Article: A preview of 2015 MLB Team Defenses.

[TheLeviathan]

 

Right, that's where the logic doesn't align.

 

The best and worst team FIPs (which for those less familiar is how many runs a pitcher should allow before accounting for luck and defense by calculating the outcomes he can control) shows a gap of 227 runs between the best AL team and the worst.

 

A gap of 140 runs between the best and worst defense is not saying that defense is the cause of 140 of those 227 runs.  They are completely separate.

 

If we are comparing the two against each other, it'd be on the net whole of them both... 227 + 140.  In that scenario, it's saying defense is 38% of the runs allowed equation while pitching is 62%.

 

Seems about right, I'd be curious to see if that lined up over multiple seasons as well.  There are known issues with single season data and it might become more reliable over the course of more seasons.  But I think that split makes sense.

[TheLeviathan]

 

Right, that's where the logic doesn't align.

 

The best and worst team FIPs (which for those less familiar is how many runs a pitcher should allow before accounting for luck and defense by calculating the outcomes he can control) shows a gap of 227 runs between the best AL team and the worst.

 

A gap of 140 runs between the best and worst defense is not saying that defense is the cause of 140 of those 227 runs.  They are completely separate.

 

If we are comparing the two against each other, it'd be on the net whole of them both... 227 + 140.  In that scenario, it's saying defense is 38% of the runs allowed equation while pitching is 62%.

 

Seems about right, I'd be curious to see if that lined up over multiple seasons as well.  There are known issues with single season data and it might become more reliable over the course of more seasons.  But I think that split makes sense.

[Brock Beauchamp]

I suppose my problems revolve around the seemingly arbitrary concept of a "run" in these metrics. They often don't align with actual runs scored on a baseball diamond. I started noticing this with WAR several years ago and it's been gnawing at me since that time.

 

If you put it in relative terms... we'll call them "Widgets Above Replacement" and said the Royals were 120 Widgets better than the Twins and Oakland was 50 Widgets better than the Twins, I'd have a much easier time acknowledging that number as accurate.

 

But when the Royals are 120 defensive runs better than the Twins while they only allowed ~150 (not looking this up, too bored) actual real-world runs less than the Twins, my eyes boggle because it simply isn't matching what happened on the field.

[Mike Sixel]

Again......the pitching and the defense both add up to that 150.

 

You seem to be arguing the following:

 

I have 4 apples today. It is impossible that someone once gave me 100 apples.

 

If team A has pitching that is 100 runs better than team B. And team B has defense that is 40 runs better than team A.....then it is NOT TRUE that defense accounts for 40/60 percent of the run differential (which is the example you are using, unless I am misunderstanding).

[jay]

 

Seems about right, I'd be curious to see if that lined up over multiple seasons as well.  There are known issues with single season data and it might become more reliable over the course of more seasons.  But I think that split makes sense.

 

Average AL FIP gap the last 4 years is 177. Average MLB FIP gap the last 4 years is 212. Using FIP, I think the MLB number is better.

 

Average UZR/DRS MLB gap the last 4 years is 147.

 

Do the math there and defense is 40% with pitching at 60%.

[Brock Beauchamp]

 

If team A has pitching that is 100 runs better than team B. And team B has defense that is 40 runs better than team A.....then it is NOT TRUE that defense accounts for 40/60 percent of the run differential (which is the example you are using, unless I am misunderstanding).

Which I've mentioned a few times... In this past season, we had the luxury of watching a bottom AL team in both defense and pitching.

 

So that's half the discrepancy shored up right there. The Twins gave up the most runs while having a bottom pitching staff and a bottom defense. Yay, Twins.

 

Inversely, the Royals had a good pitching staff and great defense. The A's had a great pitching staff and a good defense. The numbers shouldn't be completely out of whack, yet they are.

[jay]

 

Do the math there and defense is 40% with pitching at 60%.

 

And that's without accounting for any measurement error at the extremes in defense.  Reduce that 147 by, say, 10% and you're at 38%/62%.

 

Trying to drop it to something like 80 runs attributable to defense moves the split all the way to 27%/73%. 

 

Both feel within the range of reasonable, but not much higher or lower than that.

[Mike Sixel]

I don't think, for me anyway, that you've demonstrated that they are out of whack. You seem to agree that an 80 run differential is reasonable. That means, to me, within the margin of error of measurement. That also means that you should, knowing how stats work, be willing to accept that there will be outliers, and those outliers will be extreme by their nature, leading to numbers of around .75 runs per game or so (no idea what that standard deviation is.....not all that interested in looking).

[Brock Beauchamp]

 

I don't think, for me anyway, that you've demonstrated that they are out of whack. You seem to agree that an 80 run differential is reasonable. That means, to me, within the margin of error of measurement. That also means that you should, knowing how stats work, be willing to accept that there will be outliers, and those outliers will be extreme by their nature, leading to numbers of around .75 runs per game or so (no idea what that standard deviation is.....not all that interested in looking).

Absolutely, there is a margin of error... but if I get a nagging feeling that margin is up to 50%, alarm bells start ringing in my head.

