Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Primate Studies > Discussion
Primate Studies
— Where BTF's Members Investigate the Grand Old Game

Tuesday, October 08, 2002

Baseball Primer’s 2002 NL Cy Young Award

Our picks.

The Arizona Diamondbacks’ Randy Johnson pulled away from teammate Curt Schilling with a spectacular September to win his 2nd consecutive Baseball Primer Cy Young Award.

Johnson was fantastic down the stretch as Arizona wrapped up their second consecutive NL West crown. In the season’s final month the 6’ 10” lefty was 5-0, with an 0.66 ERA, allowing just 3 earned runs in 41 IP.

Johnson finished the year 24-5, with a 2.32 ERA, 334 strikeouts and 30 quality starts in 35 tries. He led the league in ERA, innings (260), strikeouts, quality starts, opponents batting average (.208) and complete games (8). He was 2nd to Florida’s A.J. Burnett (12th in the voting) with 4 shutouts and his 35 starts were second only to Tom Glavine’s 36. Johnson accomplished all of this despite hurling 21 of his 35 starts at home in a hitter’s park and playing in the toughest division in the NL.

The choice for Johnson was nearly unanimous, as he received 15 of 16 first place votes, teammate Curt Schilling finished 2nd and received the other first place nod.

If the 32 BBWAA voters agree with our staff, Johnson should emerge with his 4th consecutive NL Cy Young Award next month, and it will be the 5th of his career. He was obviously the best pitcher in baseball for the 2nd year in a row. He’s under contract for next year ($15 million if he wins the BBWAA Cy Young), he’ll be free agent in 2004.

Here’s a look at a few of the other top contenders.

2nd Curt Schilling, Arizona 141 points

Schilling was the frontrunner most of the season, due to his outstanding won/loss record (18-3 entering August) and incredible strikeout/walk ratio (212:18 entering August). Over the last two months of the season though, Schilling’s ERA was just 3.91 and he was only 5-4. Part of this was due to a shellacking in Colorado on September 20, where he gave up 9 runs (8 earned). Then, with home field advantage for the first round on the line, he was bombed again by St. Louis on September 25, giving up all 6 runs in a 6-1 loss.

All in all though, it was an outstanding year for Schilling, who finished 23-7 with a 3.23 ERA. He fanned 316 and walked just 33 in 259 1/3 IP (he and Johnson were the only NL pitchers over 233). He tossed 27 quality starts, second only to Johnson and was clearly the 2nd best pitcher in the National League again this season. Schilling is under contract to the Diamondbacks through the 2004 season ($10 million 2003, $12 million 2004).

3rd Roy Oswalt, Houston 113 points

The sky appears to be the limit for this recently-turned (August 29) 25-year old righthander. Oswalt built on his outstanding rookie season and proved that he could shoulder the load of an ace on a contending ballclub. He was third in the NL with 233 IP and tied for 4th with 34 starts. He posted the 5th best ERA in the league (3.01), despite pitching in the 2nd best hitter’s park in the league. Oswalt was also third with 19 wins (9 losses) and 5th with 208 strikeouts. He only allowed 17 HR this year as well and his 24 quality starts were tied for 4th in the NL.

Oswalt doesn’t need to improve to have a Hall of Fame career. He just needs to stay healthy and pitch like this for the next 10-11 years. He’s the best young pitcher in the National League and is now 33-12 in his career, with a 2.91 ERA. The sky is the limit for this kid and he won’t even be eligible for free agency until after the 2007 season.

4th Greg Maddux, Atlanta 105 points

Another year, another outstanding season, Maddux keeps turning these in like clockwork. This year’s version finished 2nd in the NL in ERA (2.62), with a 16-6 record in 34 starts. However, he only tossed 199 1/3 innings (5.86/start), so despite his great ERA he was dropped a few rungs in the voting. Some of his other peripherals were off a little as well. He dipped from 6.7 to 5.3 strikeouts per 9 IP. His BB/9 almost doubled from 1.04 to 2.03. He is 36 years old, so those numbers can be expected to start slipping.

Don’t get me wrong, he’s still a very good pitcher. Some wonder if he can go deep into games any more (just 3 starts of 100 pitches and no complete games this year). I think he can still go 220 IP a year. He was hurt a lot this year and Bobby Cox had his best bullpen to date, so it made sense to keep Maddux on a short leash.

The durability issue is a big one, as he’s likely to be leaving Atlanta this winter. It will be interesting to see where he ends up and how well he does next year. I think he’s still got a lot of good innings left, and he should win his 300th game in August of 2004.

5th Tom Glavine, Atlanta 87 points

Glavine jumped off to a quick start in 2002 (11-4, 2.27 ERA at the all-star break), and turned in his best season since his Cy Young campaign of 1998. He finished the year 18-11 with a 2.96 ERA (3rd in the NL), tossing 224 2/3 innings and leading the NL with 36 starts.

