THT Mailbag: Bursting with questionsby Bryan Tsao
September 26, 2007
Speed of pitch
There is a difference in baseball release speed (when the pitch leaves the pitcher's hand) and the result speed (when the pitch crosses the plate). It is usually 8-10 mph.
Question No. 1: If a typical radar gun or in-the-ballpark pitch speed measurement says the "pitch speed" as 89 mph, is this the release or the result speed?
Question No. 2: What's the reference for the answer to question No. 1?
Mahalo nu`i for your kōkua!
- Paul S.
John Beamer: The speed measured by a radar gun tends to be closer to release speed than plate speed. However, unlike the pitchF/X system spitting out great data, radar guns are far less accurate. In fact it is difficult to know exactly where the gun is pointing, but in general it will be measuring a point closer to the pitcher (ie, release speed). For more info check out my column on the topic.
Could someone please analyze the so-called "no doubles defense"? Is there enough data to establish the effect of no-doubles, i.e., how many more singles and how many fewer doubles result? My sense is that Manny Acta is way over-doing it.
- Roger C., Fairfax, Virginia
Jacob Jackson: I think that the quick answer is that the defense they're playing is more a function of their personnel and the park they play in than it is a deeply ingrained managerial or organizational philosophy.
By a "no-doubles defense," I'm assuming we're talking about playing the outfielders deeper and shading toward the line in the corners. In my opinion, to adequately answer your question of whether the Nationals are over-using this from a statistical standpoint would require us to run the numbers with each OF combination the Nationals have used this season.
My hunch is we'd find they only do this with Nook Logan in center field. He doesn't get written about much, but Nook Logan's range is simply awesome. Project him over a full year in center field, and his plays out of zone would be comparable to Andruw Jones. Meanwhile, he's also a top five guy in RZR (.910).
Here's my point: If you have a guy with that much range in center field, it makes sense to have your left and right fielders hug the corners, because Nook should be able to get the vast majority of the balls in the gaps. As for playing Nook abnormally deep in center field ... I'm certain I've seen him play less than you have, but from what I have seen (a terrific highlight reel from Washington's top plays archive on MLB.com, and about 10 Giants games), he breaks in on balls extremely well, perhaps significantly better than he does when he's going back. (Unlike Andruw, who goes back on balls extremely well, and often plays shallow).
Also, this last point is counterintuitive, but playing deep somewhat mitigates Nook's below-average arm, because he's so fast that he can charge anything in front of him and get some momentum on his throw. With his arm, he'd never throw out a baserunner if he had to back-track on a ball and throw flat-footed.
But again, to study the cost/benefit analysis of the doubles they're take away and singles they give up by using this strategy is a really complex issue, and whatever we found from this season wouldn't be definitive because of the sample size. I'd want to break down the data into each outfield combination they used, and perhaps even the types of hitters they used it against (are they doing this against pitchers? Slap hitters/speedsters? Power guys? etc.). They may even be choosing to use it only with some of their pitchers, and not with others, which would make our conclusion even less definitive.
I have no doubt, however, that if you surrounded Nook Logan with a conventional outfield alignment, and always played your corner guys at "90"—90 feet from the foul line—you'd be squandering his most valuable asset as a baseball player, because you'd have two guys getting to every ball in the gaps and nobody getting to balls down the line. For my money, that's the most likely explanation for the defensive positioning you're seeing Manny Acta employ.
Dave Studeman: Roger, I don't know the answer to your specific question, but I did take a look at how the Nationals perform on ground balls. Overall, 25.9% of the Nationals' ground balls have been hits, compared to the major league average of 25.0%. Believe it or not, this relatively small percentage difference, along with the Nationals' high error rate on ground balls, makes them one of the worst ground ball fielding teams in the majors.
Twenty-two percent of their ground balls have been singles and 3.9% of them have been doubles. The major league averages for these two figures are 21.7% and 3.3%. So, the Nats have actually allowed relatively groundball doubles than singles (18% more doubles, 1% more singles).
So, if Manny Acta is playing his corner infielders closer to the line than most managers, I wonder if it's because he's been burned a bit too much this season.
I've been thinking this over for a while, and asked a guy at work who has been around minor ball for years what an agent gets. He said 30% at least, and so I look it up and no ####. So that means some fat-assed gas bag with three or four hard working ball players is making major league coin without any chance of injury, other than biting his tongue when
he chows down on a juicy steak. Just how stinking rich are some of these agents, and what kind of percent do they tear out of the players' children's food bowls? And are they worth it?
- Joe M.
David Gassko: Agents do not get 30% -- that would be an awful deal for a player. The number is usually about 5%, often less. So if an agent can get his players 5% more money (not unlikely), then he'll be worth it.
