Actually, looking at MILB's internal stats portal (I work for a Minor League team and have access to stats that might be harder to find otherwise), it looks like Hayes intentionally walked 9 batters that season. At least, looking at the 2006 season numbers. However, when looking at his career stats (through 2008) it actually shows only one IBB. So, I'm not sure what's going on there. But, he definitely walked at least one and possibly six batters intentionally in 2006, so his rates are even better than advertised.
February 12, 2009 at 6:43 PM
Anonymous said...
The Quiz threw probably around 80 to 81. I can't say for certain because he was around before radar gun speeds were shown to television viewers. Having seen quite a bit during his best years, I'm pretty sure he hit low 80s, and maybe dipped down to 77-79 for his 'off speed' pitches.
February 12, 2009 at 7:10 PM
Anonymous said...
I like this piece because of the specific content (like all of us I am a die-hard Royals' fan) but also because it provides an important perspective on the use of statistical analysis. A point to be gleaned from Rany is that there is much to be learned about any system from its statistical outliers... from the atypical individuals in a sample compared with the typical.
The glorious world of sabermetrics provides us with detailed valuations of each player's skill set. However, in my opinion, there are two flaws to the discipline (beyond its relationship with defensive aspect of baseball). The first is a fundamental flaw of sabermetrics, the other of sabermetricians.
First, sabermetrics assigns equal value to every at bat. Any one who has played sports at any level knows this not to be ture. One plays tennis harder at deuce in a 4-4 set than when winning or losing 5-1. It is more difficult to get a hit in the third inning of a tight game than the late innings of a blow out. A good player in any sport knows when the game is on the line or MAY be on the line.
... but this is for another day.
A fundamental flaw of sabermetricians is an obsession with the usual rather than the unusual. Major league equivalents (MLEs) are the perfect example. I have an MLE Excel spread sheet that I use at the beginning of every fantasy baseball season to evaluate rookies who appear ready to make a big league roster. I type in the player's AA or AAA stats and the embedded formula, which was devised by a dedicated sabermetrician who poured over stacks of empirical data, cranks out the player's expected big league numbers for the upcoming season.
Now it is fascinating that these models work so well and so often... and it says something about the game of baseball and the difference in typical skill sets between the minors and the majors that these relationships are so often preserved. But more amazing are individuals who confound the models... or who simply have generated a few statistics that are so far to the right or left of the Gaussian distribution as to inspire wonder.
Submariners and kunckleballers, in general, may well be just such a case, as Rany observed. These players seem to have a skill which limits the utility of some statistical comparisons.
Another example are hitters who strike out exceedingly rarely, such as Alberto Calllaspo. His low K % may be more than just a statistical oddity. It may reflect some skill... some difference in approach or some biologic ability... which other players lack. It would not surprise me to read an analysis correlating visual acuity with low strikeout rate, for instance.
But this is the point. Competitive athletes (usually) are not competing to please the sabermetricians. To quote Herm Edwards: "you play to win the game." So if a statistical oddity presents itself, the oddity quite possibly reflects some talent (or more commonly, deficiency) that a player possesses relative to the group... unusually deceptive spin on a pitch, a unique inside-out swing, better performance enhancing drugs... all of these are examples of competitive advantages that baseball players possess which lead to a statistical oddity. And I believe it is a fundamental role of the sabermetrician to identify outliers and look back at the game played on the field in an attempt to explain them.
Rany is to be commended for the observation that sidearmers and knuckleballers frequently defy typical rules of projected pitcher development. And after years of reading his blog... I am certain that he has other theories to share from his approach to statistical analysis, as well.
So thanks, Rany, for another great post. Your writing has been thoroughly enjoyable of late...
WZ
p.s. you are right about the Quiz. He threw 79-81 and referred to it as his "Peggy Lee" fast ball. Hitters looked at it and said "is that all there is." (his joke, not mine.)
February 12, 2009 at 7:18 PM
Anonymous said...
walt --
one of the first things that baseball statistical analysts do, in trying to prove or disprove an effect, is to measure the persistence from year to year.
so your devastating critique doesn't really carry the punch you think it does.
February 12, 2009 at 10:58 PM
Anonymous said...
I'm pretty sure Walt was just offering his opinion, Jason, rather than a "devastating critique." I mean, we're all experts in our own opinions. Holy crap, man, drink a beer or something and freaking relax...
Rany, I guess you are too young to remember Randy Jones. Jones couldn't break 76 with his fastball, but posted back to back 20 win seasons in the mid seventies with his junkball sinker and slider. Jones won the NL Cy Young in 1976 with 315 innings and 25 complete games. I would really like to know how he pulled it off, since I was too young at the time to understand it.
February 13, 2009 at 2:29 AM
Anonymous said...
jason- i think you and i are saying the same thing.
most analysts do quantify variance from year to year and generate fantastic models based on them.
my point is that beyond crunching the numbers and making predictions based on them, it takes greater skill to identify a subset of players who do not seem to follow the typical performance trends of the other thoroughbreds... and then ask "why do these players perform differently?"
so i had hoped publically to laud rany for his analysis rather than criticize anyone. and to imply that although he has always been my favorite contemporary sportswriter (poz #2), this bit of analysis particularly tickles my baseball sensibilities.
i am sorry if the last post (which i did make at the end of a long day, with a beer in my hand) looked more like a criticism of sabermetricians in general rather support for rant.
I would sure rather have given this kid a chance than blow $9M on Farnsworth.