He’s within striking distance of 300 wins, he’s now 242-143 for his career and he’ll be 37 next year. I’d say it’s 50/50, as he’ll have to be effective (14-15 wins a year) until he’s 40 to have a shot. It’ll be interesting to see what the Braves decide to do about him this winter. The early buzz is that Glavine will stay and Maddux will go.

6th Odalis Perez, Los Angeles 57 points

Perez had a breakthrough season for the Dodgers this year, finishing 15-10 with a 3.00 ERA (4th in the NL). His numbers were propped up a little by Dodger Stadium, but it was a still a great year for the young lefty, who turned 25 in June. He only walked 38 batters all year, and tossed 222 1/3 innings. He also whiffed 155. Maybe that Sheffield trade wasn’t so bad after all? Perez won’t be a free agent until after the 2006 season.

7th Eric Gagne, Los Angeles 56 points

Gagne burst onto the national scene this year after showing signs of brilliance the last two seasons (209 K in 253 IP). He finally put it all together after a move to the bullpen. Gagne was the best closer in the NL this season, fanning 114 and walking just 16 in 82 1/3 innings. He had a 1.97 ERA and 52 saves in 56 opportunities. The 26-year old righty also allowed only 6 HR and held opponents to a .175 average. Gagne won’t be a free agent until after the 2006 season.

Honorable mention Bartolo Colon Cleveland/Montreal (11th NL, 12th AL)

Colon had his best season in 2002, but split it between two leagues, thus he didn’t do well on either ballot. In 33 starts he pitched 233 1/3 innings, finishing 20-8 (for two non-contenders), with an ERA of 2.93. His strikeout rate was down, as he only whiffed 149 batters. However, he compensated by only walking 70 batters, his best control season in the bigs. Colon has always been a pretty good, hard-throwing starter that was prone to inconsistency.

Now, it appears he’s turned the corner. He was one of the best pitchers in baseball this year, in the small pack just behind Johnson and Schilling. If he had played the whole season for the Expos, I’d have voted him 4th, I voted Johnson-Schilling-Oswalt the top in the NL, just like my brethren.

The Expos have a $6 million option on him next year, he’ll be a free agent next winter.

It’s interesting to note that John Smoltz, who some have mentioned as an MVP candidate finished 14th in our Cy Young balloting. Our 16-man panel gave him 1 eighth, 1 ninth and 1 tenth place vote. He was better than his 3.25 ERA would suggest. He allowed 8 runs in 2/3 of an inning April 6 against the Mets (score was 2-2 when he entered). What if he just been terrible that day, instead God-awful?

Let’s say after he allowed 3 runs Cox would have lifted him, which sounds reasonable. He would have a 2.69 ERA and the team’s record wouldn’t be any different. He only blew 4 saves all year. Overall, he had a pretty good year (Gagne’s road ERA was 2.63), better than most here would suggest. Because we don’t want him to be the MVP (and rightfully so), we knocked him a little more than he deserved this year.

Here’s the rest of the voting:

Player           1   2 3 4 5 6 7 8 9 10 Blts Total
Randy Johnson     15   1 0 0 0 0 0 0 0   0   16   207
Curt Schilling     1 12 2 0 0 0 0 1 0   0   16   141
Roy Oswalt       0   1 8 4 1 0 0 2 0   0   16   113
Greg Maddux       0   1 4 6 1 2 1 0 1   0   16   105
Tom Glavine       0   0 0 4 4 7 0 0 0   0   15   87
Odalis P?rez       0   0 0 0 6 1 2 0 3   2   14   57
Eric Gagne       0   1 1 1 0 4 1 1 2   1   12   56
Kevin Millwood     0   0 0 1 0 1 6 2 1   0   11   44
Matt Morris       0   0 0 0 2 1 2 2 2   0   9   35
Randy Wolf       0   0 0 0 0 0 1 3 1   2   7   23
Bartolo Col?n     0   0 1 0 1 0 0 0 0   2   4   16
A.J. Burnett       0   0 0 0 0 0 1 2 0   2   5   12
Jason Schmidt     0   0 0 0 0 0 1 1 0   1   3     8
John Smoltz       0   0 0 0 0 0 0 1 1   1   3     6
Octavio Dotel     0   0 0 0 0 0 1 0 1   0   2     6
Elmer Dessens     0   0 0 0 1 0 0 0 0   0   1     6
Chris Hammond     0   0 0 0 0 0 0 0 1   2   3     4
Byung-Hyun Kim     0   0 0 0 0 0 0 1 0   0   1     3
Matt Clement       0   0 0 0 0 0 0 0 1   0   1     2
Wade Miller       0   0 0 0 0 0 0 0 1   0   1     2
Tomo Ohka         0   0 0 0 0 0 0 0 1   0   1     2
Darren Holmes     0   0 0 0 0 0 0 0 0   0   1     1
Brian Lawrence     0   0 0 0 0 0 0 0 0   0   1     1
Vincente Padilla   0   0 0 0 0 0 0 0 0   0   1     1

Joey Numbaz (Scruff) Posted: October 08, 2002 at 06:00 AM | 17 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Mike Emeigh Posted: October 09, 2002 at 12:53 AM (#606631)
Dessens finished 6th in the NL in ERA among qualifiers (which is likely why he got a vote), and was 7-8 largely because of a lack of run and bullpen support. I haven't checked his DIPS, but given his low BB and K totals I'd guess he was extremely hit-lucky this year, as he has to be in order to survive with his stuff.

