The Twins don't use advanced metrics

[Craig Arko]

The Twins do the have #1 rated farm system. What were you hoping for?

 

A pony?

[old nurse]

But I've beaten that horse to a pulp, but it bugs me when people rip Smith for what he inherited.

 

Now Mike, really. Smith had All Stars he inherited. He had draft choices from Hunter signing elsewhere. He had top 100 prospects in the minor leagues. Yes Blackburn and Slowey were top 100 prospects at one point, as well as many others during his tenure. All these players that were top rated should have been traded for a few up and coming players. How should have Smith known this. Easy, he didn't use advanced metrics.

[TheLeviathan]

My god it seems impossible around here to just give due credit and due blame. This whole discussion wouldn't exist if just a few people got off the idea that some one person is free of blame/worthy of credit.

 

As was said earlier, the Twins are here due to the actions of both men. Their success is also largely dependent on moves both men made to Improve the team long term. You don't need any advanced metric to deduce that.

[jay]

Reviving this thread a few days later... I stumbled across this article this morning that talks about the Cardinals and how they use advanced metrics. I've derided overly praising teams that are currently successful (remember when the Twins were a model franchise?) because it can be rather cyclical. However, this article upped my level of respect for the Cardinals by another notch or two. A GM with stats books on his desk. A manager that actually gives a care about metrics. I thought I could only dream such a thing.

 

How the Cardinals use sabermetrics - SweetSpot Blog - ESPN

 

These quotes specifically are exactly what I pain for out of the Twins and I wish we had a better understanding of what they are doing behind the scenes:

 

"Most importantly, we hire a manager to make that lineup. I do think one thing that Mike and his staff have done a very good job of is embracing anything we can put together as far as advanced scouting for them. Trying to eliminate small sample sizes and make them accept larger ones for probabilities has been helpful. Mike, he is a young manager that is very interested at looking at the best ways to be successful, so that's always a good sign when you have that in an employee."

 

Oh, so you mean it doesn't actually matter when a guy is 5 for 7 lifetime off a pitcher?

 

"When evaluating a player such as outfielder Oscar Taveras, a top prospect in the Cardinals organization and one of the best in baseball, there are several prongs to making a decision on what turns a prospect into a major league player. Mozeliak says you can think of the process as legs to a table... One aspect would be the scouting, such as bat speed or a player's mechanics. A second is statistics. For the Cardinals, another important part is work ethic and character. "The character component of our decision-making does not account for anything in the analytical world," said Mozeliak. "We don't put it in the algorithm, because it is just not there,...""

 

...just the fact that they even have an algorithm. *swoon*

[old nurse]

Three of the Cardinals criteria for a player there really isn't a metric for. Work ethic, character and mechanics.

from jay's article

When you start thinking about things like the aging curve -- that is something I think 20, 30 years ago when players were signing contracts, they were signing contracts for what they accomplished -- now people are signing contracts for what you expect them to do.

Do you think that would explain Ryan's distaste for long term free agent contracts?

The GM has been in charge since 2007. The article touts his use of stats in assessing the draft candidates. They say they prefer college players. Since the new GM took over they have only had 2 impact players come through the draft, Lance Lynn and Shelby Miller ( a HS prospect). Their best prospect is a player they signed when he was 16. Metrics are important, but they are not the answer. Even the GM will tell you that.

 

They have their own statistics. They did not share what they do. No one can point to an article to what the Twins do or don't do statistics wise

[Mike Sixel]

Other than Ryan's interview here, and that they have 1 stats guy, and that no one has ever written an article about the Twins "new school" approach to anything, we do know nothing.

 

I think the article points out what many of us have been saying, it is not "either or", but both, that make a team successful.

 

I'm not sure anyone here has ever said metrics are the only answer, just as no one has said FA is the only answer. As a matter of fact, I'm certain no one here has said either of those.

[TheLeviathan]

However, this article upped my level of respect for the Cardinals by another notch or two.

 

That's a really good article, thank you for sharing.

 

The beauty of this is that it exemplifies that you don't have to throw the baby out with the bath water. Scouting is still an integral part of their process. Character is still integral. But they've also incorporated another important aspect in order to make themselves better.

 

The Cardinals should be the model for the Twins in every aspect of how they run their operation. First class, smart, and (perhaps the part we need to learn most) how to be innovative and flexible rather than antiquated and rigid.

[old nurse]

Other than Ryan's interview here, and that they have 1 stats guy, and that no one has ever written an article about the Twins "new school" approach to anything, we do know nothing.

 

I think the article points out what many of us have been saying, it is not "either or", but both, that make a team successful.

 

I'm not sure anyone here has ever said metrics are the only answer, just as no one has said FA is the only answer. As a matter of fact, I'm certain no one here has said either of those.

 

to quote you

.do people expect paragrahps in every post?