February 13, 2009 at 8:57 AM
Anonymous said...
one last thing...
most of us don;t have access to statistical databases to test our hypotheses, so we can't be experts. but as a non-experts, i would like to offer a small hypothesis:
i do believe that there are players who are slightly better than their numbers and others who are slightly worse. and i believe that good statistical tip offs on each side of the ball are range factor and low strikeout rate.
as such, i have high hopes for callaspo.
also, i would like to offer a hope that dayton is scouting pedro martinez at the world baseball classic. it is my great hope that pedro throws 100-120 innings in twenty starts for KC, then moves to the pen to set up for soria during the playoffs.
go royals.
wz
February 13, 2009 at 9:07 AM
Anonymous said...
sorry, again. i am just getting so jacked for baseball season.
i know everyone wants to say that callaspo lacks power and range, that alex gordon is somehow worse defensively than he was at nebraska, that glass won't offer pedro the two year deal it would take to outbid the market for pedro martinez.
but since it is february, please indulge me a scenario.
royals lead 4-3 in anaheim heading into the bottom of the eight in anaheim in the first game of the ALDS. meche takes the hill to open the eight with pedro and mahay throwing in the pen. the spork subs for callaspo. coco and dejesus cheat a couple of steps to their left so guillen doesn't have to cover quite so much ground... and as abreu, guerrero and hunter stare down our 6'3", 220 lb cajun... things are looking pretty good.
wz
February 13, 2009 at 9:26 AM
Anonymous said...
Happy Pitchers and Catchers Report Day Rany!!!!
February 13, 2009 at 10:37 AM
Anonymous said...
time to sign orlando hudson and start trading away the fat from the roster!
February 13, 2009 at 11:29 AM
Anonymous said...
Great initial post Walt. You may not want to ruffle the sabermetician's feathers, but I will, albeit it's a minor ruffling...
Sabermetrics ARE very accurate and a very useful tool. However, some people put TOO much stock into the numbers. The numbers cover the cookie cutter players almost flawlessly, but they DON'T account for guys like Hayes or Quizenberry. Because of this some of the sabermetricians insist that guys like Quizenberry CAN'T suceed, and there for guys like that never get a shot. There are exceptions to EVERY rule, unfortunately some of the stat heads forget that. MLB scouts certainly have.
February 13, 2009 at 11:33 AM
Anonymous said...
One of the most important things in any kind of statistical analysis is knowing your sample sizes and error bars. I think sabermetricians are well aware of this, but even then it is easy to forget in the heat of column writing. Even at BP, it occasionally turns my head how the writers draw glib conclusions from 150 PA samples.
So yeah, Wabbitkiller, as Rany pointed out, sabermetrics is limited. That shouldn't ruffle any feathers; it's a known and acknowledged fact.
There's not an exception to every rule; you just have to understand the rule.
February 13, 2009 at 7:10 PM
Anonymous said...
Hayes can't even fairly be compared with sidearm pitchers. They throw harder and rely more on strikeouts (like most pitchers). Submariners are in a class of their own. They get some strikeouts, but less than most good relievers. They give up very few walks. And then get tons of groundballs. That combination (decent number of K's, very few BB's and a very high GB%) is the profile of success for a submariner.
So, does Hayes have the kind of velocity, control, strikeout ability and groundball tendencies of guys like Bradford and Quisenberry? It's still too soon to tell, but Hayes's K and BB rates in the minors are very similar to Bradford's minor league numbers. We'll see if he can maintain the high GB numbers, particularly in the PCL. That will be a good test.
In short, I think scouts and prospect analysts are comparing Hayes to overhand and sidearm pitchers and that is like comparing a banana to a bunch of apples and oranges. Also, I don't think scouts and prospect analysts quite understand how important groundballs can be. A huge GB% and a concomitant low FB% can make up for a mediocre strikeout rate. That's how guys like Wang and many sinkerballers succeed.
February 13, 2009 at 11:01 PM
Anonymous said...
First off Rany let em just say thank you. As die hard Royals fans I am sure you are all as sick of hearing about "A-Fraud" (what a predictable moniker) through the national media as I am. Rany, your coverage of the Royals is what gets me through the off seasons.
Also Walt (and others) I wanted to say it is thoughtful responses like yours that make reading the comments section worthwhile so thanks.
Also, as I prepare for the onslaught of fantasy/rotisserie drafts this season I was wondering if maybe you could send me that spreadsheet analysis for prediciting rookies' production? Did you get this through BP?
February 14, 2009 at 1:36 PM
Anonymous said...
Sorry to digress from the sabremetrics discussion but Rany, as we approach Spring Training I was wondering if maybe you could share your thoughts on which 3 or 4players or things you think could be the keys to a successful Royals season this year. I am particularly intrigued with Luke Hochevar and Kyle Davies. It seems we know what we have with Jonah and Epic and I personally do not hold out much hope for Bannister to be more than a #5 4A type player so these two might be the critical parts of the rotation. Offensively, obviously most would point to possible breakout seasons from Butler (apparently an in-shape Billy Butler...what 22 year old pro athlete isn't in shape?) and/or Gordon for the keys to our season but what about some of the other deciding factors like who gets the ABs at catcher and 2B and Teahen's role on the team (though you have covered these previously so maybe you have other ideas)?
Also, does anyone think realistically there is any way DM does anything else even remotely significant in the free agent market?
Totally off topic here, but I see the Angels signed Ervin Santana for 4 years, $30 mil. Not sure on all the details of his contract / impending free agency status, but I'd be interested to know how Rany or others think this compares to our own recently signed Baseball Jonah.
By the way, as a Royals fan growing up in the 80s, I have great memories of Quiz. Of course, I didn't understand how special he was at the time, but he was fascinating to a young fan.
