balanced. Personally, I don't like this as it makes the ML less accurate and I happen to disagree with the track's philosophy that justifies these caps. That said, this is a service to the tracks and we have to be sensitive to their real world needs whether I agree with them or not.

3) About finding bad lines (not high ones). There is zero doubt you will find them, lots of them. Its the nature of the beast. The computer cannot be perfect. It will just be closer to perfect than nearly any human and we have the data, over 1 million starts, to prove it. Anyone can find outliers, human or computer, especially after the public weighs in and off-odds results. There are certain situations where the computer could NOT reflect something in the algorithm and humans could. So I have no doubt that in certain races a human could do a better job. For example, if everyone on the track knows a horse fell during training and had a "big knee" but was still racing, the computer could never reflect that in the ML. But that said, we have proven again and again that on average, the computer does A LOT BETTER than humans. So the folks here that pushed for someone to show mass data collections to prove the computer is lousy is singing the right song.

The logic of finding some bad races is the same to me as finding a streak where say Derek Jeter was 0-20 and had 2 errors in that 5 game stretch. I guarantee you can find this. To conclude Jeter was an awful play based on this sample size would be construed by any fan of the game to be short-sided. Similarly, a rookie that goes 3-4 with 1 homer on his first day in the big leagues would create a stir, but no one would be ready to induct him into the hall of fame even though his batting average was .750 and slugging percentage well over 1.000.

4) Finally, tracks have the ability to override the lines. This has been done rarely but has been done and when measured (I think I know of 10 adjustments) they have done more harm then good. There was a race over the summer at Vernon, the Zweig where Father Patrick and Nuncio were coupled and the track assigned odds of 5/2. At the time they were not taking the automated line which had them at 1/1 and I don't recall if there was a floor or not. They went off a .05 (1/20). These human "errors" can be found all of time as well, but as I said, you will find more variance between the humans and the computer when measured against the off-odds.

5) Lastly, while rare, humans make another type of error in lines - typos. I was so disappointed to see a horse on Jug day at the Jug with a 56/2 odds on the program. This is simply and embarrassment to the sport. Obviously that would not happen with the automated line, and of course, there are other ways to use a computer to prevent such typos.

Hope that helps clarify some points about the automated morning line. On balance, I have zero doubt that the benefits of a more accurate line, on average, is beneficial to the industry, even leaving out the other clear benefits of time, money and no chance of monkey business.

[mrroyboy]

Ok guys.
I'll put my foot into my mouth. I will try to do a revised morning line for Meadowlands when it opens Friday. As soon as I get the pp's I will ignore that ML and do my own. I will post it long before the first race. If there are scratches later I may revise it.
I of course will have free rein whatever good that is. Let's see how close I can come.
I will do it until it gets embarrassing.

[LottaKash]

Originally Posted by David Siegel

All:

As I have mentioned, this is the sole forum I post on as most of the questions and criticisms are reasonable. I do ask though that if you have questions about something, ask. It is very rare I do not address a legitimate issue. Hopefully, this post can clear up a few things.


Hope that helps clarify some points about the automated morning line. On balance, I have zero doubt that the benefits of a more accurate line, on average, is beneficial to the industry, even leaving out the other clear benefits of time, money and no chance of monkey business.

Thanks Dave for the further explanation of things...

As for the new ML's I like them.....Too bad some tracks have this "floor & ceiling" thing....It hurts the line more than the other way around...

As for horses going off much lower than their ML's, I have been seeing this for years, and especially at Yonkers...When I first started, way back then, andeven up to today, the tracks had this 5/2, 3/1 ML-fave mentality....I can remember thousands of these 5/2ml horses going off at Odds-On....I just don't know what the tracks seem to be afraid of....Especially So, when the bulk of the players now are at least fairly savvy enough to know who will be 1/9 or 5...So why all the mystery I ask, and, have always asked?...

As for the lesser tracks, thru the years and often enough, I would just crack up witnessing what is supposed to be a MOrning LIne....Thru the years I have been to so many tracks, and now thru the magic of the Internet, I have witnessed and am acquainted with so many more tracks than would be possible before the nets invention, and yet I still se some ML's that just plain stink and are laughable...They sure could use some help and guidance in that regard...

