Keith Law Top 25 Impact Prospects for 2016

[Mike Sixel]

btw, I am really being serious here.....if someone posts WAR or some other stat, do people really want that post to have a comment, every time, about that stat's limitations? Do people not get tired of this argument every time WAR is mentioned? Do some people really think that if the model does not precisely predict the future, it is a bad model (some clearly are stating that, in thread after thread)?

 

I have no issue with disagreeing with Buxton's defensive projection, and discussing that.....but I have an issue with saying that proves the model is stupid, and anyone that quotes it is wrong/bad/blinded by science in EVERY thread where WAR is mentioned.

[Mike Sixel]

I don't think I did that much, other than in response to others, but I agree, that is tedious, and shouldn't happen in every thread.

[drjim]

 

btw, I am really being serious here.....if someone posts WAR or some other stat, do people really want that post to have a comment, every time, about that stat's limitations? Do people not get tired of this argument every time WAR is mentioned? Do some people really think that if the model does not precisely predict the future, it is a bad model (some clearly are stating that, in thread after thread)?

 

I have no issue with disagreeing with Buxton's defensive projection, and discussing that.....but I have an issue with saying that proves the model is stupid, and anyone that quotes it is wrong/bad/blinded by science in EVERY thread where WAR is mentioned.

 

I personally would prefer using better analytics than WAR, but I digress.

 

I don't think you have to look very far to see how people grasp on to the small part that is wrong/misleading to try and discredit the entire system. It happens in everything.

[Brock Beauchamp]

 

I have no issue with disagreeing with Buxton's defensive projection, and discussing that.....but I have an issue with saying that proves the model is stupid, and anyone that quotes it is wrong/bad/blinded by science in EVERY thread where WAR is mentioned.

 

We had a lot of threads celebrating the Twins last year and you or Jimmer or Dave would enter those threads and say the Twins are over-performing, or being lucky and will fall to earth and start losing.  No nuance, just straight trolling. Was that useful?  

Both of these points are valid and when juxtaposed together, reveal a lot about the nature of this board.

 

10% of this board is "Math is right!!!!!"

 

10% of this board is "Math is dumb!!!!"

 

80% of this board is "Math is useful. Now will you 20% STFU for five minutes?"

 

One of the biggest problems on this board - or any other internet forum - is that those two 10% groups try to lump the 80% into their extreme black-and-white analysis whenever it's convenient for them to do so (whether that is in alliance or opposition to their viewpoint depends on circumstance).

 

Not everything needs to be strictly categorized into one camp or the other. It's acceptable to find use in math-based analytics while pointing out that they're often used incorrectly and/or wrongly presented as fact (and vice versa).

[Mike Sixel]

 

Both of these points are valid and when juxtaposed together, reveal a lot about the nature of this board.

 

10% of this board is "Math is right!!!!!"

 

10% of this board is "Math is dumb!!!!"

 

80% of this board is "Math is useful. Now will you 20% STFU for five minutes?"

 

One of the biggest problems on this board - or any other internet forum - is that those two 10% groups try to lump the 80% into their extreme black-and-white analysis whenever it's convenient for them to do so (whether that is in alliance or opposition to their viewpoint depends on circumstance).

 

Not everything needs to be strictly categorized into one camp or the other. It's acceptable to find use in math-based analytics while pointing out that they're often used incorrectly and/or wrongly presented as fact (and vice versa).

 

I'd love this post if I could.....

[drjim]

 

Both of these points are valid and when juxtaposed together, reveal a lot about the nature of this board.

 

10% of this board is "Math is right!!!!!"

 

10% of this board is "Math is dumb!!!!"

 

80% of this board is "Math is useful. Now will you 20% STFU for five minutes?"

 

One of the biggest problems on this board - or any other internet forum - is that those two 10% groups try to lump the 80% into their extreme black-and-white analysis whenever it's convenient for them to do so (whether that is in alliance or opposition to their viewpoint depends on circumstance).

 

Not everything needs to be strictly categorized into one camp or the other. It's acceptable to find use in math-based analytics while pointing out that they're often used incorrectly and/or wrongly presented as fact (and vice versa).

 

I also think there is a significant percentage of people that would just like to learn more about the analytics but aren't served too well by either this continual argument or (more especially) poor use of analytics

[Brock Beauchamp]

 

I also think there is a significant percentage of people that would just like to learn more about the analytics but aren't served too well by either this continual argument or (more especially) poor use of analytics

Also true. New, interesting metrics are constantly being revealed on this board but it's hard to sift through some of them because the way the data is presented and the way it's defended/attacked.

