It's an overlay...Is it a good bet?

[CBedo]

We've had multiple threads lately about "what is an overlay" but I was reading something yesterday that made me think more about "what is a good bet."

I was reading about a 35/1 shot winner that someone had made 10/1 on their personal line and was kicking themselves for not betting it. I have always (for a long time anyway) gone with the premise that my true contenders to be bet were the ones that I gave a win probability greater than X, where for some time now, X has been about 15% (6ish/1). I think I originally started using that number to cut down on losing streaks. Also, since I'm the first to admit that I can't tell the difference between a 8% horse and a 9% horse, the overlaid percentages get very sensitive to slight movements for these types of horses. I've just found that for me, I need to give a horse about a 6/1 fair line or better and have a 40% perceived edge to make a bet.

I thought I'd just ask all of you who make odds lines, how you decide what a good bet is? If you make a horse 10/1 and he's 35/1 is that a bet? What about a 5% shot who is 50/1, etc?

[andymays]

Just an opinion that I just put on the other Overlay topic.


An overlay is in the eyes of the beholder.

The art/science of successfully betting on Horses is really educated guessing.

The more educated your guess is the more likely you are to identify a true overlay!

[Valupix]

We've had multiple threads lately about "what is an overlay" but I was reading something yesterday that made me think more about "what is a good bet."

I was reading about a 35/1 shot winner that someone had made 10/1 on their personal line and was kicking themselves for not betting it. I have always (for a long time anyway) gone with the premise that my true contenders to be bet were the ones that I gave a win probability greater than X, where for some time now, X has been about 15% (6ish/1). I think I originally started using that number to cut down on losing streaks. Also, since I'm the first to admit that I can't tell the difference between a 8% horse and a 9% horse, the overlaid percentages get very sensitive to slight movements for these types of horses. I've just found that for me, I need to give a horse about a 6/1 fair line or better and have a 40% perceived edge to make a bet.

I thought I'd just ask all of you who make odds lines, how you decide what a good bet is? If you make a horse 10/1 and he's 35/1 is that a bet? What about a 5% shot who is 50/1, etc?
This someone may not be a fool, but he had a foolish moment. Anytime someone passes on a horse they consider an overlay is as foolish a decision as a gambler can make.

If the gambler doesn't trust himself, he has far deeper problems than this one moment. It either is an overlay in a gamblers eyes or it is an underlay and if it's an overlay there isn't a single time ever that they should pass.

If they can rationale passing on one, they MUST go back and rework what they're doing. Because they are doing something wrong. Terribly wrong!

But of course that’s just my opinion and in the end I may not have the slightest clue what the heck I’m talking about.

[arno]

If you have faith in yourself to locate the overlays don't worry about losing streaks. Seems to me you have passed many a juicy winner due to being scared of losing streaks.
Scared money doesn't win money, good handicapping and overlay prices do.

[Dave Schwartz]

CBedo,

IMHO, everthing is relative... to something.

When one considers the risk/reward scenario of a particular race it becomes obvious that a 2/1 horse with a (perceived) 20% advantage should not be bet the same as a 20/1 horse with the same advantage. The fact that these two horses may be in different races does not change that fact.

Thus, my suggestion would be that, while a "solidly profitable" 35/1 horse should be bet, it should be a small bet relative to the size of more "contender-like" bets.

As much as I do not like using Kelly for determining wager size relative to bankroll, I do like using Kelly to determine bet size relative to other bets.

Thus, if the "typical" bet this player makes is (say) 2/1 @ $2.40 $net (10% opt) this horse is a 10-unit bet.

The horse described - 10% hit rate at $72 = $7.20 $net - a huge advantage and perhaps over-estimated. This bet computes to a 7-unit play.

In my world, I would never allow a horse a $7.20 $net. I would assume that the highest possible was around $4.50, which would change the (perceived) hit rate to 6% and the units to around 3.5.


So, now we have (in my mind) a workable strategy. A profitable 2/1 horse is worth 10 units and this longshot overlay is worth about 3-4 units.


Dave

[sjk]

I bet overlays to win or on top of the exacta if I give them better than a 6% chance and on the bottom of the exacta if better than a 3% chance.

[CBedo]

CBedo,

IMHO, everthing is relative... to something.

When one considers the risk/reward scenario of a particular race it becomes obvious that a 2/1 horse with a (perceived) 20% advantage should not be bet the same as a 20/1 horse with the same advantage. The fact that these two horses may be in different races does not change that fact.

Thus, my suggestion would be that, while a "solidly profitable" 35/1 horse should be bet, it should be a small bet relative to the size of more "contender-like" bets.

As much as I do not like using Kelly for determining wager size relative to bankroll, I do like using Kelly to determine bet size relative to other bets.

