How good is the crowd (Odds vs %win)

[podonne]

I wonder what the general concensus is on how good the crowd is at estimating a horse's liklehood of winning a race? (This is not a favorites question :) )

My Odds calculation table says that a horse with 1-1 odds should win 50% of the time. Does that hold up when the public makes a horse 1-1?

Other than the favorite-longshot bias of course, im interested in the odds in between, like a 3-1 or 5-1 horse.

[ELA]

I am not a statistician by any sense. I've hear professionals say the "crowd" is pretty good at setting the odds, and I've heard others say the opposite. I sometimes look at the parimutuel pools as a "marketplace" -- in a sense. Many believe there is an efficient marketplace and others believe there is not. I believe there is ineffciency in the marketplace due to a variety of reasons. Topic for another time.

I don't know what % of the "crowd" picks up the program, form, etc. and goes with the morning line favorite or second/third choice, and what % doesn't. I also don't know if "smart money" is the driving force in the pool.

I will say that when I really like a horse and I am definitive -- the board doesn't get me off the horse. If a horse really like is so called "dead" on the board, I know what I am looking at (the physical side of the game) and I know if the horse is lame, not well, etc. But I would most likely have seen that already.

Be all that as it may, I am not a believer in the "crowd" doing well on second/third/fourth choice type horses (3-1 to 5-1).

Eric

[Overlay]

In Winning at the Races, Quirin studied the winning percentages associated with the various odds levels, based on 19,075 horses running in 2,196 races. He calculated the expected number of winners at each odds level (assuming a 17% take, and dime breakage), and compared it to the actual results. The only statistically significant variations he found were at 1-5, 1-1, 5-2 and 6-1 on the positive side, and 3-1 on the negative side. His general comment was that the 5-2 and 3-1 variances seemed to compensate for each other, and that the other three were sufficiently isolated to suggest that there was no overall pattern of deviation present. So his statistics supported the assertion that, overall, horses win in accordance with the odds that the public assigns them. (Note, however, that when take and breakage are factored in, horses don't win at the same percentage as suggested by the face value of their toteboard odds. For example, 1-1 shots would not be expected to win 50% of the time. Because the toteboard odds reflect the inclusion of take and breakage, horses with 1-1 odds after application of a 17% take and dime breakage would actually have a winning probability of (1.00 - .17)/(1.00 + 1.05), or .83/2.05, or .404878.)

[classhandicapper]

IMO, the public must be very good otherwise there would tons of winners and the very best among them would be earning super high ROIs while being very active.


All the very best players I know are either very active with a relatively low ROI or much more selective with a higher ROI. That suggests that the very active players are probably betting some underlays they should be passing and the very selective players are probably passing some marginally profitable plays they should be making. In either case, that suggests the mistakes don't come that often and aren't that huge when they do exist.

[robert99]

Most of these studies worldwide all say the same thing - that, long term, the public odds are close to expectations, less any take outs. What they do not study or tell you is how good the estimates are for individual races. So people reading the "facts" get misled into thinking they must be accurate for each and every race. The pricing errors +/- do exist in an individual race but just even out over time. Those with better information exploit those individual errors in individual races.

[TravisVOX]

I wonder, for example, if odds/payouts on the early daily double are less likely to be accurate because of the dumb-early tourist-type money. Does that make sense? In other words, the longer into the day you go, the more likely the money you'rep laying against is smart.

[DJofSD]

What they do not study or tell you is how good the estimates are for individual races. So people reading the "facts" get misled into thinking they must be accurate for each and every race. The pricing errors +/- do exist in an individual race but just even out over time. Those with better information exploit those individual errors in individual races.

Here in SoCal there is Today's Racing Digest. One thing they feature is the actual win percentages by race type. Home page for the Digest. (http://www.todaysracingdigest.com/index.php?page=main)

[BMeadow]

From our study of 200,000 races published in the January 2000 edition of Meadow's Racing Monthly:

Favorite averages 1.5 - 1 and wins 32.8%
Second choice averages 2.9 - 1 and wins 20.8%
Third choice averages 4.4 - 1 and wins 15.0%
Fourth choice averages 6.3 - 1 and wins 10.9%
Fifth choice averages 9.0 - 1 and wins 7.8%

Blindly betting any of these returns 0.79 - 0.82 per dollar.

We've also done a number of other studies which show that horses win in proportion to their odds (after adding back the takeout) up through about 20-1, when there begins a slight dropoff. These studies have involved hundreds of thousands of nationwide races.

[kenwoodallpromos]

Odds brackets on a blanket basis do not produce overlays. You have to know:
How to bet (method or angle);
When to bet (when the % of success outweighs the odds);
What to bet (how much).