No one has ever written an article documenting the old school approach either. No one really knows what steps the Twins take in selecting and scouting players. The department of statistics collection could be in the hands of the scouting department. That would make some team's statistic departments smaller. If they are considered part of scouting, that would make the stats department smaller. The biggest friend is the computer.

The pitch monitoring systems like pitch f/x are technology, not metrics. They give you the same data the scout would with a radar gun and eyes. The new technology is probably more accurate. It is not known how much the Twins use this data. I do not know but there had to be some reason for Ryan to come up with the comments he did. Specifically about Correia pitching better than his numbers and the saw something in Worhley they could improve to make his numbers better. I could only assume it meant they liked something in how the ball moves when they throw it that in their model could lead to better success for the pitchers.

 

Any of the advanced metrics are all calculations off from the basics of the game. Don't you think a good GM would inherently know that a pitcher that walks few, strikes out many and doesn't give up home runs might be better than his ERA indicates?

[jay]

They have their own statistics. They did not share what they do. No one can point to an article to what the Twins do or don't do statistics wise

 

This is true, but it does seems more reasonable to reach the conclusion the Twins aren't as deep into it as others. Unless TR is playing one heck of a poker game, his quotes (and others such as Anthony) on advanced metrics certainly don't indicate an embrace for these new things as they clearly do in the article with Mozeliak. It'd be hard to believe the one-man stats team has developed a full database and assessment algorithm.

 

I'd probably laugh if someone could say with a straight face that Gardy even *possibly* might have embraced anything near what they say Matheny has.

[Mr. Brooks]

to quote you

.do people expect paragrahps in every post?

No one has ever written an article documenting the old school approach either. No one really knows what steps the Twins take in selecting and scouting players. The department of statistics collection could be in the hands of the scouting department. That would make some team's statistic departments smaller. If they are considered part of scouting, that would make the stats department smaller. The biggest friend is the computer.

The pitch monitoring systems like pitch f/x are technology, not metrics. They give you the same data the scout would with a radar gun and eyes. The new technology is probably more accurate. It is not known how much the Twins use this data. I do not know but there had to be some reason for Ryan to come up with the comments he did. Specifically about Correia pitching better than his numbers and the saw something in Worhley they could improve to make his numbers better. I could only assume it meant they liked something in how the ball moves when they throw it that in their model could lead to better success for the pitchers.

 

Any of the advanced metrics are all calculations off from the basics of the game. Don't you think a good GM would inherently know that a pitcher that walks few, strikes out many and doesn't give up home runs might be better than his ERA indicates?

 

Given the historic lack of K's from our starting rotation, I would say that no, obviously Terry Ryan does not understand that, at all.

[Otto von Ballpark]

Three of the Cardinals criteria for a player there really isn't a metric for. Work ethic, character and mechanics.

 

Actually, there is some advanced data on mechanics. Especially for pitchers. Just look up the Trackman system, how it was originally used for golf and now baseball, tracking not just the ball but the pitcher's body.

 

ESPN recently noted that 17 MLB teams are using it. Wonder if the Twins are among those...?

[old nurse]

This is true, but it does seems more reasonable to reach the conclusion the Twins aren't as deep into it as others. Unless TR is playing one heck of a poker game, his quotes (and others such as Anthony) on advanced metrics certainly don't indicate an embrace for these new things as they clearly do in the article with Mozeliak. It'd be hard to believe the one-man stats team has developed a full database and assessment algorithm.

 

I'd probably laugh if someone could say with a straight face that Gardy even *possibly* might have embraced anything near what they say Matheny has.

 

Goins doesn't really have to do much from a maintaining database standpoint as information from games would be fed into a computer through statistics from a variety of sources. Anything specialized from a scout would be fed in through a spreadsheet like Excell. IT to write the programs is the IT department.

[old nurse]

Given the historic lack of K's from our starting rotation, I would say that no, obviously Terry Ryan does not understand that, at all.

Given the math, strikeouts are less likely to improve the ERA as limiting walks and strikeouts. Baseball prospectus on a 20 year study found the r2 value to be .25. Strikeouts important, yes. More definitive when you get above a K% of 20. Finding those pitchers is a little difficult.

[old nurse]

Actually, there is some advanced data on mechanics. Especially for pitchers. Just look up the Trackman system, how it was originally used for golf and now baseball, tracking not just the ball but the pitcher's body.

 

ESPN recently noted that 17 MLB teams are using it. Wonder if the Twins are among those...?

 

Are there any statistics relevant to movement and outcome in terms of production on the playing field?

[Mr. Brooks]

Given the math, strikeouts are less likely to improve the ERA as limiting walks and strikeouts. Baseball prospectus on a 20 year study found the r2 value to be .25. Strikeouts important, yes. More definitive when you get above a K% of 20. Finding those pitchers is a little difficult.

