Berardino: Analytics will assist Twins in June draft

[Seth Stohs]

Mike Berardino wrote a terrific article, with several quotes from Twins scouting director Sean Johnson, on how the Twins will use analytics as they prepare for an obviously very important June draft. 

 

"Assist" does appear to be the correct word. Scouting will obviously still play the largest part of the work, but now that there are analytical systems (such as Trackman or Pitch FX) in more college programs and such, that data can verify what the scouts see. 

 

 

 

“We were just kind of getting into that (under former general manager Terry Ryan) but I know in my mind we’re going to incorporate it more into our process,” Johnson said. “Broadly speaking, we’re going to use more of the metrics. We’re going to have more of the numbers, but not to drive the guys we go evaluate and who we want.”

To what purpose then?

“It will be more to keep us attuned to, ‘Are we overvaluing a guy or undervaluing a guy? Are we out of line thinking this guy is going to do a certain thing?’ ” Johnson said. “It’s just another tool.”

 

 

Does this information/article kind of align with what you thought that the draft plan would be under Falvey and Levine? How different do you think things will be under Sean Johnson (with the information he has at his fingertips) than it was under Deron Johnson (with the information he had)?

[dgwills]

My main issue is that the Twin's drafting is that they have placed to much emphasis on character in the past. I'd assume that's why they drafted Alex Kirilloff over Blake Rutherford. Nick Gordon seems like  that kind of pick as well, but picking Trea Turner might of been a reach at that time. Not that I don't like Kirilloff and Gordon, but I just don't like the strategy.

You shouldn't expect guys in their teens and early 20's to act like 40 year olds.

[Guest]

 

My main issue is that the Twin's drafting is that they have placed to much emphasis on character in the past. I'd assume that's why they drafted Alex Kirilloff over Blake Rutherford. Nick Gordon seems like  that kind of pick as well, but picking Trea Turner might of been a reach at that time. Not that I don't like Kirilloff and Gordon, but I just don't like the strategy.

You shouldn't expect guys in their teens and early 20's to act like 40 year olds.

 

Won't disagree totally, but bear in mind that the Twins have also taken chances on guys who saw their draft stock drop over supposed 'character issues', too. Stephen Gonsalves, for example.

 

I like what I read in the Berardino article. It's rediculous to ignore ANY valid and verifiable information you can get on these guys, but foolish to rely strictly on any particular component of the mix.

[dgwills]

 

Won't disagree totally, but bear in mind that the Twins have also taken chances on guys who saw their draft stock drop over supposed 'character issues', too. Stephen Gonsalves, for example.

 

I like what I read in the Berardino article. It's rediculous to ignore ANY valid and verifiable information you can get on these guys, but foolish to rely strictly on any particular component of the mix.

Good point. Gonsalves was a late pick though, so I'm not willing to count that. I agree with you about using all available information. The information is intended to get the best talent available though. Unless there is Josh Hamilton level concerns, I'm just not willing to put emphasis on character grades.

[Seth Stohs]

Character... Makeup... these things do matter. Obviously talent comes first, but let's not say that makeup doesn't matter.

[Guest]

Gonsalves wasn't a 1st round pick, but I'd hardly consider him a "late pick." Twins got him in the 4th round and that was after he had fallen a couple of rounds below where he had been projected to go originally, as I recall.

[nicksaviking]

"We’re going to have more of the numbers, but not to drive the guys we go evaluate and who we want.”

 

So that makes it pretty clear that in the past the front office was using stats to justify their already established beliefs and actions. That sounds pretty damning to me.

 

It's actually pretty encouraging that Johnson publicly acknowledged that they weren't using stats to evaluate so much as to justify. At least they should be using them with purpose now.

[Thrylos]

 

Character... Makeup... these things do matter. Obviously talent comes first, but let's not say that makeup doesn't matter.

 

Sure.  But character and makeup is an objective thing.  For example some people would think that a really religious homeschooled kid has "high makeup", some people might think that the kid has been sheltered and might have a hard time adjusting

 

Talent is talent.

[Doomtints]

I'm wondering what kind of "analytics" are available for 17-18 year old kids.

[ashbury]

It's rediculous to ignore ANY valid and verifiable information you can get on these guys, but foolish to rely strictly on any particular component of the mix.

I think this mischaracterizes what an analytics professional does, in the face of dozens of components.

[Richie the Rally Goat]

I'm wondering what kind of "analytics" are available for 17-18 year old kids.

i wonder the same about Latin American and less baseball oriented countries like say Germany for instance.

 

I would imagine some big collegiate programs have some of the ball flight equipment in place, but not many, and at most one or two high schools

[Seth Stohs]

 

i wonder the same about Latin American and less baseball oriented countries like say Germany for instance.