[robert99]

Here in SoCal there is Today's Racing Digest. One thing they feature is the actual win percentages by race type. Home page for the Digest. (http://www.todaysracingdigest.com/index.php?page=main)

DJ,

In case of any confusion, I mean an individual, single race not a race type.

[robert99]

From our study of 200,000 races published in the January 2000 edition of Meadow's Racing Monthly:

Favorite averages 1.5 - 1 and wins 32.8%
Second choice averages 2.9 - 1 and wins 20.8%
Third choice averages 4.4 - 1 and wins 15.0%
Fourth choice averages 6.3 - 1 and wins 10.9%
Fifth choice averages 9.0 - 1 and wins 7.8%

Blindly betting any of these returns 0.79 - 0.82 per dollar.

We've also done a number of other studies which show that horses win in proportion to their odds (after adding back the takeout) up through about 20-1, when there begins a slight dropoff. These studies have involved hundreds of thousands of nationwide races.

B,

In UK, from 59,112 flat races all types it is:
Favourite, Win 33%
2nd favourite 20%
3rd favourite 13%

Australia is the same.

It is an amazing universal constant in the sense that the amounts and quality of information available to the general betting public is so variable from country to country but the answer comes out almost exactly the same.

[kenwoodallpromos]

B,

In UK, from 59,112 flat races all types it is:
Favourite, Win 33%
2nd favourite 20%
3rd favourite 13%

Australia is the same.

It is an amazing universal constant in the sense that the amounts and quality of information available to the general betting public is so variable from country to country but the answer comes out almost exactly the same.
Yes, and it is also amazing how the Para-M system causes bettors to adjust to percieved overlays and underlays to make final odds ovwer the long haul turn out more or less predictable! Adding more fuel to the notion of "you can win a race" by handicapping each race individually!

[twindouble]

Those percentages don't mean anything to me. What you should be looking at is your own individual plays. What's your percentage of hitting horses other than the chalk in the races. If your hitting horses 3-1 and up that's what counts over the long haul. Break it down if you desire. What's your percentage of hitting horses over 10-1 and so on. The hell with what the crowd is doing.


T.D.

[44PACE]

Are there more favorites going off less than even money today than 15 years ago? It just seems that there are a ton of races with odds on favorites or 2 horses under 2-1.

[twindouble]

Are there more favorites going off less than even money today than 15 years ago? It just seems that there are a ton of races with odds on favorites or 2 horses under 2-1.

Pace, I'll handicap the race if the horses I come up with are 2-1 or less, I'll still figure out if I can make some money with them. That's only one race but it's part of other pools, that 2-1 just might be the key horse to making money be it picks, DD, supers or a $30 exacta I can take down a few times. I evaluate each race in that manor. If I think it's to risky, I'll move on. What I do is more important than what others do.


Good luck,

T.D.

[betchatoo]

Are there more favorites going off less than even money today than 15 years ago? It just seems that there are a ton of races with odds on favorites or 2 horses under 2-1.

There are more short priced favorites because there are more short fields than there used to be

[Overlay]

There are more short priced favorites because there are more short fields than there used to be

That also increases the importance of assessing a horse's winning chances in light of its odds, and wagering based on the presence of value, rather than pure winning probability alone.

[twindouble]

That also increases the importance of assessing a horse's winning chances in light of its odds, and wagering based on the presence of value, rather than pure winning probability alone.

I don't understand, if you have a horse that will more than likey win the race why would you bet something else just because the odds are higher? :confused: Pass on the race. Am I reading that right?

T.D.

[Overlay]

I don't understand, if you have a horse that will more than likey win the race why would you bet something else just because the odds are higher? Pass on the race. Am I reading that right?

I think you answered your own question by talking about passing a horse or race. (I assume you mean doing that when the horse you like gets bet down too low.) That's as much of a choice as betting on another horse in the same race, and the reason you would do that is because you believed that the horse you like (even though it may be the one that you think is the single best horse in the race) did not have a 100% guarantee of winning, and that the potential payoff you were going to receive from it was not going to compensate you for the risk you would be taking in betting it (that is, the chance that it might not win the race).

The flip side of that is, if no one horse has a 100% chance of winning, then multiple horses in every race have some chance of winning, and the odds on some of those horses will be higher than their actual chance of winning (overlays), while the odds on others will be lower (underlays). If you persist in playing underlays, you may cash more tickets, but the return you get from the races that you win won't compensate for the losses from races that you lose.

