Sky Sox Action

Had a chance last night to get out and see the Colorado Springs Sky Sox (Rockies AAA club) play the Sacramento River Cats (Oakland's AAA) on $2 Tuesday where all sets were just $2 (I sat right behind the plate in the first row). It was unusual game for Security Service Field as the River Cats won 3-1. The only offense for the Sox was the solo homerun by Ryan Shealy (his 14th) in the 6th inning. He hit it to the opposite field and it cleared the 386 foot sign in right-center on a high arc. Shealy was called up during interleague play and went 6-20 with a homerun, double, and three walks. His patience still seems to be an issue as he's walked now just 17 times in 267 plate appearances.

For the Sky Sox Jose Acevedo (still rehabbing) got the start and was not very effective giving up four hits in his two innings of work. His fastball clocked in from 87-91 and he also threw a curveball in the mid 70s and what looked like a slider that came in around 81. His mound opponent, 28 year-old Seth Etherton was outstanding. He's coming back from a torn labrum a few years ago and pitched in the Reds organization last season where he threw well for Louisville. In his 7 innings he gave up just 4 hits including the homeun to Shealy. His fastball topped out at 89 but he was very effective with his breaking stuff and hit all his spots. His ERA is now 2.29 and could be helpful for the A's in the second half as a spot starter or long reliever.

On the offensive side I got a chance to see Charles Thomas for the River Cats who was acquired in the Tim Hudson deal with the Braves and who has a chapter devoted to him in the new book Scout's Honor (and which I'll be reviewing soon). He hit .109 for the A's and so now he's back in AAA where he's hitting .287/.426/.378 and leading off. He could possibly be a decent fourth outfielder since he plays pretty good defense (he was in CF last night). The River Cats also have Jack Cust, who had a cup of coffee with the Rockies in 2002 and again with the Orioles in 2003. He has tremendous plate discipline and walked twice. He now has walked 68 times in 372 plate appearances to go with 13 homeruns and...get ready...95 strikeouts. He's almost a three true outcomes guy - really two true outcomes I guess. It would be interesting to see what he could do at the major league level. His AAA line is .283/.480/.415. The River Cats also have an interesting fourth outfielder kind of guy in Matt Watson (.313/.510/.399) who has seen a little action in Oakland this season. He was a 16th round draft pick in 1999 by the Expos and is now 26.

And as expected the River Cats are leading the league in walks while the Sox are dead last.

Retrosheet Studies

There were a couple of great studies updated on the Retrosheet web site by Tom Ruane recently. The first deals with clutch hitting since that's been a topic of some interest recently. I like this study because it uses much larger sample sizes than the original clutch hitting studies that simply look at year to year differentials.

Tom presents a mountain of data from 1960-2004 in a very understandable format. I was particularly interested in his data that shows that hitters do indeed perform more pooly with runners in scoring position than they do in other situations once sacrifice flies, hit batsmen, and walks are taken into account. He also breaks the data down into hitters who had above and below 3000 clutch at bats.

 1-2999        AB       H     2B    3B     HR  BAVG   SLG   BPS
Not RISP  1748765  414965  70688  9459  35513  .237  .349  .587
RISP       588724  135614  23479  3614  10466  .230  .336  .566

 >=3000        AB       H     2B    3B     HR  BAVG   SLG   BPS
Not RISP  2848124  773189 135086 19045  84334  .271  .421  .693
RISP       974948  257380  45517  7173  26300  .264  .406  .670

There are two impliciations here. First, you would expect that hitters with more clutch at bats would have a smaller differential between non-clutch and clutch performance on the theory that these hitters were "selected" for more at bats in part because of their ability to perform when needed, but this was not the case. They were better hitters of course but saw the same dropoff in clutch situations. Seconodly, this does provide support for the notion that if clutch hitting exists the defintion of a clutch hitter is more a hitter whose performance does not decline in key situations but rather remains the same or slightly improves.

Digging into this second point Tom also explores some plausible theories as to why performance might fall off in these situations including the ability for the defense to record force outs, the spike that occurs when a runners is on first due to the first baseman holding the runner, the fact that better hitters are walked more often with a runner on second or second and third, platoon advantages, quality of pitching, and even park factors. He found that platoon advantage for the hitter and higher quality of pitching tend to balance each other out in the long haul. In looking at his data I found that when excluding the at bats with a runner on first only the difference shrinks to around 11 points of BPS (batting average + slugging percentage), roughly half of the raw difference. As a result it would appear that hitters do tend to hit more poorly in clutch situations overall.

In conclusion Tom reports:

"So did I find evidence of clutch hitting? Not really. I did come up with lists of players who performed well and poorly in this area. Along the way, I presented quite a bit of data on situational hitting, platoon advantages, opposition pitching strength and park effects, and I attempted to both understand and explain what I found. At the end of all this, however, I guess I'm still not convinced that the players owe their inclusion on these lists of mine to talent rather than luck. Even when dealing with sample sizes of several thousand at-bats, the amount of random variation that I found in my simulations was very close to what I found in the real data. As I mentioned before, this doesn't necessarily mean that there isn't some real differences buried in all that noise, only that I'm not sure I found them. One could argue that the forces at work here, if they exist, must be awfully weak to so closely mimic random noise, and if they are really that inconsequential perhaps we could assume they don't exist without much loss of accuracy."

To me this tends to confirm what most sabermetricians have been saying since 1978 - clutch hitting may be a real skill but if it is it makes very little difference in the long run.

The second study is an update to an approach discussed in the 1987 Baseball Abstract. Tom applies a slightly different methodology and also includes positional adjustments and stolen base runs. He compares his method with a more straightforward linear weights method and even applies the method to pitchers. Very interesting reading.

James Part III

Here is part III of Baseball Digest Daily's interview with Bill James. In this installment James says in reference to the previous installment regarding bunting:

"To me, the arguments against the sacrifice bunt are unconvincing. Base stealing helps you at a 75% success rate, hurts you at 55%, so the attempt very often IS worth it. What DOESN’T pay often is putting a player in the lineup because he can steal bases. A player who steals 65 bases and is caught 10, 15 times...sure, that helps you. But if you put Dave Collins in the lineup rather than Oscar Gamble because Dave Collins can steal bases, then that hurts you."

And this I think is closer to the heart of the message of Moneyball and Weaver's 10 laws than simply saying you're never going to sacrifice or steal bases. When Chuck Tanner installed Omar Moreno in the leadoff spot for the Braves in 1986, he did so apparently thinking that speed in the leadoff spot by itself somehow generated runs. More specifically, he probably concluded intuitively that Moreno's speed compensated for his failure to get on base by allowing him to put himself in scoring position more often. It did not and does not because of the opportunity cost of those additional outs that Moreno consumed. Speed is a strategic weapon you can use once you reach base. It seems obvious now but it wasn't ten to twenty years ago.

James also makes some interesting comments about the recent draft and his participation for the Red Sox. Great stuff and thanks to Joe Hamrahi for sharing the interview.

Brenly a Moneyballer?

In yesterday's 5-4 12-inning loss to the Nationals yesterday I heard Cubs broadcaster Bob Brenly actually say something very interesting and very true.

He mentioned that in doing some quick math he realized that if Jerry Hairston (.360)and Todd Walker (.350) were to bat first and second instead of Corey Patterson (.274) and Neifi Perez (.296), Derrek Lee and company would have 85 to 100 more opportunities in a season with runners on base. In redoing his calculations he apparently based it on 600 plate appearances where it would yield 84 more times on base for Hairston and Walker. When you subtract homeruns that's around 70 or so but Brenly was on the right track. I also heard him over the weekend say that at least at the team level hitting well in "clutch" situations was largely luck and that it would even out over time. Is Bob a closet Moneyballer?

Today I see that Hairston is leading off against the Braves with Walker batting second.

Beyond the Box Score

I've started to post some thoughts on Beyond the Box Score. Today's topic, appropriate for July 4th is the Washington Nationals and their confounding of Pythagoras.

Sickels on the Royals

I thought Royals fans who read this blog would be interested in John Sickels Royals Top 20 in Review article from his blog.

Interesting that Sickels comments for J.P. Howell, Ambiroix Burgos, Shane Costa, and Leo Nunez all include something to the effect that they were rushed to the majors, echoing my comments of yesterday (or I guess I echoed him since he published first). I was hoping that Chirs Lubanski, the Royals number one pick (5th overall) from 2003 would be higher than 17th and have a more encouraging outlook than:

"Pulling out of early season slump. Now hitting .259/.307/.511 with 17 homers. Power improving, though some of that is due to environment, and plate discipline and contact are still major questions."

Hate to say I told you so...

Just thought I'd check in on how the Royals have been doing since they hired Buddy Bell as manager. Right out of the shoot the Royals swept the Yankees, went 1-2 with the Rangers, 2-1 against the Giants, 2-1 against Arizona, and swept the Dodgers. That put them at 10-4.

Since then they've dropped two of three to the Astros, been swept by the White Sox, swept by the Rockies, dropped two of three to the Twins, and lost the first of a three game set to the Angels. So that makes it 2-11 since June 17th for an overall mark of 12-15 and now 26-52 for the season.