 

But why is it more difficult for us than it is for all the other 29 teams? And for that matter just about any other team in the modern era of baseball?

[jokin]

Actually, there is some advanced data on mechanics. Especially for pitchers. Just look up the Trackman system, how it was originally used for golf and now baseball, tracking not just the ball but the pitcher's body.

 

ESPN recently noted that 17 MLB teams are using it. Wonder if the Twins are among those...?

 

I'm guessing the Twins passed on that option as they probably thought Trackman had something to do with Pacman, determining that it too, is no longer compatible with their Atari 2600 system.

[ThePuck]

I'm guessing the Twins passed on that option as they probably thought Trackman had something to do with Pacman, determining that it too, is no longer compatible with their Atari 2600 system.

 

hey now, nothing wrong with Atari 2600... :-)

[old nurse]

But why is it more difficult for us than it is for all the other 29 teams? And for that matter just about any other team in the modern era of baseball?

Ghost of homerdome?

[Mr. Brooks]

Ghost of homerdome?

 

Could you please clarify? Maybe I'm just dense, but I'm not quite sure where you are going here.

[jokin]

Given the historic lack of K's from our starting rotation, I would say that no, obviously Terry Ryan does not understand that, at all.

 

This is an old POV, but I think still lingers in the Twins way of thinking- it goes along the lines in the theory that the Twins originally embraced P2C for economic reasons as their only fiscally viable way to remain competitive during, and right after, the contraction era. K-heavy pitchers are generally more in demand, thus driving up their price of retention, and riskier, as they break down more easily.

 

Recent history should raise alarm bells, league average K/9 rates have steadily risen since 2005, from 6.38K/9 then, to 7.56K/9 today. Meanwhile, the Twins peaked in their K/9 rate in 2006 at 7.28 and have steadily and continually dropped in the waning years of the 00s, to finally cratering to the worst in the league over the last 3 years at right around 6.0K/9.

 

And as Puck pointed out, the K/9 gap in 2013 between 29th place and 30th place is yawning, and an even more distinct point of distinction for the Starting Pitchers. In 2006, the Twins SP had a K/9 of 6.83 vs. a League Average K/9 of 6.21. In 2013, the Twins SP K/9 is TWO FULL Ks lower/9IP, at 4.83, vs a League Average K/9 of 7.15.

 

Hope for trends changing in the minors? Yes, Meyer, May and Darnell have nice K/9 numbers this year. But in taking a look at the top 3 Twins farm clubs and strikeouts and K/9 rates, the AAA club ranks last, and both the AA and A+ clubs are 9th (out of 12).

[jorgenswest]

It seems that the Twins shifted directions on strike outs with the 2012 draft and then the following winter with the acquisitions of May and Meyer. I wonder of those added to the minor leagues starting with that draft tend to be closer to their league averages in strike outs.

 

Let me acknowledge that the signings of Correia and Pelfrey do not fit this plan. They also aren't in the long term outlook. Correia had the reputation of staying healthy which was a need given that every starter had been injured in the previous year. He has stayed healthy and been the only constant this year. Pelfrey was a gamble that didn't work out.

 

My interest would be studying the acquisitions minor leaguers in the draft and trades. Are we seeing more guys with above their league average strike out totals. Is it a coincidence or change in direction?

[Otto von Ballpark]

Are there any statistics relevant to movement and outcome in terms of production on the playing field?

 

This is proprietary stuff, but yeah, I would guess that being able to quantify a pitcher's mechanics and "stuff" could easily help identify good pitchers, healthier pitchers, good prospects, and optimal roles for them.

 

This is the kind of thing that should be a no-brainer when you have the worst starting staff in the league. What do you have to lose by trying to gain a little edge in this fashion?

[old nurse]

This is proprietary stuff, but yeah, I would guess that being able to quantify a pitcher's mechanics and "stuff" could easily help identify good pitchers, healthier pitchers, good prospects, and optimal roles for them.

 

This is the kind of thing that should be a no-brainer when you have the worst starting staff in the league. What do you have to lose by trying to gain a little edge in this fashion?

 

That would only be helpful if you could run prospects through some sort of combine to utilize the technology to identify players. To detect potential injuries you would have to run someone through an MRI that works with movement. That technology to my knowledge does not exist. The greatest potential for obtaining the best players is still through the draft and international signing. Nobody has shown a system of advanced metrics works in identifying those players.

Trackmaster to me is best in developing your scouting book on what pitch players hit and where it goes. I think the information would be easily downloaded into your computer.

[Oxtung]

The greatest potential for obtaining the best players is still through the draft and international signing. Nobody has shown a system of advanced metrics works in identifying those players.

