Thread: Article: Where are the Twins at with statistical analysis?

Goin has expressed interest in affixing Pitch F/X cameras through the team’s affiliates to better capture their prospects’ developments. Of course, the cost of installation and an annual fee for having the system may prohibit this from becoming a reality.

How many other organizations have Pitchf/x in all their affiliate parks? It would seem something that owners, and players would agree on, to install Pitchf/x in as many minor league ballparks as possible and share all the data. That would allow orgs to cut down on scouting costs while allowing lesser known players to be more visible to other teams. Are the startup costs such a barrier, or what?

I'm actually completely comfortable with St. Peter's comments to you, ABG. I don't feel like the Twins owe it to anyone to give away their inside secrets. Just because Billy Beane was willing to let an author follow him around and see pretty much everything doesn't mean that's how it should be done. Don't get me wrong. I too want to know all of the inner workings of the front office, but it would be silly for them to do that.

I think it's silly to think that HR, RBI and ERA and Wins are all that the team looks at. If Goin was willing to talk about Pitch F/X, that likely means there are about a dozen or more things that he didn't want to talk about.

I won't pretend to understand the Correia signing, but I'm also not going to pretend I'm smarter than the people in the Twins front office either. I have to believe that they make their decisions based on many different factors, some statistical.

Maybe that's naive, and since they don't share their secrets it's obviously debatable with every transaction/decisions, but again, I have no problem with them keeping their secrets to themselves.

Originally Posted by Seth Stohs

I'm actually completely comfortable with St. Peter's comments to you, ABG. I don't feel like the Twins owe it to anyone to give away their inside secrets. Just because Billy Beane was willing to let an author follow him around and see pretty much everything doesn't mean that's how it should be done. Don't get me wrong. I too want to know all of the inner workings of the front office, but it would be silly for them to do that.

I think it's silly to think that HR, RBI and ERA and Wins are all that the team looks at. If Goin was willing to talk about Pitch F/X, that likely means there are about a dozen or more things that he didn't want to talk about.

I won't pretend to understand the Correia signing, but I'm also not going to pretend I'm smarter than the people in the Twins front office either. I have to believe that they make their decisions based on many different factors, some statistical.

Maybe that's naive, and since they don't share their secrets it's obviously debatable with every transaction/decisions, but again, I have no problem with them keeping their secrets to themselves.

And yet many people posting on TD seem to believe that TR only looks at HR, RBI, ERA, and Wins. If anybody out here really believes that, I have a bridge to sell them.

The article was interesting, but Seth hits it on the head about keeping secrets. We know of one guy with an assistant providing "statistical analysis" to the Twins. Do we really believe a business that size with only 2 guys providing this type of data? We have no clue what they really do in this area. Whatever his faults, TR is extremely intelligent. He could care less that the fans think him a country bumpkin. He would actually probably prefer it.

That Goin and his <strike>staff</strike> intern have to dumb down their analysis for the scouts and translate their findings into scoutting grades suggests that they are still relying primarily on the older model of evaluation. That they haven't invested in pitchf/x equipment in their affiliates suggests that implementing and expanding into new models isn't a priority. Heck, their recent history of pitching signings and extensions suggests at the very least that the decisionmakers don't value the same statistics that other front offices value, the Cardinals included.

It is encouraging to know that they consider the new data including Pitch F/X.

Perhaps analysis led them to believe that pull hitters would be more effective at Target Field and signing Willingham followed the analysis.

Perhaps recent analysis has led them to see the value in power pitchers with the acquisition on May and Meyer.

Perhaps they have chosen to disregard the current research about the value of defense and defensive metrics. While the other teams strive to improve their defense, the Twins have loaded the roster with some of the poorest defenders in baseball in Willingham, Doumit, Plouffe and Parmelee. Their bats may make them look like a bargain. However if the Twins want to get back any value in trade, they have to find another team that doesn't value defense. Doumit and Willingham might be a tough sell to NL teams.

Perhaps analysis has led them to believe that Doumit and Willingham will not decline and instead maintain their current level of play through the end of their contracts in 2014.

Perhaps their analysis has led them to a gem and elite fielder in Florimon where no other team saw the value. No one wanted Mastroianni? The Twins see a starting center fielder and lead off hitter.

