The Meadows

[traynor]

I model races by category, with a very specific delineation between "select" and "non-select" races. The latter category culls all races for NW1, NW2, and what are typically referred to as "rag races"--NW of $X in Last X, and so on.

The "select" races are NW3, and races open to winners--standard claiming fare, with open, preferred, etc. included.

The current model for trainers in the select category of trot racea at The Meadows:
Mea - In 165 Trot Races WIN% for TOP RANK was 32.12 % Mean Mutuel 7.59
Mea - In 165 Trot Races PLACE% for TOP RANK was 6.06 %
Mea - In 165 Trot Races WIN% for SECOND RANK was 9.09 % Mean Mutuel 6.11
Mea - In 165 Trot Races WIN ROI for TOP RANK was 1.22
Mea - In 165 Trot Races WIN ROI for SECOND RANK was .30
Mea - In 165 Trot Races TOP THREE Trainers WON 47.27 % ROI 0.59
Mea - In 165 Trot Races AB Exacta ROI was 3.19 13.33 %
Mea - In 165 Trot Races AC Exacta ROI was .43 2.42 %

This is one of the very few circumstances in which the "top 2" provide a generous return--the top trainer to win, the second-best to place. In this type of race at The Meadows, it is not especially useful to agonize over whether Horse A has a pace advantage over Horse B, or whether Horse C is higher class than Horses A and B. Ignore the horses. Look at the trainers.

Exacta figures are corrected for outliers. Actual return is higher.

[LottaKash]

Traynor, this post caused my eyes to "bulge"....:ThmbUp: :ThmbUp: :ThmbUp:

[RaceTrackDaddy]

I model races by category, with a very specific delineation between "select" and "non-select" races. The latter category culls all races for NW1, NW2, and what are typically referred to as "rag races"--NW of $X in Last X, and so on.

The "select" races are NW3, and races open to winners--standard claiming fare, with open, preferred, etc. included.

The current model for trainers in the select category of trot racea at The Meadows:
Mea - In 165 Trot Races WIN% for TOP RANK was 32.12 % Mean Mutuel 7.59
Mea - In 165 Trot Races PLACE% for TOP RANK was 6.06 %
Mea - In 165 Trot Races WIN% for SECOND RANK was 9.09 % Mean Mutuel 6.11
Mea - In 165 Trot Races WIN ROI for TOP RANK was 1.22
Mea - In 165 Trot Races WIN ROI for SECOND RANK was .30
Mea - In 165 Trot Races TOP THREE Trainers WON 47.27 % ROI 0.59
Mea - In 165 Trot Races AB Exacta ROI was 3.19 13.33 %
Mea - In 165 Trot Races AC Exacta ROI was .43 2.42 %

This is one of the very few circumstances in which the "top 2" provide a generous return--the top trainer to win, the second-best to place. In this type of race at The Meadows, it is not especially useful to agonize over whether Horse A has a pace advantage over Horse B, or whether Horse C is higher class than Horses A and B. Ignore the horses. Look at the trainers.

Exacta figures are corrected for outliers. Actual return is higher.

You got me really interested Traynor . Let me ask you if my understanding about the AB exacta is correct or not.

If I bet the Top trainer in the race I am betting, over the second best trainer in that trotting race at the Meadows 100 times. I will win 13.33% of my wagers and lose 87.64% of the time but I would end up more than three times my initial investment? If that is what I am seeing, that is quite an impressive statistic and will become part of my arsenal.

[traynor]

You got me really interested Traynor . Let me ask you if my understanding about the AB exacta is correct or not.

If I bet the Top trainer in the race I am betting, over the second best trainer in that trotting race at the Meadows 100 times. I will win 13.33% of my wagers and lose 87.64% of the time but I would end up more than three times my initial investment? If that is what I am seeing, that is quite an impressive statistic and will become part of my arsenal.

