Getting a Line on the Early Odds

[hyipro]

Getting a Line on the Early Odds



By Joe Kristufek

As Arlington Park's official morning-line odds maker my job is to predict, as accurately as possible, how the betting public will wager on each race. The actual odds change minute by minute based on how much money is bet on each horse.

The intention of the morning line is to give the wagering public a general idea of what the odds on each horse is expected to be at post time. It's a subjective system, and is merely my educated guess, based on my 18 years of handicapping Illinois horseracing professionally.

Keep in mind that the morning line odds do not reflect the odds maker's selections, they are merely a prediction of how the betting public will wager on the race. Although I do my best to calculate accurate early odds, some races are tougher to assign odds values on than others.

Read the rest here:

http://www.arlingtonpark.com/node/10222

[Overlay]

A useful tutorial on how a morning linemaker does his job. But if a player uses valid mathematical or statistical values/procedures to make his own 100% line, it eliminates guesswork and the need for juggling figures to make sure that they come out right. It also allows for direct comparison of the line with the toteboard to determine whether a horse is offering betting value or not.

Also, although I would tend to agree with Mr. Kristufek's views about the potential value offered by horses with public odds that are higher than their morning line, I'm not sure that it's possible to conclude as broadly as he apparently does that value is present in all such instances. It takes a case-by-case evaluation to determine whether a horse with low morning-line odds but higher actual odds is "dead on the board" or not.

He advises avoiding looking at the morning line until after you've done your own handicapping. By that time, however, if you have confidence in your own line, there would no need to consult the morning line. The only thing that would matter would be how your line compares with the actual public odds.

Finally, he openly acknowledges the upper and lower constraints (30-1 and 4-5, respectively) placed on the morning linemaker with respect to odds, even though in any given race, one or more horses may merit actual odds that fall outside those limits (especially on the high side). That further reduces the usefulness of the morning line as a genuine handicapping tool, regardless of how conscientious or accurate the linemaker may otherwise be.

[Ocala Mike]

A very informative write-up or article. I still say, however, that for M/L makers at every track, there is at least some consideration given to "selecting" one horse over another rather than just flat-out predicting what the public will do.

I submit as evidence of this (and it's probably done because the tracks want it this way) the fact that you will NEVER see a reputable M/L maker list the EXACT same M/L price on the FAVORITE(S) in a given race. He will ALWAYS take a stand and make one, say, EVEN money and one 6/5 rather than list them both at the same price. Also, many tracks still note the lowest 4 M/L odds horses as "Selections" at the bottom of the program page. Again, I believe this is pretty much what that particular track wants to maybe assist novice bettors.

[JolyB]

A very useful and informative article. There is one portion that left me scratching my head.

The article states:

"A true morning line must "balance". If there wasn't any takeout from the ‘win' pool, the base for the point system would be 100. But the line maker must factor in takeout to his formula to arrive at a mathematically correct morning line.

The takeout in the win pool in Illinois is 17 percent. Therefore, by adding 17 to a base of 100, we arrive at 117 points. Then, by designating an additional point per horse, the morning line will generally balance between 125 and 129 points for fields consisting of eight to 12 horses."

My question is as follows: If the takeout of 17% is required to cause the ML to balance, why should the linemaker add an additional 8 to 12 points to end up with a number of between 125 and 129? Why not leave it at 117? The only thing I can think of is breakage, but that doesn't seem to be quite that large.

Any explanations would be greatly appreciated.

[Overlay]

My question is as follows: If the takeout of 17% is required to cause the ML to balance, why should the linemaker add an additional 8 to 12 points to end up with a number of between 125 and 129? Why not leave it at 117? The only thing I can think of is breakage, but that doesn't seem to be quite that large.

Any explanations would be greatly appreciated.
To account for dime breakage in making a morning line in the manner described, I would add a flat 5% to the percentage of the take, since I figure that the breakage would average out to that amount over time. (I'm not sure where Mr. Kristufek got his procedure of adding 1% to the take for each horse in the field.) Uncertainties like that are another reason that I make a 100% line for personal use, which removes breakage from the equation.

[cj]

To account for dime breakage in making a morning line in the manner described, I would add a flat 5% to the percentage of the take, since I figure that the breakage would average out to that amount over time. (I'm not sure where Mr. Kristufek got his procedure of adding 1% to the take for each horse in the field.) Uncertainties like that are another reason that I make a 100% line for personal use, which removes breakage from the equation.

Pretty sure 5% is going too far, closer to 1% if I remember correctly.

Example:

Horse A is in a pool with 17% take and dime breakage. $10,000 total in pool, $2,000 bet on Horse A. What does he pay?

10,000 after takeout becomes 8,300. Payout comes out to 8,300 / 1000 = 8.30 which equals 8.20 after breakage.

So, if there were no breakage, what takeout gives the same breakage?

18%, or a 1% increase.

8,200 / 1000 = 8.20.

Of course it varies from race to race, but long term I think 1% is just about right.

[Dave Schwartz]

JolyB,

Your math process is incorrect. The correct calculation is to divide 100 by .83, giving you 120.48.

BTW, you do not actually need to know the takeout, instead, do this:

1. Take one day of charts for a track.
2. For each race...
for each betting interest in the race...
compute 1/(odds+1)
3. Sum the horses in the race to get the points.
4. At the end of the day you will have 8,9 or 10 races. Create an average of the points reflected in each race.

This will accurately give you the points unique to this track without actually knowing the takeout.

[Overlay]

Pretty sure 5% is going too far, closer to 1% if I remember correctly.

Example:

Horse A is in a pool with 17% take and dime breakage. $10,000 total in pool, $2,000 bet on Horse A. What does he pay?

10,000 after takeout becomes 8,300. Payout comes out to 8,300 / 1000 = 8.30 which equals 8.20 after breakage.

So, if there were no breakage, what takeout gives the same breakage?

18%, or a 1% increase.

8,200 / 1000 = 8.20.

Of course it varies from race to race, but long term I think 1% is just about right.
You're absolutely right. I was somehow thinking of it in terms of its percentage of the basic $2.00 bet, rather than of the average mutuel. Thanks. (One more reason -- from my perspective -- for making a 100% fair-odds line that can be directly compared to the minimum payoff represented by the tote odds, where breakage has already been taken into account, and doesn't have to be further considered in the making of the line.)

[Ocala Mike]

Rather than fool around with actual takeouts, in constructing a M/L for ANY track you're not going to be too far off using 120 as the target value.