 

As for the numbers, Fangraphs states that the Royals were 113 defensive runs (DRS, 40 to -73) better than the Twins in 2014, yet on the field they allowed an actual 153 less runs (777 to 624) on the season. Do we really believe the Royals pitching staff was worth only 40 runs over the Twins' staff in 2014?

[jimmer]

The Royals pitching staff had the 15th best FIP and the Indians had the 2nd best FIP. Royals, as a team were average in their team FIP. Indians were great.

 

The Royals starting rotation has the 19th best FIP, the Indians had the 2nd best FIP.  Below average rotation for the Royals, great starting pitching from the Indians.

 

Based on that, you would think that the Indians starters would have pitched more innings that the Royals starters, especially when you consider Yost loved to use those three reliever so much, but they didn't. They were only 18th in IP, while the Royals were 8th.

 

Since the Indians defense was so bad, their better starting rotation got knocked out the games earlier because of the defense allowing more runs than a defense like the Royals would have. Overall the Royals gave up less runs even with the much inferior rotation.

 

P.S., Indians ended up with the 19th best rotation ERA while the Royals were 11th.

[jay]

 

As for the numbers, Fangraphs states that the Royals were 113 defensive runs (DRS, 40 to -73) better than the Twins in 2014, yet on the field they allowed an actual 153 less runs (777 to 624) on the season. Do we really believe the Royals pitching staff was worth only 40 runs over the Twins' staff in 2014?

 

Again, those figures can't be accurately compared against each other. There are many more variables that went into their actual differential and it's skewing the picture. Use FIP or something else agnostic to compare pitching staff value to defensive value.

[Brock Beauchamp]

Again, those figures can't be accurately compared against each other. There are many more variables that went into their actual differential and it's skewing the picture. Use FIP or something else agnostic to compare pitching staff value to defensive value.

And my point is that if they're using the words "runs" in the metric and attempt to quantify runs saved, the numbers should be close to what actually happens on the field in runs allowed.

 

Again, if they want to call them Widgets, they can use any number they like... But they don't. They're trying to put a run value to the metrics and the numbers churned out are screwy. I could handle a few margin points here and there but not that kind of discrepancy.

[jay]

 

Do we really believe the Royals pitching staff was worth only 40 runs over the Twins' staff in 2014?

 

FIP says their staff was 45 runs better.

 

Royals FIP 3.69 (15th best in MLB)

Twins FIP 3.97 (26th best in MLB)

[jay]

 

And my point is that if they're using the words "runs" in the metric and attempt to quantify runs saved, the numbers should be close to what actually happens on the field in runs allowed.

Again, if they want to call them Widgets, they can use any number they like... But they don't. They're trying to put a run value to the metrics and the numbers churned out are screwy. I could handle a few margin points here and there but not that kind of discrepancy.

 

FIP says the Royals staff was 45 runs better.

DRS says the Royals defense was 113 runs better.

Combined, 158 runs.

 

Their real-world runs allowed difference was 153.

 

Where's the screwy?

[jimmer]

 

 

FIP says their staff was 45 runs better.

 

Royals FIP 3.69 (15th best in MLB)

Twins FIP 3.97 (26th best in MLB)

And the gap is even closer when looking at rotations.  Royals 19th, Twins 24th. It's amazing how much a great defense can help a rotation since their ERAs were so far apart.

[Mike Sixel]

 

Absolutely, there is a margin of error... but if I get a nagging feeling that margin is up to 50%, alarm bells start ringing in my head.

 

As for the numbers, Fangraphs states that the Royals were 113 defensive runs (DRS, 40 to -73) better than the Twins in 2014, yet on the field they allowed an actual 153 less runs (777 to 624) on the season. Do we really believe the Royals pitching staff was worth only 40 runs over the Twins' staff in 2014?

 

Lots of things that seem to be inconceivable turn out to be true. I understand your skepticism, I have no idea what the real number is. But, people much smarter than me put a lot of time into this. I'm giving them the benefit of the doubt on this one.

 

Again, the issue for me isn't if the Twins are losing half a run a game, or .75 runs a game.....its that they are not gaining runs per game.

[TheLeviathan]

 

Lots of things that seem to be inconceivable turn out to be true. I understand your skepticism

 

The problem is that the skepticism doesn't seem to be understood nor appreciated.  (A huge problem IMO) With something like this it should ALWAYS be a welcome part of the process.  Just look at what it has lead to -  a discussion with a lot of moving parts to it and a lot of interesting analyses.

 

All amongst people that put a high emphasis on defense - we should embrace that skepticism, clarification, and exploration.  Not automatic, dogmatic acceptance.

[mikelink45]

Does defense matter? Of course, especially with the old pitch to contact Twins.  What factors out the defense is strikeouts.  The fewer times a ball hits a glove the less defense is involved. And even more important, the less a ball is well stroke, the less a good play is required.  Defense makes pitchers and pitchers make defense. 