Exactly, with the obvious caveat that "past performance is no guarantee of future results." Understand that the values are for "select" races as defined above, not "all" races.

To clarify, I make three separate models for every "strategy." Select races, non-select races, and "all" races. The reasons are highly pragmatic (defined as $$$$), not theoretical. I have found those designations to be much more useful than "class" levels, purses, or any number of other ways to categorize race conditions.

[RaceTrackDaddy]

Exactly, with the obvious caveat that "past performance is no guarantee of future results." Understand that the values are for "select" races as defined above, not "all" races.

To clarify, I make three separate models for every "strategy." Select races, non-select races, and "all" races. The reasons are highly pragmatic (defined as $$$$), not theoretical. I have found those designations to be much more useful than "class" levels, purses, or any number of other ways to categorize race conditions.
Thank you Traynor. Just off hand, I have recalled that most betting strategies have worked much better for the lower classified horse than the claiming or higher end races like the open and invitational. Most of my personal success wagering (no matter the breed of animal) has always been with the maidens and Nw Other Than conditions. Would you say that is similar in your studies on wagering statistics of winners?

[traynor]

Thank you Traynor. Just off hand, I have recalled that most betting strategies have worked much better for the lower classified horse than the claiming or higher end races like the open and invitational. Most of my personal success wagering (no matter the breed of animal) has always been with the maidens and Nw Other Than conditions. Would you say that is similar in your studies on wagering statistics of winners?

"Similar" is a loaded term. What I have found is that specific strategies work most effectively at different categories of races. A strategy that works well on maiden and NW races will (generally) flop when applied to straight claimers and "higher end" races. That is one of the reasons why so much research comes up short--it tries to lump everything into some metacategory or other, the analysis of which generates little more than semantic noise.

I use completely different strategies (for example) for high end claimers at Yonkers than I use for NW of X$ in Last 5 at Northfield. It took awhile to code an app "smart" enough to make the distinctions, but the effort is worthwhile.

[RaceTrackDaddy]

"Similar" is a loaded term. What I have found is that specific strategies work most effectively at different categories of races. A strategy that works well on maiden and NW races will (generally) flop when applied to straight claimers and "higher end" races. That is one of the reasons why so much research comes up short--it tries to lump everything into some metacategory or other, the analysis of which generates little more than semantic noise.

I use completely different strategies (for example) for high end claimers at Yonkers than I use for NW of X$ in Last 5 at Northfield. It took awhile to code an app "smart" enough to make the distinctions, but the effort is worthwhile.
Thanks again Traynor, this old dog still is willingly to learn new tricks. I definitely will looking at races at the Meadows in a different perspective now in the trotting ranks. Boy I love the data, it is the implementation into a wagering plan that drives me nuts.

[traynor]

Thanks again Traynor, this old dog still is willingly to learn new tricks. I definitely will looking at races at the Meadows in a different perspective now in the trotting ranks. Boy I love the data, it is the implementation into a wagering plan that drives me nuts.

One detail I omitted above (or failed to adequately clarify). NW races for lifetime earnings are "select" if the amount in the condition is at least four times the amount of the purse tonight. Meaning a NW of 20K LT with a 10K purse is NOT, but the same race with a 5K purse would be.

[RaceTrackDaddy]

One detail I omitted above (or failed to adequately clarify). NW races for lifetime earnings are "select" if the amount in the condition is at least four times the amount of the purse tonight. Meaning a NW of 20K LT with a 10K purse is NOT, but the same race with a 5K purse would be.

I can't see the causation for the results you discovered as naive as I am about certain things I would expect the higher purses to be more instead of less select. Did it strike you as strange or just an anomaly? Did it hold true before the slot money came in bumping up the purses in 2006?

This is becoming a real interesting thread.

[traynor]

I can't see the causation for the results you discovered as naive as I am about certain things I would expect the higher purses to be more instead of less select. Did it strike you as strange or just an anomaly? Did it hold true before the slot money came in bumping up the purses in 2006?