Buddy may be a teacher but he's not a miracle worker.

As a Royals fan I'm not so disappointed that the Royals stink since that was a foregone conclusion since I saw them in Spring Training. I'm most disappointed that they've somewhat recklessly brought guys up to the majors who have no reason to be there including Burgos (21 yrs old), Nunez (21 yrs old), Howell (21 yrs old), Huber (23 yrs old), and Costa (23 yrs old). These guys might form a good nucleus in Omaha, especially when you team them with Denny Bautista (24 yrs old) and Ruben Gotay (22 yrs old) but all of these guys now have major league service time and other than Gotay are not getting at bats on a regular basis. What would be the harm in letting these guys mature as a group in Omaha and then bringing them up together next season in a true youth movement that includes Greinke, Teahen, and DeJesus? As it is the Royals have a mix of young guys with a few veterans like Terrance Long, Matt Stairs, and Emil Brown. Can you really justify giving at bats to these guys over developing Huber and Costa? I sense some grasping at straws to try and salvage something from this season at the major league level. I say, be patient and bigger dividends will be paid down the road.

Getting Rid of Batting Average

Kansas City Star columnist Joe Posnanski had a column in Thursday's paper called "Believe it: It’s time to retire batting average". Since the column requires registration I'll report that he, with the help of Bill James, makes a humerous case for why batting average is so misleading by comparing it to rating a novel by the number of words per page and rating doctors by the number of patients they see per day among others.

He also uses the much more serious example of Ken Harvey, a guy whose entire value is tied up in his batting average and the contra-example of Jimmy Wynn to illustrate that batting average does not tell much of the story. He goes on to say,

"There are a lot of players like Wynn who never were appreciated enough because of relatively low batting averages — Bobby Grich, John Mayberry, Bob Allison, Darrell Evans, Boog Powell, Toby Harrah and Roy White, to name a few. They got on base. They hit for power. Those things matter. Batting average doesn’t.

It’s time. I’m not sure which statistic we should use in its place. The trendy statistic now is OPS — on-base percentage plus slugging percentage — but that has its own flaws. There is “secondary average,” which takes into account total bases, walks, stolen bases. There is plain old on-base percentage, which I like. There are plenty of options."

I have seen OPS appear in a television broadcast, I just can't remember where. As a single number I'd prefer it because of its simplicity, comparative ability, and correlative value. It does a much better job of conveying the run producing ability of a player than any of its three component parts (AVG, SLUG, and OBP) do and could be used effectively in shorthand venues such as television screens. That said, I'd rather see more stats on the top and bottom portions of the screen instead of out of town scores and miscellaneous sports news. Then we could see the standard AVG/OBP/SLUG format that conveys even more information than OPS.

The Fog is Lifting

The February 2005 issue of By The Numbers, the newsletter of the SABR statistical analysis committee has been published (for those readers who are not members of The Society for American Baseball Research, click here to join).

Of particular interest were articles by Jim Albert and Phil Birnbaum, both of which relate to Bill James' article in the 2005 Baseball Research Journal titled "Underestimating the Fog".

The Fog Reviewed
I wrote a short review of James' article back in March and at that time noted that:

"James' argues that some of the best known negative sabermetric conclusions should not really be viewed as conclusions at all, but rather simply as non-answers to questions under study.

James goes on to criticize the common technique employed in various sabermetric studies that typically cited to 'prove' these conclusions - for example Dick Cramer's famous 1977 Baseball Research Journal article on clutch hitting and James' own look at platoon differentials in the 1988 Baseball Abstract. That technique involves the search for recurrence or persistence of the phenomena being studied. In other words...studies were done that attempted to determine if the effect...persisted across seasons. In each case repeated studies have shown that it doesn't - therefore the effect is, in the words of James, 'transient' and not 'persistent'. That which is not persistent is then assumed not to be real.

James then argues that in many of these cases the negative conclusion - the phenomena is not real - is flawed because there is too much instability in the data used to make the conclusion...The randomness involved in such a small sample size tends to swamp the differential itself, thereby making the results meaningless. James notes that Cramer's original study of clutch hitting was flawed for the same reason."

Albert's Comments
In his article, titled "Comments on 'Underestimating the Fog'", Albert brings a bit more clarity to what James was saying and talks specifically about the amount of luck and the amount of ability in various baseball statistics (a topic that he and co-author Jay Bennett discussed in their excellent book Curve Ball). He notes that given a set of players with 500 at bats, about 50% of the variation in batting average will be due to luck while the other 50% can be attributed to differences in their hitting ability. He then notes, as does James, that the amount of luck will vary with sample size. He has also discussed this topic in another online article that I referred to in in a post on the general topic of luck and batting average a few months back as well.

Albert goes on, however, to disagree with James in his assertion that using comparison offshoots (difference between two statistics) in making comparisons between players has the effect of summing the luck component of the individual statistics. He says this notion is simply "nonsense" and uses a simple example to show why. Albert then illustrates his methodology of testing against a simple model and concludes that there is evidence for a platoon difference (using SO/AB) as he and Bennett concluded in chapter 4 of Curve Ball. Platoon differences are one of the sabermetric conclusions that James called into question in his original article.

In conclusion Albert notes that what James is simply a restatement of the logic statisticians use when they test models meant to approximate reality.

"When we assess the goodness of fit of a model, there are two possible conclusions: either there is significant evidence to reject the model or there is insufficient evidence. Saying there is not enough evidence to reject the model doesn’t say the model is true, but it does say that we just can’t provide evidence to say that it is false. Statisticians don’t prove models are true – that’s why we are careful to say things like "we have insufficient evidence to reject a model.'"

His opinion, then, is that phenomena like those mentioned by James including clutch hitting may indeed exist, but it is likely they have a very small effect and are difficult to pick up with a statistical test.

Birnbaum's Regressions
The more controversial of the two articles in By The Numbers is Phil Birnbaum's "Clutch Hitting and the Cramer Test". In this article Birnbaum attacks James' notion that "random data proves nothing - and it cannot be used as proof of nothingness" and says it is "certainly false". His position is that if one can show that their study is well designed, then a conclusion of "no effect" can reasonably be offerred.

Birnbaum goes on to apply the method used in Cramer's 1977 article to data from 14 pairs of seasons from 1974 through 1990 and shows the correlations in the table reproduced here.

         r     r^2   f
74-75  .0155 .0002 .86
75-76  .0740 .0055 .37
76-77  .0712 .0051 .40
77-78  .0629 .0040 .44
78-79 -.1840 .0339 .02
79-80  .0038 .0000 .96
82-83 -.0250 .0006 .75
83-84  .0456 .0021 .60
84-85  .0222 .0005 .79
85-86  .0728 .0053 .38
86-87  .0189 .0004 .82
87-88  .0034 .0000 .97
88-89  .0829 .0069 .33
89-90  .0373 .0014 .67

He then creates a model where clutch hitting exists and compares the model's results with the actual results. His conclusion is that while the Cramer test can easily determine whether clutch hitting is a real phenomenon if the ability were normally distributed with a standard deviation of 30 points, the method starts to fail once the standard deviation falls below about 7.5 points. In other words, he is in agreement with Albert that if clutch hitting is real, its effect is small.

He then combines the data for his 14 seasons into a single regression and concludes that the Cramer test could pick up the effect at a standard deviation of 15 points but begins to fail at a standard deviation of about 10 points.

In his set of conclusions Birnbaum acknowledges that the Cramer test cannot completely disprove the existence of clutch hitting, but it does put a "strong lower bound on the magnitude of the possible effect". That lower bound is that the effect is certainly less than 15 points of batting average; "that is, at least two-thirds of all players are expected to clutch hit within .015 of their non-clutch batting average." And so his conclusion is that given the small effect it is probably impossible to distinguish good clutch hitters from bad.

Two Responses
These article then generated a response from Bill James called "Mapping the Fog" which he posted to the SABR Statistical Analysis Committee Yahoo group. That was quickly followed up by a response from Phil Birnbaum published in the same forum. Both responses have now been published along with a lively discussion by Chris Dial on The Baseball Think Factory.

In the meat of his response James creates a model where clutch hitting exists (a model league where 20% of the hitters are clutch performers at the level of 25 points) and then uses the Cramer test to try and detect the effect. The test does not reliably detect it (it would do so only 65% of the time) and so he concludes that even under these ideal circumstances the Cramer test has nothing to say as to whether clutch hitting is a real phenomena or not.

"Even when we know that the clutch effect does exist within the data, even when we give that effect an unreasonably clear chance to manifest itself, there is still a 35% chance that it will entirely disappear under this type of scrutiny."

James goes on to respond specifically to statements made in both articles and notes that in his opinion "there is an immense amount of work to be done before we really begin to understand this issue."

In his response Birnbaum response is critical of the way James applied the Cramer test to his model, using a simpler sign based test rather than an actual regression. He goes farther, however, and actually uses James' model with the regression analysis and concludes that using 14 separate seasons of data the Cramer test works in that it would pick up clutch hitting using James' model.

