Hey!

Welcome to Hey Bill, where Bill answers questions from his subscribers almost every day. Visitors can read the most recent Hey Bill's on this page.  Subscribers can ask Bill a question directly and also view our archive of questions and answers.

 

The fifteen most recent questions are listed here and will change almost every day.

Hey Bill

FAQ Categories

15 Most Recent Questions

"what about evaluating pitchers just by the three true outcomes... Or has that already been done?"  
 
The reader is referring to DIPS advanced by Voros in the early 2000s, or my shortcut version FIP.  While I agree that we shouldn't exclude the other 70% of the outcomes, doing do still gets us 90% of the where we want to be.  I posted career totals of active pitchers in FIP and ERA here: http://www.twitter.com/tangotiger/status/1136784215838736385
Asked by: tangotiger

Answered: 6/18/2019
 Noted.  I didn't read it that he was referring to FIP, only that FIP was included within the area he was asking about.  

 

About Milwaukee County Stadium in the 1953-1965 era:  
 
The outfield fences were changed a few times when additional seating was built beyond the outfield. The dimensions Chuck quoted are not quite right. The power alleys were 377 or so. The stadium had additional distances posted, 355 (or 362) halfway between the power alleys and the foul lines, and 394 (or 392) halfway between the alleys and centerfield. His mistake comes from the Seamheads ballpark database, which mis-reports the distances.  
 
In the mid-1950s NL there were a bunch of good to great HR parks. County Stadium in 1953 was probably the toughest HR park in the NL (when Forbes still had the short fence up for Kiner). The other parks were Wrigley, Ebbets, Sportsman's, Shibe, Crosley, and the Polo Grounds.
Asked by: stevebogus

Answered: 6/18/2019
 Thanks.  There's an interesting "trick of perception" thing there, granting that there are interesting trick of perception things everywhere.   But because the Braves had SO MANY great home run hitters in that era (Aaron, Mathews, Adcock, to a lesser extent Crandall, Wes Covington).  .because the Braves hit so many home runs, people perceive it as a home run park, although the data clearly shows that it was not.    I think Adcock was actually the best home run hitter of the group, on a per-at-bat basis.   

 

There were a couple of errors in the estimate of Khris Davis having the same batting average for 4 straight years.  The first error is that to get the probability of his hitting the same for two years, we should just ask what is P(hits .247 in year 2); not P(hits.247 in year 1) times P (hits .247 in year 2).  So for Davis, the "only" remarkable thing is that he hit .247 again in three more years, not four...Against that, if he ever had, say, 552 AB, then it's impossible for his BA to round to .247.  After the first year, the probability of his getting a suitable number of AB in a given year is about 0.6.  Factor this in and I make it 1 in about 80,000 (not 365,000).  Still, I think that means that we would expect only about one player in MLB history to do it.
Asked by: PeteRidges

Answered: 6/18/2019
 Thanks; good point about the three seasons rather than four, and I should have spotted that myself.  THe point about it being impossible to hit .247 in 552 at bats, it seems to me, would only be relevant if .247 was an especially difficult or especially easy point to land on.   In other words, it would be relevant if the number was .250, because .250 is (I would assume) a relatively "easy" target to hit, and it would be relevant if it was .249, because .249 is (I would assume) a target that cannot be hit from a larger-than-normal number of at bats.   But it probably would not effect the calculations at .247 because .247 is probably no more difficult to land on than an average number.  
 
But I'm not SURE that is right. . . .would have to do simulation to help me think it through completely.  Thanks.  

 

Hey Bill,  
I was looking over fence distances from stadiums in the past and noticed that County Stadium in Milwaukee had relatively short fences during its National League days (1953-65). Just taking one season at random- 1959 - that park had the 2nd shortest LF pole (320 ft), the shortest Left-center (355), an average CF (402), a right-center (355) nearly the shortest (St Louis' was 354), and a 320 distance to the RF pole. The fences were not high, and Milwaukee's elevation is 617 feet. Any idea why it played like such a tough home run park?
Asked by: chuck

Answered: 6/16/2019
 Weather.   It played as a hitter's park in June and July, but as SUCH a pitcher's park in April, May and the second half of September that the balance was pitcher-friendly.   I studied it once in one of the Abstracts.  That's been a long time ago, 30-some years, and it would be easy to study it much better today; you might find something different.   But that was what I found at that time, that the weather was the dominant factor.  
 
