Tangotiger Blog

Monday, August 27, 2018

Statcast Lab: Fielder Roles, part 2 (roles and functions)

?As I showed two weeks ago, we will be adding "Role" designations to fielders (in addition to maintaining  their official position).  It takes just a small amount of effort to understand the Fielder Roles.  They are grounded in the traditional positions, with additional identifiers of ".1" (left) and ".2" (right).  There are a few other slices or zones, and you can kind of figure out the scheme.

Now that we know the role for each player, for each play, and we have assigned each play to one fielder (based on the proximity of the ball to the player), we now need to know their FUNCTION, the landing point of the ball.  It's going to be more exciting when I show it for infielders, but let me describe it for outfielders, since it's easier.

(Click to see larger image.)

?

For a fielder who plays in the traditional LF spot (role 7.0), and the ball lands in that same area (landing 7.0), 76.6% of those plays are converted into outs.  And if you look at landing 7.1, meaning the ball lands in the slice that is toward the LEFT side  of the field, they converted 72.9% of those plays into outs.  For balls that land toward the  right side of  the field (landing 7.2... and remember we are still looking at role 7.0), they convert 70.6% of those plays into outs.

If you look at each of the green boxes, you will see that balls that land in the same slice as the fielder is standing, the conversion rate is highest, and the more the ball lands away from that slice, the  fewer outs are made per play. (For the most part anyway.)  This is fairly obvious in terms of DIRECTION.  What we see here is the MAGNITUDE that this is true.

I'll let you digest this for a bit, then I'll post the infielder data.

Sunday, August 26, 2018

Zero-point of positional adjustments (part 5 of 6)

?A recap of outfielders, infielders, and catchers can be found in part 4 with links to the other 3 parts.

Now the DH.  Remember that we've introduced TWO types of positional adjustments: one for defense and one for offense.  

• Intra-outfield, we only have defensive adjustments.  
• Intra-infield (2b, ss, 3b), we also only have defensive adjustments.
• Inter-IF-OF, it's still a defensive adjustment, but based on the offensive value of their replacement pool.
• Catcher has two distinct adjustments, one for defense and one for offense.

DH is very similar to catcher.  It is much harder to hit as a DH than as a non-DH.  So, we need to apply an offensive adjustment.

Now, about defense.  Obviously, a DH does not "defend" anything.  But you have to do... SOMETHING.  Take for example Frank Thomas, who has some of the worst  stats as a fielder for 1B.  Let's say he is -10 runs per season as a fielder for 1B.  In the years as a DH, he obvious didn't help or hurt.  Or did  he?  He is taking up a spot as DH and so forcing his team to deploy a 1B.  If that fielder as a 1B is -5 runs per season, that fielder has more defensive value than Frank Thomas does.

In the end, the best way to make sense of this is to look at it from the point of view of "defensive value" rather than "fielding value". Fielding means actually fielding.  Defense is at a higher level that encompasses more than fielding.  It can include pitching. And it can include how to deploy the players, such that someone is at DH.

And the position we are taking with regards to the "defensive value" of a DH is that it is equivalent to the fielding value of a poor-fielding 1B.  That is, Frank Thomas, whether as a fielder or as a DH, has the same defensive value.

I know this is not necessarily the clearest of all the positional adjustments we have.  However, when you create a model you try to represent reality.  And the reality is best represented when the defensive value of Frank Thomas matches as 1B and as DH.  

Sunday, August 19, 2018

Zero-point of positional adjustments (part 4 of 6)

?With outfielders, we treated the players as one big pool, without needing a positional distinction.  We can directly calculate the defensive positional adjustment.

With infielders, we kept the three positions as distinct-but-related, with the three positions bridged by common players.  So, we can infer the defensive positional adjustment.

In comparing the pool of outfielders to the pool of infielders, we use the replacement pool for each to value the pools the same.  And therefore, the gap in offense among the replacement pool (not among the average player in those pools) is balanced by the gap in defense.  And that gap is a defensive positional adjustment.