My Response
I find this entire discussion important and fascinating since it comes so close to the heart of how many sabermetric studies have been done - the most recent and famous being the conclusions drawn in favor of DIPS.

For the most part I have to side with Albert and Birnbaum and agree that while these statistical tests can never prove that clutch hitting or platoon differentials do not exist, using data over a number of seasons (after all, the number of true clutch plate appearances is small, around 50 per 600) or for an entire career they can surely indicate that if they exist, their effects are small and may even fall within the normal variation of the statistics that are used to measure them. As a result - and here is the actionable thing - they needn't be considered when making personnel decisions over the long haul.

As an aside I thought the most interesting comment made in the discussion of this topic on TBTF was that it could be the case that clutch hitting is actually a "negative ability". In other words, it could be the case that most hitters actually perform worse in clutch situations and that good clutch hitters can be defined as those who maintain their performance level. It would be interesting to see what a model would look like to test for this.

One Run Games

And here is a good summation of thoughts about one-run games and the White Sox from The Hardball Times...

Blowouts and Pythagoras

A questioner asks regarding the Pythagoras and the White Sox article, how much does a blowout (say a 16-0 win instead of a 2-1 victory) make in the various projections?

On April 7, 2004 the Rangers beat the A's and Barry Zito 2-1. What would the projections look like if instead the Rangers had won 16-0? To find out I looked at the Pythagorean and the Pyth+Actual predictions at each point in the season. The following table shows the differences between the actual number of wins and the expected number of wins on each date when using the 2-1 versus the 16-0 result. In other words, on April 10th by including the 16-0 game the Pyth projection is 22.5 wins worse than using the Pyth with the 2-1 win. By May 30th, the impact of inserting that 16-0 game goes down to 4.4 wins (in that case the Pyth projection was almost exact projecting the Rangers to win 89 games while using the blowout had them at 93.4 wins).

 Pyth   Pyth+Actual
10-Apr   22.5    21.2
15-Apr   17.4    16.4
30-Apr    8.7     7.5
30-May    4.4     3.2
30-Jun    2.8     1.5
30-Jul    2.1     0.9
30-Aug    0.9     0.4
5-Oct    -1.4       0

Interestingly, by the end of the season using the blowout projected the Rangers to win 88.9 games while not using it projected them to win 87.5 games. They actually won 89 of course.

More Pythagoras

Bruce Cowgill, a member of the SABR Statistical Analysis Committee was kind enough to send me another formula used to predict a team's final winning percentage given their actual winning percentage at a point during the season. I don't have the original site where it was posted (by an analyst with the name or initials DBOZ) but here is the formula:

Final W% = (0.5 x (1 - GamesPlayed / 162)^2.25) + (CurrentWin% x (1 - (1 - GamesPlayed / 162)^2.25))

I applied this formula to the data I had for 2004 and added yet another column to the two tables I posted in my previous post.

                Pyth            Actual          Pyth + Actual   DBOZ
        Avg G   AvgE  StdDev    AvgE  StdDev    AvgE  StdDev    AvgE  StdDev
10-Apr     5    18.2    11.9    25.5    18.0    18.2    12.0    11.1    6.6
15-Apr     9    15.8    14.5    17.8    14.6    15.5    14.5    11.0    6.0
30-Apr    22    11.5     9.3    12.9     7.1    11.2     9.0     8.6    6.2
30-May    49     8.0     5.6     7.3     5.6     7.5     5.0     7.2    5.1
30-Jun    76     6.7     5.0     6.5     4.8     6.3     4.5     6.6    4.8
30-Jul   102     5.1     3.3     3.8     3.0     4.0     2.7     4.1    3.1
30-Aug   130     3.6     3.0     2.3     1.8     2.2     1.8     2.3    1.8
5-Oct    162     3.1     2.8     0.0     0.0     0.0     0.0     0.0    0.0

Because this formula is a "regress to the mean" formula much like using the Pyth+Actual method, the average error ends in 0 at 162 games. It matches pretty closely with Pyth+Actual but does a better job very early in the season. In looking at the raw data, the reason is that it tends to lump everybody together, thereby reducing the standard error but reducing the correlation coefficient as well.

        AvgG  PythW%  ActualW%  Pyth+ActualW%    DBOZ
10-Apr    5   0.414    0.270      0.411         0.255
15-Apr    9   0.300    0.347      0.306         0.348
30-Apr   22   0.530    0.649      0.559         0.641
30-May   49   0.719    0.771      0.757         0.771
30-Jun   76   0.780    0.797      0.812         0.799
30-Jul  102   0.890    0.930      0.934         0.929
30-Aug  130   0.937    0.975      0.976         0.976
5-Oct   162   0.951    1.000      1.000         1.000

High Schoolers

Here's a quick tidbit from a recent Tracy Ringolsby column.

"High school players accounted for 529 of the 1,501 players drafted, 35.2 percent, which was up from 33 percent last year, ending 10 consecutive years of decline. The all-time record for high school drafts was 63.7 percent in 1969. The all-time low was 25 percent in 1985."

Pythagoroops

Several folks have emailed me (and my own brother took me to task) regarding my article Pythagoras and the White Sox on The Hardball Times and asked why I didn’t include a measure of how well the Pythagorean method predicted a team’s final number of wins given the games that they had already won. In other words, use the Pythagorean method to predict the number of games a team will win in the remainder of the season and then add that to the number they had already won. That result could then be compared with the other methods. After all, a team already has those wins in the bank and so they are in economic terms a “sunk cost”. Quite frankly, I missed this entirely but it makes perfect sense. So I did in fact do the calculations and here they are:

                Pyth            Actual          Pyth + Actual
        Avg G   AvgE  StdDev    AvgE  StdDev    AvgE  StdDev
10-Apr     5    18.2    11.9    25.5    18.0    18.2    12.0
15-Apr     9    15.8    14.5    17.8    14.6    15.5    14.5
30-Apr    22    11.5     9.3    12.9     7.1    11.2     9.0
30-May    49     8.0     5.6     7.3     5.6     7.5     5.0
30-Jun    76     6.7     5.0     6.5     4.8     6.3     4.5
30-Jul   102     5.1     3.3     3.8     3.0     4.0     2.7
30-Aug   130     3.6     3.0     2.3     1.8     2.2     1.8
5-Oct    162     3.1     2.8     0.0     0.0     0.0     0.0

This table now includes the Pyth + Actual columns that show how well the Pythagorean method does when incorporating the current number of wins. As you can see the average error and standard deviation are a bit lower than using the actual winning percentage through April and then the average error basically tracks with the actuals all the way through the end of August while the standard deviation remains a bit lower. I was a bit surprised that even as late as August 30th the Pythagorean + Actual method was slightly more accurate (2.3 to 2.2 in average error). As a result, I wouldn’t be hesitant to use it at any point in the season.

I must also here admit to an error in the original article. The fourth table I showed that included the correlation coefficients for the Pythagorean wining percentage and the actual winning percentage computed against the final winning percentage was in error. Here is the corrected table along with a third column for the Pyth + Actual.

        AvgG  PythW%  ActualW%  Pyth+ActualW%
10-Apr    5   0.414    0.270      0.411
15-Apr    9   0.300    0.347      0.306
30-Apr   22   0.530    0.649      0.559
30-May   49   0.719    0.771      0.757
30-Jun   76   0.780    0.797      0.812
30-Jul  102   0.890    0.930      0.934
30-Aug  130   0.937    0.975      0.976
5-Oct   162   0.951    1.000      1.000

Of course, this clears up my surprise in the original article that the PythW% correlations were so strong (all around .8 or higher). As might be expected the actual winning percentage does a better all the way through of predicting a teams relative rankings and hence number of wins. However, using the combination of Pythagorean winning percentage and actual wins does a better job through mid June until the end of the season. That conclusion validates the results of the previous table and means that using the Pyth+Actual method is a valid way of predicting season outcomes at anytime.

Harball Times

I'm pleased to report that my first article, Pythagoras and the White Sox, has been published on The Hardball Times. Thanks to Dave Studeman and Aaron Gleeman for inviting me to submit my strugglings...

Mere Christianity

Awhile back I outlined C.S Lewis' main points in books one and two of Mere Christianity for my own uses. It is one of my favorite books and so although I probably won't do it justice, I'll reproduce it here for anyone who hasn't read the book and is interested in the kind of direction it takes.