I would imagine some big collegiate programs have some of the ball flight equipment in place, but not many, and at most one or two high schools

 

It's a fair question. Obviously most high schools and international sites won't have Trackman or Pitch F/X or that kind of thing. But I would guess that a lot of the locations for the showcases and the national circuit probably do by now. And the Dominican and Venezuelan players often travel to the states for showcases and tournaments as well. So, it would be limited, but it would be something. 

[bird]

 

"We’re going to have more of the numbers, but not to drive the guys we go evaluate and who we want.”

So that makes it pretty clear that in the past the front office was using stats to justify their already established beliefs and actions. That sounds pretty damning to me.

It's actually pretty encouraging that Johnson publicly acknowledged that they weren't using stats to evaluate so much as to justify. At least they should be using them with purpose now.

 

 

I don't interpret his statement in that way. He's not saying they used statistics to create a justification of either an opinion of a player or a decision about a prospect, is he? I think he's saying they plan to use more data. The purpose will be to verify the story on a prospect, not justify a decision, as you're suggesting. I agree, that would be damning, but I disagree that it ever happened.

 

I doubt any organization drives its decisions about who to draft by pouring a bunch of stats onto a spreadsheet and sorting things by some statistical criteria. I believe they all eyeball the prospect, sniff around and ask questions, study video and available stats, analyze the kid's background, and all that good stuff, and that triggers the confidence to render an opinion. Isn't Sean Johnson simply saying they're going to use more metric data as a tool for greater scrutiny of the opinions they're forming? From what I see it's all good. They're just taking steps to take advantage of the available information better than was previously done.

 

[Guest]

I'm wondering what kind of "analytics" are available for 17-18 year old kids.

I expect they're not talking about FIP or zone rating, but about things like spin rate, batting plane, arm angle, etc. Remember, this is a team that implemented video in the minors only a few years ago.

[mlhouse]

I have suggested a much better approach than "analytics". Instead of drafting in June based on HS and college performance, invite draft prospects to play in the rookie leagues. Each team would operate a team with its current rookie league managerial staff and specialists, and potentially could have a preliminary draft so they could more closely evaluate the prospects they are most interested in.

 

Each team would be fostered with enough players and pitchers so they could play serveral games a week and double headers, The normal short season lasts around 60 games, so with extra double headers this would give scouts the opportunity to evaluate players over 60-90 games playing against invite players only.

 

After the short season then teams would conduct the player entry draft. They would not be required to draft only players from the invite league, and not all the players in the invite league would be guaranteed to be drafted.

 

THe players would receive a stipend and allowance for playing just like rookie league players today. It might be worked out that teams could bring in their previously drafted players to compete in these leagues, or perhaps they would form additional teams in these leagues for more competition and development.

 

The value to the league is that they get to evaluate the draft able players using wooden bats and facing select competition, with a reasonable sample size and hands on work with the players.

 

There is a lot of value to most of the draftable players too. A hole in the system would be that I believe the top level players in the draft would skip the league, but players that would be non-first round picks and middle round picks could see lots of value proving that they are elite prospects. Think about Justin Mourneau, from a Canadien high school being allowed to show case his skills over 300 plate appearances. Instead of being a 3rd round pick, he would have probably moved up to the first round, and the guy the Twins took 5th overall in that draft, and gave a $1.5 million bonus, B.J. Garbo, would probably have fell a couple of rounds.

 

In the end, it makes a lot of sense because the biggest problem in evaluating players is the differences in the competition level they play in. When there are thousands of high schools, tens of thousands of high school players, hundreds of college programs, and thousands of college players, the independent variables are out of control for a solid analytical approach. But, let's say you invite 600 players and you pay them $1,000 a month stipend, plus living costs for two months. The total cost would be under $2 million, divided between the 30 teams, makes the cost under $70,000 per teams.

[beckmt]

 

I have suggested a much better approach than "analytics". Instead of drafting in June based on HS and college performance, invite draft prospects to play in the rookie leagues. Each team would operate a team with its current rookie league managerial staff and specialists, and potentially could have a preliminary draft so they could more closely evaluate the prospects they are most interested in.

Each team would be fostered with enough players and pitchers so they could play serveral games a week and double headers, The normal short season lasts around 60 games, so with extra double headers this would give scouts the opportunity to evaluate players over 60-90 games playing against invite players only.

After the short season then teams would conduct the player entry draft. They would not be required to draft only players from the invite league, and not all the players in the invite league would be guaranteed to be drafted.