Now, the catchers.  You COULD do the same for catchers.  You could treat them as its own pool.  You could take the replacement pool, and compare that to the replacement pool among infielders and outfielders.  And you could presume the gap in offense  among the replacement pool is balanced by the gap in defense.  You COULD do that.  Except catchers have a tougher time hitting.  And so PART of the  gap in offense and defense is purely attributed to this constraint catchers face.  Ideally, we'd separate it out  so that part of the positional  adjustment for catchers is because a .350 wOBA by a catcher is not the same as a .350 wOBA by a non-catcher.  That's because it's REALLY hard  for a catcher  to do that.  There's a catcher-penalty to hitting, much like there's a SP penalty to pitching (or a RP bonus if you prefer).  So, we SHOULD have an offensive positional adjustment AND a defensive positional adjustment for catchers.

In  the end, it doesn't REALLY matter because it comes out in the wash.  Overall, nothing changes.  But in terms of isolating the offensive and defensive production, it matters a whole lot.

***

We'll  talk  about DH next time.

Zero-point of positional adjustments (part 3 of 6)

In part 1, we looked at outfielders.

In part 2, we looked at infielders.

Now we need to look at infielders compared to outfielders.  Let's step away from baseball for moment and think of football.  We would never presume that an average QB = average OT = average TE.  There's nothing inherent about any of that.  And certainly, we wouldn't make a QB play OT or TE in order to value him as a QB.

The same thing applies in life, with regards to any product or service.  How do you evaluate these things?  In the end, it's how much you pay for it.  And if you had ten dollars, you've decided how much water and juice and salad and legal advice you will pay for, regardless of how similar the utility of those things are.

In sports, we look at the bubble players, those guys who are paid at or close to the league minimum, regardless if they are a QB or a goalie or a SS or a LF.  

It is not common for a player to switch between infield and outfield.  Indeed, virtually all of these position switchers is unidirectional, going from infield to outfield. See, within the infield (2B, SS, 3B) and within the outfield, those players are all part of the same pools.  There is no such thing as a pool of 2B.  Those guys are not only 2B, but also SS, because any player that is in the SS pool is automatically part of the 2B pool.  To some extent, the guys in the 2B pool may (but not necessarily will) be part of the SS pool. So we have an infield pool.

When it comes to IF to OF comparisons, it is closer to comparing a forward and defenseman in hockey than it is to comparing a winger and center.  As a result, we need to find a different bridge than we found for infielders.  And that bridge is the pool of players who are just hanging on.  And those players, if you look at the offense AND defense, will be equals.  How do I know?  Because they are all being paid the same, regardless of how much water or legal advice they provide.

And once you have established the pool of players that are equals, and once you have calculated their offensive value, the remaining value is their defensive value.  And if you have the UZR of these players in the infield and in the outfield, you bridge their UZR value to their defensive value through a positional pool adjustment (infield and outfield pools).

***

Next up: catchers and DH.

Saturday, August 18, 2018

Zero-point of positional adjustments (part 2 of 6)

?We looked at outfielders. Let's now look at infielders, specifically 2B, SS, 3B.  As we saw a bit earlier with outfielders, metrics like UZR will "force" the average 2B, SS, 3B to all be "0".  However, since each of these positions are their own universe, a 0 at 2B does not mean the same thing as a 0 at SS.  This is most clear in high school where it would be impossible for the better fielder to be at 2B instead of SS for any team.  League-wide the avg SS would be far better than the avg 2B.  But if you had a system like UZR, it would force both to have an avg of 0.

So, what to do?  Unlike outfielders, each position has its own responsibilities, so we can't (easily) compare what a 0 at 2B matches at SS.  We are NOT asking "what would a 0 at 2B do at SS".  We are NOT asking "what would a 0 at SS do at 2B".  What we are trying to do is find some common baseline to bridge the two universes.  In other words: all other things equal, what would you trade a 0 at 2B for if you wanted a SS?  And what would you trade a 0 at SS for if you wanted a 2B?  Or, if you paid a 0 at 2B X number of dollars, what kind of fielding at SS would you need to pay the exact same X number of dollars.

And one way to get there is to find a low paid infielder who spends alot of time at both positions.  Indeed, you will find dozens of such players.  These guys are paid the league minimum, they spend as much time on an MLB roster as they do in the minors, they plug holes at 2B, SS, and 3B.  Those players provide a common baseline.  We know how much they are worth, and we know how well they compare as fielders at multiple positions.  These guys are the bridge.  

And so we can compare Altuve and LeMahieu and Schoop to this bridge.

And over in the SS universe, we can compare Correa and Simmons and Lindor to this same bridge.

And so, without our two universes of fulltime 2B and SS ever intersecting, we can compare those universes by a bridge of players who have the same value whether these players are playing 2B or SS.  And these guys we know their fielding values at 2B and we know their fielding values at SS. 