The beauty of the pari-mutuel system is that it allows for value-shopping, rather than offering fixed, adverse odds on every bet like most wagering games. I think the main thing you have to avoid is getting hung up on the thought that, since only one horse can win a particular race, it doesn't make sense to talk about each horse in the race having a certain percentage chance of victory. The fact that each horse is assigned odds of winning is a recognition in every race of the fact that all of them have some chance of winning (however likely or unlikely that might be).

[twindouble]

I think you answered your own question by talking about passing a horse or race if the horse you like gets bet down too low. That's as much of a choice as betting on another horse in the same race, and the reason you would adopt that course of action is because you believed that that horse (even though it may be the one you think is the single best horse in the race) did not have a 100% guarantee of winning, and that the potential payoff you were going to receive from it was not going to compensate you for the risk you would be taking in betting it. The flip side of that is, if no one horse has a 100% chance of winning, then multiple horses in every race have some chance of winning, and the odds on some of those horses will be higher than their actual chance of winning (overlays), while the odds on others will be lower (underlays). If you persist in playing underlays, you may cash more tickets, but the return you get from the races that you win won't compensate for the losses from races that you lose. The beauty of the pari-mutuel system is that it allows for value-shopping, rather than offering fixed, adverse odds on every bet like most wagering games. I think the main thing you have to avoid is getting hung up on the thought that, since only one horse can win a particular race, it doesn't make sense to talk about each horse in the race having a certain percentage chance of victory. The fact that each horse is assigned odds of winning is a recognition in every race of the fact that all of them have a chance of winning (however likely or unlikely that might be).

Overlay, you can sure make a case for betting a horse you don't like. The only time I say to myself "what the hell lets see what happens" I've got a 6 horse box in the super or I went 7 deep in the picks.

T.D.

[Overlay]

Overlay, you can sure make a case for betting a horse you don't like.

In the context of wagering value, "like" changes from an absolute to a relative term. The horse you like becomes the one with odds that compare most favorably to its winning chances, and it becomes possible to like any horse as long as it is enough of an overlay.

[Que]

Here are some results of how accurate the public is in estimating the odds. The test sample was for all North American Thoroughbred races between 1/1/1997 to 10/20/2006. Coupled entries were not included in the sample. I think these results are accurate, with the exception of the estimated odds of the horses at 1-20 odds which appear to have an error. Personally, I think the publics' ability to accurately estimate the horses actual chances of winning, when taken over a large sample, is amazing given all the variables. This doesn't mean that you can't find errors in the public odds in specific races, but when averaged over a large sample the publics' ability to accurately estimate the correct odds is incredible. For example, over 193,000 horses started at odds of 7-1 in the sample, and the actual win% was 9.60% and the publics average estimated odds was also 9.60%--you can't get much better than that. Howerver, there are some differences, specifically horses at low odds are slightly underbet, i.e. you could have made a slight flat bet profit betting every horse at 2-5 odds to place--in a 16,000 horse sample. In addition, horses at high odds are overbet, and the higher the odds the more the horse is overbet, i.e. the ROI for horses at 30-1 or greater is significantly lower than the tracks take and breakage (652,731 starts, with an ROI to win, place, and show at 0.56. 0.54, and 0.46 respectively).