February 14, 2009 at 4:29 PM
Anonymous said...
chris- i have been using MLEs that may be outdated (because i'm too lazy and cheap to find new ones). If you are just looking to get MLEs on a few players, though, this website is helpful.
http://www.minorleaguesplits.com/mlecalc.html
if you really want my little spread sheet, i will be happy to email it you. send me a note at wez2002@hotmail.com. (i rarely use that adress but will look at it at least once before mid march.)
john- in answer to your question, i think there is a good chance. dayton will acquire another pitcher.
below are the # quality start (QS) by team with AL rank in parentheses: 1 - chicago (93) 4 - minnesota (86) 5 - cleveland (84) 8 - KC (78) 12 - detroit (67)
i really believe dayton thinks he can win the central in 2009 and i also believe that to do so, the back end of the royals' rotation will need to improve.
here's what the KC staff looked like in terms of QS/GS per starting pitcher in 2008: greinke 23/32 meche 21/34 banny 11/32 hochevar 8/22 davies 8/21 rest of staff 7/21
assuming the starters average 5.8-5.9 innings per game (pulling those figures from my *ss), we would hope to get 1000 combined innings from our starters. meche and greinke can't really do much better between the two of them than they did last year (at least in terms of quality starts) so we need other guys to step up. davies and hochevar have the stuff to make a big jump from last year and one of them will have to if KC is to contend. in my opinion, this is the MOST IMPORTANT thing for the Royals' chances this year.
so for the sake of discussion, say KC gets the following production from the top 3 in the rotation (basically a full season of the same performances they got in the last month of 2008 when they were the best team in the AL):
greinke 23/32 meche 21/34 davies or hochevar 16/28
then what happens with the last two spots? there are still 58 starts left (more than 1/3 of the season!!) and we would still need half of them to be quality. our candidate contributors include hochevar, bannister and horacio ramirez (ouch on that last one). if you assume that either davies OR hochevar will succeed and the other will fail, that leaves bannister and horacio.
so this is a long winded way of justifying my opinion that KC needs one more starter to be successful... and that said starter doesn't need to be a cy young candidate but rather a slightly above average major leagues who can soak up 20 starts (giving about 20 each to banny and hocehvar and leaving horacio in AAA as an insurance policy).
that being said, i do not think pedro martinez would come play for the royals, even if offered two years, $14 million.
my other pipe dream is that KC will trade butler, teahen and a prospect to SF for matt cain. the giants need to fill both corner IF spots and they are flush with young pitching. but again, i recognize this is a pipe dream.
later.
walt
February 14, 2009 at 7:40 PM
Anonymous said...
correction...
AL rank with #QS in parentheses. tough to proofread in this little blog box. sorry
wz
February 14, 2009 at 7:41 PM
Anonymous said...
Buster Olney says Dayton is trying to sign O-Dawg. Hell yeah, Dayton, get it done!
February 14, 2009 at 8:11 PM
Anonymous said...
I wanted to mention Randy Jones some more (I see someone else had him in mind). I remember an article in SI about him, and they were interviewing his catcher. He said how Pete Rose (maybe someone else but I'm pretty sure it was Rose) would be at the plate, cursing Jones to throw the ball.
Not just Quiz -- I also remember Ted Abernathy when the Royals first got started, and he was effective for us. I've wondered before if facing these weirdo, slow pitchers helps the rest of the staff - kind of like facing knuckleballers messes up the timing in the rest of the series. It was almost unfair when the Astros could pitch Nolan Ryan and JR Richards with Joe Niekro sandwiched between them.
February 14, 2009 at 10:13 PM
Anonymous said...
In unrelated news:
"Royals' minor-league outfielder Jarrod Dyson was suspended for 50 games Saturday after testing positive for an amphetamine."
"Investigators first became suspicious of Dyson's amphetamine use last season, when he stole 39 bases in just 12 attempts."
You have me convinced about Chris Hayes. Bring him up, he has to be better than Peralta.
February 16, 2009 at 4:45 PM
Anonymous said...
Speaking of knuckleballers, one of the longest standing questions I've ever had without an answer is: Why are there no left-handed knuckleballers? Surely the oddity of it would be enough to be successful if the guy could just throw strikes. If someone has a theory or interesting reading on this topic, let me know by email... at hagertyp@slu.edu
February 19, 2009 at 5:14 PM
Anonymous said...
My knuckleballers can't throw hard and need to develop a "trick" pitch to hang around the league. You can be a soft tossing lefty and still have a good career like a Jamie Moyer. There is no need for a lefthander to develop a knuckleball, even though I can see how it would be beneficial.