And that's how we bridge the UZR values of Altuve and Correa, without either one ever playing another position, nor do we need to entertain the idea of them playing another position.

***

Next installment, it's the catchers and DH.

Zero-point of positional adjustments (part 1 of 6)

?Positional adjustments do two things, maybe three(*), which we wrap into one.  When we do that wrapping we lose sight of those two or three things.

(*) I'll tell you if it's two or three after I finish writing these blog posts.

Let's take the easy one.  When you look at something like UZR, those metrics establish a zero-point at the positional level.  In effect, each position is its own universe.  You CANNOT compare the zero-point across positions.  More specifically, since the average SS in UZR is 0 and the average LF in UZR is 0, these two 0 are not equal to each other, even though they both show 0.  0 <> 0.  That's because one is 0 in the SS universe and the other is 0 in the LF universe.

However, in Catch Probability, LF, CF, and RF belong to the same universe.  In UZR, the avg CF is 0 and the avg LF is 0 and the avg RF is 0; in no way does UZR, by itself, allow us to directly compare CF to LF to RF.  Catch Probability however DOES allow us to compare the three because  it doesn't treat the three as three positions, but rather one: outfield.  And it  compares each outfielder to ONE common standard. And so the average OF  = 0.  What is the average CF?  This is important: in Catch Probability it does not "force" the average to be anything.  We can actually let the model tell us what is an average CF.  If for example, for some reason, Hamilton and Buxton and Inciarte and Kiermaier and the rest of them all became full time LF, then the average LF would end up being better than the average CF.  But Catch Probability doesn't force that.  It  simply uses one baseline for the entire outfield.  And we can then determine  the quality of the avg LF, avg CF, and avg RF based on the results of  the system.  

As it turns out, these days, the avg CF is 5 runs ahead of RF who is 5 runs ahead of LF.  But certainly in prior decades LF = RF, and potentially LF > RF in the really early days. That's not the world of today however.

Coming up: we'll talk about infielders, catchers, and DH.

Monday, August 13, 2018

Statcast Lab: Fielder Roles

?As we know, in this day and age where Javy Baez can play anywhere on the field, even pitch to pitch, the designation of "2B" and "3B" is not very helpful, analytically.  And it is especially not helpful when an infielder is part of a 4-man outfield, yet maintains his infielder designation.  I generate a warning report when an infielder makes a putout deep in the outfield... only to find that it might be Kris Bryant who officially maintained his 3B position.  His ROLE however was quite different.

So as to maintain some semblance of  continuity with the 1-9 position designation we've come to know and love, but to enhance it to make use of how the players are placed, we are working to create ROLES. This is a first pass.

Read More

Saturday, August 11, 2018

Nola and his fielding adjustment

?This is just a collection of my tweets from yesterday and today.  The basic point is that Nola (a) has one of the best BABIP and (b) plays with one of the worst fielding team.  And so, it boils down to: how do you adjust for his context?

And more specifically: do we treat his fielders as having the expectation to play at their typical fielding level FOR THE SEASON or ON THOSE PARTICULAR TIMES WHEN NOLA IS ON THE MOUND?  It's a nuanced distinction that has a very specific implication to Nola.

Here they are:

Read More

Friday, August 10, 2018

Statcast Lab: Catch Probability: CF v Corner OF (updated)

?There was a pretty fun thread last year on this issue.  I suggest you spend 10-20 minutes reading all that first.

...

Back already?  Ok, so just to followup with current data.

1. Is there a bias in catch prob based on CF and corners?

Using the same matching method noted in the main thread, I have 211 outfielders with a total of 11907 plays (roughly 32 162-game seasons), and these players were +19 outs above average in CF and the same  players are +32 outs above average in the corners.  Pro-rated down to a single season, and that works out to:

• +0.6 outs in CF
• +1.0 outs in corner

Therefore, we can conclude that catch probability is not biased by CF/corners.  To the extent that it is, we are talking about less than half an out per season.

***

2. Having established that we have no positional bias in the metric, what are the outs above average (per play) by position, league-wide for all players?