 

Of course we don't know of any advanced metrics that work. No team is going to come out and brag about their algorithms and given their competitors an advantage. You seem to think that these teams are hiring dozens of statisticians, some extremely highly regarded, just to look at fangraphs daily. These guys are creating their own metrics and computer models and we won't know what those are for years. If the twins wait that long to jump on board.... well they'll reap the rewards and I don't think any of us will like it.

[TheLeviathan]

The greatest potential for obtaining the best players is still through the draft and international signing. Nobody has shown a system of advanced metrics works in identifying those players.

 

Sounds like an excellent place to gain an advantage if we can be the first to do that. Could reap enormous rewards for a relatively small investment.

[Otto von Ballpark]

That would only be helpful if you could run prospects through some sort of combine to utilize the technology to identify players. To detect potential injuries you would have to run someone through an MRI that works with movement. That technology to my knowledge does not exist.

 

I think you are confusing this Trackman system with plain Pitch/FX (which mainly tracks the pitch and not the pitcher's full mechanics) and thus selling it short.

 

The ESPN magazine article on it focused on its ability to quantify a pitcher's mechanics, and the consistency/repeatability of those mechanics. It's still a young system, so you'd have to build up your data set, but I could easily see that being helpful in identifying quality pitchers (and the best roles for pitchers -- how much does a guy's mechanics change after X number of pitches?). And while it's not going to predict injuries, mechanics repeatability is thought to be a pretty big factor in a pitcher's injury potential -- quantifying it would be a huge help. Remember, the Twins weren't even relying much on pre-draft medical info until recently -- they could use all the help they can get!

 

Also, the whole Trackman system is now small and portable, meaning it could be set up anywhere by scouts or coaches and used to analyze potential amateur and international prospects. So it gives you Pitch/FX-type capability everywhere, for any pitcher, and adds mechanics analysis on top of that. Find a guy somewhere with similar mechanics and pitch movement as Glen Perkins and you've really found something, regardless of his stats or if he's even pitched much before.

 

I pretty much agree that metrics like WAR, FIP, etc. aren't exactly huge helps over other previously available stats, and they don't do anything for amateur prospects.

 

That's why the place to look is proprietary technology, behind the scenes -- I can even imagine a system similar to the above to analyze a batter's swing. Find an amateur with a repeatable swing like Joe Mauer's and you may have a prospect on your hands, regardless of his stats.

 

17 MLB teams are using Trackman on pitchers, and I'm sure some of those are also working on new technologies too. There's really no reason the Twins shouldn't be one of those.

 

EDIT TO ADD: And if the Twins are or become one of those, they should be disclosing that fact to fans as much as possible. Keep the proprietary info close to your vest, of course, but tell the fans you're striving to find new ways to get better, faster. And for heaven's sake, manage the PR better than the Jack Goin "stats guy" nonsense.

[old nurse]

Trackman measures ball speed, location and spin

http://baseball.physics.illinois.edu/trackman/Rob%20Ristagno.pdf

[Otto von Ballpark]

Trackman measures ball speed, location and spin

http://baseball.physics.illinois.edu/trackman/Rob%20Ristagno.pdf

 

Sorry I don't have a link, but from a recent ESPN Magazine article:

 

"TrackMan measures the pitcher's arm extension and release height to the trajectory, spin rate, and speed of the ball."

 

Further:

 

"Now TrackMan has gone mobile, allowing scouts to use a portable radar to measure major league players and prospects wherever they find them."

 

It's tricky because the data isn't public etc. but these kind of developments are out there. I am actually really curious if the Twins are into this, moreso than publicly available "metrics"

[old nurse]

You seem to think that these teams are hiring dozens of statisticians, some extremely highly regarded, just to look at fangraphs daily.

 

Please show me where you derived such an idea? I have no idea how other clubs operate. I have no idea where the clubs get their statistical information from. I would assume it comes from the mlb office for official stats and the clubs plug those spreadsheets into their computers to print the numbers they want to look at.

I have never said clubs hire statisticians. You said that. Any article that has ever mentioned the player development people into statistics, there isn't a one of them with a statistics degree. I hope the entire Twins front office are avid golfers taking lessons from a professional instructor in hopes that one of them can see the benefits of some of the technology and has brought in on board.

[TheLeviathan]

Please show me where you derived such an idea? I have no idea how other clubs operate. I have no idea where the clubs get their statistical information from..

 

Arguments from ignorance aren't much of an argument. We have a variety of clues, partial evidence, and tendencies to reach a conclusion. If you are waiting for a line by line description of metric use by team to form an opinion, it won't happen. Not to mention how easy it is to question the accuracy of public disclosures. (the forum is ripe with this in regards to comments made by Ryan on many things)

 

So I'm not really sure what the aim is here. The preponderance of the evidence that is available (and it's sufficient) indicates we are behind many teams. Playing the ignorance argument just isn't valid.