March 4, 2009 at 10:45 AM
So, is Chris Hayes a prospect? The experts have spoken. Chris Hayes is nowhere to appear in Kevin Goldstein’s rankings of the Top 11 Royals prospects released today ($$), not that this is surprising at all. What might be surprising is that he is nowhere to be found in Baseball America’s listing of the Royals’ Top 30 Prospects, and in fact, his name has never appeared in the pages of that august publication. One minor league expert who has not completely ignored Hayes is my friend John Sickels, whose The Baseball Prospect Book 2009 is now on sale. Hayes warranted an entry in this year’s book: Chris Hayes went to NorthwesternUniversity in Illinois and was quite successful during his junior and senior seasons. But scouts don’t like him very much, as he is a submarine pitcher with below average velocity. He signed up with WindyCity in the Frontier League in 2005, pitched well, then signed with the Royals as a free agent in 2006. He continued to pitch well in Double-A last year, throwing strikes and getting ground balls. I don’t like the low strikeout rate, but the other numbers are sharp. If he keeps this up in Triple-A, he could sneak into the major league pen. Stranger things have happened. Grade C. Grade C is the lowest grade that Sickels gives out to a player worthy of mentioning in his book. Saying “he could sneak into the major league pen” and “stranger things have happened” is not exactly an unqualified endorsement. So the experts have spoken, and what they’re saying is that Chris Hayes is, at best, a marginal prospect, a future 11th man if he’s lucky. I am not a minor league expert, though I played one in the pages of Baseball Prospectus for several years. Nonetheless, I can say with some confidence that I think the experts are wrong and that Hayes is very much a prospect, and that he has a very real chance to enjoy a long and successful career in the major leagues. I believe the reason why my evaluation of Hayes’ potential differs so much stems from the fact that I don’t think you can evaluate Hayes the same way you evaluate overhand pitchers. Much like the rules that apply to traditional pitchers don’t necessarily work when discussing a knuckleball pitcher, I think that we ought to think of low-arm-angle pitchers as their own breed as well. Many years ago – I think in one of the “Baseball Books” he published from 1993 to 1995 – Bill James made the point that the sidearm pitchers in the major leagues almost all had the same two characteristics: 1) they came from very humble origins and 2) they were almost universally effective pitchers. This was after the heyday of sidearm/submarine pitchers like Tekulve and Quisenberry, but you still had pitchers like Mark Eichhorn, who had a career ERA of 3.00 on the nose, and in 1986 had one of the most underrated rookie seasons of all time, with a 1.72 ERA in 157 innings. He famously declined the Blue Jays’ offer to let him start on the final day of the season so that he could get the five innings he needed to qualify for the ERA title. If he had taken them up on the offer, not only would he have wrested the title from Roger Clemens, he would have posted the lowest ERA of any AL pitcher in the last 40 years. Going to college in Baltimore in the early 90s, I watched up close as Todd Frohwirth gave the O’s three terrific seasons after he was let go by the Phillies. There always seems to be a half-dozen guys in the majors who throw from a low arm slot, and they’re almost invariably overlooked because they’re slotted in middle relief, but they almost all do good work. Chad Bradford, obviously, was made famous in Moneyball precisely because of his lack of recognition. Cla Meredith was a throw-in to the Padres from Boston in the Doug Mirabelli trade. Brian Shouse briefly pitched for the Royals in 2002, then – naturally – came into his own with the Rangers as soon as he left, at age 34. I had never heard of Pat Neshek before he was called up to Minnesota. I only knew about Brad Ziegler because he was one of Goldstein’s favorite minor leaguers – all he did was set a major league record with 39 consecutive scoreless innings to start his career. None of these guys was well-regarded as an amateur (none were drafted in the first five rounds). None of them were well-regarded in the minor leagues – I am almost 100% certain that none of them ever appeared on Baseball America’s Top 100 Prospects list. And yet all of them have been excellent relief pitchers in the majors – sometimes, like Neshek and Ziegler, immediately, but in each case in their first extended opportunity. (Baseball teams may finally be wising up to the market inefficiency regarding sidearm pitchers – Joe Smith was a 3rd-round pick out of college by the Mets, and immediately put on the fast track, throwing just 42 innings in the minors before getting called up.) The biggest knock against Hayes – and I say “biggest knock” in the way that you might say that the biggest knock against Muggsy Bogues was his height – is that he doesn’t throw hard. While we don’t have any Pitch f/x data, the best estimation is that his fastball runs about 79 mph. And if Hayes was a conventional pitcher, the discussion would end there. The list of conventional pitchers who have succeeded in the majors with a fastball under 80 is pretty much zero. (The closest I can think of would be John Tudor, who in his final season (1990) threw maybe 81-82 mph, if memory serves. He had a 2.40 ERA in 146 innings.) There are two types of pitchers that can survive, and even thrive, in the majors with a fastball in the 70s: knuckleballers and submariners. Among pitchers who threw even 10 innings last season, the slowest fastball (with an average speed of 72.9 mph) was thrown by Tim Wakefield (he only threw his FB 13% of the time; his knuckler averaged 65.1 mph, and his curveball – his curveball? – averaged 59.5.) The second-slowest fastball, at 79.6 mph, was Chad Bradford’s. Third was Brian Shouse, at 80.5 mph. Interesting, all the other sidearmers came in at 85 or higher – Meredith at 85.1, Ziegler at 85.3, Neshek at 89.0, Joe Smith at 89.4. This goes with a theory I have that the ability for a pitcher to succeed at lower velocities correlates with just how low his arm angle is. Whether it’s the deception that comes with the fact that the ball looks like it’s shooting out of the pitcher’s mound, or the bizarro movement on the pitches, or simply the novelty of facing an underhand pitcher, it appears that throwing submarine-style is worth about 5 mph compared to a sidearm pitcher, and throwing sidearm is worth about 5 mph compared to an overhand pitcher. As Hayes pointed out in his interview, Bradford is really the only true submarine pitcher in the majors. (Shouse throws only slightly harder than Bradford, but he’s a lefty, and lefties always seem to thrive with fastballs a few mph slower than their right-handed counterparts.) I’m old enough to remember Quisenberry on the mound, but not old enough to remember what the consensus was on how hard he threw. Some of you older hands help out – how hard did Quisenberry throw? I want to say around 80 mph, but it could have been slightly faster – or slightly slower. The point is, no one seems to take Hayes seriously as a prospect because of his velocity, when in fact we have a lot of evidence that throwing 79 mph is not an impediment to success for a submarine pitcher. On the contrary – 79 mph is exactly in the range of what the best submarine pitchers in recent memory have thrown. So the scout-based objections to Hayes’ prospect status appear overblown. That leaves us with his performance record. The biggest black mark on Hayes’ stat sheet is his strikeout rate – 137 whiffs in 203 career innings (6.06 K/9) is not particularly