Gamblers Fallacy
 

Have never had a database that made this sort of query easy, just wondering if anyone had ever looked into it with a very large sample size.

Any evidence of the immediately prior race outcome affecting next race betting pools (basically the just ended event causing underestimation or overestimation of repeat occurring)?

i.e.

-----------------------------------------------------------------------------

the roi of projected front runner after prior race front running winner.
vs
the roi of projected front runner following win by closer.

------------------------------------------------------------------------

the roi of pp following that pp win.
vs
the roi of pp following that pp loss.

------------------------------------------------------------------------

the roi of jockey following jockey win.
vs.
the roi of jockey following jockey loss.

-------------------------------------------------------------------------

the roi of trainer following trainer winner.
vs.
the roi of trainer following trainer loss.

------------------------------------------------------------------------

the roi of favorite following favorite prior race winner.
vs.
the roi of favorite following non-favorite prior race winner.

------------------------------------------------------------------------
the roi of longshot horses after longshot prior race winner.
vs.
the roi of longshot horses following favorite prior race winner.


I'd expect them to be weak in most cases, you could string two duplicate priors together to make them stronger I'd guess.

I understand what you're saying, but would such behavior necessarily be a "fallacy"? Are the outcomes of the events truly independent in the same sense that they would be in a game of pure chance? For example, perhaps the outcome of the prior race is indicative of a track bias that would be equally applicable to the next race, and it would be perceptive (not fallacious) to pick up on that, and to take it into account when handicapping and wagering.

I've often wondered this and also don't know if I can actually research it from my DB.

I think it would take more than one race to affect the collective betting psyche, such as when the favorite loses the first 3 or 4 races. People will react differently, some (eroneously) thinking the fav is 'overdue' and others thinking that the track is not playing true to Speed Ratings (FT). As Overlay says the 2nd group may not be wrong.

The Wall Street Journal had an article (a little bit) related to this with investing.

http://online.wsj.com/article/SB1000...toWhatsNewsTop

I don't think the sentiment of the crowd and a streak has an inverse relationship. For example if Baze/Hollendorfer have a losing streak, I don't believe the crowd rationalizes that the next one needs to be bet. I think the crowd will feel the pain and avoids the pair, and that's when a Baze/Hollendorfer play gets +5-1. When I express it this way, I as the individual am falling prey to the fallacy.

The crowd itself doesn't feel the gamblers sentiment. The crowd goes along with the trend, and avoids the bets when on losing streak, and bets when on a win streak. I believe the crowd will always overbet on a win streak or underbet on a losing streaks.

Gamblers Fallacy
 

Back in the days of WPS and DD betting only (no gimmicks), the last race of the day at NYRA would often produce an underbet favorite winner. It was usually a Starter Handicap race, and quite formful; only way to "get out" for the day for many people was to plunge on a longshot, so there was a bias away from the chalk.

Don't have any "database" to prove this, but I guess this comes in as "anecdotal."


Ocala Mike

I agree, that's partially why I think it's interesting. Most dbases are built to treat races as independent events, it's more difficult to get them to answer questions like this. If they aren't truly independent, the win% might change if there really is something meaningful happening, I'm curious how the roi changes (if bettors are correct in whatever their perception about it is or not).

I believe Bill Quirn touched on this in his book, I know this isn't the first time I have heard this about a favorite paying more in the last race of the day.

I have a feeling that these kind of relationships have probably broken down (or weakened, anyway) in this era of simulcasting where you are not necessarily thinking about the last race at a particular track when you've played 4 other tracks in between races. However, if the first four races of the day are all won wire-to-wire on the rail or something like that, people certainly take notice. But does that then make it a good idea to bet the closer? Probably not.

There are times when a major event will influence subsequent events.

Gamblers Fallacy
 

Amen. Not only on the racetrack, i.e., did anyone think the Cubs had even a chance to beat the Marlins in their 2003 Game 7? Preordained.


Ocala Mike