Dave Studeman: Ever met any hedge fund managers?
In discussing the merits of the "clutch" concept with a fellow baseball fan, he told me that "clutch" does not exist, but "unclutch" does exist. Personally, I think the statement is logically fallacious, but he said he'd like to see a study done of players who perform worse in "clutch" situations only (I presume since most studies focus on players who are "clutch").
First, is the statement actually logically fallacious as I believe it to be?
And secondly, would such a study actually prove anything, in your opinion? If it would, is there any chance THT would do a study on a topic at which most sabermetricians scoff?
- Bill B., Philadelphia
John Walsh: I believe the subject of clutch hitting is one of the most intensely researched subjects in baseball in the last few years. You can read up on a bunch of these here.
I believe the general consensus thus far is this: There may be players who consistently do better than expected in clutch situations, but any effect is so tiny that it is hard to discern and indeed it's hard to identify such players (if there are any).
Chris Jaffe: Thanks for writing. I've always been intrigued by this notion as well.
Is it logically false? Hmm ...in the general population maybe, but professional athletes are a self-selecting sample size. Thus even if both extremes exist, you could get one but not the other in MLB.
Then again, by that logic, your friend's notion of one existing but not the other can be right, but he'd likely be picking the wrong one. A player who can't handle the stress would be likely to flop out well before making it to the majors, eliminating the unclutch from the sample size.
Even that could be overthinking it though. One person might not mentally melt to be "unclutch." He might feel a twinge more pressure that knocks down his game just enough to be noticable when the game's on the line. Under that circumstance, if he's good enough in general, his numbers could be unclutch.
As for doing a study on clutch and unclutch, I plan on writing (actually, I've got the rough draft in the can) an article for this year's THT Baseball Annual along those lines. It's tentatively titled, "The gods and dogs of garbage time."
I'm looking at everyone with 1,000 games played from 1957-onward and checking their splits at Baseball Reference. In particular, I'm looking at how they did in games where their team was ahead/behind by five or more runs compared to how they normally did. If a guy is "unclutch" he should hit far better then. The theory is that he isn't under as much pressure and can relax. And if he's worse then, that could aid a clutch reputation.
Short version: There are some very remarkable players in either direction, and a few who become very different players once the at bat doesn't matter. Based on the sample size of hitters, it could be fluke, but in a few cases there are some players who really transformed once the game was out of reach.
Does it prove anything? I dunno. But it certainly pertains to value.
John Beamer: Further to Chris' answer let me add a couple of things. One thing I'd caution you over is how you identify what is good vs. bad "unclutch" play. For instance, in a 15 run blow out the fact that a hitter strike's out is irrelevant as LI is effectively zero. Do you care that he struck out? Is that good or bad "unclutch" play?
For me the way to define unclutch is to look at the Leverage Index of the play and if it is below a certain number then bucket it "unclutch". Chris' method for the 2008 Annual is also very valid and I'll be keen to read his conclusions. Studies like that always yield fascinating insight, but Chris what mentions probably predicts little.
I have a question about UZR. A lot of sites have Troy Glaus' zone rating at .737, which is the worst in the American League. That makes sense seeing that he has been hobbled by plantar fasciitis this season. However, Hardball Times has his UZR at .706, which is among the best in the AL. He also has more balls fielded out-of-zone than most players, which makes his range look like the best in the AL other than Brandon Inge. I thought that UZR was just ZR separated into two different components. How could it give such a different impression of a players' range?
- Jonathan G.
David Gassko: First off, I think you're confusing two statistics here: UZR and RZR. I know, the acronyms can be confusing. UZR is "Ultimate Zone Rating," a statistic created by Mitchel Lichtman which uses play-by-play data and a large number of adjustments based on that data to spit out how many runs a fielder has saved versus average. A short explanation of UZR (and some other defensive metrics) can be found here.
RZR is "Revised Zone Rating," a statistic which we run here on The Hardball Times which measures the percentage of opportunities in his assigned zone that a fielder has converted into outs. Among qualified players, Glaus is fourth in the American League in RZR. Our statistics are provided by Baseball Info Solutions.
However, ESPN and Sports Illustrated buy their statistics from Stats Inc., which keeps a different definition of Zone Rating and shows Glaus with both more chances and a significantly worse performance. Who's right? I don't know, but if you want to read more about the differences between the two systems, I suggest reading this article by Sean Smith.
Just looking at the Win Shares today. I think you still have Ichiro Suzuki as a right fielder in your Win Shares calculations. He has only played center field all season. His defensive Win Shares value is only about half of what it should be, consistent with a corner outfielder.