impressive for a minor league pitcher, and last season his K rate was just 5.35 per nine. Once again, though, it’s not entirely fair to evaluate Hayes’ strikeout rate as if he’s a conventional pitcher. The whole point of throwing underneath is that, if you do it well, you don’t need to strike batters out in order to be successful. Quisenberry struck out 3.27 men per nine innings in his career. (Read that sentence again.) Bradford’s career K rate is 5.47 per nine. The reason why strikeouts hardly matter for a submariner is that it seems to be a near-universal law that submarine pitchers keep the ball on the ground. That’s not necessarily the case for sidearmers – while Meredith and Ziegler have great sinkers, Neshek is a flyball pitcher who has given up 15 homers in 121 career innings. Byung-Hyun Kim was a sidearmer who threw exceptionally hard, and his penchant for fly balls led to some of the most dramatic baseball moments of this decade. But submarine pitchers, even in the high-offense era we’re in, are relatively immune to the home run. They also tend to have well above-average control. And if you can limit walks and home runs enough, it takes three hits in an inning to score a run. Even if you don’t strike anyone out, the odds are slim that the opponent will turn five groundballs into three hits. (And even if they do, they still might not score if you can mix in a GIDP.) Quisenberry walked the unfathomably low total of 0.79 batters per nine innings when you strip out the IBB, which as Hayes remarked is the lowest figure since 1926 (and I think he was including the IBB’s – take them out and Quiz might have the best control of any major league pitcher since four balls became a walk.) He gave up just 59 homers in 1043 career innings. I’m not exaggerating when I say that he would have been a successful pitcher if every one of his strikeout victims was given a mulligan and allowed to bat until they made fair contact. Bradford isn’t Quisenberry – no one is – but his career rates of 1.73 unintentional walks per 9 innings, and 0.48 HR per 9, are both outstanding, and the reason why he has a career 3.24 ERA even though he’s allowed exactly one hit per inning in his career. A lot has been made of the fact that Hayes’ terrific season in 2008 was the result of a BABIP of just .241, which is not likely to be sustained. That’s certainly true, but Hayes was so successful last year that he could easily regress to .300 and still be effective. Keep in mind, not only did he have a 1.64 ERA last season, but given his peripherals (less than a baserunner an inning, just four homers allowed in 66 innings), his expected ERA was probably even less than that. In 2006 and 2007, his BABIP was in the normal range, he gave up about a hit an inning – and he still had ERAs of 2.78 and 3.10, because he did everything else so well. For his career, Hayes has walked 43 batters (at least 6 intentionally; I don’t have data for 2006) in 203 innings, a UIBB rate of 1.64 per nine. He’s allowed just 7 homers, for a HR rate of 0.31 per nine. Those numbers are even better than Bradford’s minor league rates: Bradford walked 2.31 per nine (although I don’t have any IBB data for him), and allowed 0.37 HR per nine. I expect Hayes’ HR rates to go up as he spends more time in the high minors, but even so his performance record is every bit as impressive as Bradford’s was. (For the record, Quiz walked 1.74 batters per nine in his minor league career, and struck out 4.39 per nine. I only have homer data for 1978 and 1979; in those two seasons he allowed just 2 homers in 99 innings.) There’s one other reason why I think that Hayes can maintain his success in the major leagues. Allow me to quote him directly here: My advantage in an at-bat is how different I am from your average pitcher. I have yet to see a professional hitter taking batting practice off a batting practice pitcher throwing submarine-style. An experienced hitter has taken a swing at a “normal” back-spinning fastball thousands of times. It's been a while since I've hit, maybe it's more like millions. They see thousands of 90 mph-plus fastballs per season. Then, enter the “weirdo” stage right, and I throw that all upside down (literally and figuratively). A guy just out of college in A-ball may be used to guys throwing 85 mph on average, and they haven't honed their swing as much as a more salty veteran. The higher up the levels I go, I believe the bigger advantage I have, because I'm that much different from the norm. For years I’ve made essentially this argument with knuckleball pitchers: that the typical decline in performance that all players endure when they go from Triple-A to the majors is not completely applicable to knucklers. If the typical pitcher with a 4.00 ERA in Triple-A (assume a neutral park, etc.) can be expected to post a 5.50 ERA in the majors, a typical knuckleball pitcher with a 4.00 ERA might expect his ERA to be 4.75 in the majors. Why? In a nutshell – because major league hitters are not selected for their ability to hit the knuckleball. Sports are the ultimate Darwinian process – at every level, the players who show the most ability advance, and at each higher level they then face players with more ability, and so on. If you can hit an 80 mph fastball in high school, you’ll get the chance to hit 85 mph fastballs in college, then 90 mph fastballs in the minors, then 95 mph fastballs in the majors – provided you succeed at each level. And therefore hitters at each higher level have been selected for their proven ability to succeed against inferior pitchers. But the knuckleball is such a unique pitch, a pitch that works not because of its velocity but because of its unpredictable and late movement, that the best knuckleball hitters in the world may not be in the majors at all. There might be a guy down in A-ball who can’t catch up to the fastball and can’t lay off the slider, but can follow a 69-mph floater right up until the last moment. And there might be guys in the majors who could hit Sidd Finch if his fastball was straight enough, but are helpless against the mystical powers of the butterfly pitch. (I’ll pause here to tip my hat to Dave Nilsson, who might have been the greatest hitter of my lifetime against the knuckleball. The Australian catcher was 16-for-25 lifetime against Tim Wakefield, with a slash line of .640/.719/.960. He was just 2-for-10 against Dennis Springer, but with a homer and three walks.) I made this argument on the internal BP mailing list years ago, and Clay Davenport looked into my theory a little. What he found was that after a promotion to the majors, knuckleball pitchers did appear to require less of a translation to their numbers than the average pitcher. The sample size was small and not necessarily significant, but the theory had promise. Just as submariners can be lumped in with knuckleballers in the sense that they can survive without velocity, I think that like knuckleballers, their pitching style is so far from the norm (and there are so few of them to practice against) that major league hitters are not much likelier to hit them than their minor league counterparts. Does that mean Hayes is the next Bradford, or even Quisenberry Jr? Not necessarily. But the Royals owe it to themselves to give him every opportunity to prove otherwise. Some team is going to get lucky and find the next great submarine pitcher. The Royals were lucky enough to sign Hayes out of a tryout camp, but if luck is the residue of design, it’s the team that has designs on Hayes as a major league pitcher that’s going to really get lucky. The White Sox didn’t know what they had with Bradford; the Red Sox didn’t know what they had with Meredith. Let’s hope the Royals know what they might have with Hayes.