THe players would receive a stipend and allowance for playing just like rookie league players today. It might be worked out that teams could bring in their previously drafted players to compete in these leagues, or perhaps they would form additional teams in these leagues for more competition and development.

The value to the league is that they get to evaluate the draft able players using wooden bats and facing select competition, with a reasonable sample size and hands on work with the players.

There is a lot of value to most of the draftable players too. A hole in the system would be that I believe the top level players in the draft would skip the league, but players that would be non-first round picks and middle round picks could see lots of value proving that they are elite prospects. Think about Justin Mourneau, from a Canadien high school being allowed to show case his skills over 300 plate appearances. Instead of being a 3rd round pick, he would have probably moved up to the first round, and the guy the Twins took 5th overall in that draft, and gave a $1.5 million bonus, B.J. Garbo, would probably have fell a couple of rounds.

In the end, it makes a lot of sense because the biggest problem in evaluating players is the differences in the competition level they play in. When there are thousands of high schools, tens of thousands of high school players, hundreds of college programs, and thousands of college players, the independent variables are out of control for a solid analytical approach. But, let's say you invite 600 players and you pay them $1,000 a month stipend, plus living costs for two months. The total cost would be under $2 million, divided between the 30 teams, makes the cost under $70,000 per teams.

You cannot pay these players until they turn pro, amateur rules.

It is a good thought if their was a way to set up an amateur tournament to showcase these players. 

[HitInAPinch]

 

I expect they're not talking about FIP or zone rating, but about things like spin rate, batting plane, arm angle, etc. Remember, this is a team that implemented video in the minors only a few years ago.

VHS or Blu-Ray?

[HitInAPinch]

 

I have suggested a much better approach than "analytics". Instead of drafting in June based on HS and college performance, invite draft prospects to play in the rookie leagues. Each team would operate a team with its current rookie league managerial staff and specialists, and potentially could have a preliminary draft so they could more closely evaluate the prospects they are most interested in.
...

 

 

You cannot pay these players until they turn pro, amateur rules.

It is a good thought if their was a way to set up an amateur tournament to showcase these players. 

Already happening.

 

Independent Leagues:

 

An independent baseball league is a professional baseball organization located in the United States and Canada that is not operated in conjunction with either a Major League Baseball team or an affiliated Minor League Baseball team. Being independent allows teams to be located close to major league teams without their consent. Such leagues have been around for many years and were once known as "outlaw leagues" due to their position outside the rules of affiliated minor league baseball.

 

 

[mikelink45]

Old and skeptical.  I will wait to judge.  Baseball draft is way behind basketball, hockey, football in  ability to get the best - they all fail at times, but the fact is we are looking at high school players, international players, and we are still guessing at the development of body and skills, attitude and adjustment.  For the most part universities and colleges have supported the other three major sports and given a showcase for performance.  

 

I would love to know what the analytics are at this point unless they are college draftees I question them.

[Richie the Rally Goat]

Already happening.

 

Independent Leagues:

 

An independent baseball league is a professional baseball organization located in the United States and Canada that is not operated in conjunction with either a Major League Baseball team or an affiliated Minor League Baseball team. Being independent allows teams to be located close to major league teams without their consent. Such leagues have been around for many years and were once known as "outlaw leagues" due to their position outside the rules of affiliated minor league baseball.

indeed local to Twins Territory (Geographically) is the Northwoods League. Alumni of the Northwoods League include Andre Ethier, Ben Zobrist, and Austin Barnes.

 

It's fun baseball to watch

 

http://northwoodsleague.com/about-nwl/alumni-in-the-majors/

[Seth Stohs]

 

Sure.  But character and makeup is an objective thing.  For example some people would think that a really religious homeschooled kid has "high makeup", some people might think that the kid has been sheltered and might have a hard time adjusting

 

Talent is talent.

 

Correct... but obviously that's where the scouting comes in... Since you're obviously referring to Kirlloff, not only would the area scout be able to talk to his high school coach and teammates, but he can talk to the family and members of their church... and others can also check out how he interacts with teammates and coaches in the national circuit that he plays in and such. 

 

Talent is talent, but if a kid can't handle social interaction, it greatly reduces his likelihood of getting to the big leagues. We've see a couple of examples of that in recent years.

[Oldgoat_MN]

 

Sure.  But character and makeup is an objective thing.  For example some people would think that a really religious homeschooled kid has "high makeup", some people might think that the kid has been sheltered and might have a hard time adjusting

 

Talent is talent.

I absolutely agree, Thrylos.

Talent is talent and people will see the same thing different ways.

 

Still, I'm sure they want to have an idea of 'character' before they give a 18-22 year old a check for $4m.

[Guest]

I absolutely agree, Thrylos.