Might also consider the altitude.  I assume that the park was near sea level, because it was near the lake, but I don't actually know.  

 

Hey Bill...what about evaluating pitchers just by the three true outcomes? Something like k/9 + bb/9 + hr/9? Maybe not exactly that formula but something along those lines. Wouldn't that be a much better indicator of a pitcher's ability than e.r.a or w-l? Or has that already been done?
Asked by: manhattanhi

Answered: 6/16/2019
 
I my view, we can't simply decide that everything not measured by the three true outcomes is not real and not significant.   Pitchers certainly have differing abilities to control the running game, different vulnerabilities to the running game.   Pitchers have different abilities to get a double play ball when they need a double play ball.  Pitchers may have. . . .certainly do have to some extent. . .different abilities to pitch to the situation, different abilities to give the hitter what the situation allows them to give the hitter.   They may react differently to defensive mistakes behind them.   They have different levels of game endurance, and probably have different levels of inning endurance. . .meaning some pitchers may be fine with the 20th to 25th pitches in an inning, whereas other pitchers may be in trouble in a long inning.   
 
The essence of the issue is whether we focus on the ELEMENTS of production, one thing at a time, or whether we focus on the SUM TOTAL of production.   Historically, old school, pre-sabermetric analysis, oddly, focused all evaluation of hitters on elements, and ignored the sum total, whereas evaluation of pitchers focused entirely on the outcomes--wins, losses, and runs allowed--and ignored the elements.   Historically, there were no counts of stolen bases allowed/caught stealing by pitchers, of double play support by pitchers, of RBI even against a pitcher, of doubles or triples allowed by a pitcher, whereas these things WERE counted for hitters, but hitters had no bottom-line summary such as runs created, OPS, or offensive WAR.  THe first challenge of sabermetrics was to balance the analysis by isolating the details for pitchers and creating a bottom line for hitters.
 
But to just focus on three elements and ignore everything else that makes a pitcher effective or ineffective--whether a hitter or a pitcher--would just, to me, seem like an obvious error.   You would be, in essence, deciding to pay no attention to 40%  of the game.   I just don't see how that is smart.   

 

Hey Bill.  This is about the comments by MarisFan61 re the old SI Runs Produced metric.  You commented that it's unfairly biased against Home Run hitters because HR are subtracted from the total.  
 
Without saying anything about the usefulness of the stat (Runs Created killed it off IMO), it's appropriate to subtract HR because what the metric was measuring was runs 'participated in' (for lack of a better description) by the batter.  Because a HR confers credit for both the run scored AND the run driven in, the HR is subtracted because only the one run scored as a result of the hit.  If the hitter instead hit a triple, and was then driven in, he would be credited with only the 1 run scored (ignoring possible RBI on the triple, of course).  
 
I see no bias against HR hitters given the aims of the metric.  Am I missing something in your opinion?
Asked by: myersb

Answered: 6/16/2019
 I did some research about this. . .meant to make it into an article but have not gotten that done yet.   

 

Runs Produced: I did a study some 10 years ago, similar in spirit to what you did in the 1983 Abstract, but on a larger scale.  And the results make it clear that the HR should be subtracted.  Not only does the empirical data support that, but so does the theoretical model.  The article is here: www.tangotiger.net/runsproduced.html  And given the abundance of data, I invite your readers to do something similar.
Asked by: tangotiger

Answered: 6/6/2019
 I'll try to look at it.  Thanks.  