"The Future of Disco."
32 Comments -
Wow, very interesting. Coming from someone not old enough to see Quisenberry's work, I'm very interested to see what a submariner can do.
Great follow-up, Rany.
February 12, 2009 at 6:06 PM
Actually, looking at MILB's internal stats portal (I work for a Minor League team and have access to stats that might be harder to find otherwise), it looks like Hayes intentionally walked 9 batters that season. At least, looking at the 2006 season numbers. However, when looking at his career stats (through 2008) it actually shows only one IBB. So, I'm not sure what's going on there. But, he definitely walked at least one and possibly six batters intentionally in 2006, so his rates are even better than advertised.
February 12, 2009 at 6:43 PM
The Quiz threw probably around 80 to 81. I can't say for certain because he was around before radar gun speeds were shown to television viewers. Having seen quite a bit during his best years, I'm pretty sure he hit low 80s, and maybe dipped down to 77-79 for his 'off speed' pitches.
February 12, 2009 at 7:10 PM
I like this piece because of the specific content (like all of us I am a die-hard Royals' fan) but also because it provides an important perspective on the use of statistical analysis. A point to be gleaned from Rany is that there is much to be learned about any system from its statistical outliers... from the atypical individuals in a sample compared with the typical.
The glorious world of sabermetrics provides us with detailed valuations of each player's skill set. However, in my opinion, there are two flaws to the discipline (beyond its relationship with defensive aspect of baseball). The first is a fundamental flaw of sabermetrics, the other of sabermetricians.
First, sabermetrics assigns equal value to every at bat. Any one who has played sports at any level knows this not to be ture. One plays tennis harder at deuce in a 4-4 set than when winning or losing 5-1. It is more difficult to get a hit in the third inning of a tight game than the late innings of a blow out. A good player in any sport knows when the game is on the line or MAY be on the line.
... but this is for another day.
A fundamental flaw of sabermetricians is an obsession with the usual rather than the unusual. Major league equivalents (MLEs) are the perfect example. I have an MLE Excel spread sheet that I use at the beginning of every fantasy baseball season to evaluate rookies who appear ready to make a big league roster. I type in the player's AA or AAA stats and the embedded formula, which was devised by a dedicated sabermetrician who poured over stacks of empirical data, cranks out the player's expected big league numbers for the upcoming season.
Now it is fascinating that these models work so well and so often... and it says something about the game of baseball and the difference in typical skill sets between the minors and the majors that these relationships are so often preserved. But more amazing are individuals who confound the models... or who simply have generated a few statistics that are so far to the right or left of the Gaussian distribution as to inspire wonder.
Submariners and kunckleballers, in general, may well be just such a case, as Rany observed. These players seem to have a skill which limits the utility of some statistical comparisons.
Another example are hitters who strike out exceedingly rarely, such as Alberto Calllaspo. His low K % may be more than just a statistical oddity. It may reflect some skill... some difference in approach or some biologic ability... which other players lack. It would not surprise me to read an analysis correlating visual acuity with low strikeout rate, for instance.
But this is the point. Competitive athletes (usually) are not competing to please the sabermetricians. To quote Herm Edwards: "you play to win the game." So if a statistical oddity presents itself, the oddity quite possibly reflects some talent (or more commonly, deficiency) that a player possesses relative to the group... unusually deceptive spin on a pitch, a unique inside-out swing, better performance enhancing drugs... all of these are examples of competitive advantages that baseball players possess which lead to a statistical oddity. And I believe it is a fundamental role of the sabermetrician to identify outliers and look back at the game played on the field in an attempt to explain them.
Rany is to be commended for the observation that sidearmers and knuckleballers frequently defy typical rules of projected pitcher development. And after years of reading his blog... I am certain that he has other theories to share from his approach to statistical analysis, as well.
So thanks, Rany, for another great post. Your writing has been thoroughly enjoyable of late...
WZ
p.s. you are right about the Quiz. He threw 79-81 and referred to it as his "Peggy Lee" fast ball. Hitters looked at it and said "is that all there is." (his joke, not mine.)
February 12, 2009 at 7:18 PM
walt --
one of the first things that baseball statistical analysts do, in trying to prove or disprove an effect, is to measure the persistence from year to year.
so your devastating critique doesn't really carry the punch you think it does.
February 12, 2009 at 10:58 PM
I'm pretty sure Walt was just offering his opinion, Jason, rather than a "devastating critique." I mean, we're all experts in our own opinions. Holy crap, man, drink a beer or something and freaking relax...
February 13, 2009 at 12:48 AM
Rany, I guess you are too young to remember Randy Jones. Jones couldn't break 76 with his fastball, but posted back to back 20 win seasons in the mid seventies with his junkball sinker and slider. Jones won the NL Cy Young in 1976 with 315 innings and 25 complete games. I would really like to know how he pulled it off, since I was too young at the time to understand it.