• Humans have a common standard of behavior and realize they cannot live up to it. Lewis calls this The Law of Nature or the Moral Law
• The moral law is not an instinct, it is what chooses between instincts. The moral law is not social convention because morality can be compared to a standard. If it were social convention, then the idea of moral reformers would not exist
• The moral law is not an evolved adaptation to living in groups (i.e. decent behavior benefits society). Why should individuals care about society?
• In the one case where we can try and see if there is anything more than matter to the universe we find that there is. This is our notion of the moral law.
• If there were a power outside the universe we would not expect it to be one of the things inside the universe but rather an influence (the moral law), which is what we find.
• If indeed there is a power behind the universe and this power instituted the moral law, then it must be dissatisfied with us since we fail to live up to that law. That is the fix we are in
• Atheism is too simple since it posits that the universe has no meaning. To discover it has no meaning, one would have to have an idea of justice, which wouldn't exist without the moral law
• Only Christianity and Dualism deal with the reality that men find much of the world to be bad and some meaningless. Christianity claims a good world spoiled while dualism claims equal and opposite powers on a battlefield. Dualism doesn't work, however, since you need to introduce a third thing to judge between the two powers. Christianity rings true since in our experience badness is always goodness pursued by the wrong means or twisted
• The conclusion that evil is derived immediately raises the issue of God's all powerfulness. If He is so, then why is there evil? Here enters Free Will. God apparently thought it worth giving creatures like us the freedom to reject him. But God left men with several witnesses; a) conscience, b) myths of dying Gods ("good dreams"), c) a chosen people to bring his message and ultimately the answer to man's rebellion. And Jesus claimed to be God and to forgive sins. Ultimately you must decide between Lord-Liar-Lunatic (the trilemna)
• So what did Jesus come to do. He came to reconcile our since to God. How he did this is not important, only that he did. He became the perfect penitent because we in our sins could not repent fully
• So how does a man appropriate the atonement? By belief and the sacraments (in CSL's view). So then how are those who haven't heard of Christ saved? We don't know the complete answer. Why did God come in such a disguised fashion? To preserve free will

There's obviously alot more there but I was trying to distill the main points.

Two Firsts

I was privy to a couple of first tonight in the Rockies 12-4 win over the Royals at Coors Field. First, Justin Huber, obtained by the Royals in July of last season from the Mets, picked up his first major league hit and RBI. He was called up on 6/21 and was hitting .332/.420/.517 with 17 doubles and 9 homeruns in the Texas League (AA). He hit a fly ball down the right field line that hit the scoreboard in the 6th inning that scored Ruben Gotay. It'll be interesting to see how Huber develops. He'll turn 23 on July 1st and although he started out as a catcher he'll likely end up at first base in the long run. I'm not sure why, once again, Allard Baird decided he needed to play in the majors and start his service time clock running but he's here now. There must be something that Allard doesn't like about Omaha...

Also tonight I saw Jimmy Gobble strike out the side in the 6th inning. He was topping out at 90mph but was very aggressive in the inning getting Atkins, Helton, and Wilson in order. That was the first time Gobble had struck out the side in his major league career. It's surprising because he has such a low career strikeout rate - just 81 Ks in 201.7 career innings coming into tonight. I wrote last year about how pitchers with such low strikeout rates don't generally have long careers. In one game I scored last year Gobble did not have one swinging strike in 5+ innings of work. That may be a record as well. It would have been nice if the Royals could have unloaded Gobble last season but who knows, maybe he'll look decent in a long relief role and that will generate some value that Royals can trade on.

Youngsters

Just getting set to score the Rockies/Royals game tonight for MLB.com at Coors Field. These are the two youngest teams in baseball with Royals coming in at 27.11 years old and the Rockies at 28.19 years. The Royals come in having gone 12-9 since Buddy Bell took over. The Royals are also 11-9 in June and if they play well this weekend could lock up their first winning month since July of 2003 when they went 15-11.

Here's an interesting Rockies stat. The Rockies have held the lead in 27 of their 48 losses (56%). That leads the majors. They are also 7-12 in one-run games and 6-10 in two-run games. The Rockies troubles also continued on the 1-8 road trip they just finished where they hit .224 and went 11 for 66 with runners in scoring position. They are also now 5-31 on the road, the worst since the Twins went 5-32 to start the 1982 season.

Virtually Playing

When I lived in KC I never got a chance to go to a T-Bones game (the independant Northern League franchise in Shawnee) but it looks like their promotion on July 16th will be a first. I always knew a computer gamer could make a difference in this world.

Kile at Coors

Here is an excellent article posted on Viva El Birdos on the third anniversary of Darryl Kile's death. The article deals with Kile's time with the Rockies, a subject I'm becoming more interested in.

What interested me was that Kile had pitched fairly well during the first half of 1998

"as the season reached its midpoint, kile stood 5-10 with a 4.40 era, but let's parse that: on the road, 85 innings with a 3.19 era; at home, 40 innings, 6.97 era. moreover, four of the losses should rightfully have been wins, and at least two no-decisions also could/should have been wins as well. kile pitched well enough to end the first half at 10-6 or so, rather than 5-10"

I had always thought Kile just didn't pitch well for the Rockies at all. Of course he never pitched well at Coors perhaps because his curveball was flat but also because his control wasn't great in those days which will get you in trouble at altitude. Things really didn't fall totally apart until June of 1999 when he lost a 9-1 lead to the Cubs who went on to win 13-12. After that his days in Denver were numbered.

Park Effects at the K

A curious commenter wondered how Kauffman Stadium was playing this season in terms of parks effects. The table I had developed and updated at the end of last season now includes 2005 up to today.

 Home    Away   Index  
 Games Royals Opp Opp%+ Games Royals Opp Opp%+ Royals Opp Overall
2005 36 144 186 29% 35 167 210 26% -16% -14% -11%
2004 80 338 426 26% 82 382 479 25% -9% -9% -9%
2003 80 433 512 18% 82 403 355 -12% 10% 48% 28%
2002 81 434 505 16% 81 303 386 27% 43% 31% 36%
2001 81 382 485 27% 81 347 373 7% 10% 30% 20%
2000 81 451 488 8% 81 428 442 3% 5% 10% 8%
1999 80 441 449 2% 81 415 472 14% 8% -4% 2%
1998 80 353 492 39% 81 461 407 -12% -22% 22% -1%
1997 80 387 434 12% 81 360 386 7% 9% 14% 11%
1996 80 372 369 -1% 81 374 417 11% 1% -10% -5%
1995 72 285 346 21% 72 344 345 0% -17% 0% -8%
1994 59 325 287 -12% 58 249 245 -2% 28% 15% 22%
1993 81 370 354 -4% 81 305 340 11% 21% 4% 12%
1992 81 314 346 10% 81 296 331 12% 6% 5% 5%
1991 81 344 378 10% 81 383 344 -10% -10% 10% -1%
           
91-94 302 1353 1365 1% 301 1233 1260 2% 9% 8% 9%
95-03 715 3538 4080 15% 721 3435 3583 4% 4% 15% 9%
04-05 116 482 612 27% 117 549 689 26% -11% -10% -11%

So you can see that 2005 has turned out much like 2004 thus far with scoring down 15%. In the new era at the K in which the fences were moved back there has been 11% less scoring translating into a park factor of about 95.

Oh Say Can You Sing?

Well, here's something new. A company called Good Sports Recordings, Inc. has made a CD called "Oh Say Can You Sing?" which features 11 current and former major leaguers singing and performing various songs. Here's the list.

Ben Broussard - Cleveland Indians - Vocals & Guitar on U2's With or Without You
Sean Casey - Cincinnati Reds - Vocals on Toby Keith's How Do You Like Me Now?
Jeff Conine - Florida Marlins - Vocals on Stone Temple Pilots' Plush
Coco Crisp - Cleveland Indians - Rap on his original track We Got That Thing
Matt Ginter - Detroit Tigers - Banjo on The Dillard's Dooley
Aubrey Huff -Tampa Bay Devil Rays - Vocals on John Michael Montgomery's Letters From Home
Scott Linebrink - SD Padres - Vocals & Guitar on Pat Green's Wave on Wave
Jimmy Rollins - Philadelphia Phillies - Rap on his original track Wish List
Ozzie Smith - Hall of Famer - Vocals on Sam Cooke's Cupid
Omar Vizquel - SF Giants - Vocals & Drums on the Goo Goo Dolls Broadway
Kelly Wunsch - LA Dodgers - Vocals & Guitar on John Mellencamp's Hurts So Good

I'm not sure that either music or baseball fans will be enamored of these recordings but you gotta love that entrapenurial spirit.

Luck Ball

I know that others have written more completely on this topic but I just had to get my two cents in after watching some of the White Sox 4-3 victory over the Dodgers last night.

What raised my ire were Joe Morgan's comments on ESPN following the bottom of the 8th inning rally that allowed the Sox to score two runs and take the lead. In that inning following a walk the Sox elected to sacrifice. A misplay by the Dodgers and a bad call by the first base umpire allowed the batter to reach first safely. Once again the Sox sacrificed, this time resulting in runners on second and third with one out. A base hit followed that plated the two runs that were the difference in the game. In that inning Morgan made repeated references to "smart ball" - Sox manager Ozzie Guillen's supposed new offensive strategy focused on speed and situational hitting - and how that strategy is paying big dividends and is a key factor in their 46-22 record.

The problem with that theory is that it is demonstrably not true and it's not really even a matter of opinion.