• CF: +0.013 outs per play
• RF: -0.003 outs per play
• LF: -0.014 outs per play

On a seasonal basis, and applying the opportunity of each position we have:

• CF: +6
• RF: -1
• LF: -5

***

3. What is  that in runs and what about the arm?  

You can multiply the above number by a bit over 0.8 to get the runs.  You can also figure that the ARM is about +1 for RF, 0 for CF and -1 for LF.  So, we'd have these presumed run values:

• CF: +5
• RF: 0
• LF: -5

***

4. So what does that  say about the positional adjustment?

You can make the case that this that we use in WAR:

• +2.5 CF
• -7.5 RF
• -7.5 LF

Should be this:

• +1 CF
• -4 RF
• -9 LF

Sunday, July 22, 2018

Statcast Lab: Outfield fielding components of Reaction, Route, Burst, Speed

?I've been saying that the pinnacle of sabermetrics is convergence of scouting and performance analysis for 15 years. Here's one such exchange back in 2006:

I think people like to associate "numbers" and performance analysis to sabermetrics, and relegate scouting and observation as some ugly duckling. Sabermetrics is about the search for truth about baseball. And, at its core, baseball is about the physical and mental abilities of its players, which manifest themselves in explosions a handful of times in a game. Since we have limited samples in which to evaluate a player by his performance, we need to supplement that with some keen observations. The pinnacle of sabermetrics is the convergence of performance analysis and scouting.

And for the last 15 years I've been showing that belief in running the Fans Scouting Report, nicely hosted at Fangraphs, and un-nicely on my site. There's not a single name in that top 30 list that you could think "nah, way too high". And I think the reason this project works is because I broke it up into components:

?

And if you follow the headers left to right, you can see what I did: I followed the path of the ball. The blue is before contact, the green is post-contact, and the purple is post-catch. So rather than ask the fan, to aggregate 400 plays of a fielder of which they may have seen 100 of them into one number, I instead asked the fan to focus on components. Not only did the focus on components free them from the potential bias in advanced stats like UZR, it also gives us a glimpse in a player's fielding profile that we'd otherwise not get.

Until now. This is what we're working on.

?

We already have each fielder's overall performance on Savant. What we're working on is HOW. The above represents the top fielding performances on 2+ star plays. We can see the unsurprising names (especially if you've been following closely).  

The metric is in terms of feet (relative to the outfield average). Why feet, instead of time? That's a good question. In constructing such a metric, or any metric, you have to approach it multiple ways to understand the benefits and costs of doing it each way. And in this case there are two avenues:

• set your thresholds based on time, and measure your metric in distance
• set your thresholds based on distance, and measure your metric in time

Both are valid, both are reasonable. There are three reasons the PRIMARY view of the metrics will be based on distance:

1. The range in feet is going to be wider than in seconds, so that we don't have to go to two-decimals like we would for time. We could potentially go just with integers with feet. Would you prefer to see a range of feet of +/- 5, or a range in seconds of +/- 0.20?
2. You can SEE distance. When an outfielder misses a play by "that much", we see "that much" as having missed the play by say 2 feet or 5 feet. We do not think that he missed it by 0.08 seconds or 0.20 seconds.
3. Route. If you wanted to measure the indirectness of a route, would you think in terms of his indirect route added 6 feet to his run, or that it added 0.25 seconds? Especially if you tie it in to #2. His indirect route added 6 feet, he missed it by 5 feet. There's the story.

Anyway, that's what's in the lab, it's what I'm working on right now. And you can see early returns on twitter. Or keep following along in my blog. Eventually we'll encapsulate the entire play to include positioning (so that the reader can decide whether to include it for the fielder or for the team), as well as post-catch results (throwing). And as we are ready to rollout, you'll see it all on the Savant pages. More to come...

Monday, July 02, 2018

StatCast Lab: Infielder Spray on Broadcast

?This weekend we rolled out a new snazzy graphic with our broadcast partners:

The idea, here anyway, is straightforward: show the batter's spray pattern over his last 100 batted balls that landed within 200 feet of home plate.  As we've previously discussed, the positioning of the fielders to distinguish between what is an infielder and what  is an outfielder is at 220 feet.  Therefore, in order to look at the landing spot of the ball, it would have to be less than 220 feet.  Following a similar process, this is how many plays initiated by infielders and outfielders based on the landing spot of the ball.

?

The criss-cross point is 211-212 feet.  If we focus on the point where 80% of  the plays are made by the  infielder, that's right around 200 feet for the landing distance. And so, that's where we settled.