Many words have been written about the analysis that led them to the current construction of the starting rotation.

I hope that the Twins are correct. The fastest way out of this hole is to find contributing players among those rejected by other teams. Do you trust the Twins management to lead us out of the hole?

I think "dumbing it down" is perhaps a bit strong, but yeah... Goin does have to take the analysis and communicate the results in a way that Terry Ryan and others in senior management with the Twins understand. That's hardly unusual. My staff does the same thing for me because I don't understand all of the details of the analysis they do in their areas of expertise. Similarly, I take what's important and communicate it to my boss in ways he'll understand. If Ryan didn't need someone to translate the data in to ideas he can understand, he could simply run the data himself. The tough part isn't running data, it's figuring out what it means AND communicating the meaning to those who need it.

If Goin isn't very strong with those translation and communication skills, his difficult task becomes impossible.

I generally think using statistical analysis when evaluating players is very overrated, especially by fans on chat boards. I found it interesting that the example he used was pitch f/x which doesn't strike me as stats per say but more of a way to enhance scouting analysis. Statistics at their best work to confirm or build upon what is noted through traditional scouting. I would be quite surprised if any team in baseball used statistics as a leader for analyzing players.

I don't really find the examples cited by people all that convincing in suggesting that the Twins are ignorant of stats when it comes to evaluating players, more that they are cheap, and only willing to spend a limited amount in free agency. This is of course a different issue.

To me, the area where the Twins should be more scrutinized is general strategy and use of assets. The example I would use is the draft. The team had a unique chance, based upon the amount of picks they had, to outspend all other teams significantly last draft but they chose to pass and leave some money on the table. The team wasn't cheap in the draft, they outspent every other team, but did miss an opportunity to spend more to acquire talent (perhaps by selecting some additional overslot guys after round 10). Other teams, such as Toronto/Boston/Houston seemed to be more strategic in how they allocated funds. The reason I think people are frustrated with the Twins is because they tend to be more conservative and methodical as opposed to some other teams that appear more progressive and creative. Both styles can be successful (the Twins had a pretty damn good run for a decade), but obviously one is generally more appealing to chat boards like this. There are strengths in the Twins way but it is not nearly as sexy.

After everything is stripped away it still comes down to evaluating talent. The Twins are conservative in statistical analysis but I think there is a lot of derp derp in the complaints. They use all information that they can, but probably trust their scouts above all else.

As a side note I like that he cited the Cardinals as the example, they are model franchise in my mind.

Originally Posted by Shane Wahl

Any guesses on who the pitcher (and his sinker) under scrutiny was?

B-Webb???????

Pelfrey probably...


Interesting stuff.

The big question is: Who is in the driver's seat and who is in a supporting role:

- Are the scouts driving and the analytics people supporting them?
- Is it 50-50%?
- Are the analytics folks driving and the scouts supporting them?

Methinks that (based on the number of the staff - ie. the "dedication" of the organization to analytics - a guy and an intern-kid, compared to that of the number of scouts and the background of the people in the decision making positions) this is all for show: They got a guy because they did not want to be seen that they did not have one like the other organizations do. Unless they invest in it, data analytics capabilities do not grow on trees.

Originally Posted by Andrew Bryz-Gornia

This seems to agree with what I've been told by Dave St. Peter... the Twins believe they are middle of the road when it comes to advanced statistics.

One thing I had St. Peter tell me is that the organization does not want other teams to know what they use. He doesn't mind that the fans think this team's sabermetrics is counting home runs and looking at ERA, because he feels that if the team has an advantage, they are not going to broadcast it to the world a la Billy Beane and Moneyball.

I wouldn't mind this if this were in evidence by most of their moves. However, last year's presented a number of red flags in this area from the Jason Marquis signing (and now Correia) evaluation of players using spring training numbers (ex. Jeff Gray)....

Pitch F/X when I have looked at it on fangraphs is not so much about a statistics as much as it looking at location,or speed of pitch. That is a much more usefull thing to know than FIP as it would show you the quality of pitches thrown.

Brooks Baseball · Home of the PitchFX Tool - The Front Page

A great pitchf/x resource.

Originally Posted by Willihammer

How many other organizations have Pitchf/x in all their affiliate parks? It would seem something that owners, and players would agree on, to install Pitchf/x in as many minor league ballparks as possible and share all the data. That would allow orgs to cut down on scouting costs while allowing lesser known players to be more visible to other teams. Are the startup costs such a barrier, or what?