This is becoming a real interesting thread.

What I do is a reflection of current reality. Simply stated, the processes and procedures that worked (some quite well) five years ago are currently (at best) marginal. Most of those complaining of poor returns are trying to use old ways (essentially obsolete processes and procedures) to handicap new races.

For comparison purposes, I just ran "all" races for the same segment as above. The exacta returns are still nice, but the frequency dropped considerably.

Mea - In 416 Trot Races WIN% for TOP RANK was 25.72 % Mean Mutuel 5.88
Mea - In 416 Trot Races PLACE% for TOP RANK was 13.70 %
Mea - In 416 Trot Races WIN% for SECOND RANK was 14.66 % Mean Mutuel 5.24
Mea - In 416 Trot Races WIN ROI for TOP RANK was 0.86
Mea - In 416 Trot Races WIN ROI for SECOND RANK was 0.43
Mea - In 416 Trot Races TOP THREE Trainers WON 50.48 % ROI 0.55
Mea - In 416 Trot Races AB Exacta ROI was 3.16 7.69 %
Mea - In 416 Trot Races AC Exacta ROI was 1.45 3.37 %

[traynor]

I can't see the causation for the results you discovered as naive as I am about certain things I would expect the higher purses to be more instead of less select. Did it strike you as strange or just an anomaly? Did it hold true before the slot money came in bumping up the purses in 2006?

This is becoming a real interesting thread.

I spent (and spend) a LOT of time analyzing races by purse values--to the extent of parsing purse ranges at each track individually, to see if certain factors were (are) more predictive at specific odds ranges. I assumed (incorrectly) that higher grade races would be more predictable using current form, etc. That is definitely not the case.

[traynor]

And as long as we are on the topic of trainers in select races at The Meadows, I may as well post the (current) stats for pace races:

Mea - In 187 Pace Races WIN% for TOP RANK was 34.76 % Mean Mutuel 4.15
Mea - In 187 Pace Races PLACE% for TOP RANK was 19.79 %
Mea - In 187 Pace Races WIN% for SECOND RANK was 12.30 % Mean Mutuel 6.98
Mea - In 187 Pace Races WIN ROI for TOP RANK was 0.91
Mea - In 187 Pace Races WIN ROI for SECOND RANK was 0.43
Mea - In 187 Pace Races TOP THREE Trainers WON 58.29 % ROI 0.66
Mea - In 187 Pace Races AB Exacta ROI was 2.51 13.37 %
Mea - In 187 Pace Races AC Exacta ROI was 0.70 1.60 %

Again, this model is restricted to the select races as defined above.

[traynor]

Again, for comparison purposes, these stats are for "all" races:

Mea - In 494 Pace Races WIN% for TOP RANK was 28.34 % Mean Mutuel 5.25
Mea - In 494 Pace Races PLACE% for TOP RANK was 20.65 %
Mea - In 494 Pace Races WIN% for SECOND RANK was 13.77 % Mean Mutuel 6.54
Mea - In 494 Pace Races WIN ROI for TOP RANK was 0.89
Mea - In 494 Pace Races WIN ROI for SECOND RANK was 0.48
Mea - In 494 Pace Races TOP THREE Trainers WON 52.43 % ROI 0.62
Mea - In 494 Pace Races AB Exacta ROI was 1.35 7.09 %
Mea - In 494 Pace Races AC Exacta ROI was 1.01 1.82 %

It is reasonable to assume that the "positive ROI" for AB and AC exactas in "all" races is simply the results of the "select" races diluted by wagers on NW races.

[RaceTrackDaddy]

This is so interesting. You have to keep up with the stats as things can change all the time.

One thing I do believe still remains true over all this time is that the public has a harder time betting on trotters than pacers.

Your stats on the pacers had the same high being the AB Ex on both the trot and pace but the ROI's on the pace are lower.