As for the number of batted balls, it was clear that early in the season, we couldn't rely on season-to-date, and so, we approached it on the basis of a rolling average.  In terms of the spray tendency, the amount of regression required was fairly low, so using 100 is right where we were comfortable to showing the signal. I'll update this post in a bit with that research.  

Anyway, back to the charts, they look pretty sweet as it reflects the open slice matching the low spray numbers among the five slices.  In other words, we approached it on the  idea that we have 4 infielders, and so the question for the manager  is "where do I give  up the hole?".  And in those two images, you can see why the manager did what he did.

Naturally, we prefer doing one based on "true talent", and so a "spray forecast", that includes the batter and pitcher, and potentially the base-out situation.  But, for  broadcast purposes, simply going with "what has the batter done" was the preferred approach.  As much  as possible, we should stick to the facts, when the facts can tell most  of the story.

Sunday, June 03, 2018

MGL on testing catcher framing

?One of the virtues of Twitter is the ease we can post things.  And because of that, it's easy for some great posts to get lost under the avalanche of other tweets.  Here's what MGL said:

Group I pitchers pitched with a good framing catcher in year 1 and a random catcher on another team in year 2. 

Group II pitched with a bad framing catcher in year 1 and a random catcher in year 2. 

How did both groups' RA9 change from year 1 to year 2? 

If catcher framing were a thing, we'd see Group I go up a lot more than group II. 

Group I, the pitchers with good-framing catchers in yr 1 (by an avg of .12 runs per game) saw their RA9 go up in year 2 by .28 runs. 

Those with bad framers (avg of .11 runs per game), saw their RA9 go DOWN in year 2 by .09 runs. QED.

In other words, MGL focused on team-switchers, and broke them up into those pitchers who were paired with either a good framing catcher (group 1)  or bad framing catcher (group 2).

And what he found is that when the pitcher no longer had the benefit of his good framing catcher, his RA/9 went up by 0.28 runs, while the pitcher who no longer was impeded by his bad framing catcher has his RA/9 go down by 0.09 runs.  That's a swing of 0.37 runs.  

(The framing numbers would have suggested only 0.23 runs, so if anything the framing numbers seem muted, though of course with a small study, random variation will impact results to some extent.)

Note that these numbers are in-line with previous tests by both MGL and Max Marchi several years back.

***

On a related note, MGL used the same process to test how pitcher do when they go from a good to bad fielding team (and the opposite). And the results were similar: the range of the delta in RA/9 was 0.25 runs whereas we would have expected 0.18 runs. Again, this shows  that UZR is indeed capturing some fielding effect, just as catcher framing numbers are capturing fielding effect as well.

***

This is one of the simplest if not clever ways to test a system.  And you rarely see it  being done.  And MGL just tossed it out there without hardly anyone noticing. 

Wednesday, March 28, 2018

When is an infielder an outfielder?

I checked to see where all OF played, and where the three IF (4,5,6) played.  Beyond 222 feet, it was 100% outfielders.  Less than 213 feet it was 100% infielders.  In-between, this is what it looks like:

?

In this gray zone of 214-220 feet, there were 17 fielders of which 8 were IF and 9 were OF. (This is 2017 only.)

Another way to look at it is purely based on counts.  And here we see that under 200 feet and over 240 feet, the counts are noticeable.  So, the gray area is somewhere around 200-240, in terms of deciding what does it mean if an IF plays at 200-220 and an OF plays at 220-240.  At that range, is an OF actually now part of "the shift"?  Is an IF in that range actually now a "4th outfielder".  So, that's the gray area to discuss, and where we can decide to set the IF/OF line.  I have it at 220, but I can certainly see how it can be moved anywhere between 210 to 230. (Image below is clickable.)

?

We can even say that you are an "infielder" if you have a reasonable chance of getting the batter out on a force play at 1B.  If you can't, then you are really an outfielder, there to catch balls on the fly.  To that end, maybe an infielder has to be shallower, like at 190-210, and what we see as deep "shifts" is in fact not a shift, but a 4th outfielder.  Anyway, lots to think about.

What say you?

Sunday, March 25, 2018

Shift v NoShift Preliminary Statcast Research

Here's my preliminary research:?

Shift v NoShift (pdf)

Below you will find MGL's initial response to this research, followed by my response to him.

Read More

Tuesday, January 23, 2018

Statcast Lab: Catch Probability: CF v Corner OF

?There were 180 players who played CF and one of the corner OF positions in 2016+2017 (treating that time period as a single time period).  Marcell Ozuna is the best example of this, as most of his CF playing time was in 2016 and most of his LF time was in 2017.

Ozuna was -0.007 outs per play in CF and +0.014 in the corners.  On the flip side was Aaron Hicks, +0.022 in CF and +0.003 in the corners. 

Note that Catch Probability doesn't distinguish between CF and the corners in establishing a baseline.  So, Ozuna/CF is being compared to the same generic player as Ozuna/LF.  When we add up all 180 players, weighting on the lesser of their playing time of CF/corner, all of these outfielders were +0.005 outs per play in CF and +0.003 outs per play in the corners.  With about 300-400 plays in a full season, this comes out to roughly being a difference of less than 1 out per season.  To the extent that there is some sort of bias in using the same baseline for CF and the corners, it is a very unnoticeable one.  

Therefore, we can use Catch Probability to compare between the outfielder positions.  Over the 2016-17 time period, the average CF was +6.2 outs above the generic outfielder, while the average corner OF was -3.1 outs, for a gap of CF to corner of 9.3 outs.

The corner OF was actually -4.8 outs in LF and -1.5 outs in RF, showing a gap of 3.3 outs in performance between the two corner positions.

Friday, December 15, 2017

Statcast Lab: How much does Buxton’s speed help in the outfield?

?As followers of Catch Probability know, Byron Buxton led MLB in Outs Above Average, with 25 more outs than an average outfielder.  His most impressive play was this five-star catch:

?

Where he had to cover 112 feet (compared to the RF who only needed to cover 103 feet) in 5.4 seconds of opportunity time.  An average outfielder, given that amount of time to run, would have covered it in 103 feet.  So, we can see Buxton needed an extra 9 feet to cover.  Since he ran at pretty much his fastest speed, this play is an easy determination to flag as a "above average speed was used" play.

Buxton had 25 plays where he needed to cover more ground than an average outfielder could cover AND he made the out.  Of those, only three times did he run at below-league-average speed.  So, we can say that 22 times out of 25, he needed his speed, and in the other 3 times, he needed... something else.

We're still in the process of defining the "something else", but it would essentially be related to either his out-of-the-gate quickness, or his burst of speed during the acceleration phase of his run.  In these three cases, he was barely below-league-average in his sprint speed, and the extra distance needed to cover was less than 1 foot.  Every foot counts, and since his lack of speed here is costing him about a foot, then he would need to make up for it in quickness or acceleration to ultimately have made the catch in these three cases.

Now how about #2 guy, Ender Inciarte, with 19 outs above average.  He is actually a barely-above-average outfielder in terms of his Sprint Speed.  Indeed, he's below average for a CF.  We've identified 24 times where he needed to cover more ground than an average-speed-outfielder when he caught the ball.  Only 7 times did he run faster than an average outfielder to make the play.  In the other cases, he made plenty of short-run catches, relying on his quickness and his acceleration to make the plays.

So, this is where we will be headed this off-season, identifying every catch (and eventually potential catch), to determine which of his quickness, acceleration, and speed played a determining role in making the play (and which he was lacking if he missed on a viable play).  And leaderboards to follow.

Tuesday, October 24, 2017

Fielding Positional Adjustments, part X of N

?Pizza does a good job of laying out some of the considerations and challenges.  He helpfully links to a good thread I had on the subject, so that saves me time in looking for it.  I could probably create five separate subthreads on what he wrote.  Let me just highlight a few things:

***

We’re not REALLY asking how would Smith do at 2B and at SS.  We are asking how would Smith/2B do compared to “generic fielder” (let’s call him Hubie Rice) at 2B and how would Smith/SS do compared to the same generic fielder Hubie Rice.  So, we aren’t converting Smith’s skillset.  We are simply comparing him to some generic guy who can.  It’s a bit nuanced, but, it keeps the argument away from “Jeter can’t play 3B, but ARod could”.  We don’t care.  That’s not the question.  I've contributed to the confusion by relying on the multi-position players.  But, the intent there was to use that as a guideline, rather than proof.

***

I then found players who had experience with at 

least 100 such ground balls at second base and another 100 at shortstop. Their “success rates” correlated at a mere r = .237. That’s … not what we might expect.

Actually, we haven't establish what to expect.  This is one of those things that we've discuss many times in the past regarding correlation and the number of trials.  Imagine for example you take 200 groundballs at SS, and split them up into two halves.  What would the correlation be between the two halves?  I don't know, but it won't be 1 and it won't be 0.  Could it be r=.237?  Maybe, in which case the above finding would actually be impressive.  