Talent is talent and people will see the same thing different ways.

 

Still, I'm sure they want to have an idea of 'character' before they give a 18-22 year old a check for $4m.

Character as a negative factor (does he do drugs, disrespect his coach) is clearer than character as a positive factor, which is often based on more subjective factors, can be a "beauty is in the eye of the beholder" type of thing, and can often just be a cliche like "plays the game the right way" that is used to justify a decision that may otherwise be hard to explain.

[Richie the Rally Goat]

It's a fair question. Obviously most high schools and international sites won't have Trackman or Pitch F/X or that kind of thing. But I would guess that a lot of the locations for the showcases and the national circuit probably do by now. And the Dominican and Venezuelan players often travel to the states for showcases and tournaments as well. So, it would be limited, but it would be something.

how significant of a sample size can you get from the showcases and tournaments these players participate in? Would it be closer to the equivalent of 30 games or 50, more?

[Doomtints]

Here's my skeptical view of this.

 

Analytics for the draft are going to be untrustworthy unless drafting college players. High school data and foreign data is going to be incomplete and untrustworthy. For drafting these types of players, scouting is essential. No player at this level should be drafted without being scouted.

 

Now, where analytics comes into play is when drafting college players or analyzing players who are in the minors. There is data there to trust, and the data is going to be complete. These players should still be scouted, but data is more important and analyzing the data the right way can make a huge difference in helping to target players that you should watch.

 

My feeling is that this is a feel-good puff piece written to smooth over relationships with fans who think the Twins are behind with analytics. Because, literally, the idea of pre-draft analytics of non-college players making a difference is laughable.  

[drivlikejehu]

Clubs are allowed to invite prospects to work out. So measurements could be made using Target field equipment during such a work out, and analytics applied to the results. So even for high school players, more advanced means can play a role.

[ScrapTheNickname]

 

My main issue is that the Twin's drafting is that they have placed to much emphasis on character in the past. I'd assume that's why they drafted Alex Kirilloff over Blake Rutherford. Nick Gordon seems like  that kind of pick as well, but picking Trea Turner might of been a reach at that time. Not that I don't like Kirilloff and Gordon, but I just don't like the strategy.

You shouldn't expect guys in their teens and early 20's to act like 40 year olds.

Matt Garza comes to mind as a human being devoid of positive character traits. Anyway, that's how he came off to me when I watched him pitch in Elizabethton TN when he was starting out. A cocky robot with poor manners. So the Twins must have been looking at the radar gun, etc., first on him. Just an example. Okay, I'm ranting now, but he's earned $67 million dollars to win 1/3 of his MLB starts. Glad he's long gone.

[nicksaviking]

 

Clubs are allowed to invite prospects to work out. So measurements could be made using Target field equipment during such a work out, and analytics applied to the results. So even for high school players, more advanced means can play a role.

 

And while I agree that advanced measurements are likely much more difficult and thus less reliable for HS kids, I would also think that some kind of analytics could be used to compare and asses players to similar draft picks previous. Someone could study results of players based on things that can be measured or assesed like velocity, geography, level of competition, speed, position, age, etc.

[Seth Stohs]

 

how significant of a sample size can you get from the showcases and tournaments these players participate in? Would it be closer to the equivalent of 30 games or 50, more?

 

The sample size for any high school player isn't going to be much maybe 100 at bats. So the summer circuits may get them an extra however-many games... but it's games against other top-end talent, so it's better for evaluation than their normal high school or Legion season. But either way, the sample is small. 

 

Of course some of the top talent has been on that stage since they were 13, and had scouts seeing them on multiple occasions.  Again, sample still small, but there are more data points. 

 

Of course, for any pitcher the sample size is probably going to be small too even the college guys. 

[mlhouse]

 

You cannot pay these players until they turn pro, amateur rules.

It is a good thought if their was a way to set up an amateur tournament to showcase these players. 

 

 

Well, the fact is that they would be professional players but to preserve the amateur nature you can easily structure any payment as a stipend.  

 

THe problem with "amateur" tournaments is that you could not get enough sample size to make it effective.  How many games would an "amateur" be willing to spend playing as a showcase?  WHo would pay for their transportation and living expenses?  

 

Think of the league I mention as unaffiliated minor leage players in the short season rookie leagues.  The Twins would essentially run 3-4 teams:  Their Elizabethton rookie league team stocked with players drafted in previous years, a Rookie Plus team of draftable players, and a Rookie Minus team of draftable players, and perhaps a GCL team of players drafted in a previous season.   There might even be a short season A league setup for more advanced players but I think 2 levels is more than enough.

 

The improved way of scouting these players is much more value than any potential cost.