Odds Results Actual Estimated Return On Investment (1$)
Range (Starts - Win - Place - Show) Win% | Plc% | %Shw Win% | Plc% | Show% Win | Place | Show
1-20 (143 - 98 - 18 - 7) 68.53 | 81.12 | 86.01 48.72 | 60.62 | 63.57 0.7048957 | 0.7364863 | 0.7354545
1-10 (399 - 313 - 50 - 19) 78.45 | 90.98 | 95.74 63.33 | 83.16 | 87.80 0.8611543 | 0.9807922 | 0.9791183
1-5 (4623 - 3276 - 771 - 299) 70.86 | 87.54 | 94.01 58.66 | 82.96 | 90.97 0.9277939 | 1.017027 | 1.002957
2-5 (16055 - 9860 - 3103 - 1409) 61.41 | 80.74 | 89.52 53.06 | 80.04 | 91.36 0.8964878 | 1.051587 | 0.9823563
3-5 (31706 - 16658 - 6787 - 3400) 52.54 | 73.94 | 84.67 47.69 | 75.79 | 89.84 0.8707611 | 0.9989013 | 0.9641072
4-5 (46036 - 21095 - 10267 - 5755) 45.82 | 68.12 | 80.63 42.98 | 71.18 | 87.21 0.8488227 | 0.9705297 | 0.9523295
1-1 (53570 - 22057 - 12135 - 7098) 41.17 | 63.83 | 77.08 38.86 | 66.60 | 84.01 0.8442595 | 0.9677746 | 0.9420589
6-5 (60214 - 22368 - 13928 - 8542) 37.15 | 60.28 | 74.46 35.46 | 62.52 | 80.91 0.8356458 | 0.9638834 | 0.9396271
7-5 (66530 - 22645 - 15055 - 10032) 34.04 | 56.67 | 71.75 32.60 | 58.91 | 77.96 0.8332869 | 0.9703178 | 0.9312666
8-5 (70967 - 22537 - 15646 - 11049) 31.76 | 53.80 | 69.37 30.24 | 55.79 | 75.32 0.8403454 | 0.9264773 | 0.924005
9-5 (75584 - 22250 - 16752 - 11960) 29.44 | 51.60 | 67.42 28.15 | 52.87 | 72.65 0.8378243 | 0.9282244 | 0.922362
2-1 (203638 - 52755 - 43205 - 33374) 25.91 | 47.12 | 63.51 25.15 | 48.47 | 68.37 0.8264451 | 0.9125828 | 0.9054583
5-2 (203021 - 45407 - 41085 - 33390) 22.37 | 42.60 | 59.05 21.81 | 43.33 | 63.05 0.8238991 | 0.8842628 | 0.8857743
3-1 (192992 - 37316 - 36785 - 32302) 19.34 | 38.40 | 55.13 19.24 | 39.12 | 58.40 0.8089843 | 0.852015 | 0.8679422
7-2 (178090 - 30769 - 32192 - 29611) 17.28 | 35.35 | 51.98 17.21 | 35.64 | 54.33 0.8076168 | 0.8532904 | 0.8566626
4-1 (161556 - 25191 - 27518 - 26571) 15.59 | 32.63 | 49.07 15.56 | 32.71 | 50.74 0.8069213 | 0.8309042 | 0.8450419
9-2 (146193 - 20861 - 23908 - 23498) 14.27 | 30.62 | 46.70 14.21 | 30.18 | 47.46 0.8096881 | 0.8226615 | 0.8415053
5-1 (254587 - 31862 - 38786 - 40221) 12.52 | 27.75 | 43.55 12.60 | 27.08 | 43.29 0.8005998 | 0.8250639 | 0.8299121
6-1 (219985 - 23843 - 30295 - 33662) 10.84 | 24.61 | 39.91 10.90 | 23.74 | 38.60 0.8013988 | 0.8083635 | 0.8178841
7-1 (193666 - 18594 - 24934 - 28188) 9.60 | 22.48 | 37.03 9.60 | 21.09 | 34.72 0.804517 | 0.8117227 | 0.8079869
8-1 (171515 - 14889 - 20611 - 23664) 8.68 | 20.70 | 34.49 8.59 | 19.00 | 31.58 0.8141084 | 0.8260845 | 0.8002198
9-1 (152902 - 11721 - 17006 - 20339) 7.67 | 18.79 | 32.09 7.77 | 17.26 | 28.90 0.7949234 | 0.7971302 | 0.7868616
10-1 (134685 - 9458 - 13914 - 17237) 7.02 | 17.35 | 30.15 7.09 | 15.83 | 26.68 0.798054 | 0.7855626 | 0.7731073
11-1 (120226 - 7783 - 11452 - 14825) 6.47 | 16.00 | 28.33 6.53 | 14.61 | 24.76 0.8005359 | 0.7730889 | 0.7602435
12-1 (107232 - 6337 - 9843 - 12520) 5.91 | 15.09 | 26.76 6.04 | 13.56 | 23.06 0.7897094 | 0.7765263 | 0.7530004
13-1 (96587 - 5236 - 8297 - 10836) 5.42 | 14.01 | 25.23 5.62 | 12.65 | 21.59 0.7776598 | 0.7618844 | 0.7356888
14-1 (86976 - 4500 - 7124 - 9525) 5.17 | 13.36 | 24.32 5.26 | 11.86 | 20.31 0.794769 | 0.7639111 | 0.7319151
15-1 (78312 - 3727 - 6079 - 8337) 4.76 | 12.52 | 23.17 4.94 | 11.16 | 19.17 0.7755196 | 0.7650631 | 0.7332938
16-1 (70988 - 3172 - 5283 - 7365) 4.47 | 11.91 | 22.29 4.65 | 10.53 | 18.12 0.7739155 | 0.7593975 | 0.7228513
17-1 (64500 - 2677 - 4573 - 6325) 4.15 | 11.24 | 21.05 4.40 | 9.98 | 17.22 0.7602314 | 0.7376796 | 0.7036571
18-1 (59799 - 2403 - 4065 - 5744) 4.02 | 10.82 | 20.42 4.18 | 9.48 | 16.36 0.774395 | 0.748497 | 0.7034233
19-1 (55716 - 2096 - 3552 - 5226) 3.76 | 10.14 | 19.52 3.97 | 9.03 | 15.62 0.7611325 | 0.7265959 | 0.6836939
20-1 (51260 - 1823 - 3202 - 4651) 3.56 | 9.80 | 18.88 3.79 | 8.61 | 14.91 0.7546201 | 0.7384276 | 0.6847326
21-1 (47932 - 1618 - 2829 - 4051) 3.38 | 9.28 | 17.73 3.62 | 8.23 | 14.26 0.7493951 | 0.7082772 | 0.6577171
22-1 (45147 - 1450 - 2547 - 3780) 3.21 | 8.85 | 17.23 3.47 | 7.90 | 13.70 0.7489657 | 0.7112234 | 0.6612862
23-1 (42487 - 1321 - 2274 - 3433) 3.11 | 8.46 | 16.54 3.32 | 7.57 | 13.14 0.7531281 | 0.7055861 | 0.6501368
24-1 (39688 - 1167 - 2068 - 3266) 2.94 | 8.15 | 16.38 3.19 | 7.28 | 12.65 0.7364945 | 0.6953598 | 0.6439793
25-1 (168693 - 4357 - 8050 - 12269) 2.58 | 7.35 | 14.63 2.88 | 6.57 | 11.44 0.7237384 | 0.6876593 | 0.6205702
30-1 (129135 - 2661 - 5046 - 8092) 2.06 | 5.97 | 12.23 2.44 | 5.59 | 9.75 0.6729513 | 0.641869 | 0.571569
35-1 (100049 - 1709 - 3382 - 5592) 1.71 | 5.09 | 10.68 2.12 | 4.86 | 8.49 0.647462 | 0.6283016 | 0.543077
40-1 (145153 - 2017 - 4039 - 6904) 1.39 | 4.17 | 8.93 1.79 | 4.11 | 7.19 0.6213843 | 0.5931291 | 0.499829
50-1 (95241 - 909 - 2051 - 3650) 0.95 | 3.11 | 6.94 1.47 | 3.36 | 5.89 0.5209129 | 0.5279222 | 0.4446416
60-1 (63402 - 489 - 1042 - 1970) 0.77 | 2.41 | 5.52 1.24 | 2.84 | 4.98 0.4892724 | 0.471551 | 0.3933643
70-1 (42074 - 252 - 572 - 1063) 0.60 | 1.96 | 4.48 1.08 | 2.47 | 4.32 0.4313292 | 0.4244659 | 0.3420263
80-1 (27634 - 127 - 310 - 601) 0.46 | 1.58 | 3.76 0.95 | 2.18 | 3.81 0.3791525 | 0.3907932 | 0.3087908
90-1 (18643 - 60 - 142 - 313) 0.32 | 1.08 | 2.76 0.85 | 1.94 | 3.39 0.2953521 | 0.28575 | 0.2452286
100-1 (31400 - 93 - 208 - 392) 0.30 | 0.96 | 2.21 0.67 | 1.54 | 2.70 0.3163201 | 0.2928042 | 0.2122354

