Sabermetrics

[Actor]

After seeing Moneyball I'm wondering if there is anything in Sabermetrics that could apply to handicapping? Or does Sabermetrics simply apply handicapping methods to baseball? Is there a good book on Sabermetrics?

[Robert Fischer]

good question

[Sapio]

Originally Posted by Actor

After seeing Moneyball I'm wondering if there is anything in Sabermetrics that could apply to handicapping? Or does Sabermetrics simply apply handicapping methods to baseball? Is there a good book on Sabermetrics?

Hi Actor

The commonality between Sabermetrics and Horse Racing is Bayesian Hierarchical Models.

Thomas Sapio

[BettinBilly]

I have a guy whom will definitely know. He's my S.O.'s Brother. He's a highly functional Autistic Savant, and Handicaps Baseball to the 100th power. I recently just turned him on to Thoroughbred Racing as well - told him there are more stats in Racing than there are in Baseball. That did the trick.

He is sometimes hard to reach. Especially if he's Handicapping. If no one has answered this question by the time I reach him, I'll post what he tells me. If there is a book on the subject, he will know, and most likely have read it.

[TravisVOX]

There are lots of books that deal with baseball statistics... some are really intense mathematical texts. Moneyball - the book itself - is pretty good. Another that is interesting is Trading Bases: How a Wall Street Trader Made a Fortune Betting on Baseball by Joe Peta.

[thaskalos]

http://www.amazon.com/gp/product/078...=ATVPDKIKX0DER

I bought this book a couple of days ago...and I am eagerly awaiting its arrival.

[Robert Fischer]

SABR =Society for American Baseball Research

so technically we need a new acronym.

SAHRR? Is it contagious? not as cool sounding as a saber?


Another thing about the Sabermetrics is that they tried to make it a rule to only use objective data.

[dannyhill]

Originally Posted by Robert Fischer

SABR =Society for American Baseball Research

so technically we need a new acronym.

SAHRR? Is it contagious? not as cool sounding as a saber?


Another thing about the Sabermetrics is that they tried to make it a rule to only use objective data.

Actually that is what horse racing needs, a saber to the heart.
Only thing that will save this game is starting over!

[pandy]

I found the movie and the theory interesting. There are certain things about sports, particularly baseball, that I never thought made sense. For instance, sacrificing a runner to second when there are no outs and the batter is a regular player. I can understand when the pitcher is hitting, but many managers let a good hitter bunt in that situation when it's late in the game. How can you possibly have a better chance of scoring if you have 2 outs to work with instead of three?

[Dark Horse]

Even today, with Moneyball to explain their success, the Oakland A's continue to be one of the best bets in baseball (check under 'money'):

http://www.covers.com/pageLoader/pag...standings.html

A good book is actually called the Book. And Tango Tiger has a website with a run estimator based on that research.

[raybo]

Originally Posted by pandy

I found the movie and the theory interesting. There are certain things about sports, particularly baseball, that I never thought made sense. For instance, sacrificing a runner to second when there are no outs and the batter is a regular player.

With no outs, if you can scarifice a runner to 2nd then you take the double play out of the picture and you have at least 2 at bats to get an outfield hit to get the runner on 2nd home. Basic fundamental baseball strategy.

Originally Posted by pandy

How can you possibly have a better chance of scoring if you have 2 outs to work with instead of three?

Because the double play is removed from the equation. Any ground ball, or line drive to an infielder could start the double play, then you have no men on base and only 1 out to go. Again, just basic strategy and playing the probabilities. The key to manufacturing runs is to get a runner in scoring position first, regardless of how you do it. With no outs you have the best possibility of scoring because you have more chances and more options. Of course much of it depends on what point you are in the batting order, if one of your best hitters is up, or your power is up, then you probably let them hit away, depending on the situation on the scoreboard and what point in the game you are at.

[raybo]

I loved the movie and have watched it 5 or 6 times, at least. Great acting and lots of baseball stuff that you don't normally get exposed to. Baseball was always my number 1 game so I love almost any movie related to baseball. I also, was very interested in the technical maneuvering and computer analysis involved in putting the right players on the field and manufacturing runs. The two main characters were both right, the key to winning is producing runs, and the way you produce runs is by getting men on base. That's fundamental baseball, at its core. Everything else is just attempts to enhance that fundamental core, or trying to reduce the same fundamental core degree in your opponents, through pitching and defense, many times to the detriment of your bottom line and your placing at the end of the season.

Can you buy championship teams? Yes, of course, but at a very high price, that is always in jeopardy to injuries, slumps, egos, and bad luck, etc..

[reckless]

As a long suffering Mets fan, if sabermetrics could be applied to handicapping the ponies I want to be the first to nominate Sandy Alderson as the worst handicapper of all time.

[raybo]

I think the big difference between using stats in baseball (or other team sports) and stats in horse racing, is that the individual participants in baseball have many more "events" from which to collect data. In horse racing that is not the case, so the ability to define percentages for different categories of performance is much lower than a sport where the participants can be involved in 162 regular season "events", plus post season play, versus 10 or 12 "events" in horse racing, per year. And lifetime stats are even more in favor of baseball, as the players can play many years, while horses are racing fewer and fewer years as time goes by.

So, what we are forced to do in racing is group similar horses together in order to collect large enough sample sizes to be meaningful. But, by grouping them together you dilute the individual performances of individual horses, and it is individual horses that we usually bet on.

Thus, in baseball, you have teams that have many individual participants, with many "events" each, which allows for a great amount of meaningful data collection, both individually, and as a team. Not so for horses. Night and day.

[Cratos]

Originally Posted by Sapio

Hi Actor

The commonality between Sabermetrics and Horse Racing is Bayesian Hierarchical Models.

Thomas Sapio

I believe that the BHM is appropriate, but not for the everyday horseplayer.

The BHM needs an approach to horseracing the way Nate Silver approached political elections predictions with sabermetrics.

However I believe the major obstacle in constructing this model will be data integrity.