"or what" is more like it. The startup costs are < $50-100 K a ballpark and, even though most affiliates will whine about it, the MLB clubs can foot the bill (and write it off) in a jiffy.

You mention "owners" and "players". The big issue is the party you don't mention and the one that will get the most exposed by it: Umpires.
PitchF/X knows the strike zone and can show how bad home plate umpires suck
It is all politics

This article shows the twins still know nothing about advanced statistics.Pitch f/x is a tool just like a radar gun. But it has nothing to do with advanced metrics.

Embarrassing.

Originally Posted by clutterheart

Pitch f/x is a tool just like a radar gun.

Its a little more than that. Here is the pitchf/x data for one pitcher for one game

http://www.brooksbaseball.net/pfxVB/...700&v_size=500

edit: and here is a key: http://webusers.npl.illinois.edu/~a-...b/tracking.htm

The SABR stuff has become glamorized. It's a bit overrated by us chatters.

Your run-of-the-mill data analysts DO practically grow on trees these days. We're all proof of that. We have two dozen contributors to TD who are pouring through mountains of statistics, using all the latest software technology, and then communicating the results quite succinctly to anyone who will listen. For free.

Statistics verify the story. They don't tell it. Give me the qualitative assessment, often hugely enriched by experience-based intuition, from scouts. Verify their story with a look at the statistical stuff.

I mean, you guys are great and all, but...

Originally Posted by birdwatcher

The SABR stuff has become glamorized. It's a bit overrated by us chatters.

Your run-of-the-mill data analysts DO practically grow on trees these days. We're all proof of that. We have two dozen contributors to TD who are pouring through mountains of statistics, using all the latest software technology, and then communicating the results quite succinctly to anyone who will listen. For free.

Statistics verify the story. They don't tell it. Give me the qualitative assessment, often hugely enriched by experience-based intuition, from scouts. Verify their story with a look at the statistical stuff.

I mean, you guys are great and all, but...

Yes. I think the reason it is very overrated by us chatters is that it allows us to tell the story of a player when we have watched him only a handful of times at most as opposed to scouts who will watch many repetitions of the same thing.

I find it pretty funny that just because the Twins front office doesn't speak in the same language that writers here do that all of a sudden the Twins don't know baseball or understand what they do. Statistical analysis basically breaks down everything to this:
Pitching - maximize Ks, minimize BBs, shoot for more GBs and less FBs
Hitting - don't swing at pitches outside the strike zone and put a good swing on those in the strike zone - then determine appropriate tradeoffs between contact rate and pop
Defense - large amounts of range, the ability to catch the ball, and strong, accurate arms = good
Baserunning - take extra bases when possible at an appropriate risk

I am actually much more interested that the Twins are using pitch f/x and the like to enhance scouting operations than I would be in finding out they have multitudes of people crunching numbers to find new and creative ways to say the above.

Well hell, if that's all there is to it why would any team ever bother?

Originally Posted by clutterheart

This article shows the twins still know nothing about advanced statistics.Pitch f/x is a tool just like a radar gun. But it has nothing to do with advanced metrics.

Embarrassing.

What "advanced" statistic has predictive value. They only describe what happened for that year.

Thus to you it is insignificant to know what the pitches thrown do. It is far more useful to know the skills of the player than just the outcome.

Embarrassing is that people still do not get the notion that the front offices do not talk about the methods they use to analyze players. People like Beane do it for whatever reason. The 14 people on this board who complain have no clue if the Twins use metrics or not. They make the assumption that because they don't talk about it it is not done.

Originally Posted by snepp

Well hell, if that's all there is to it why would any team ever bother?

I still think there is huge value in it that is being exploited by more progressive clubs in terms of in game strategy such as defensive shifts, matchups, etc. This is where I think the Twins have great room for improvement, though I'm not sure if there is even thought of using more advanced statistics in that matter, which is a shame.

I also think that it is useful as an enhancement to scouting (which is how the Twins seem to most be utilizing it), but I would still say it is vastly overrated as a primary method of player evaluation - which is where much of the weeping and gnashing of teeth as it concerns the Twins front office is found.