Que.

[speculus]

Most of these studies worldwide all say the same thing - that, long term, the public odds are close to expectations, less any take outs. What they do not study or tell you is how good the estimates are for individual races. So people reading the "facts" get misled into thinking they must be accurate for each and every race. The pricing errors +/- do exist in an individual race but just even out over time. Those with better information exploit those individual errors in individual races.

Well said!:ThmbUp:

[speculus]

From our study of 200,000 races published in the January 2000 edition of Meadow's Racing Monthly:

Favorite averages 1.5 - 1 and wins 32.8%
Second choice averages 2.9 - 1 and wins 20.8%
Third choice averages 4.4 - 1 and wins 15.0%
Fourth choice averages 6.3 - 1 and wins 10.9%
Fifth choice averages 9.0 - 1 and wins 7.8%

Blindly betting any of these returns 0.79 - 0.82 per dollar.

We've also done a number of other studies which show that horses win in proportion to their odds (after adding back the takeout) up through about 20-1, when there begins a slight dropoff. These studies have involved hundreds of thousands of nationwide races.

Thanks BMeadow!

Abraham Lincoln was once traveling with a friend who spotted a beautiful horse grazing in a roadside field, and exclaimed, "Here! Look at that beautiful black horse!"

Lincoln glanced over, and replied, unexcited," Well, it's black on THIS side."

If viewed from the other side, the same findings can be stated this way:

67.2% races are won by horses with average odds way over 2.9 to 1.
46.4% races are won by horses with average odds way over 4.4 to 1.
31.4% races are won by horses with average odds way over 6.3 to 1.
20.5% races are won by horses with average odds way over 9 to 1.
12.7% races are won by horses with average odds way over maybe 15 to 1?