So, this is something that aspiring saberists can tackle.  Just make sure that the number of trials of the SS/sample1 to SS/sample2 and the SS/sample3 to 2B/sample4 has the same number of trials.  I'd love to see the results from you researchers out there.

Remember, outside  of systematic bias, you can make any two things with the most modicum of relationship achieve r=1 by having number of trials approach infinity.  This is why reporting a correlation without reporting the number of trials is telling an incomplete story.

Thursday, September 21, 2017

2017 Fans Scouting Report, voting now hosted at Fangraphs

?Fangraphs was kind enough to host the balloting.  Given the huge reach they have, this is a welcome change.  My thanks to David and his team, who are as topnotch as they come.

Help me, help you, help everyone else, and vote for your team:

http://www.fangraphs.com/fanscouting

Sunday, August 27, 2017

Statcast Lab: Catch Probability: Outs Above Average

?Coming soon, to a Savant site near you.  

Catch Probability is determined primarily based on the proximity of the fielder to the (eventual) landing spot of the ball, relative to where the fielder was at the release of pitch.  This becomes the "opportunity space" of all nine fielders.  

Unlike a batter, who we KNOW is at the plate, and who we KNOW has a strike zone that is fairly fixed and fairly similar to other batters, fielders are completely different.  First, there are nine possible fielders, so for any batted ball, we don't know before hand who is mostly responsible.  Secondly, the "catch zone" is unique for every batted ball.  Not only is the distance to cover different for every batted ball, but so in the time in which the fielder has to make the play.  Contrast that to a strike zone where the batter has a set time and set area to cover (more or less).

So, this is how Catch Probability works, for outfielders.  For every single batted ball, we decide who is the responsible fielder.  

• If an outfielder makes the putout, or the error, we assign that batted ball to him.  

• If there is a base hit, we assign whichever fielder was closest to the eventual landing spot of the ball (based on his starting position at time of pitch release).  The exception to the base hit rule is any batted ball that hits the fence at least 8 feet above the ground, or we have an outside the park HR: we throw those out.  We deem those unplayable.

Using distance-time, we take a (perfectly) smoothed version of this chart.  For any batted ball that is deemed "is-back" we lower the catch probability by 0 to 25% points.  Is-back is any batted ball that is +/- 30 degrees from straight-back.  The simplest way to think about the difficulty it is that we remove and extra 4 or 5 feet and an extra 0.1 to 0.2 seconds to the play.  It gets more involved, but that's the basic way to think about it, that we acknowledge that when you run straight back, you really don't have the same opportunity as otherwise.

The exception to the above is the wall-balls.  Any basehit that is within 8 feet of the base of the fence, or would have otherwise landed beyond the fence (if not for the fence) is given a catch probability of 0.  While this may seem otherwise generous to the outfielder who allows the ball to drop in for a hit, also consider the treatment of putouts, where the outfielder doesn't get the extra benefit of making a play at the wall.  Basically, these two treatments are in balance so we have no bias.  In the off-season, we'll be looking at treating wall-balls with more subtlety.  In the end, whatever the treatment, whther the simple solution implemented, or the more complex method we'll explore, the end result won't be more than a couple of outs above average.

Since everyone will ask about Fenway Park, the outs above average for leftfielders:

• Home fielders at Fenway are minus 5.7 outs, while Away fielders at Fenway are +1.4
• Redsox fielders at home are minus 5.7 outs, while Redsox fielders on the road are -2.1

It would seem that the treatment of Redsox LF is fair, or at least, not unduly biased.  In any case, the Catch Probability Leaderboards you will see will allow you to do breakdowns based on:

• outfielder (without regard to position, team, park)
• park, home-away, position
• team, home-away, position

We're getting into the top of the third...

Tuesday, April 25, 2017

What is it with those LF anyway?

We've talked quite a bit over the last several years as to how the LF is a very different position than the others.  One of the Straight Arrow readers suggested that it might be because the NL is treating the LF as a quasi-DH.  Pizza has some great data here.  It would be good to see if there's any kind of AL/NL distinction.

And from what I remember, the average starting LF is a worse fielder than his backups, and this was the only position that this was true.  That was a good 5-6 years ago I think, so it would be another thing to look at.