[Mike Sixel]

For my part, I'm just going to do my best not to defend math anymore, and just let people keep posting how much they hate it (despite sitting on a computer, a device completely built on math), and not respond. 

[Craig Arko]

 

For my part, I'm just going to do my best not to defend math anymore, and just let people keep posting how much they hate it (despite sitting on a computer, a device completely built on math), and not respond. 

I love Maths. I am not a huge fan of sloppiness. Mea culpa.

[70charger]

 

For my part, I'm just going to do my best not to defend math anymore, and just let people keep posting how much they hate it (despite sitting on a computer, a device completely built on math), and not respond. 

 

Gotta say. You're not being very realistic here, just spitting out tropes. Yes, computers are "built on math." But if the programmer puts in bad data, guess what you get out?

[Mike Sixel]

 

Gotta say. You're not being very realistic here, just spitting out tropes. Yes, computers are "built on math." But if the programmer puts in bad data, guess what you get out?

 

Sigh. like that was my point at all......but I do agree with the point you are making.

[drjim]

 

Gotta say. You're not being very realistic here, just spitting out tropes. Yes, computers are "built on math." But if the programmer puts in bad data, guess what you get out?

 

I think the other part of this is that people aren't necessarily against math as a tool, but perhaps skeptical of certain results.

 

And implying things that certain outcomes represent "math" (or my favorite "the debate is settled") don't exactly help the conversation.

[Mike Sixel]

 

I think the other part of this is that people aren't necessarily against math as a tool, but perhaps skeptical of certain results.

 

And implying things that certain outcomes represent "math" (or my favorite "the debate is settled") don't exactly help the conversation.

 

Well, imo, most of the people that say WAR sucks talk about how it isn't precise, therefore the whole model sucks and is junk. That, imo, is an argument against the basic principles of math and statistics. Maybe not, maybe they are being nuanced in their posts about how stupid WAR or BABIP or whatever is......

[Mike Sixel]

If you don't stop replying, I'm not sure I can stop myself, damnit.

[drjim]

 

Well, imo, most of the people that say WAR sucks talk about how it isn't precise, therefore the whole model sucks and is junk. That, imo, is an argument against the basic principles of math and statistics. Maybe not, maybe they are being nuanced in their posts about how stupid WAR or BABIP or whatever is......

 

I always thought it was more about the nebulous concept of replacement player, the unreliability of defensive metrics, the issues with converting (unreliable) defensive metrics into the same unit as offensive numbers for one all-encompassing number, and the fact that the two main sources of the stat don't agree with each other.

 

But I'm sure precision is part of it too. And hatred of math.

[Dman]

 

For my part, I'm just going to do my best not to defend math anymore, and just let people keep posting how much they hate it (despite sitting on a computer, a device completely built on math), and not respond. 

 

First let me say I am not a big WAR believer.  It is used a lot to measure performance now but I don't buy everything about it.  Math and analytics are obviously important to the game of baseball.  We use it to compare players and a players performance all the time. There are, however, things that math doesn't tell us.  It doesn't tell us that players are hurt and that is why they had a down month or down year.  It doesn't take into account lot's of human elements it just cruelly spits out the numbers of performance.

 

Can we glean information from those numbers? Sure.  We can create averages and figure out how certain players will perform at certain levels year in and year out? Sure, but over time the math just doesn't project very well.  Players decline sometimes rapidly and the math lied to us.  Sometime players over perform for other various reasons and the math didn't see it coming.  So IMO math is not the be all end all for measurement but it is very important at giving us an idea of what value a player might have over another player over time.

 

Projection systems based on math are tough to swallow because there are so many things that can happen in a season that it is hard to take them seriously.  The beauty of this sport is that anything can happen any given game.  Even if a great pitcher makes a bad pitch at a bad time then game over.

 

I like the math stats that people bring out on this board.  Often times it changes my perception about a player or situation.  And even though math experts will say the numbers don't lie I think when used incorrectly they at the very least mislead.

[Otto von Ballpark]

FYI, nobody here actually drew any strong conclusions from Fangraphs's defensive projection for Buxton.  Somebody took a shot at it after positive comments were made about his defense.

 

On other threads, I don't think anyone has drawn any strong conclusions from the Fangraphs projections except about the general range of our team-level projection, of which Buxton's defense is a relatively small part.