Looked at it this way, the truth emerges VERY CLEARLY. Your chance of beating this game depends on whether sufficient number of your winning bets come from these?

In other words, the value of any angle, theory, method or whatever you use to handicap the horses depends on how often your top choice is NOT THE CHALK and how ofen your non-chalk top choice WINS.

[twindouble]

Thanks BMeadows!

Abraham Lincoln was once traveling with a friend who spotted a beautiful horse grazing in a roadside field, and exclaimed, "Here! Look at that beautiful black horse!"

Lincoln glanced over, and replied, unexcited," Well, it's black on THIS side."

If viewed from the other side, the same findings can be stated this way:

67.2% races are won by horses with average odds way over 2.9 to 1.
46.4% races are won by horses with average odds way over 4.4 to 1.
31.4% races are won by horses with average odds way over 6.3 to 1.
20.5% races are won by horses with average odds way over 9 to 1.
12.7% races are won by horses with average odds way over maybe 15 to 1?

Looked at it this way, the truth emerges VERY CLEARLY. Your chance of beating this game depends on whether sufficient number of your winning bets come from these?

In other words, the value of any angle, theory, method or whatever you use to handicap the horses depends on how often your top choice is NOT THE CHALK and how ofen your non-chalk top choice WINS.

Speculus, it's good to see percentages posted that are worth while, that's what I meant when I said your better off knowing your % of wins on horses 3-1 and up. It's also important to recognize with today's wagering menu the chalks can make you money as well, it's fool hearty to toss them when they figure to win or hit the board.

Good post,

T.D.

[speculus]

[/b]

Speculus, it's good to see percentages posted that are worth while, that's what I meant when I said your better off knowing your % of wins on horses 3-1 and up. It's also important to recognize with today's wagering menu the chalks can make you money as well, it's fool hearty to toss them when they figure to win or hit the board.

Good post,

T.D.

You need not toss them out, but you can surely draw a line for minimum accetable odds depending upon your personal betting style, money management strategy and handicapping skill. The worst thing (which most gamblers end up doing) is betting MORE when playing chalk, and less when playing at higher odds.

[twindouble]

You need not toss them out, but you can surely draw a line for minimum accetable odds depending upon your personal betting style, money management strategy and handicapping skill. The worst thing (which most gamblers end up doing) is betting MORE when playing chalk, and less when playing at higher odds.

I agree, the only time I step out is when I have good value and good reason to bet the horse.

Good luck,

T.D.

[Que]

I want to correct the figures in my post above regarding betting horses at 2-5 odds to place. While the win, place, and show figures should still be correct, the ROI values were skewed due to errors in my database for races run at El Comandante Racetrack. When I excluded those races from the query, the new ROI for horses at 2-5 odds when bet to place were 0.902, 0.969, and 0.933 to win, place, and show respectively. While this is certainly above the track's take, its not a flat bet profit. BTW, I don't necessarily recommend betting these horses...

Que.

[PriceAnProbability]

That also increases the importance of assessing a horse's winning chances in light of its odds, and wagering based on the presence of value, rather than pure winning probability alone.

Which of course requires you to be smarter than the morning line maker and the public at setting the true odds.

Care to explain how to do THAT?

[DJofSD]

Answer here - follow the link! (http://www.100monkeys.org/)

[speculus]

I agree, the only time I step out is when I have good value and good reason to bet the horse.

Good luck,

T.D.

Stepping out for good value, non-chalk horse is indeed the best betting strategy.

I conduct handicapping classes to teach what I call Dynamic Handicapping, a theory or method which took more than 20 years of empirical reserach to develop to its present "practicable" form.

Three years ago, a very renowned bookmaker (bookies are 'legal' in my country, India) enroled for the two-day course, but kept on insisting that he wanted to see me first and talk before the course began. Curious, I called him over to my office.

He said he was very busy, and could not possibly devote two full days to learning, that he had heard there is a special two-hour session that I conduct to discuss how to use Dynamic Handicapping to spot whether the fav is legit or vulnerable or outright false, and that he was interested in learning only that part. He also added he did not mind paying the full fees.

I had a hard time convincing him that the session will go over his head if he did not learn the methods in the first place. But he did not budge. He kept on repeatedly arguing that he would be anyway paying the full fees, so I should allow him this special privilege.

I finally told him he need not pay the fees, he was most welcome to join as my guest only for that two-hour session, on the condition that he will not be allowed to raise any query if he did not understand what was being discussed, as that would amount to “wasting” other participants’ time who may not share his ignorance. He agreed.