The Sox of 2005 are winning because their pitching staff has given up just 3.84 runs per game which leads the American League. Their offense is performing well at 4.75 runs per game which ranks 5th in the league but clearly their strength is in run prevention not in run production. In 2004 in fact the Sox scored 5.33 runs per game (3rd in the AL) while tying for the leage lead in homeruns at 242. The problem last year was that they gave up 5.13 runs per game (10th in the AL). While you can attribute some of their improvement in run prevention to an improved defense, there is simply no way that it comes even close to accounting for one and a quarter runs per game. It should also be noted that while the Sox do lead the AL in stolen bases (71) and sacrifice hits (25), they also are third in the league in homeruns with 82 and 25% of their runs have scored on homeruns (4th in the AL).

Trading away Carlos Lee for Scott Posednik and letting Maglio Ordonez go are not the reason the White Sox are in the first place. Starters Mark Buerhle, Freddy Garcia, John Garland, and Jose Contreras, along with relievers Dustin Hermanson, Cliff Politte, and Neal Cotts are the reason. These seven pitchers have thrown 75% of the White Sox innings, made 54 of 67 starts, and recorded a 3.36 ERA.

A second reason for the Sox success Morgan mentioned is their 20-8 record in one-run games and he seemed to link "smart ball" to this ability to win close games. Once again the historical data just don't support the idea that a team or manager has the ability to win such a large percentage of their one-run games. Using this article by Bill James the Sox this year would have been expected to win 40 or 41 games based on their ratio of runs scored to runs allowed and 15 of their 28 one-run games. So even with their excellent pitching they've still outperformed expectations by 5 or 6 wins overall and 9 wins in one-run games. They're also playing lots of one-run games and leading the league in that category when they would be expected to have played just 15 one-run games using the formula in James article.

Simply put, I just don't see the White Sox .676 performance being sustainable. Not only are some of those seven pitchers likely to fall back towards their historical norms as the season progresses, but their record in one-run games should start to regress towards the mean and their offense is already producing about as well as you would expect (although the return of Frank Thomas should help). That doesn't mean they won't win their division but right now they're relying on a style of play I'd rather call "Luck Ball".

Time

Here's a quote that hits home for me from time to time by C.S. Lewis....

"The great thing if one can, is to stop regarding all the unpleasant things as interruptions of one's 'own life' or 'real life'. The truth is of course that what one calls the interruptions are precisely one's real life - the life that God is sending one day by day."

Lewis also puts this same sentiment in the mouth of Screwtape in the The Screwtape Letters when he coaches Wormwood to help inculcate the attitude "my time is my own" into his patient. That attitude is especially tempting to me as the weekend approaches.

Now that I've shamed you all into two days filled with selfless action, have a nice weekend!

Road Kill

With last night's extra inning loss to the Indians at Jacobs Field the Rockies fell to 4-25 on the road, the worst record in baseball behind even Tampa Bay. As has historically been true the problem is both offensive and defensive but the offense has been particularly anemic averaging just 3.31 runs per game on the road - last in the majors. The pitching has been bad with an ERA of 6.07 but at least its only 28th ahead of Cincinnati and Tampa Bay.

As of now the Rockies project to win just 11 road games all season (a .138 winning percentage) which would break the record for futility on the road topping the 1935 Braves who had just a .167 winning percentage.

This morning I re-read the 2003 Baseball Prospectus article on the Rockies which argued that the Rockies don't need to learn how to win at altitude - they already do that - they simply need to learn how to win on the road. And learning how to win on the road means getting better players, especially pitchers. It's interesting that the article references The Hidden Game of Baseball where Palmer and Thorn discuss what it takes to win championships in chapter 12. I grabbed my copy and re-read the relevant portion. The following were their conclusions.

"Yet to win a pennant with an extreme Park Factor, a team must construct its talent to take maximum advantage of what its home park hinders, not what it helps....A team whose home park favors hitters to an extent 10 percent above average (PF 110) or more cannot have a won-lost record 10 percentage above average at home and win a pennant - at least it has never been done by any of the 94 teams which have played in hitters' parks. So for such a team to win a pennant, it must (a) have exceptional pitching and (b) win big on the road. The 14 pennant winners from hitters' parks have produced 11 league-leading pitching staffs and have played .627 on the road - 170 points, or 37 percent, above average."

Walks Followed by Homeruns

A week or so ago someone on the SABR-L list asked the question as to whether the theory that homeruns are more frequent following walks was true. The reasoning behind this theory, which the questioner had heard an announcer posit, is that the pitcher often tries to re-establish the strike zone after a walk and hitters end up taking advantage.

To see if this is true I calculated the frequency of homeruns following a variety of events in 2004. Here are the results sorted by frequency.

Event     HR    Events/HR
Homerun  197       27.7
Walk     430       34.5
K        920       34.6
Double   240       37.2
Single   751       39.0
Triple    22       40.8
I-Walk    30       46.0
HBP       39       47.4
Out     2584       47.5

From this it doesn't seem like the theory probably holds as any kind of general rule. Homeruns occur more frequently after other homeruns and are not signficantly more frequent following walks than strikeouts or doubles.

Note: This analysis doens't include homeruns that followed errors, fielder's choice, leadoff homeruns, and 118 homeruns that followed other plays that were not batter events out of the 5,451 homeruns in 2005.

Blake Street Bombers Redux?

Saw this article from the NY Times on winning at altitude over at Baseball Musings. Not too much that hasn't been said before. Nobody knows how to win blah blah blah...throw everythinig out the window blah blah blah. What I found interesting however, was the following:

"Going into last night's games, Colorado's earned run average of 5.68 was tied for last with Tampa Bay among the 30 major league teams. It has never been better than 24th. That was in 1994, when there were 28 teams.

When Colorado made its only playoff appearance, as a wild-card team in the strike-shortened 1995 season, its E.R.A. was 4.97, the third worst in baseball."

What the author fails to realize is that since the pitcher park factor (PPF) at Coors in 1995 was 128 the Adjusted ERA for that team was 108, in other words 8% above average. Meanwhile Rockies hitters in 1995 had an adjusted OPS of 99 due to the batter park factor (BPF) of 128 - actually below average. So their pitching was their strength in 1995, not their weakness as implied in the article.

Recently I've heard lots of discussion on talk radio in Denver on this topic and caller after caller falls in line with the idea that the only way to win at Coors Field is to stack the lineup with hitters ala the "Blake Street Bombers" and simply outscore your opponents both at home and on the road. Even David at Baseball Musings said:

"The more I think about the Coors problem, the more I believe the Rockies can only win by being a super offensive team. They can't be content with hitters who look great at home but stink on the road. They need to find four of five superstars who will just pound the ball and push the team toward 8 runs a game at home. Then just hope the pitching can hang on. They need an infield of A-Rod, Tejada, Kent and Helton."

While I hesitate to differ with someone as knowledgable as David, it seems to me that pitching, specifically relief pitching, and not hitting is the key. I say that because the year the Rockies were most successful (not 1995 but 2000), their pitching was the key. As shown in the table below, in that year the Rockies pythagorean record was 87-75, their best ever.

    OPS+ ERA+ pW pL  W  L
1993  89  91  63 99 67 95
1994  95  96  53 64 53 64
1995  99 108  72 72 77 67
1996 100  97  81 81 83 79
1997 105  99  82 80 83 79
1998 102 102  78 84 77 85
1999  90  95  72 90 72 90
2000  90 112  87 75 82 80
2001 102  98  82 80 73 89
2002  86  94  70 92 73 89
2003 100  91  78 84 74 88
2004  96  91  73 89 68 94

In 2000 their ERA+ was second in the league at 112 while their OPS+ was second to last in the league at 90 and yet they still were projected to win 87 games. That team had five very solid relievers including Jose Jimenez, Mike Myers, Gabe White, Mike DeJean, and Julian Tavarez who started 12 games and relieved in 39. In the starting rotation Pedro Astacio and Brian Bohanon turned in good performances although the other three primary starters were pretty average.

And to me it makes sense that a strong and deep bullpen should be the focal point in an environment where lots of runs are scored for two reasons. First, in games at Coors Field your starters will tend to throw lots of pitches and therefore won't be able to get into the 7th and 8th innings very often. That means that often your bullpen will be called upon to pitch three and four innings per night. On most teams the first reliever in the game in the 5th inning is a guy who is a marginal major leaguer. In Coors Field I think you can take advantage of other teams relievers by having a quality guy in that spot that can hold the oppostion while you pile up runs. For the Rockies and only the Rockies I think it may make sense to carry 12 pitchers. For this reason it seems that someone who plays a role like Brooks Keishnick would be useful to the Rockies since they'll typically need more pinch hitters.

Secondly, for whatever reason, the Rockies hitters perform terribly on the road. Some have contended that this is due to a "hangover effect" where hitters have trouble adjusting to breaking balls at lower altitudes. While this seems logical there really isn't any evidence to support it as I've tried to document here. Others suggest that the Rockies have simply had poor hitters who look good at altitiude. I'm not discounting that theory entirely but in my quick look at the theory its not obvious that's what's happening. In any case, despite the bombers of the past the Rockies have consistently ranked at or near the bottom in runs per game on the road.

     rdR/G Rank
1993 3.32  14
1994 4.27  11
1995 4.17  12
1996 3.74  14
1997 4.67  10
1998 3.70  15
1999 4.12  15
2000 4.14  16
2001 4.56  11
2002 3.46  16
2003 4.15  15
2004 4.16  14

So in order to win road games the Rockies need to keep those games close by, for example, pulling starters earlier than other teams and shutting down the opposition in the late innings.