 

It's a little annoying when folks go well out of their way like this to take shots at stuff.

[jimmer]

 

And yet Fangraphs somehow has Buxton projected to provide negative value on defense.

By Fangraphs do you mean ZiPS? For the record, they just get that from Dan S. who does a chat for Fangraphs but isn't an actual member of Fangraphs.

 

In regards to why Law thinks Buxton is better than what another projection model projects (not predicts), Law has his own opinion.  Not sure what the point of pointing out what 'Fangrapghs" projects when looking at what Law says. Law isn't affiliated with Fangraphs (or ZiPS).

[TheLeviathan]

So Keith Law was teaching a math class?

[Seth Stohs]

What does all this have to do with Keith Law's Impact Prospect List? Keep it on topic, please.

 

If people want to whine about stats and analytics and how other people choose to read and interpret them, do it somewhere else... But play nice... 

[Craig Arko]

What does all this have to do with Keith Law's Impact Prospect List? Keep it on topic, please.

 

If people want to whine about stats and analytics and how other people choose to read and interpret them, do it somewhere else... But play nice...

 

Sorry about that, Chief.

[drjim]

Axioms of Probability (and it's stepchild Statistics).

 

http://statistics.about.com/od/ProbHelpandTutorials/a/What-Are-Probability-Axioms.htm

 

The important thing to come away with is that likelihood is a value between 0 and 1, inclusive. Not 0 or 1, exclusively.

 

Apply this to evaluating arguments. Most of them are grey, not black or white. If you intend to argue about the value of mathematics, learn some first.

Posts like this make me realize that people are mostly talking past each other, and not even about the same thing.

[drjim]

What does all this have to do with Keith Law's Impact Prospect List? Keep it on topic, please.

 

If people want to whine about stats and analytics and how other people choose to read and interpret them, do it somewhere else... But play nice...

I think there is value in better understanding what different projection systems for said prospects actually represent.

[jimmer]

 

Hard to see why this conversation goes nowhere with statements like this.

As opposed to all the other rational posts on this thread coming at it from the other side that aren't deemed as conversation stoppers?  Kind of my point.

 

In any event, like I said, most teams are putting a lot of money and effort into analytical teams and making their decisions based on them.   They seem to believe it helps them more than they way it was done before.

[Craig Arko]

Posts like this make me realize that people are mostly talking past each other, and not even about the same thing.

Let's just say a pet peeve was scratched. However, I'll delete some of the more egregious posts so nobody else has to. My position is well-known.

[jimmer]

 

What benefit comes from meeting the irrational with more smug irrational?

perhaps it just comes from frustration that advanced stats get slammed because they aren't 100% perfect yet the alternative is the eye test which also clearly isn't 100% perfect.  

 

This extends to projection systems as well. I'm guessing Law uses some form of that too.  I don't think he randomly guesses.

[Seth Stohs]

 

I think there is value in better understanding what different projection systems for said prospects actually represent.

 

True enough... but there should be no reason that our moderators deem things worthy of hiding and handing out warnings... 

[drjim]

perhaps it just comes from frustration that advanced stats get slammed because they aren't 100% perfect yet the alternative is the eye test which also clearly isn't 100% perfect.

I would personally differentiate between people who ignore/slam advanced stats out of hand and those who are either unclear, not up to speed, and/or skeptical of some amount stats.

 

The first group I would look at, laugh a little, and then mostly ignore. Otherwise you are just banging your head against the wall. This strikes me as a very small part of this board, and baseball fandom in general, though is unfortunately kind of loud about it.

 

The second group you might have something. Either an interesting conversation, a chance to educate, or perhaps even a chance to challenge your own thinking and learn something.

 

While it is probably obvious, there are different degrees of accepting and application of advanced stats, even among the experts on teams. It is not a monolithic thing and there is no perfect stat.

 

Part of the fun/interesting part for me is exploring all the stats and advanced metrics and trying to sort out what is good and useful, and what I find less so. It also feeds and challenges my more skeptical side.

 

I find this type of conversation more interesting than becoming entrenched in an extreme and letting it fly.

[drjim]

True enough... but there should be no reason that our moderators deem things worthy of hiding and handing out warnings...

Concur on that completely. This conversation can actually happen civilly.

[drjim]

I do acknowledge the thread got a little hijacked and I apologize for my role, but I do find the conversation, when done civilly, is an interesting one. And there was something from the beginning that did inspire this direction.