That just may be why over time I migrated to the trotters. It just felt that they paid better when I won an exotic wager with basically the same odds on the winning horse and like for the remainder of the ticket under part.

Thank you for this thread it is enlightening to say the least. One of the most informative of any thread I have come across. If noting else, I will be adjusting my betting patterns w/respect to purses and try to come up with a classification that works for me to create the select groupings.

[RaceTrackDaddy]

I really, really dislike giving people bum advice. Being bored, J have been watching the trainer choices at The Meadows. In race after race (for the past couple of days) the top trainer and second best trainer have entered some real losers--with odds appropriate to their performance. Not being foolish by nature, when I see a horse that I bet going off at 50-1 or so, it makes me wonder if I might be missing something.

So I re-ran the model, this time ignoring any horse entered by the top and second trainer in any given race that went to post at odds greater than 15/1. With all due respect to those who equate betting longshots with self-confidence, they are not advising other people what horses to consider. Specifically, I culled the long priced entries and rebuilt the model.

Mea - In 178 Trot Races WIN% for TOP RANK was 32.58 % Mean Mutuel 7.60
Mea - In 178 Trot Races PLACE% for TOP RANK was 6.74 %
Mea - In 178 Trot Races WIN% for SECOND RANK was 9.55 % Mean Mutuel 7.22
Mea - In 178 Trot Races WIN ROI for TOP RANK was 1.23
Mea - In 178 Trot Races WIN ROI for SECOND RANK was 0.34
Mea - In 178 Trot Races TOP THREE Trainers WON 48.31 % ROI 0.60
Mea - In 178 Trot Races AB Exacta ROI was 5.45 12.92 %

The win frequency is pretty much the same. However, culling the entries of the top trainer and second trainer that went off at odds of 15/1 or greater dramatically increased the AB exacta return, and also culled out enough losers to show a nice ROI for nothing more complex than betting the entry of the top trainer in select races (defined above) at The Meadows in races in which that entry goes to post at odds LESS than 15/1.

I will let the heroes and windmill jousters swing for the bleachers and bask in glory. I am only interested in the money.

thank you for your diligence.. you have opened my eyes and will be using all this valuable information

[traynor]

thank you for your diligence.. you have opened my eyes and will be using all this valuable information

Don't bet on it yet--I am still working on it!

[RaceTrackDaddy]

I will hold my fire....lol

[traynor]

My (fortunately brief) moment of hysteria was in vain. Corrected values below:

Mea - In 178 Trot Races WIN% for TOP RANK was 32.58 % Mean Mutuel 7.60
Mea - In 178 Trot Races PLACE% for TOP RANK was 6.74 %
Mea - In 178 Trot Races WIN% for SECOND RANK was 9.55 % Mean Mutuel 7.22
Mea - In 178 Trot Races WIN ROI for TOP RANK was 1.23
Mea - In 178 Trot Races WIN ROI for SECOND RANK was 0.34
Mea - In 178 Trot Races TOP THREE Trainers WON 48.31 % ROI 0.60
Mea - In 178 Trot Races AB Exacta ROI was 5.45 12.92 %

When I changed the test algorithm, I only changed it for pace races. So I had to re-run it for trots.

[grant miller]

ron burcke has had the hot hand in p-burg (medows) bet aginst him & palone?

[traynor]

Mea_May16_R09_EXACTA #7 IndoorVoices ==> #4 YoureCatchingOn_Trainer EXACTA 164.80

YoureCatchingOn paid 27.40 to place. Off at 55/1.

[RaceTrackDaddy]

Mea_May16_R09_EXACTA #7 IndoorVoices ==> #4 YoureCatchingOn_Trainer EXACTA 164.80

YoureCatchingOn paid 27.40 to place. Off at 55/1.

Way to go Traynor!

[traynor]

Way to go Traynor!

I hope you see why I quickly deleted the posting (that you so graciously quoted before I could catch it and delete it) about high odds horses in exactas.