- Ken A.
[editor's note: This question was originally submitted on Sept. 14]
Dave Studenmund: Actually, the Win Shares system doesn't reference which specific outfield position a player has played. All outfielders are put in one bucket. That's because Win Shares are built for historical purposes, and we don't have specific outfield positions played for all of baseball history.
Win Shares recognizes the extra worth of center fielders by awarding "range bonus plays" to all players, and this has a particular impact giving center fielders more credit than their outfield peers. At this stage, Ichiro has 92 Range Bonus Points. Among other top AL center fielders, Coco Crisp has 108, Curtis Granderson has 100, Gary Matthews Jr. has 64, Torii Hunter has 63, Grady Sizemore has 49, Vernon Wells has 48 and David DeJesus has 11 (gulp).
The issue, as you've pointed out, is that all of these guys have more fielding Win Shares than Ichiro. Like you, I wondered how this could be?
Ichiro's "problem" is the Seattle outfield has the least Win Shares claim points of any outfield in the AL. Because Win Shares is a top-down system, that means that Ichiro has the opportunity to receive fewer fielding Win Shares than any other AL outfielder, despite the fact that he has some very good fielding stats (and he seems to spend a lot of time covering for the rest of Seattle's outfield, too).
Like you, my sense is that the top-down nature of Win Shares is handicapping Ichiro's standing in fielding Win Shares, perhaps by three or four fielding Win Shares. Given that he's already third in the league in Win Shares, a boost to his fielding total would bring him pretty close to A-Rod in the Win Shares MVP race.
I always struggle with this issue. I've toyed with the idea of totally revamping Win Shares for our site, making use of the most recent advancements in baseball stats (such as PBP fielding stats), changing the name, etc. But then you would lose the historical continuity that Win Shares gets you, and which many of our readers value. Also, it would be a lot of work, and I'm inherently lazy.
Still, perhaps we'll tackle this issue some day.
I don't know that this is really a question for the mailbag or just an interesting research project to look into where you can do a better job than I could.
I don't think there's any question anymore about whether or not Barry Bonds took steroids. Even amongst Bonds supporters (of which I have been one), his own testimony indicates that he did. Whether or not he knew what they really were or was duped by people he trusted is another issue that we may never know, but he definitely took them.
Still, does anybody think that he's still juicing? I mean this story broke in 2003 and so this is four years later now. Back when he was in the midst of those superhuman seasons all of his detractors said that the only reason he was doing so well was because of the juice he was on. He's likely retained some additional strength that he wouldn't have had, but otherwise by now any effects of the drugs should be long gone. Way long gone. Isn't his season so far somewhat reminiscent of those though? Considering just how much better than everybody else his numbers are? Especially considering he's now that much older? Doesn't what he's doing now maybe make one take a step back and wonder just how much of what he's done was drugs and how much of it was Bonds?
I can't tell you how many times I've seen the "Well, he'd never hit more than 49 home runs in a season before hitting 73" argument thrown around. But he hasn't hit more than 49 in any season since then either now has he? 2001 is the one outlier in number of home runs, not every year since 1999. I've also seen how his at bat per home run numbers are so much better now than they were in the '90s so that must prove it. They are better. From 1992 through 2000 he averaged one homer every 12.5 at-bats, with a low of 9.8 in 2000 (which wasn't much different than the 10.6 in 1994 by the way). In 2001 he dropped to 6.5, and has gone 8.8, 8.7, 8.3, 8.4, 14.1 (injured last year), 8.0 (so far this year). That's
a very noticeable and marked difference (also very consistent), however, this is also the same period of time when Bonds has been getting walked at an unbelievable rate.
His home run per plate appearance numbers are not far off where they always were: 14.4/homer in 1993, 15.9/homer in 1996, 12.6/homer in 1999, 13.1/homer in 2002, 11.9/homer in 2003, 13.4/homer in 2004, and 11.4/homer so far this year. His numbers are maybe a tick better now than before, but I believe the difference is in line with the difference in the league as a whole over that time. In 2001, by the way, his number was 8.9/homer, so once again only 2001 is an outlier. (Note: I only added walks in for plate appearances, I don't figure the other modifiers happen enough to make much of a difference and it was the difference the walks made that I wanted to note.)
I guess what I'm getting at is that I've recently seen speculation that without drugs he'd likely be approaching 600 now and this whole 755 thing wouldn't even be an issue. I'm just wondering if maybe the truth is that he'd be approaching 700 or have already past it and be approaching Ruth now with people looking to next year maybe being the year, and with his production levels saying it's likely to happen. I mean unless people think he's still juicing, at what point can we look at what he's doing and say, "Man, this guy Bonds is pretty good."