It would be nice to see this point of view raised a bit more on the front range anyway.

La Russa and Moneyball

Here's an interesting review of 3 Nights in August that contains comments on Moneyball. Haven't read the former book but plan to based on this review.

What I found interesting was the following:

"in a passage numerous critics have focused on, Bissinger probes La Russa’s thoughts on on-base percentage, the golden stat of the Moneyballers. La Russa, one of the game’s most successful managers, sees OBP 'as akin to the latest fashion fad oversaturated, everybody doing it, everybody wearing it, until you find out the hard way that stretch Banlon isn’t quite as cool as originally perceived.' La Russa combats the rise in OBP by urging his players to 'play the scoreboard.' La Russa wants to see aggression from a leadoff hitter and his RBI men. He doesn’t want to see players taking pitches down the plate in an effort to draw a walk to boost that OBP."

I think there are two ways of looking at these comments. The first is that La Russa is simply wrong if in his comment about Banlon he's actually saying that OBP is not as important as "stat heads" think. The structure of baseball - not performance analysis - dictates the importance of getting on base and since OBP reflects that ability, there really isn't alot of wiggle room for those who denigrate its importance. A comment like that is akin to saying that homeruns are just a fad and they aren't really as important as people think. No, the structure of baseball is such that homeruns are the most efficient way to score runs and that's the way it will always be. End of story.

However, what I think La Russa was really getting at is that if you make OBP the most important thing in all situations, then you are not playing the game strategically. There are times when sacrificing an out is important just as there are times when swinging at a 3-0 pitch is beneficial to the team. That doesn't mean that one should sacrifice outs willy-nilly or swing away all the time.

To me, the lessons of Moneyball are those strategies and approaches that can be applied over a large number of trials and that reflect the "house odds". What La Russa and others are typically referring to are exceptions to those strategies in very specific situations. Both are correct and baseball should have room for both.

Pitch Outs

So how often do pitchouts succeed?

In looking at 2003 and 2004 data I found the following:

Pitchouts:       232 
Stolen bases:    100 (43%)
Caught Stealing: 124 (53%)
Passed Ball:       2 (>1%)
Wild pitch:        5 (2%)
Pickoff:           1 (>1%)

It surprised me a bit that the stolen base percentage was as high as 43% when a pitch out occurred. When you add the other positive outcomes the percentage raises to 46%. That's not bad considering that the defense guessed right in these 232 instances.

Neifi an All-Star?

Ok, so I've never been a Neifi fan and I still cringe every time I see him come up in an important situation. That said, I have to grudgingly admit that he's been very solid thus far and without him the Cubs would be in even worse shape (why they still hold on to Jose Macias and Enrique Wilson is another matter).

But does that make him an All-Star? Derrek Lee and Nomar Garciapara seem to think so. In this article on MLB.com it was noted:

"Neifi's been unbelievable all year," Chicago first baseman Derrek Lee said. "In that leadoff spot or No. 2 spot, he's been on base the whole time. He makes the plays, he's won a Gold Glove. The way he's swinging the bat, he's an All-Star right now."

I hate to disagree with Derrek but Neifi's OBP thus far is .331, hardly impressive in the leadoff spot where he's walked just 7 times in 232 plate appearances. He's been more impressive with his .449 SLUG.

Derrek and Nomar also noted...

"He deserves it," Lee said. "Look at his numbers. I don't think there's a shortstop with better offensive numbers."

"There should be a Cubs shortstop there, and it's the guy who's playing there right now," Garciaparra said of Perez. "He's been unbelievable. He definitely deserves it."

Well, in looking at the numbers Neifi is 9th among NL shortstops in OBP, 4th in SLUG, and 5th in batting average. In OPS he's fourth at 779 behind Felipe Lopez (917), Clint Barmes (886), and Bill Hall (848). If I were voting for NL shortstops I'd have to go with Lopez at this point.

Pocket PCs and MLB

I missed this article on Pocket PCs and baseball from last September written by The Numbers Game author Alan Schwarz.

The Baseball Same Game

Recently Stephen Lombardi was kind enough to send me a copy of his recently published book The Baseball Same Game: Finding Comparable Players from the National Pastime (224 pages, iUniverse Books). I got a chance to read it over the holiday weekend and give you a glimpse into the methodology and structure of the book.

In the book Stephen presents 65 "cases", each comparing two players who had very similar career statistics. The statistics, or as he points out better called "performance data", he uses for position players are:

• Games played
• Plate appearances
• Runs Created Above Average (RCAA)
• Offensive Winning Percentage (OWP)
• OPS vs. League
• Runs Created per Game (RC/G) vs. League

Of course, what I like about this approach is that he uses sabermetric performance measures to make his comparisons rather than the traditional runs, RBIs, homeruns, and batting average. These measures serve as a much better foundation since they more closely track with producing runs and therefore winning baseball games. In the case of two of the measures that use Runs Created as a basis (RCAA, OWP) they are also adjusted for the park the player played in thereby creating a level playing field when players find themselves playing in extreme parks. I also like that when he uses OPS and Runs Created he places them in the context of the league in which the player played.

By including games played and plate appearances he contextualizes the other stats so that the comparisons are meaningful (for example two players with an OWP of .560 are not really comparable if one had 1,000 career plate appearances and the other 10,000). All together then, these measures do a great job of finding comparable offensive players.

While I do think this methodology is on the right track, I have two criticisms. First, OPS vs. League and RC/G vs. League are not park adjusted. As I’ve discussed previously OPS can easily be normalized for both park and league to be a more accurate measure. I’m not sure why RC/G vs. League could be adjusted as well.

Second, within the cases themselves the measures are presented as follows as in case #1.

                Roy Campanella Sixto Lezcano
G                       1215        1291
PA                      4816        4814
RCAA                     135         146
OWP                     .586        .606
OPS vs. League          .107        .084
RC/G vs. League         1.27        1.17

As you can see the measures against league are presented as the difference between the league average and the player measures. In this case Campanella created 1.27 runs per game more than the league average and had an OPS .107 higher than the league average.

In thinking about this presentation it occurred to me that it would tend to help players who played in eras when more runs were scored. This would be the case since the run environment dictates the cost (in runs) of additional wins. For example, using the Pythagorean formula a team that gives up 700 runs and scores 700 runs will obviously win 81 games in a run environment where 4.32 runs per game are scored. To get to 82 wins, they need to score 710 runs. Therefore the cost of that additional win was 10 runs (they could also get that extra victory by giving up five fewer runs and scoring five more). If, however, a team gives up 600 runs and scores 600 (a run environment of 3.70 runs per game) that 82nd victory can be purchased at the cost of eight runs. As a result, a player that created 1 run more per game in the 4.32 run per game environment would not be as valuable as a player who created 1 run more per game in a 3.70 run per game environment. A better comparison would be to divide the players RC/G by the league average and multiply by 100 to produce a value centered around 100. In fairness to Stephen this is essentially what offensive winning percentage (OWP) does by taking the RC/G and dividing it by the sum of the square of the RC/G and the league average of the runs score per game. The same argument can be used for OPS where a difference of .107 in a lower scoring league (1968) means more than the same difference in a high scoring league (2001).

On the pitching side he uses:

• Innings Pitched
• Runs Saved Above Average (RSAA)
• ERA vs. League
• K to BB ratio vs. League
• Base runners allowed per 9 IP vs. League
• K per 9 IP vs. League

Of these, RSAA (along with RCAA) is a sabermetric measure created by Lee Sinins of Sabermetric Baseball Encyclopedia fame. RSAA measures the number of runs the pitcher saved over and above an average pitcher given the same number of innings and league and is park adjusted.

As with the offensive measures that are compared to the league average, these have the same issue of relative importance given the run, base runner, or strikeout and walk rates for the league in question.

As for the cases themselves there are some very interesting ones. I especially enjoyed seeing the similarities between Tom Seaver and Christy Mathewson, Mark McGwire and Johnny Mize, George Brett and Sam Crawford, Fergie Jenkins and Eddie Plank, and Barry Larkin and Jim Rice. These are especially interesting when the players made their contributions in very different ways as in the case of Larkin and Rice and even seen in Mize and McGwire. In each case Stephen provides a brief synopsis of the two careers and often uses the chance to discuss the relative merits of one or the other player for the Hall of Fame or makes a point about some other aspect of their careers that contributes to a fan's perception of these players - for example, in the case on Willie Hernandez and Jeff Reardon discussing the value of the Save statistic. I enjoyed the readability of the book and found lots of nuggets of info I hadn't heard before.

In some of the cases, such as Campanella and Lezcano above, Larkin and Rice just mentioned, and cases like Bruce Sutter and Smokey Joe Wood, I tend to question the value of the comparison not because the numbers don’t line up but because the players in question played different positions. Stephen fully realized that this would cause a bit of controversy and noted on his web site that

“my point is that once you step into the batter's box, you're a hitter, and you should not be given extra credit (or lose something) because of the position you play in the field when you are not batting.