 

Watching Mike Yastrzemski batting last night, I wonder if handedness is an inherited trait. Can you think of any examples of a MLB player who batted or threw the other way from his father or his grandfather? (Both Yazes bat lefty, which was the only thing that they have in common, far as I can tell.)
Asked by: Steven Goldleaf

Answered: 6/6/2019
 The Bondses would seem to be a pretty obvious example, if I understand your question.  

 

Since I know you like good names, did you know there is a player in the Nats organization named Joan Baez?    
 
I believe he's Dominican and it is a variation on Juan, but still ...
Asked by: pablo

Answered: 6/6/2019
 It's a shame Bill Singer will never get to pitch to him.  

 

Hey Bill,  
 
Diamondbacks outfileder got hit by the pitch for the 10 time this season.  In less tan 100 plate appearances.  
 
This would seem to be unusual.  Is it?  
 
Thanks  
 
Steve
Asked by: SteveN

Answered: 6/6/2019
 It would be very unusual, yes.  Who's the Cleveland guy, forget his name, who had the phenomenally high HBP rate a couple of years ago.  
 
Brandon Guyer.  Had to look it up.  Hit by pitches 31 times in 345 plate appearances in 2016.  

 

I was thinking about the historical increase in strikeouts and the apportioning of defensive Win Shares between pitching and fielding...  I wonder if maybe the apportioning is should be more sensitive to strikeouts than it currently is.  
 
For example, a team which is league average in all respects and strikes out 4.5 per 9 would have 67.5% of defensive WS go to pitchers, which would be 85 WS.  
 
If this average team instead strikes out 9.0 per 9 innings, the pitchers' share goes up to 69.2%, which works out to 87 WS.  
 
It seems to me the extra 4.5 * 162 = 729 outs (16.7% of total outs, roughly) ought to get pitchers more than just 2 Win Shares.  
 
Having said all this, I'm not sure what data one would look at to change the sensitivity to strikeouts.  Maybe Statcast can help?  
 
Anyway, I guess my question is... do you think that the sensitivity of pitcher WS to strikeouts should be larger than you initially laid it out to be?
Asked by: jgf704

Answered: 6/6/2019
 I can't tell you that it is absolutely right the way that it is.   It was a process that I set up to estimate something that we don't actually know.   If you can make a more accurate estimate of what it should be, I'd certainly listen.   

 

On manhattanhi's question:  
 
Hey Bill...any idea what the odds are of a player having the same batting average four consecutive years. Khris Davis hit .247 four straight years and this year he's hitting .248.  
 
Saw this in a different forum...  
 
One way to answer is to first figure out the probability that a "true" .247 hitter would actually hit .247 in the number of AB that Davis had in each season.  This is given by the binomial distribution:  
 
p(97 H | 392 AB) = 0.04664021  
p(137 H | 555 AB) = 0.03924708  
p(140 H | 566 AB) = 0.03883194  
p(142 H | 576 AB) = 0.03853069  
 
And since they're independent, we can multiply the four probabilities to get an overall probability.  
 
0.04664021 x 0.03924708 x 0.03883194 x 0.03853069 = 0.000002738822  
 
2.74 in a million  
1 in 365,121  
 
Asked by: jgf704

Answered: 6/6/2019
 Thanks.  

 

Could you re-state your position on power hitters batting leadoff or second in the order? On the one hand, they get more at-bats but there aren't as many runners on. Is it a good trade off?
Asked by: manhattanhi

Answered: 6/6/2019
 I don't believe I have ever taken a position on that.  

 

Hey Bill,  
 
Was wondering if you knew of any source where historical active rosters are stored by day/game?  
 
Thanks in Advance,  
Eoin.
Asked by: emclaverty

Answered: 6/6/2019
 No, I don't.  

 

Hey Bill...any idea what the odds are of a player having the same batting average four consecutive years. Khris Davis hit .247 four straight years and this year he's hitting .248.
Asked by: manhattanhi

Answered: 6/4/2019
 I think it has never happened before, if that's a clue.   I guess you could figure it.  You'd have to make some assumptions.  

 

 
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy