New MLB data collection/analysis

[JB_Iowa]

http://www.mlb.com/news/article/mlbam-introduces-new-way-to-analyze-every-play?content_id=68514514&partnerId=as_mlb_20140301_19352324&ymd=20140301

 

This sounds pretty amazing .... and exciting that TF is one of the first.

[amjgt]

Time to get a Tampa Bay Ray's-size analytics team.

[Thrylos]

Data is always good to have, and the more the better, especially in the depth this effort seems to be. However, the biggest issue is to know what questions you need to answer with this data and how to crunch it to answer them. I think that this would be a challenge.

[biggentleben]

Time to get a Tampa Bay Ray's-size analytics team.

 

Thrylos mentioned it as well, but a lot of people want to go from 0 to 60 in analytics. It's just not that simple. You can hire the perfect team and find that it makes all of about a half game of difference until you find exactly the information you need and craft a roster to meet what you find. The Pirates started making some big strides about 4 years ago in analytics, and while they had improvements before last year, last season was their first season about .500 after making many moves to bring in players that fit what they thought they could do. The Pirates were actually the team that shifted the most in 2013.

[ashbury]

a lot of people want to go from 0 to 60 in analytics.

 

The same argument holds for teams that think they will hang back until the dust settles. They won't be able to go from 0 to 60 when they decide to.

 

Now, if we agree that nobody is proposing to stay at 0, and nobody is proposing to aim at 60 right now, maybe there is some middle ground to discuss. An increased investment in an analytics department doesn't mean you suddenly run the team by computer autopilot.

[Willihammer]

"Just on the field, with the coaching staff and the manager -- when you start to look at positioning, and you start to see the exit velocity of the ball coming off the bat, and is he late or is he ahead of a lot of pitches, and then you move your infielders and outfielders accordingly," Duquette said. "Along with the spray charts that are already used. A ball goes into the gap and an outfielder picks the ball up, sometimes with his bare hand, and he throws it into the infielder, and the infielder gets rid of the ball as quickly as he can, and makes a strong, accurate throw to home plate to nail a runner. You can evaluate and measure each of those points in the relay.

 

Will be interesting to follow what the data says about relays vis a vis teams that shift and those that dont. That has been one of Bill James' criticisms - that shifts take players out of position to relay a throw.

[biggentleben]

The same argument holds for teams that think they will hang back until the dust settles. They won't be able to go from 0 to 60 when they decide to.

 

Now, if we agree that nobody is proposing to stay at 0, and nobody is proposing to aim at 60 right now, maybe there is some middle ground to discuss. An increased investment in an analytics department doesn't mean you suddenly run the team by computer autopilot.

 

Yeah, I'm not saying they should stay back and do nothing, but simply hiring 100 staffers for the point of hiring 100 staffers is not exactly progress for a team. The Pirates 12 months ago were in the midst of being bashed for odd training methods as they were trying different things along the way. Interesting no one mentioned that as they made their first playoff appearance since Jim Leyland was managing them.

[ashbury]

hiring 100 staffers

 

Not sure that was quite the middle ground I was hoping for.

[The Wise One]

What I gather from this is thata it will allow a team to better analyze defensive capabilities. Reaction time can accurately be measured, Judgemnt on what balls the player should get to based on the player's speed and quickness to recognize what is happening on the play. It could make defensive metrics more meaningful.

Velocity on the ball being hit has been able to be tracked by Trackman for quite some time, 2008 or 09. The new system sounds like it can track the players movement and speed in conjunction with the ball. That is something Trackman couldn't do.

In terms of staffing for data analysis. A computer interfaced with the data collection device and a programmer to write the software. You need someone who understands what the data means. Why would you need 100s? Even 10?

[Seth Stohs]

This is pretty remarkable... great information... however, I don't see this need to hire a ton of people. Most of the data is out there already. It just needs to be analyzed and then put into action. The Twins have at least 3, and probably a few more, stats people... I don't know how many more they would need.

[biggentleben]

What I gather from this is thata it will allow a team to better analyze defensive capabilities. Reaction time can accurately be measured, Judgemnt on what balls the player should get to based on the player's speed and quickness to recognize what is happening on the play. It could make defensive metrics more meaningful.

Velocity on the ball being hit has been able to be tracked by Trackman for quite some time, 2008 or 09. The new system sounds like it can track the players movement and speed in conjunction with the ball. That is something Trackman couldn't do.

In terms of staffing for data analysis. A computer interfaced with the data collection device and a programmer to write the software. You need someone who understands what the data means. Why would you need 100s? Even 10?

 

I agree, but there are arguments out there that "ZOMG, there are only 5 people looking at these stats, but 55 scouts on staff?! Obviously Team X hasn't bought into analysis!" That sort of thinking isn't educated in understanding the nuances of what's going on, but for the casual fan, it's what they see.

[JB_Iowa]

I thought that this column was timely: http://msn.foxsports.com/north/story/molitor-working-to-improve-twins-defensive-positioning-030214

 

While there's no indication that the new analysis played into what Molitor saw, it is clear that Goin and his team will be able to pull up even more video and other data for Molitor and other coaches.

 

In the long run, that's what this is about.

[cmathewson]

I read this article, and I am a data scientist. But I don't understand how this is different from what we already have. Does it introduce new data (like the much needed batted-ball vector info.)? Or is it just a new way of presenting existing data to the audience? The article didn't make this clear.

[snepp]

http://www.fangraphs.com/blogs/mlbam-announces-new-data-stream-the-future-is-almost-here/

[eLee612]

I agree, but there are arguments out there that "ZOMG, there are only 5 people looking at these stats, but 55 scouts on staff?! Obviously Team X hasn't bought into analysis!" That sort of thinking isn't educated in understanding the nuances of what's going on, but for the casual fan, it's what they see.

 

For the casual fan, what they see is the bottom line: is it getting us anywhere positive? This is obviously going to take time to implement, analyze, and then figure out what the heck to do with it.

 

In the mean time… the rest of us can sit back and number crunch all we want. (Considering that we probably would have been anyways, with or without the new data.)

[Oxtung]

I'm excited about this new data. With a players acceleration, top speed and reaction time we will finally have some concrete numbers to work with that don't depend on a human's best guess. If a players glove to throw time (like pop to pop for catchers) is available and their velocities on their throws we can finally start to get a very clear understanding on just how valuable defense actually is.

 

As the FanGraph's article says, we don't know how public this information will be. I sure hope it becomes publicly available. I think both the teams and the fans gain when that happens.