Giving someone ‘extra’ or an ‘adjusted’ offensive value is his relative batting results because of his position in the field implies that just playing that position in the field provides a batting benefit to his team.”

While that reasoning makes perfect sense my preference would have been to make comparisons by ranges in the defensive spectrum so that the comparisons would have been a bit more meaningful. For example, while Larkin and Rice may have had similar offensive numbers, Larkin was far more valuable because he played a more demanding defensive position. And I’m not sure that comparisons between starters and relievers (Sutter and Wood for example) make much sense since it has been shown that reliever’s ERAs are not really equivalent to starter's ERAs and therefore should be adjusted based on something like Component ERA.

And as Stephen says, these are his cases and you could find your own, but I'm glad he did the work. Overall, it's an interesting book that I'd recommend.

Gyroball

Ok, I'm a little late to the game but just read Will Carroll's description of the "gyroball" or "shooto" as it is called in Japan. It is supposed to be a pitch with screwball action but without the elbow and wrist strain. As Carroll puts it, a "set it and forget it pitch". Carroll claims that a Japanese pitcher actually throws it although there is some disagreement on this point and whether the gyro is the same as the shooto.

I'm not sure I follow his description of the arm motion and I haven't seen it. Apparently, Carroll taught the pitch to an Indiana high school pitcher who used it in his opening start this season.

Could be the next big thing in the evolution of pitching following the curveball, spitball, knuckball, slider, split-fingered fastball...

Rockies Draft Analysis

Waiting for the Rockies/Tigers game today at Coors Field and going over the 2005 Rockies Draft Choices. Here's how they break down.

52 players drafted
27 pitchers (19 right-handed)
13 infielders
8 outfielders
4 catchers
17 high schoolers
35 college and junior college

The balance was pretty good between pitchers and position players and was fairly heavy on college players (as opposed to the Angels who went out of their way to select high schoolers). The first pick (7th overall) was Troy Tulowitzki (TWO-lo-witz-kee), a 6'3" 205 pound shortstop from Long Beach State University who bats and throws right-handed. He was selected after his junior season in which he hit .349 with 8 homeruns and 29 RBI in 39 games. He was a two-time All American and three-time All-Big West selection. He'll turn 21 in October and has agreed to terms with the Rockies including a $2.3M signing bonus, working out with the team yesterday. He'll report to class A Modesto on June 17th.

He reportedly has a good arm although since he's pretty big for a shortstop Walt Weiss, who works for the Rockies in the front office will be watching his footwork.

"Primarily, a guy's footwork will tell you a lot about him. If you're going to play in the middle of the infield, it's all about your feet."

Baseball America ranked him as the fourth best prospect before the draft. I'm not sure why he slipped to 7th.

In the supplemental round with the 52nd pick overall the Rockies chose Chaz Roe, a 6'5" right-handed pitcher from Lafayette Kentucky High School. The comments on MLB.com on Raz are:

"Loose, quick arm. Shows no major mechanical problems. Tough FB that bores into RHH hands. Bat-breaker in wood bat leagues. Throws a knee-buckling CB for his out-pitch. Comes at batters w/ a hard, tight 3/4 tilt w/ late bite."

In the Bill James interview I blogged about the other day James noted that in a high school pitcher what they look for is a "clean delivery" which is I assume analagous to "no major mechanical problems".

No Patience

While shooting the breeze with my brother during the latest Royals drubbing we got to discussing offensive ineptness. On that subject he mentioned that Mariano Duncan once had a lower on base percentage than batting average. My first reaction was of course that this wasn't possible but then he reminded me that on base percentage is calculated by including not only walks and hit by pitch but also sacrifice hits and sacrfice flies. So here was Duncan's 1995 season with the Phillies.

 AB  H BB HBP SF SH  AVG  OBP
196 56  0  1   3  1 .286 .284

The question is, has anyone been more inept? As hard to believe as it is the answer is yes.

                          AB  H  BB HBP SF SH  AVG  OBP
Alfredo Griffin 1984 TOR 419 101  4  1   4 13 .241 .240
Rob Piccilo     1980 OAk 271  65  2  0   0  8 .240 .238

James Again on BDD

Just got a chance to read part II of Baseball Digest Daily's interview with Bill James. I commented on part I of the article here. Several more interesting comments on the importance of defense and scouting.

First, James comments on learning plate discipline.

"I think it is easier to learn plate discipline than it is to learn speed or to develop a strong throwing arm—but not much easier. A player who lacks plate discipline at age 18 will usually lack plate discipline at age 30. But not always; some players can adapt well to the challenge of learning to lay off certain pitches."

To me this makes sense given my recent reading on pitch recognition and the Go/No Go reaction time. There are obviously certain physical skills that a good hitter is born with like dynamic visual acuity and there are some skills that can likely be improved with practice like Go/No Go. Hitter's walk rates increase with age I assume because of a combination of improvement in some of these skills coupled with experience.

A comment that surprised me a bit was this one in response to a question about the coachability of high school versus college players.

"There is an argument that the best and only place to play major league baseball is in the major leagues, and also there is a strong argument that one's ability to learn is inversely proportional to one's age. Therefore, the further you can advance a player at a young age, the better."

At first glance this sounds like it goes against a bit of conventional sabermetric wisdom that college players are better investments in the draft. But on second glance the key phrase is the last one. If you can advance a player at a young age you'll be better off but the odds of doing so are reduced because of the injury gauntlet that high school players, especially pitchers, must run. James explains his view on this topic in response to another question.

"What I think happens is, between the ages of 18 and 21, a large number of pitchers are going to hurt their arms during the process of learning to pitch. If they go to college, they’ll hurt their arms in college. If they enter pro ball, they’ll hurt their arms in pro ball.

So which side of the bridge do you want to stand on? Do you want to choose pitchers AFTER they have been thinned by the injuries that occur at ages 18 to 21, or do you want to choose pitchers before they have been thinned by the injuries that occur from 18 to 21?"

So James' view is that in most cases injuries might happen because injuries simply happen. Therefore major league teams who think that drafting a high school player will protect them from injuries because of training at the professional level are probably just kidding themselves. James previously articulated this view in his book The James/Neyer Guide to Pitchers which I highly recommend.

Which of course leads us to the Cubs first round (20th pick) draft choice which they used to select Mark Pawelek, a left-handed high school pitcher out of Springville Utah. He's 6'3" 190 pounds and struck out a Utah high school record 476 batters in his prep career. This season he struck out 132 in 63 innings while not giving up an earned run. The Cubs apparently liked him because they feel he can add strength given his frame, has good command, and he already throws 90 to 96 mph. When asked about taking a high school pitcher Cubs scouting director John Stocksill said, "We just felt the upside of Mark outweighed who was on the board from college and more ready." Baseball America ranked Pawelek third among high schoolers in terms of being closest to the majors.

The full scouting report shown on MLB.com is:

"LHP w/ ideal pitcher's frame. Tall and long w/ lean muscle. Explosive low-90s FB that sinks and runs. Comes out of a quick, loose whip-like delivery. Curveball is out pitch, w/ 1-7 break that comes near the plate."

You can also see a scouting video on Pawelek here.

So is this a bad pick? Sure Pawelek is a long way from the major leagues but there is just the chance that he has enough old-pitcher skills to progress quickly and barring injury perhaps we could see him in Wrigley Field in three years or so. We can always hope anyway.

Overall, the Cubs chose 10 pitchers in the first 18 rounds, 12 college players, 3 high schoolers, and 4 from community or junior colleges.

Ryan Shealy

Last weekend I had a chance to go see the Sky Sox and take a first hand look at Rockies' prospect Ryan Shealy. Shealy was drafted in the 11th round of the 2002 draft out of the University of Florida. He's a 6'5" 240 pound guy and looks every bit of it on the playing field.



He's a right handed hitter and thrower and this season is now 24 years old and in his fourth minor league season. Last year in AA at Tulsa he was named player of the year in the Texas League and hit 29 homeruns and 32 doubles in 530 plate appearances with a 995 OPS. Interestingly, he has a track record of getting hit with pitches and was hit 48 times from 2002-2004 (I can't find his HBP this season) - indicating that perhaps the book on him is try and jam him.

On the down side his walk rate has declined precipitously this season:

      PA/BB
2002  5.62
2003  9.12
2004  8.69
2005 18.46

And although he's still hitting .318/.364/.536 in 203 plate appearances, the massive park effects in Colorado Springs have much to do with this. His strikeout rate has also inched up this season:

     PA/K
2002 5.40
2003 5.32
2004 4.31
2005 5.49

But his biggest weakness is that he projects to be far on the left-side of the defensive spectrum. In the game I witnessed he misplayed two balls at first base, not charging one ball aggressively enough and then dropping another.

Hopefully he can improve his plate discipline, cut his strikeout rate a bit, and become, if not good, an adequate first baseman.