[jorgenswest]

FIP isn't truly defense independent. Catching does not factor into FIP. Certainly the framing and pitch sequencing does not. If the result of that defense is fewer strikes and more balls, that impact will not be seen in FIP. If framing is costing or saving a team runs, the percent of impact has to come from the portion attributed to pitching.

[jay]

FIP isn't truly defense independent. Catching does not factor into FIP.

That's true. Using FIP in this instance is still going to be a far superior method compared to simply applying the positional defensive values to actual runs allowed.

 

The intent is to get closer to the pitching staff's true expected performance. Do you have a better figure to propose?

 

Better yet, do you have a better breakdown of defense vs pitching vs any other variables for assessing runs allowed?

[Mike Sixel]

 

The problem is that the skepticism doesn't seem to be understood nor appreciated.  (A huge problem IMO) With something like this it should ALWAYS be a welcome part of the process.  Just look at what it has lead to -  a discussion with a lot of moving parts to it and a lot of interesting analyses.

 

All amongst people that put a high emphasis on defense - we should embrace that skepticism, clarification, and exploration.  Not automatic, dogmatic acceptance.

 

Who isn't skeptical? Even the creators admit there are error bars and confidence levels, and that the numbers aren't precise. There is a difference, imo, between skepticism, and dismissing something because it doesn't make "common" sense. Or some of the other phrases people have used here, frankly, that seem a bit demeaning, imo, of people who are "believers".

[TheLeviathan]

 

Who isn't skeptical? Even the creators admit there are error bars and confidence levels, and that the numbers aren't precise. There is a difference, imo, between skepticism, and dismissing something because it doesn't make "common" sense. Or some of the other phrases people have used here, frankly, that seem a bit demeaning, imo, of people who are "believers".

 

No offense, but have you read the tone of some people responding to the skepticism?  No one is dismissing anything - every step of this thread has been more about questioning (the heart of skepticism) and it has been met with derision.  

 

Thank goodness a few like jay have tried to keep the dialogue productive and interesting rather than dogmatic and dismissive

[Mike Sixel]

I've read them all, even the deleted ones........It has mostly been a good conversation. There have been slips in that on both sides of the argument, imo. Some tone, some in the choice of words.

[TheLeviathan]

 

I've read them all, even the deleted ones........It has mostly been a good conversation. There have been slips in that on both sides of the argument, imo. Some tone, some in the choice of words.

 

I agree and look, I have a hard time expressing my response to the numbers here as anything but common sense-defying.  You pick a term that is less offensive to you I guess.  If I call it counter-intuitive than that gets shot down.  The point is, something feels off about that number.  It feels like an extreme outlier being used to prop up the case and being portrayed as common.  Whatever approach you want to take something about it doesn't mesh with me.

 

Am I wrong?  Maybe, jay has made a few compelling arguments.  Some I still struggle with and much of it stems from my own misgivings about the confidence people have in defensive metrics, but it's not an unfair question.

[TheLeviathan]

 

No offense, Levi, but the tone didn't get bad on this until you decided to say that believing in the 120 run difference lacked basic common sense.

 

I would argue it was already off track on the first page of this thread when there was some healthy skepticism expressed and then patronizingly unaddressed.

[jimmer]

 

I would argue it was already off track on the first page of this thread when there was some healthy skepticism expressed and then patronizingly unaddressed.

I have no doubt you would argue that.  I would argue that continuing to use the argument that some people take metrics as gospel is pretty insulting especially when the people accused of taking metrics as gospel keep saying that no one is actually doing that.

[TheLeviathan]

 

I have no doubt you would argue that.  I would argue that continuing to use the argument that some people take metrics as gospel is pretty insulting especially when the people accused of taking metrics as gospel keep saying that no one is actually doing that.

 

It would help then if you approached the issue without that coming across.  I have no doubt that saying it defied common sense sounded like an insult to people who took the numbers at face value - that was never my intent.  But I can't think of a term that conveys that feeling when I read it any better than has been expressed more than a few times.

 

I'm glad some others helped flesh this out a bit more because it's been enlightening.  No matter where you fall on this particular metric, that should be seen as a good thing.  Anyone willing to ask a question or express skepticism is a person trying to embrace the value of the metric or find out where it is weak or strong.

 

Baseball metrics are where they are now because lots of people have said "No...I don't buy that, we can find a better way to measure that"  And now we have a whole host of interesting ways of measuring the game and I think that's a good thing.

[Mike Sixel]

Levi, I agree, statistics are better due to skepticism. All of life is better because of that. However, I don't fee the other side is actually offering a better alternative, and until you can offer a better alternative, I am not ready to dismiss the best alternative we have. 

 

I'm not offended by your words at all, as I've said here many times, this is just entertainment for me, and I don't take the comments personally.

[TheLeviathan]

Levi, I agree, statistics are better due to skepticism. All of life is better because of that. However, I don't fee the other side is actually offering a better alternative, and until you can offer a better alternative, I am not ready to dismiss the best alternative we have. 

 

I'm not offended by your words at all, as I've said here many times, this is just entertainment for me, and I don't take the comments personally.

There is not a better statistical alternative so feel no pressure to dismiss anything. I'm just not on the bandwagon due to skepticism.