February 13, 2009 at 2:29 AM
jason-
i think you and i are saying the same thing.
most analysts do quantify variance from year to year and generate fantastic models based on them.
my point is that beyond crunching the numbers and making predictions based on them, it takes greater skill to identify a subset of players who do not seem to follow the typical performance trends of the other thoroughbreds... and then ask "why do these players perform differently?"
so i had hoped publically to laud rany for his analysis rather than criticize anyone. and to imply that although he has always been my favorite contemporary sportswriter (poz #2), this bit of analysis particularly tickles my baseball sensibilities.
i am sorry if the last post (which i did make at the end of a long day, with a beer in my hand) looked more like a criticism of sabermetricians in general rather support for rant.
wz
February 13, 2009 at 8:54 AM
typo- "rany" at the end of the last, obviously.
February 13, 2009 at 8:55 AM
I would sure rather have given this kid a chance than blow $9M on Farnsworth.
February 13, 2009 at 8:57 AM
one last thing...
most of us don;t have access to statistical databases to test our hypotheses, so we can't be experts. but as a non-experts, i would like to offer a small hypothesis:
i do believe that there are players who are slightly better than their numbers and others who are slightly worse. and i believe that good statistical tip offs on each side of the ball are range factor and low strikeout rate.
as such, i have high hopes for callaspo.
also, i would like to offer a hope that dayton is scouting pedro martinez at the world baseball classic. it is my great hope that pedro throws 100-120 innings in twenty starts for KC, then moves to the pen to set up for soria during the playoffs.
go royals.
wz
February 13, 2009 at 9:07 AM
sorry, again. i am just getting so jacked for baseball season.
i know everyone wants to say that callaspo lacks power and range, that alex gordon is somehow worse defensively than he was at nebraska, that glass won't offer pedro the two year deal it would take to outbid the market for pedro martinez.
but since it is february, please indulge me a scenario.
royals lead 4-3 in anaheim heading into the bottom of the eight in anaheim in the first game of the ALDS. meche takes the hill to open the eight with pedro and mahay throwing in the pen. the spork subs for callaspo. coco and dejesus cheat a couple of steps to their left so guillen doesn't have to cover quite so much ground... and as abreu, guerrero and hunter stare down our 6'3", 220 lb cajun... things are looking pretty good.
wz
February 13, 2009 at 9:26 AM
Happy Pitchers and Catchers Report Day Rany!!!!
February 13, 2009 at 10:37 AM
time to sign orlando hudson and start trading away the fat from the roster!
February 13, 2009 at 11:29 AM
Great initial post Walt. You may not want to ruffle the sabermetician's feathers, but I will, albeit it's a minor ruffling...
Sabermetrics ARE very accurate and a very useful tool. However, some people put TOO much stock into the numbers. The numbers cover the cookie cutter players almost flawlessly, but they DON'T account for guys like Hayes or Quizenberry. Because of this some of the sabermetricians insist that guys like Quizenberry CAN'T suceed, and there for guys like that never get a shot. There are exceptions to EVERY rule, unfortunately some of the stat heads forget that. MLB scouts certainly have.
February 13, 2009 at 11:33 AM
One of the most important things in any kind of statistical analysis is knowing your sample sizes and error bars. I think sabermetricians are well aware of this, but even then it is easy to forget in the heat of column writing. Even at BP, it occasionally turns my head how the writers draw glib conclusions from 150 PA samples.
So yeah, Wabbitkiller, as Rany pointed out, sabermetrics is limited. That shouldn't ruffle any feathers; it's a known and acknowledged fact.
There's not an exception to every rule; you just have to understand the rule.
February 13, 2009 at 7:10 PM
Hayes can't even fairly be compared with sidearm pitchers. They throw harder and rely more on strikeouts (like most pitchers). Submariners are in a class of their own. They get some strikeouts, but less than most good relievers. They give up very few walks. And then get tons of groundballs. That combination (decent number of K's, very few BB's and a very high GB%) is the profile of success for a submariner.
So, does Hayes have the kind of velocity, control, strikeout ability and groundball tendencies of guys like Bradford and Quisenberry? It's still too soon to tell, but Hayes's K and BB rates in the minors are very similar to Bradford's minor league numbers. We'll see if he can maintain the high GB numbers, particularly in the PCL. That will be a good test.
In short, I think scouts and prospect analysts are comparing Hayes to overhand and sidearm pitchers and that is like comparing a banana to a bunch of apples and oranges. Also, I don't think scouts and prospect analysts quite understand how important groundballs can be. A huge GB% and a concomitant low FB% can make up for a mediocre strikeout rate. That's how guys like Wang and many sinkerballers succeed.
February 13, 2009 at 11:01 PM
First off Rany let em just say thank you. As die hard Royals fans I am sure you are all as sick of hearing about "A-Fraud" (what a predictable moniker) through the national media as I am. Rany, your coverage of the Royals is what gets me through the off seasons.
Also Walt (and others) I wanted to say it is thoughtful responses like yours that make reading the comments section worthwhile so thanks.
Also, as I prepare for the onslaught of fantasy/rotisserie drafts this season I was wondering if maybe you could send me that spreadsheet analysis for prediciting rookies' production? Did you get this through BP?
February 14, 2009 at 1:36 PM
Sorry to digress from the sabremetrics discussion but Rany, as we approach Spring Training I was wondering if maybe you could share your thoughts on which 3 or 4players or things you think could be the keys to a successful Royals season this year. I am particularly intrigued with Luke Hochevar and Kyle Davies. It seems we know what we have with Jonah and Epic and I personally do not hold out much hope for Bannister to be more than a #5 4A type player so these two might be the critical parts of the rotation. Offensively, obviously most would point to possible breakout seasons from Butler (apparently an in-shape Billy Butler...what 22 year old pro athlete isn't in shape?) and/or Gordon for the keys to our season but what about some of the other deciding factors like who gets the ABs at catcher and 2B and Teahen's role on the team (though you have covered these previously so maybe you have other ideas)?