And Even More on Bunting

I think I've caused folks some confusion (most of all me) in my post on bunting. I took a few minutes today and tried to recast the numbers so it is clear.

Successful sac bunts: 1,731
Bunts that are potential sacs: 466
Bunt strikeouts in sac situations: 162
Positive outcomes from potential sacs: 241
Negative outcomes from potential sacs: 389 (includes strikeouts)

Upper limit of sac percentage (includes all bunt singles): (1731+241)/(1731+241+389) = 83.5%

Lower limit of sac percentage (no bunt singles): (1731+18)/(1731+18+389) = 81.8%

Outcomes by percentage for all sacrific attempts where Lower does not include singles and upper does:

Negative         Lower   Upper
Double play:      1.4%   1.2%
Force out:        8.3%   7.5%
Fielder's Choice: 1.0%   0.9%
Strikeout:        7.6%   6.9%
Total:           18.2%   16.5%


Positive                 Lower   Upper
Error:                    0.6%   0.5%
Single + Error:           0.0%   1.2%
Single:                   0.0%   8.2%
Fielder's Choice + Error: 0.2%   0.2%
Force out + Error:        0.1%   0.1%
Successful sac:          81.0%  73.3%
Total:                   81.8%  83.5%

Sammy in Fenway

Courtesy of SABR-L, with Sammy Sosa's homerun in Fenway Park on June 1 he becomes the player with the longest time span between homeruns in a ball park and probably the player with most homeruns in between (577).

Here is a list of others with long spans from David Vincent:

Sammy Sosa            5824 06/21/1989 06/01/2005 Fenway Park
Luke Appling          5810 06/04/1933 05/01/1949 Sportsman's Park III
Sam Rice              5520 05/05/1919 06/15/1934 Shibe Park
Early Wynn            5445 06/03/1944 05/01/1959 Comiskey Park I
Fred Clarke           5399 07/04/1896 04/17/1911 West Side Park II (Chicago)
Gary Sheffield        5367 10/01/1988 06/12/2003 Network Associates Coliseum
Jimmy Dykes           5331 10/02/1921 05/07/1936 Griffith Stadium
Pee Wee Reese         5116 08/02/1941 08/05/1955 Wrigley Field
Ralph Winegarner      5095 08/11/1935 07/23/1949 Sportsman's Park III
Luke Appling          5075 08/26/1932 07/19/1946 Fenway Park
Gary Sheffield        5070 06/28/1990 05/15/2004 Yankee Stadium
Buck Martinez         5052 08/05/1969 06/05/1983 Memorial Stadium
Ron Fairly            5050 07/21/1963 07/21/1963 County Stadium
Rafael Palmeiro       5008 10/01/1988 06/18/2002 Wrigley Field

Recognition Reaction Time

Recieved an interesting abstract of an article from unca related to baseball the other day titled "Intensive baseball practice improves the Go/Nogo reaction time, but not the simple reaction time" published in the journal Cognitive Brain Research and written by Kida N. Oda and S. Matsumura M.

"Baseball hitters are required to make decisions whether to swing or not as quickly as possible. Therefore, we can assume that skilled baseball players have a quicker response. To verify this hypothesis, we assessed the effect of baseball experience or skill levels on simple reaction times and Go/Nogo reaction times in 82 university students (22 baseball players, 22 tennis players, and 38 nonathletes) and 17 professional baseball players. Also, to clarify whether this ability was innate or acquired, we examined the effects of long-term practice for baseball hitting in 94 senior high school students (26 baseball players and 68 non-baseball players).

There were no differences in simple reaction time either for sports experience or for skill levels. On the contrary, the Go/Nogo reaction time for baseball players was significantly shorter than that of the tennis players and nonathletes. The Go/Nogo reaction time of higher-skill baseball players was significantly shorter than that of lower-skill players, while that of the professional baseball players was the shortest. The professional players showed the highest (almost linear) correlation between the Go/Nogo reaction time and simple reaction time. A longitudinal study showed that 2 years of hitting practice improved the Go/Nogo reaction time, while the simple reaction time remained constant. A cross-sectional study of high school non-baseball players showed no differences either in the simple or Go/Nogo reaction times in relation to school year. In conclusion, intensive practice, including Go or Nogo decision making, improved the Go/Nogo reaction time, but not the simple reaction time."

So from this study it appears that simple reaction time - the ability to respond to a stimulus which includes the mental processing time and movement time - is innate and cannot be improved with practice. However, identifying pitches and making the decision as to whether or not to swing based on location, speed, and spin, is categorized differently as "recognition reaction time" here called the Go/Nogo reaction time. It appears that this ability can be improved with practice although I'd assume that baseball players in general are self-selected for this ability. This is similar to the conclusion that dynamic visual acuity (the ability to perceive information in moving objects - is unrelated to static visual acuity. Together they provide two clues as to the physical attributes required of professional hitters.

In wondering how the study might have been conducted I ran into this short description.

"In Type C, or 'recognition,' reaction time, there are multiple possible signals but only one response. In this case, the responder makes the response when one stimulus occurs but withholds response when the other(s) appears. The standard lab version of this paradigm has a subject with his/her fingers on 1 telegraph key and waits for one of x different lights to flash. When the signal light occurs, s/he releases the telegraph. If one of the nonsignal lights occurs, then the subject should make no response. This is sometimes called the 'go, no-go' paradigm. Reaction times are invariably longer than for simple reaction time."

A Royal Buddy

I know that's a lame title for a post but it's hard to think of a pithy title for everything you write.

Anyway, the Royals have hired a manager and selected....Buddy Bell?

I guess I'd have to side with Rob and Rany and think that perhaps Allard Baird pulled the trigger a bit too fast and should have instead been concentrating on the draft. After all, it certainly couldn't have hurt to have Bob Schaefer play the interim role for a bit longer during this train wreck of a season.

That said, I find it hard to believe that Bell was the best available candidate from a tactical perspective. After all, he hasn't exactly distinguished himself in the past with a record of 352-458 in his tenure with the Rockies and Tigers. His best season was an 82-80 finish in 2000 with the Rockies courtesy of Baseball Reference.

 Year    League   Team     Age    G     W    L    WP   Finish
 1996 AL East     Detroit   44   162   53  109   .327      5
 1997 AL East     Detroit   45   162   79   83   .488      3
 1998 AL Cent     Detroit   46   137   52   85   .380      5

 2000 NL West     Colorado  48   162   82   80   .506      4
 2001 NL West     Colorado  49   162   73   89   .451      5
 2002 NL West     Colorado  50    22    6   16   .273      4

                  Colorado       346  161  185   .465
                  Detroit        461  184  277   .399

      TOTAL                      807  345  462   .428

So apparently it isn't his results on the field that led Baird to hire him. In fact, if you look at Brad Doolittle's WAMMER ratings Bell ranks 96th out of 100 managers at -15.6 - meaning that his teams underperformed by almost 16 wins during his tenure based on a combination of their pythagorean record given their offensive and defensive elements and their record in one-run games. I'm not completely sold on Doolittle's methdology since Bill James has shown that winning percentage in one-run games is a pretty ephemeral thing to try and measure, let alone stick completely on a manager. In addition, I think the methodology penalizes managers of bad teams by assuming that their teams should play .500 ball in one-run games. James showed in the article just mentioned that that wouldn't be the case. Bad teams should be expected to lose more one-run games than good teams.

What it was that led Baird to Bell, according to this story on Yahoo was his combination of experience, teaching ability, and toughness.

"We're very excited about this," general manager Allard Baird said. "If you look at his career -- 18 years in the big leagues -- he's a teacher, a communicator.

"He's been there and has experienced it as a player and as a teacher. He has patience but is extremely demanding."

So there you go. My hope is that Baird sets some philosophical direction in the front office and that Bell has bought into it but I'm not holding my breath. The Royals gave him a contract through 2007 but I don't know the amount.

Interestingly, the Yahoo article also mentioned that owner David Glass said that there is room to expand the payroll next year as well above the $39M level of 2005. I'm wondering how that fits in with the youth movement given that the two biggest offseason acquisitions, Eli Marrero and Terrance Long are huge busts up to this point. Does it mean that they spend enough to sign a quality free agent or keep spending a couple of million on mid-level players? It seems to me Baird would be better off using the free talent pool of replacement level players to fill positions until the young players are actually ready. That means Gotay, Burgos, Nunez, Bautista, et. al. get some more seasoning in the minors before being asked to step into a major league role. Of course, what I'd like to see Glass do with that money is draft the best player available in a couple weeks and pay the signing bonus. That would probably be their best investment.

Low Score Coors

Just getting set to score the Rockies/Cardinals game for MLB.com tonight at Coors Field. In looking over the press notes I noticed that last night's 2-1 victory by the Rockies was the third lowest scoring in Coors Field history. There have been two 2-0 games (one in 1995 and one in 2002) and now eight 2-1 games, the most recent being September 23, 2004 when the Rockies beat the Diamondbacks in 10 innings.

Jason Jennings became the career leader in victories at Coors Field overtaking Pedro Astacio with 25 after last night and ran his record here to 25-16 lifetime.