Baseballs Secret Formula

[sgtcluels]

Just stumbled on this today. Apparently it originally aired in 2006, but it's new to me . It's a show on the Science Channel about the stats and numbers (Mainly Sabre metrics) and it's effect on the game. Interesting stuff, especially about Sac Bunts and it's Runs Scored Probability effects and stealing bases.

 

Plays again at 1am

[trwi7]

Thanks for the heads up. I'll record it and post a recap if anybody wants one.

[Ennder]

I'd enjoy a recap.

[trwi7]

I'd enjoy a recap.

 

I'll try to stay up until 2 this morning but if I can't just hope that it records. If it does record then I'll post a recap sometime tomorrow. http://forum.brewerfan.net/images/smilies/smile.gif

[1992casey]

As a sneak preview, one of my favorite parts of the show was the illustration that (strongly) suggested that a team's #1, #2, #4, and #5 hitters should all be better than its #3 hitter.

[trwi7]

Okay here's my recap. Anybody can feel free to add anything on that I might have missed.

 

-It all started when Bill James was circumspect about the emphasis scouts put on the visible performance of players saying that approach alone is too subjective.

 

-James and other statisticians looking at the biggest possible picture.

 

-James got started with developing Sabermetrics working as a night watchman. He said he would go around the plant once an hour and the rest of the time he would be looking at box scores out of Sporting News.

 

-James' theories were shunned at first until the general managers of the Oakland A's and Boston Red Sox noticed them and started to use them.

 

-Sandy Alderson, now the CEO of the Padres but formerly of the Oakland A's starter using Sabermetrics back in the 1980's. He was a lawyer that was involved in contract negotiations before a front office shake up gave him the GM job. Alderson said something along the lines of he was glad he wasn't involved in baseball at the time because he might have been influenced by the old school theories and then may have never used Sabermetrics.

 

-James estimates that managers make 11,000 decisions in a season.

 

-Doesn't like the sacrifice. He belives that the out you lose may be three times as valuable as the base you gain.

 

-They were showing run expectancy per inning. If you have a runner on 1st with 0 outs you average .953 runs per inning, if the next batter sacrifices the runner to 2nd the run expectancy drops to .725 runs per inning, if he bunts into a double play the run expectancy drops to .117, however, if the batter gets a hit the run expectancy would jump to at least 1.573 runs per inning.

 

-When talking about stolen bases James said if a team hits 10 more doubles, draws 10 more walks or hit two points higher in batting average those all connect to winning except stealing bases.

 

-He uses linear weights formula for trying to see how much stealing bases matters.

 

-It studies the cost of making an out stealing bases when you don't need to. Shows that stealing bases is a "worthless strategy" even at the average of 70% success rate.

 

-James doesn't like intentional walks, even to the best hitters in baseball. He said any batter who gets on base, no matter how has a higher percentage chance of scoring.

 

-They then go on to talk about closers. Alan Schwartz I believe his name is said that if you're winning by two runs in the 9th inning you already have 95% chance of winning no matter what. They went on to say your best reliever would most likely be best used in the 7th inning.

 

-Talked about "similarity scores" which is used to compare players. They compare them by using whatever factor(s) are relevant to the study. That could include batting average, games played, runs, at-bats etc. If the comparison player's score was 850 points or greater they are similar to the benchmark player. Anything below 800 the players are not similar.

 

-As more teams begin to use Sabermetrics advantages are lost and teams are always looking for new ideas (this leads to the new fielding statistics.)

 

-James believes that fielding statistics are the last bastion of Sabermetrics.

 

-James says fielding statistics are so primitive and badly organized that we don't know how many runs a players defense might have cost the team.

 

-They went to Baseball Info Solutions to interview a few of their analysts and to uncover how they determine their fielding statistics. They track every pitch, the location, speed, and trajectory of every hit and fielding play and assess individual responsibility for the results.

 

-The information is coded on specifically developed software that can instantly break out all relationships and averages between balls hit and fielded.

 

-The idea is once each fielder's misplays, errors and successes are calculated, then averaged in comparison to league standards, they would be able to say who was better than the norm and who was worse.

 

Bill James also gave his all-time team. This is what it was.

 

C- Yogi Berra- Because of his consistency in offensive contribution. He also noted that Josh Gibson could have been the best catcher of all-time but they don't have enough data to see.

 

1B- Lou Gehrig- Because he was a better hitter than anybody else.

 

2B- Joe Morgan- Because he had a ton of power and walks. Called Morgan te best percentage player baseball ever had.

 

3B- Mike Schmidt- Because of his power and walks. Also because he was the best defensive 3rd baseman ever.

 

SS- Honus Wagner- "Extremely obvious."

 

LF- Ted Williams- "Extraordinary offense outweighs his shortcomings as a left fielder, base-runner and human being."

 

CF- Willie Mays- Tremendous power, speed and defense.

 

RF- Babe Ruth- Because of his power and walks and his .348 batting average created more run per game, per at-bat than anybody else who ever played the game.

 

SP- Roger Clemens- In his opinion what he accomplished on the mound is more impressive than what anybody else in history has accomplished.

[Ennder]

Thank you. Very interesting stuff.

 

I strongly disagree with him that fielding is the last bastion though. Pitch f/x data is going to be every bit as important over the next 10 years. It lets you break the game down at the pitch level instead of the AB level and that is going to make a huge difference and opens a whole new level of data for sabermetrics to work off of.

[rluzinski]

Bill James readily admits that he hasn't kept up with the latest sabermetric studies, so it's odd that he's always offering his opinion on the matter.

[sgtcluels]

It was released in 2006 Russ. So that might have something to do with it, might have been filmed in 05. Was Mrs. James more active in Sabermetrics at the time?

[Fatter than Joey]

Bill James readily admits that he hasn't kept up with the latest sabermetric studies, so it's odd that he's always offering his opinion on the matter.

 

His opinion is still his opinion -- It's probably like John Madden commenting on what it's like coaching the athletes of today.

[TerraceReserved]

Thank you. Very interesting stuff.

 

I strongly disagree with him that fielding is the last bastion though. Pitch f/x data is going to be every bit as important over the next 10 years. It lets you break the game down at the pitch level instead of the AB level and that is going to make a huge difference and opens a whole new level of data for sabermetrics to work off of.

 

 

I agree. Pitch f/x has so many opportunites, like pitcher/hitter evaluations, and more broad questions about different pitches and stuff. The work done by John Walsh at THT has been particularly fascinating.

[rluzinski]

It was released in 2006 Russ.

 

Naaa, he's still doing that:

 

http://www.insidethebook.com/ee/index.php/site/article/the_future_of_sabermetrics/

 

He's done more for field of sabermetrics than anyone else, so I certainly respect him.