This guy showed up precisely for that session, and just listened to the discussion and debate as promised. Interestingly, the class discussed the next day’s card and singled out a horse as “extremely vulnerable” which turned out to a 2 to 5 fav the next day. We also discussed the dynamic profile of another horse in the same race that was way superior to the 2-5 horse. This horse was at 7 to 1. The fav lost and the 7-1 horse won.

A week later, I received an envelope that contained cash and a note from him that said two of his assistants would be enrolling for the next batch, that I should adjust part of the cash towards their fees, and the balance (which was a small fortune, I must say) I was free to donate to a charity of my choice. The note also explained that he was doing what he felt he ought to do, because he had not only laid against the discussed fav (something he would have anyway done as part of his bookmaking business), but he had also loaded a sizable bet on the 7-1 winner, because, he said in the note, “I was convinced I was risking a part of the profit that would anyway come to me after the race. As a bookie I have observed many times that consistently good handicappers have most often been right about their assessment of the fav, especially if they don’t like it.”

Without knowing it, this man taught me my most profitable lesson, which I have used to great advantage when similar situation presented again. The lesson can be summed up thus: Whatever your handicapping method, if you are VERY SURE the fav is false or extremely vulnerable, first, you MUST lay against it (presuming it’s legal in your country or if you have an account with a betting exchange), and then, part of the projected profits (what part depends on your money management ideas) you MUST risk on your top choice. After a while you will love to see most “peaks” on your graph of growth of capital correspond to these situations.

[twindouble]

Without knowing it, this man taught me my most profitable lesson, which I have used to great advantage when similar situation presented again. The lesson can be summed up thus: Whatever your handicapping method, if you are VERY SURE the fav is false or extremely vulnerable, first, you MUST lay against it (presuming it’s legal in your country or if you have an account with a betting exchange), and then, part of the projected profits (what part depends on your money management ideas) you MUST risk on your top choice. After a while you will love to see most “peaks” on your graph of growth of capital correspond to these situations. Quote; Speculus

I didn't realize I was doing something "dynamic" for so many years. Never bet through an Exchange and if it wasn't for Youbet I'd have no record of what I do, even with that what I put through OTB isn't accounted for on a spread sheet. Keeping track of my bankroll is simple, either it's growing or it isn't. Anything race related comes out of it, plus I have no desire to rehash losing streaks or winning ones, every day is a new one to me with different conditions and opportunities or lack of. I only have to keep track of one figure, that's what my bankroll is when I started and I don't put unreasonable restrictions on it. Most inportant, I'm willing to "gamble" on myself. That's me and I enjoy swapping horse stories.

Good luck,

T.D.

[speculus]

I didn't realize I was doing something "dynamic" .........
T.D.
Well, this particular "betting strategy" angle is "explosive", not "dynamic".;)

[Overlay]

Which of course requires you to be smarter than the morning line maker and the public at setting the true odds.

Care to explain how to do THAT?

PriceAnProbability:

Check your PA messages.

[PriceAnProbability]

PriceAnProbability:

Check your PA messages.

He means this:

For me, a statistically-based fair-odds line for each race rolls all four "quarters" together nicely. Using the probabilities associated with a well-distributed mix of fundamental factors simultaneously ranks the horses; produces fair odds through the ratio of the final product generated for each horse to the sum of products for the field; provides a greater degree of both confidence and discipline in wagering than would be the case if decisions were based only on opinion or guess; and facilitates both record-keeping of results, and isolating the cause of any significant variations or changes in performance.


So how can I use this to generate a line for a real-live race?

[sjk]

You?

Up to you to figure out.

P.S. Very doable. Just takes some hard work.

[sjk]

Hate to reply to my own post but I was impolite above.

If you have a power rating which makes a profit based on top rated horse you already have the key to the mint. If you have gotten that far making it into a good odds line should be no more difficult than a good crossword puzzle.

I would conjecture that those who have had the most success making power ratings, odds lines and real money wagers are probably the most credulous as to whether other posters have enjoyed the success they claim. I take it from your other posts that you are doing very well and congratulate you on your good efforts.

[Overlay]

He means this:

For me, a statistically-based fair-odds line for each race rolls all four "quarters" together nicely. Using the probabilities associated with a well-distributed mix of fundamental factors simultaneously ranks the horses; produces fair odds through the ratio of the final product generated for each horse to the sum of products for the field; provides a greater degree of both confidence and discipline in wagering than would be the case if decisions were based only on opinion or guess; and facilitates both record-keeping of results, and isolating the cause of any significant variations or changes in performance.


So how can I use this to generate a line for a real-live race?