David Gassko: My favorite method for measuring a player's home run power is to divide home runs by at-bats minus strikeouts, which is the same as batted balls. Here are Bonds' numbers in that category over the past 10 years:
1998 - .080
1999 - .116
2000 - .122
2001 - .191
2002 - .129
2003 - .136
2004 - .136
2005 - .139
2006 - .082
2007 - .099
You can see that there are basically three distinct numbers here. Between 1999 and 2005, and again this year, Bonds has hit a home run about 13% of the time he puts the ball in play. In 1998 and 2006, that number was 8%, and in 2001, it was an astounding 19%. Now what does that all mean?
First of all, 2001 is a clear outlier. Yes, steroid use may have helped Barry hit more home runs, but steroid use alone did not propel him to the single-season record. With or without steroids, 2001 would have been a career year for Bonds.
Secondly, we see that Bonds' home run rate last year was more similar to his home run rate before he allegedly started taking steroids. This year, it's around 10%, which is up, giving us more evidence that Bonds' steroid use helped him hit a lot of home runs.
But I guess the pertinent question is where would Bonds be without steroids? Assuming his steroid use lasted from 1999 to 2005, let's adjust his numbers in those seasons in three ways.
First, let's lower his walk rate because pitchers wouldn't be pitching around him as much if Bonds wasn't taking steroids. I'm going to assume his true walk rate was about one walk per five plate appearances, which is how often he walked in '98. That has the effect of giving him 294 more at-bats over that time period.
Next, we'll adjust Bonds' strikeout numbers to reflect the extra at-bats by multiplying his strikeout rate in each year by his new number of at-bats. That adds 39 strikeouts to his ledger.
And finally, we will multiply Bonds' new at-bats minus strikeouts number by 8%, except for in 2001, where we will up that by about 60%. In all, Bonds loses 86 home runs, which would put him at 687 now. That jives with other estimates I've seen, and what it tells us is that Bonds was certainly a great hitter with steroids or without, but he would need to play healthy through 2009 to challenge Hank Aaron.
Walks and runs
Watching a Phillies-Cubs game, the commentator said that in almost all innings wherein teams score more than one run, there is a walk involved. Is there any statistical evidence for that, or is it just something he said?
John Walsh: While, it's true that getting walks certainly helps scoring multiple runs per inning, the announcer you mention was exaggerating. This table shows the number of times a certain number of runs was scored in an inning (in 2006). For each run total, I also give the fraction of those innings where at least one walk occurred:
+------+-------+----------+ | runs | num | walk_pct | +------+-------+----------+ | 0 | 30665 | 0.2191 | | 1 | 6768 | 0.3503 | | 2 | 3157 | 0.4881 | | 3 | 1557 | 0.6256 | | 4 | 726 | 0.7107 | | 5 | 293 | 0.7406 | | 6 | 120 | 0.8000 | | 7 | 51 | 0.8235 | | 8 | 25 | 0.7600 | | 9 | 13 | 0.8462 | | 10 | 5 | 0.8000 | | 11 | 3 | 1.0000 | +------+-------+----------+
I can also show you what happens when you divide all innings into less than 2 runs scored and 2 or more runs scored:
+----------+-------+----------+ | runs_inn | num | walk_pct | +----------+-------+----------+ | <2 | 37433 | 0.2428 | | >=2 | 5950 | 0.5753 | +----------+-------+----------+
Obviously, there are more walks in the higher scoring innings, but unless you consider 57% of the time "almost always", the announcer was indulging in some hyperbole.
I think I have a good question, if a player starts out the year and then is moved up to the big team, how do they handle his contract? Obviously he has a contract for say Triple-A ball but what happens when he is moved up? Is there something in their contract that stipulates what additional monies they would get if they are moved up? Also what about for a big team player that is sent down to the minors, any pay adjustments there?
- Michael D.
Dave Studeman: The key question is whether the player in question was on the 40-man major league roster. Every player on the 40-man roster has a major league contract, which includes his salary for the amount of time he is on the 25-man active roster. For virtually all minor league players on the major league roster, there are typically two pay scales outlined in the contract: one for his time in the majors and another for his time in the minors. If a player is called up from Triple-A and has to be added to the 40-man roster, he signs a major league contract at that time.
For big team players that are sent to the minors: if he has a minor league pay schedule in his contract, then that schedule will apply. Otherwise, he'll continue to get paid at the major league level. Most players who have been in the majors at least one year have only major league salaries in their contracts.
Bryan Tsao is the editor of The Hardball Times website. He welcomes comments, questions, and suggestions for both himself and the site via email.