Race odds as a factor of horses in race

[thearmada]

Does anybody have an average sum of odds for races with 5, 6, 7, ..., and 20 horses respectively?

For example, SAR R12 8/25/12 - Travers Stakes
Horse Odds
-----------
3 33.5
6 2.4
11 32.5
7 8.1
10 3.3
8 4.9
4 23.2
1 52.5
9 7.4
2 44.5
5 6.7

Sum of odds = 219

I am looking into creating my own odds for races and I am looking for the sum of odds to ensure I am making odds for the race within reason.

I hope that makes sense. If not, please let me know,

[thaskalos]

Originally Posted by thearmada

Does anybody have an average sum of odds for races with 5, 6, 7, ..., and 20 horses respectively?

For example, SAR R12 8/25/12 - Travers Stakes
Horse Odds
-----------
3 33.5
6 2.4
11 32.5
7 8.1
10 3.3
8 4.9
4 23.2
1 52.5
9 7.4
2 44.5
5 6.7

Sum of odds = 219

I am looking into creating my own odds for races and I am looking for the sum of odds to ensure I am making odds for the race within reason.

I hope that makes sense. If not, please let me know,

It works better when you convert the odds into win percentages.

[davew]

Originally Posted by thaskalos

It works better when you convert the odds into win percentages.

especially when a couple in the race should be over 200/1


you need to convert to probabilities
1 / (odds+1) = probability to win

[thaskalos]

Originally Posted by thearmada

Does anybody have an average sum of odds for races with 5, 6, 7, ..., and 20 horses respectively?

For example, SAR R12 8/25/12 - Travers Stakes
Horse Odds
-----------
3 33.5
6 2.4
11 32.5
7 8.1
10 3.3
8 4.9
4 23.2
1 52.5
9 7.4
2 44.5
5 6.7

Sum of odds = 219

I am looking into creating my own odds for races and I am looking for the sum of odds to ensure I am making odds for the race within reason.

I hope that makes sense. If not, please let me know,

I apologize for not giving you a more complete reply; I was otherwise occupied at the time.

The percentages are calculated according to the formula that Dave provided...and they should total about 120 -- to account for the takeout extracted from the win pool.

[mountainman]

I was taught years ago to use 100 as the obvious starting point, add 1 point for each percentage point of takeout, and one additional point for each starter. Thus the target total for a 7 horse field at a track with a 17% skim would be 124.

But since all price-points are rounded down-and never upward-the actual total derived from real-time odds can vary a great deal, and in an unpredictable manner. The cardinal sin for a track odds-maker is to come in low, thus promising the public more in potential return than is likely to exist. I still hate to come in high, though, because, to me, it's indicative of an amateurish lines-maker unwilling to make tough decisions. The guys who are really lousy at it almost always come in high.

[TJDave]

When I made odds I only used my contenders. Also make adjustments based on field size and PP.

[Dave Schwartz]

Armada,

I would respectfully suggest that you change your methodology.

Instead of summing the odds, sum the odds / (odds +1). This will be more meaningful.


Regards,
Dave Schwartz

[MightBeSosa]

Originally Posted by Dave Schwartz

Armada,

I would respectfully suggest that you change your methodology.

Instead of summing the odds, sum the odds / (odds +1). This will be more meaningful.


Regards,
Dave Schwartz

I think you mean 1/(odds+1)

[MightBeSosa]

Originally Posted by mountainman

I was taught years ago to use 100 as the obvious starting point, add 1 point for each percentage point of takeout, and one additional point for each starter. Thus the target total for a 7 horse field at a track with a 17% skim would be 124.

But since all price-points are rounded down-and never upward-the actual total derived from real-time odds can vary a great deal, and in an unpredictable manner. The cardinal sin for a track odds-maker is to come in low, thus promising the public more in potential return than is likely to exist. I still hate to come in high, though, because, to me, it's indicative of an amateurish lines-maker unwilling to make tough decisions. The guys who are really lousy at it almost always come in high.

Actually, there's no justification for adding the field size. Not sure where that came from, maybe its an attempt to mitigate scratches?

The 'book' for a 5 horse field and a 14 horse one is exactly the same.

[davew]

I think Mark (mountainman) is talking about making a Morning Line that track handicappers make as a 'prediction' of what they feel the odds will close at.

If you make a probability line, it doesn't matter how many horses are in the race as the sum total should be near 1.000 - although maybe a touch higher depending on what you do with dead heat wins.

[Dave Schwartz]

I think you mean 1/(odds+1)

Thank you!

You are right, of course.

[MightBeSosa]

Originally Posted by davew

I think Mark (mountainman) is talking about making a Morning Line that track handicappers make as a 'prediction' of what they feel the odds will close at.

If you make a probability line, it doesn't matter how many horses are in the race as the sum total should be near 1.000 - although maybe a touch higher depending on what you do with dead heat wins.

Yes, I know exactly what he is saying. And again, let's do a mind experiment.

If you knew exactly how the public would bet, you'd make the morning line the same as the closing odds, which would be the same % total.

So if you are making a precision ML, why would you do anything differently. Now, I can tell you, in the case of AE's, there's no much a ML can do to perfect the line. That's why I think some guys may have added points to a large field, to assume a scratch or two. Other than that, it makes little sense to me.


In the case of a dead heat, no, same book. The DH payoffs have nothing to do with the closing odds.

[thearmada]

This is why I love this site. You ask for something, and people even give you more than you expect. I never thought about the 1/(odds+1).

Thanks to all for their input. I will start trying this out and see how it works before I bet.

Well I couldn't bet anyway seeing as I can't use my ADW account. Illinois at its prime.

[mountainman]

Originally Posted by MightBeSosa

Yes, I know exactly what he is saying. And again, let's do a mind experiment.

If you knew exactly how the public would bet, you'd make the morning line the same as the closing odds, which would be the same % total.

So if you are making a precision ML, why would you do anything differently.

No, the morning line can never precisely reflect actual closing odds-nor should it attempt to. And the reason should be obvious: the program can't list horses at 3.30-1 or 10.6-1 etc. Such limitations aside, no oddsmaker could be that accurate, which is why very few track linesmakers will even use odd-numbered integers above 5 (ie 7-1, 9-1 etc). To do so would imply a precision no linesmaker can manage.

And as far as adjusting the total points as per field size, i do that because, as stated earlier, price points are never rounded upward on the oddsboard, thus each additional wagering interest brings the potential for a higher point total. The formula isn't infallible, but it's the golden rule for track oddsmakers. A salient point to remember, sosa, is that unlike precise closing odds such as listed in the charts, the tote board odds aren't a zero sum game.

[MightBeSosa]

You are correct, by 'lowballing' the actual odds, the total DISPLAYED % will increase, and the more horses there are, the more opportunity there is for that to happen. Of course its only an illusion. Never dawned on me that a linemaker would adjust for the posting anomalies of the tote. Not sure I like the idea, but in the grand scheme of things, its not a big deal.

I imagine when we finally get ACTUAL decimal odds this will become a moot point.

Thanks for the explanation.