Thus, if the "typical" bet this player makes is (say) 2/1 @ $2.40 $net (10% opt) this horse is a 10-unit bet.

The horse described - 10% hit rate at $72 = $7.20 $net - a huge advantage and perhaps over-estimated. This bet computes to a 7-unit play.

In my world, I would never allow a horse a $7.20 $net. I would assume that the highest possible was around $4.50, which would change the (perceived) hit rate to 6% and the units to around 3.5.


So, now we have (in my mind) a workable strategy. A profitable 2/1 horse is worth 10 units and this longshot overlay is worth about 3-4 units.


Dave
This makes total sense, and is what I do to a large extent. I use an edge to odds type calculation to determine relative bet sizes.

I also agree on the smoothing. I change the numbers from time to time, but have been using about a $4.00 net max, and also have done the same on the flip side. If I hate a favorite, I assume that overall, it's probably not as bad as I think and upgrade his chances to about a 1.20 net at the worst. The high and low numbers really don't come from a solid data study, just my gut feel.

Dave, from the webinars I've seen of yours, you seem to limit yoru possible bets to the top three or four contenders. So in some sense, aren't you doing what I'm doing and not betting the lower win percentage horses even if they're overlayed? I guess in some weird situations, a horse who has a 8% chance to win might show up in my top 3, but not often.

[DeanT]

Chris,

This is just my 2 cents - I think if you are properly money managing you should make all of those bets. After some record keeping one knows if they are good on horses that are 15-1 and over, and know if they have an edge. The hit rates are low, but if you bet them properly it wont matter one iota. One should never pitch a potential ROI positive bet into the scrap heap, imo.

Example: if a threshold for a bet (the inner risk of the player) is $200 and he/she can not stand to lose more than that, use it is the guide and count it as a unit, say when betting even money shots. So if an even money shot is up and you think it is a good overlay you would bet $200 to win $200. Conversely if a longshot a player likes at 20-1, he would bet $10 on that to win $200.

Imo (and I have been guilty of this for a long time, at stretches) we overbet longshots we like too much and we go on a losing streak and stop, eventhough if we bet sized properly they would be a really good play in our arsenal. Bet sizing longshots properly, for the 'non computer' player can make the game much more fun and one does not have to eliminate them from their playing.

Quite honestly when I see a horse I like at 20-1 to 100-1 I bet them, but I make sure I play them within the bankroll specifications highlighted above. I know I will never be able to measure an edge on those, especially with late odds drops and so forth, so I think Kelly is a bad way to go about such bets. But at the end of the year I would hope an edge was apparent, and using MM to soften the blows, it is something one can do betting LS's year after year, without losing their mind.

[fmolf]

Chris,

This is just my 2 cents - I think if you are properly money managing you should make all of those bets. After some record keeping one knows if they are good on horses that are 15-1 and over, and know if they have an edge. The hit rates are low, but if you bet them properly it wont matter one iota. One should never pitch a potential ROI positive bet into the scrap heap, imo.

Example: if a threshold for a bet (the inner risk of the player) is $200 and he/she can not stand to lose more than that, use it is the guide and count it as a unit, say when betting even money shots. So if an even money shot is up and you think it is a good overlay you would bet $200 to win $200. Conversely if a longshot a player likes at 20-1, he would bet $10 on that to win $200.

Imo (and I have been guilty of this for a long time, at stretches) we overbet longshots we like too much and we go on a losing streak and stop, eventhough if we bet sized properly they would be a really good play in our arsenal. Bet sizing longshots properly, for the 'non computer' player can make the game much more fun and one does not have to eliminate them from their playing.

Quite honestly when I see a horse I like at 20-1 to 100-1 I bet them, but I make sure I play them within the bankroll specifications highlighted above. I know I will never be able to measure an edge on those, especially with late odds drops and so forth, so I think Kelly is a bad way to go about such bets. But at the end of the year I would hope an edge was apparent, and using MM to soften the blows, it is something one can do betting LS's year after year, without losing their mind.
good advice ...personally i would not have bet the horse because i do not consider any horse as a prime bet contender if i have assigned him odds higher than 8/1....i would however make a small watch bet if i had no other contender in the race and thought tha 35/1 seemed inordinately high....it is in my opinion that most long shots 25/1 and higher are underlays and should be higher....

[Valupix]

It either is an overlay in a gamblers eyes or it is an underlay Beyond the fact I am a horrible typist, and in such a hurry I rarely spell check...I realized this isn't completely true. There are horses that are hard to read due usually to the fact there are multiple unknowns about certain factors with the horse.

But I would think most line players wouldn't even pin an overlay/underlay price point to such horses to begin with.