So, me, I like the new look & feel of Trackmaster's ML's, and I applaud the tracks that now take those lines...I wish some tracks would just leave them alone tho, as just like Junkies on some dope, they just can't seem to let go of old crappy ML making, and they must seem to think that they still need some adjustments to aid in the betting...Which they don't, imo...I say card more competitive races, and the odds will take care of themselves...

The New Trackmaster ML's, even if not totally accurate to a fine degree, are still way better, and the projected odds are much more believable and very relative on a scale to each other, imo at least...

P.S.

DAVE could you please repost the Harness Tracks that are now using your new an improved ML's...

Thx ahead..

[mrroyboy]

Welcome back John
How about some posts on Pace Handicapping?

[David Siegel]

I am happy to oblige. First some thoughts about the last couple of posts.

I agree with your position on ceilings and floors, and think when a 15/1 goes off at 50/1, it makes the track look silly. Tracks have also explained to me that owners and trainers don't like to see their horses at 50/1. Their "solution" of a 15/1 morning line to me is a short-term mask that ultimately hurts in the long term. A 50/1 off-odds horse points to a horse that most likely is in the wrong class. That could be the fault of the trainer, the owner, or just a horse population issue. But "masking" it, very temporarily, with the morning line is not the best answer.

Harness fans, horsemen and management are amongst the slowest and most resistant to adopting new technologies and ideas. In fact, it is somewhat ironic to hear such folks say something like, "yeah, things are not very good and we need some real increases in handle to make things better. And yeah, we need change to make things better for sure. Just make sure not to change anything that affects me".

What can I say. Some tracks have been open to the morning line and other improvements and over time, new things will be adopted. It just takes a lot of time in this industry.

The tracks that are currently using the line (or that have closed for the season and used the lines are as follows):

Dover
Saratoga
Cal Expo
Plainridge
Rosecroft
Bangor
Freehold
Philly
Yonkers
Aces
Springfield
DuQuoin

Dave

[The Inside Scoop]

Originally Posted by David Siegel

I am happy to oblige. First some thoughts about the last couple of posts.

I agree with your position on ceilings and floors, and think when a 15/1 goes off at 50/1, it makes the track look silly. Tracks have also explained to me that owners and trainers don't like to see their horses at 50/1. Their "solution" of a 15/1 morning line to me is a short-term mask that ultimately hurts in the long term. A 50/1 off-odds horse points to a horse that most likely is in the wrong class. That could be the fault of the trainer, the owner, or just a horse population issue. But "masking" it, very temporarily, with the morning line is not the best answer.

Harness fans, horsemen and management are amongst the slowest and most resistant to adopting new technologies and ideas. In fact, it is somewhat ironic to hear such folks say something like, "yeah, things are not very good and we need some real increases in handle to make things better. And yeah, we need change to make things better for sure. Just make sure not to change anything that affects me".

What can I say. Some tracks have been open to the morning line and other improvements and over time, new things will be adopted. It just takes a lot of time in this industry.

The tracks that are currently using the line (or that have closed for the season and used the lines are as follows):

Dover
Saratoga
Cal Expo
Plainridge
Rosecroft
Bangor
Freehold
Philly
Yonkers
Aces
Springfield
DuQuoin

Dave

What is the cost to a track to have this done for them ?
Does the track pay per card ?
Do they deduct money off the payment when they have to change the odds ?
Is it cheaper for a track to have this system make the morning lines for them than it is to pay a live person to do it ?

[David Siegel]

TrackMaster has been a long standing partner of the USTA and provides a number of service at no charge to member tracks. The automated line is something I felt strongly could be used to assist tracks for a variety of reasons and is provided at no fee to the racetracks.

[The Inside Scoop]

So it is done for free. Not surprising.

[LottaKash]

Originally Posted by David Siegel

TrackMaster has been a long standing partner of the USTA and provides a number of service at no charge to member tracks. The automated line is something I felt strongly could be used to assist tracks for a variety of reasons and is provided at no fee to the racetracks.

Thx for the update Dave...Keep on.. The lines are not 10's, as I don't believe that any line could be, but a solid 8+, and a noticeable improvement over the replaced lines, imo...