Also, does anyone think realistically there is any way DM does anything else even remotely significant in the free agent market?
February 14, 2009 at 1:50 PM
Totally off topic here, but I see the Angels signed Ervin Santana for 4 years, $30 mil. Not sure on all the details of his contract / impending free agency status, but I'd be interested to know how Rany or others think this compares to our own recently signed Baseball Jonah.
By the way, as a Royals fan growing up in the 80s, I have great memories of Quiz. Of course, I didn't understand how special he was at the time, but he was fascinating to a young fan.
February 14, 2009 at 4:29 PM
chris-
i have been using MLEs that may be outdated (because i'm too lazy and cheap to find new ones). If you are just looking to get MLEs on a few players, though, this website is helpful.
http://www.minorleaguesplits.com/mlecalc.html
if you really want my little spread sheet, i will be happy to email it you. send me a note at wez2002@hotmail.com. (i rarely use that adress but will look at it at least once before mid march.)
john-
in answer to your question, i think there is a good chance. dayton will acquire another pitcher.
below are the # quality start (QS) by team with AL rank in parentheses:
1 - chicago (93)
4 - minnesota (86)
5 - cleveland (84)
8 - KC (78)
12 - detroit (67)
i really believe dayton thinks he can win the central in 2009 and i also believe that to do so, the back end of the royals' rotation will need to improve.
here's what the KC staff looked like in terms of QS/GS per starting pitcher in 2008:
greinke 23/32
meche 21/34
banny 11/32
hochevar 8/22
davies 8/21
rest of staff 7/21
assuming the starters average 5.8-5.9 innings per game (pulling those figures from my *ss), we would hope to get 1000 combined innings from our starters. meche and greinke can't really do much better between the two of them than they did last year (at least in terms of quality starts) so we need other guys to step up. davies and hochevar have the stuff to make a big jump from last year and one of them will have to if KC is to contend. in my opinion, this is the MOST IMPORTANT thing for the Royals' chances this year.
so for the sake of discussion, say KC gets the following production from the top 3 in the rotation (basically a full season of the same performances they got in the last month of 2008 when they were the best team in the AL):
greinke 23/32
meche 21/34
davies or hochevar 16/28
then what happens with the last two spots? there are still 58 starts left (more than 1/3 of the season!!) and we would still need half of them to be quality. our candidate contributors include hochevar, bannister and horacio ramirez (ouch on that last one). if you assume that either davies OR hochevar will succeed and the other will fail, that leaves bannister and horacio.
so this is a long winded way of justifying my opinion that KC needs one more starter to be successful... and that said starter doesn't need to be a cy young candidate but rather a slightly above average major leagues who can soak up 20 starts (giving about 20 each to banny and hocehvar and leaving horacio in AAA as an insurance policy).
that being said, i do not think pedro martinez would come play for the royals, even if offered two years, $14 million.
my other pipe dream is that KC will trade butler, teahen and a prospect to SF for matt cain. the giants need to fill both corner IF spots and they are flush with young pitching. but again, i recognize this is a pipe dream.
later.
walt
February 14, 2009 at 7:40 PM
correction...
AL rank with #QS in parentheses. tough to proofread in this little blog box. sorry
wz
February 14, 2009 at 7:41 PM
Buster Olney says Dayton is trying to sign O-Dawg. Hell yeah, Dayton, get it done!
February 14, 2009 at 8:11 PM
I wanted to mention Randy Jones some more (I see someone else had him in mind). I remember an article in SI about him, and they were interviewing his catcher. He said how Pete Rose (maybe someone else but I'm pretty sure it was Rose) would be at the plate, cursing Jones to throw the ball.
Not just Quiz -- I also remember Ted Abernathy when the Royals first got started, and he was effective for us. I've wondered before if facing these weirdo, slow pitchers helps the rest of the staff - kind of like facing knuckleballers messes up the timing in the rest of the series. It was almost unfair when the Astros could pitch Nolan Ryan and JR Richards with Joe Niekro sandwiched between them.
February 14, 2009 at 10:13 PM
In unrelated news:
"Royals' minor-league outfielder Jarrod Dyson was suspended for 50 games Saturday after testing positive for an amphetamine."
"Investigators first became suspicious of Dyson's amphetamine use last season, when he stole 39 bases in just 12 attempts."
February 15, 2009 at 3:57 AM
This comment has been removed by the author.
February 15, 2009 at 12:17 PM
This comment has been removed by the author.
February 15, 2009 at 12:17 PM
My brother was really great friends with Brad Ziegler and a big fan of yours. He would have been very excited to read your column.
February 15, 2009 at 12:17 PM
Here's an awesome idea for the bullpen: Oil Can!
http://www.mlbtraderumors.com/2009/02/cafardos-late-2.html
Now that's my choice for Rany's next interview...
February 15, 2009 at 7:36 PM
You have me convinced about Chris Hayes. Bring him up, he has to be better than Peralta.
February 16, 2009 at 4:45 PM
Speaking of knuckleballers, one of the longest standing questions I've ever had without an answer is: Why are there no left-handed knuckleballers? Surely the oddity of it would be enough to be successful if the guy could just throw strikes. If someone has a theory or interesting reading on this topic, let me know by email... at hagertyp@slu.edu
February 19, 2009 at 5:14 PM
My knuckleballers can't throw hard and need to develop a "trick" pitch to hang around the league. You can be a soft tossing lefty and still have a good career like a Jamie Moyer. There is no need for a lefthander to develop a knuckleball, even though I can see how it would be beneficial.
March 4, 2009 at 10:45 AM