No, the post I meant was this (#201 on that thread):

"In response to Formula 2002's request, consider the following value matrix as an example of specific factors that could be combined through multiplication to produce an effective fair-odds line, while keeping the total of variables involved to a workable number. (The three impact values listed on each line corrrespond to dirt sprints, dirt routes, and turf routes, respectively.)(Sorry for the formatting problems.)

Overall Factors:

Best Speed in Last 30 Days (Speed/Condition)
Number of “Good” Races in Last Three Starts (Condition/Consistency)
Quirin Speed-Point Total (Early Speed/Pace)
Two-Year Earnings (Class)
Jockey (Connections/Intent)

Best Speed in Last 30 Days (Rank in Field):

First 2.08 2.07 1.83
Second 1.52 1.45 1.55
Third 1.18 1.09 1.39
Front Half .97 1.06 1.04
Rear Half .59 .59 .60


Number of “Good” Races in Last Three Starts

3 1.54 1.53 1.45
2 1.29 1.25 1.19
1 1.03 .93 .86
0 .67 .61 .64

Quirin Speed-Point Total:

8 1.58 1.58 1.22
7 1.54 1.34 1.22
6 1.52 1.26 1.12
5 1.24 1.15 1.01
4 1.20 1.02 1.01
3 1.10 .90 1.01
2 .93 .89 .90
1 .91 .88 .90
0 .65 .65 .90



Two-Year Earnings (Rank in Field):

First 1.65 1.57 1.60
Second 1.35 1.35 1.39
Third 1.16 1.13 1.19
Front Half 1.05 1.01 .92
Rear Half .72 .74 .76


Jockey:

Top 5 at Meeting 1.50 1.37 1.37
“Good" Race on Horse 1.17 1.07 1.07
Neither of the Above .58 .64 .64"

[JackS]

We always have to consider the crowd favorite and I personally place this horse into one of three catagories. The horse WILL win.....The horse SHOULD win......The horse MIGHT win. The first two catagories are easily possible plays and especially the first.
Mentally I make the WILL win 50/50 or 1to1. This is the equivilent of a loaded coin that flips heads-up 80% of the time. Play this horse at 1/1 or above every time.
The SHOULD win horse probably plays to the crowd win average of 2/1 and could also be a play at anything at 3/1 or above which happens but not that often.
The MIGHT win horse and favorite should be played against every time. My estimation is that these horses will win about 20% of their races which is never going to produce a profit over the long run. A plus with this type is that other horses in the race are probably overlayed because of him. If he happens to be a false favorite, his odds of winning are even lower and others are even higher.
For exoctic plays, the WILL on top only. The SHOULD 1st and 2nd, the MIGHT 2nd ,3rd,4th or out ,your choice.

[twindouble]

We always have to consider the crowd favorite and I personally place this horse into one of three catagories. The horse WILL win.....The horse SHOULD win......The horse MIGHT win. The first two catagories are easily possible plays and especially the first.
Mentally I make the WILL win 50/50 or 1to1. This is the equivilent of a loaded coin that flips heads-up 80% of the time. Play this horse at 1/1 or above every time.
The SHOULD win horse probably plays to the crowd win average of 2/1 and could also be a play at anything at 3/1 or above which happens but not that often.
The MIGHT win horse and favorite should be played against every time. My estimation is that these horses will win about 20% of their races which is never going to produce a profit over the long run. A plus with this type is that other horses in the race are probably overlayed because of him. If he happens to be a false favorite, his odds of winning are even lower and others are even higher.
For exoctic plays, the WILL on top only. The SHOULD 1st and 2nd, the MIGHT 2nd ,3rd,4th or out ,your choice.

I like those categories, they are common expressions at the track. Like many they mean different things to some players. I've used "should win it" many times but with no odds attached to it, could be 2-1 or 30-1. I've got in my mind, baring any trouble in the race or getting caught up in a slow pace. Not a "key" horse in my opinion or one you'll bet he house on, a good percentage of them come from behind, good exotic plays as you say.

The "will win" with me again they have no specific odds attached, this one is my key horse and I feel regardless of the pace he stands above the rest. A sound win bet and to be used whereever I can make money.

"Might win" horses, have vulnerabilities but given the ideal race could upset or in most cases are bet down with no value. These are the ones you use deep in the exotics or not at all and don't qualify for a serious win or exacta play.


I never use the expressions "sure thing", "shoe in", "lock", "in the bank" or unbeatable for obvious reasons. Don't get me wrong, when someone takes a strong stand on a horse and it wins, it is a confidence builder and I enjoy watching their reactions but I'm not playing to build confidence, that should be there to begin with, well to the extent racing will allow you to have it. :D

Good luck,

T.D.