Favorites-Longshots Bias - Horse Racing Forum - PaceAdvantage.Com

BMeadow · 06-28-2002, 01:05 PM

While this used to be the case--favorites lost much less than the track take--it is no longer true.

I think we definitely answered this in the January 2000 issue of Meadow's Racing Monthly (index at www.trpublishing.com). Jim Cramer reported the results of more than 200,000 races with these results:

1st Choice 0.81
2nd Choice 0.82
3rd Choice 0.80
4th Choice 0.79
5th Choice 0.79
6th Choice 0.78
7th Choice 0.76

While it's true that extreme longshots (e.g., 11th or 12th choice) did far worse than the track take, for the most part it's a fair statement to say that the old-time favorites-longshots bias is dead.

Triple Trio · 06-28-2002, 01:32 PM

This is interesting. The pools in Hong Kong have always been very large and the favorites-longshots bias has never existed here. If it's true that the bias doesn't exist in the US anymore, my speculation is that simulcasts have caused the pools to grow larger, making the betting markets more efficient in the process and thus the bias eliminated.

so.cal.fan · 06-28-2002, 01:37 PM

Very interesting facts, Barry and TT.
We all need to think about this. I know I am.

Derek2U · 06-28-2002, 05:44 PM

cant you, barry m., stop advertising your web site & that
quasi newsletter? and as far as cramer & his database of
450 billion races go, so what. He's never written anything
useful either and is only good because (a) he's a nice & honest
guy and (b) he replaced that peewee brain from sport stat.

Rick · 06-28-2002, 06:39 PM

Barry,

It would appear that looking for solid 2nd choices would be the place to find value now. I've seen similar results for 2nd morning line choices as well.

Derek,

What have you contributed that you would consider "useful". Stop being so juvenile and people will respect you more.

BIG HIT · 06-29-2002, 08:59 PM

Hi guys only thing I noticed is if there were a lot of favorites one day. The next day there will be long shots goes like this first day nothing over six to one next day three of the 10 or nine raceswill be long 8 to 1 or higher. For the life of me I can't figure out why but most times it does happen.

ranchwest · 06-30-2002, 05:38 PM

Quote:

Originally posted by BIG HIT
Hi guys only thing I noticed is if there were a lot of favorites one day. The next day there will be long shots goes like this first day nothing over six to one next day three of the 10 or nine raceswill be long 8 to 1 or higher. For the life of me I can't figure out why but most times it does happen.

A lot of times, the racing secretary writes it that way.

Aussieplayer · 07-14-2002, 07:54 PM

Barry/all,

Been meaning to get around to this - this is good research.

Firstly, although this would be from several tracks, what was the average track take?

Secondly, one thing that has me a little confused - surely if you have odds ranges that lose MORE than the track take (in your case it looks like the 4th choice onwards), doesn't there HAVE to be a group that loses LESS?????

Thirdly, we have declining figures in Oz just like you guys. But the first choice loses much less (maybe half the takeout), the second choice a bit less (maybe 75% of the takeout), third choice loses the takeout and then it gets worse from there.

I wonder if that's why oz "pros" advocate finding "true" favs. and having substantial bets on them, as opposed to looking for prices?

A question arises from that thought: are we (I) better off in some way by sticking to the section of the odds that lose less than the takeout? ie. the first couple of favs.? Or, is that one of those elusive concepts that don't actually cut it in "real life?"

Cheers
AP

BMeadow · 07-14-2002, 09:03 PM

"Firstly, although this would be from several tracks, what was the average track take?

Secondly, one thing that has me a little confused - surely if you have odds ranges that lose MORE than the track take (in your case it looks like the 4th choice onwards), doesn't there HAVE to be a group that loses LESS?????

Thirdly, we have declining figures in Oz just like you guys. But the first choice loses much less (maybe half the takeout), the second choice a bit less (maybe 75% of the takeout), third choice loses the takeout and then it gets worse from there.

I wonder if that's why oz "pros" advocate finding "true" favs. and having substantial bets on them, as opposed to looking for prices?

A question arises from that thought: are we (I) better off in some way by sticking to the section of the odds that lose less than the takeout? ie. the first couple of favs.? Or, is that one of those elusive concepts that don't actually cut it in "real life?" "

The tracks are nationwide. The overall return is approximately 0.80, so favorites and second choices and third choices finish about equal to the track take or slightly better, while fourth choices and worse finish slightly worse.

There is no "favorable" odds area. Everybody loses somewhere close to the track take if played indiscriminately.

As always, it comes down to identifying horses which are underbet in comparison to their "true" chances.

Aussieplayer · 07-14-2002, 10:30 PM

Thanks for the reply Barry. One thing I still don't get: if some higher odds lose more than the track take, then surely some odds group MUST lose less????

Cheers
AP

Aristides · 07-16-2002, 01:29 AM

Hi,

First time poster.

The documented favorite/longshot bias that I am familiar with applied only to "heavy favorites." I think about 2/5 or less. The public has always been shown to produce a solid odds line (if you factor in take out) just as your recent data shows. It was only in the case of extreme favorites and longshots that any strong bias or inefficiency was found.

By any chance, could you take a look at the "heavy" favorites using your database and see if that bias still exists.

It would be interesting to see if it does.

Thanks for sharing the data.

BMeadow · 07-16-2002, 01:26 PM

0.20--71.9%W--0.86 ROI
0.40--61.3#W--0.86 ROI
0.60--51.6%W--0.83 ROI
0.80--45.1%W--0.81 ROI

These are EXACT odds, so that if a horse went off at 0.30 - 1, it is not listed above.

There is no odds range close to profitability, though there is some evidence that extreme favorites are underbet and extreme longshots are overbet.

Rick · 07-16-2002, 01:42 PM

Does anyone have figures on last race favorites to place or show? It's pretty well documented that the win ROI improves, but carrying it one step farther to the lower place and show payoffs might prove interesting.

Aristides · 07-16-2002, 11:12 PM

Barry,

Thanks. I don't think it has ever been profitable. If you factor in "breakage," which is more signficant for low-priced favorites, it does look like there is still a bias. Of course, that and and $5 will get you a cup of coffee at Starbucks...

Thanks again.

Chris

entropy · 08-16-2002, 05:44 PM

aussieplayer twice asks the question if some horses lose more than the track take there must be some who lose less.

aren't the returns listed below ALL less than the track take -- country wide.

1st Choice 0.81
2nd Choice 0.82
3rd Choice 0.80
4th Choice 0.79
5th Choice 0.79
6th Choice 0.78
7th Choice 0.76