-- MWE
   2. User unknown in local recipient table (Craig B) Posted: October 09, 2002 at 12:53 AM (#606632)
There has been a great deal of discussion on how to account for Colon. I have, after many years, decided on how to settle (for myself) the issue of players who switch leagues. I simply use their performance in both leagues to decide, but only vote for them in one league (the league they ended up in, excpt where that produces absurd results). I would do this for RoY, Cy, and MVP awards.

My rationale is simple (probably too simple for most people)... that NL Cy (for example) goes (in a very general way) to the best pitcher in the National League. Colon, at the time of voting, is in the National League even though some of his performance is in the AL. It's all major league performance, in my view, so I just treat Colon as one player isntead of two. As I said, this is likely too simplistic a view for most people.

I wouldn't do this where a pitcher goes 20-10 in the AL and then is involved in a waiver deal in mid-September finishing 1-1 in the NL. There, I'd probably identify his performance enough with the AL to vote him there... but I would use the NL performance.

In the end, taking Cleveland and Montreal performance together meant I had Colon fourth (my IBA ballot was Johnson-Schilling-Oswalt-Colon-Perez with great apologies to Maddux)
   3. MattB Posted: October 09, 2002 at 12:53 AM (#606633)
I'm surprised Octavio Dotel didn't appear on more than two ballots.

Perhaps in the Gagne/Smoltz debate, people lost track of the fact that Dotel was better than both of them? His ERA was lower than Gagne's in almost 20% more innings, and more (Holds+Saves) than anyone who was not their team's designated closer. He had more strikeouts than any other non-starter, his walk rate was a tad higher, but his OPS against was 511, lower than both Gagne and Smoltz.
   4. SM in DC Posted: October 09, 2002 at 12:54 AM (#606638)
Just trying to help....</B>

Am I clutch or what?
   5. Rob Wood Posted: October 09, 2002 at 12:54 AM (#606639)
Very nice article Joe. It looks like there is a slight glitch in the vote tally table. The last three guys (Holmes, Lawrence, Padilla) should each be credited with a 10th place vote.
   6. Walt Davis Posted: October 09, 2002 at 12:54 AM (#606640)
OBN=on-base percentage on balls not-in-play
SLN=slugging on balls not-in-play
OPN=OPS not-in-play
OBI=OBP in-play
SLI=SLG in-play
OPI=OPS in-play

Pitcher 1
224.2 ip, 936 bf, 210 h, 21 hr, 78 bb, 127 k
HR rate= .093/ip, K rate= 5.1/9, BB rate = 3.1/9
OBN=438, SLN=568, OPN=1006
OBI=266, SLI=341, OPI=607

Pitcher 2
178 ip, 737 bf, 173 h, 24 hr, 49 bb, 93 k
HR rate= .139, K rate= 4.7/9, BB rate= 2.5/9
OBN=440, SLN=821, OPN=1261
OBI=261, SLI=322, OPI=583

There's no particularly big difference between these pitchers other than IP. #1 is Glavine, #2 is Dessens.

But to return to a theme I introduced on the SF/Atl Game 2 thread when discussing Rueter, it's time we let go of the notion that pitchers letting balls in play is bad. DIPS does not say that keeping the ball out of play is good ... it says that pitchers have little/no control over what happens to BIP, therefore what distinguishes pitchers is what happens on balls out of play. That doesn't mean that keeping balls in play is bad, it means that striking people out is good, walking people and letting them hit HR are bad.

Generally speaking, only the true greats have OPN numbers lower than OPI numbers. For most pitchers, it seems they'd actually be better off if they kept more balls in play (assuming their distribution of BnIP among K, HR, and BB remained the same).

To wit:

ATL average on BIP:
Total BF: 6,137
BIP: 4,360
OBI: 270
SLI: 340
OPI: 610

CIN average on BIP:
Total BF: 6,304
BIP: 4,545
OBI: 292
SLI: 372
OPI: 664

Would anyone here be unhappy if their pitchers put up a 664 OPS against? (and yes, Dessens was a bit hit-lucky this year)

By the way, to have an OBN of 290, you'd have to have a K/BB ratio better than 2.5 (how much better depends on how many HR you give up). Since pitchers with K/BB ratios of 2.5 or higher are rare, strictly from an OBP angle, you'd rather your pitchers put the ball in play. To maintain an SLN of 370, you have to K about 10 per HR allowed, which means you'd better strike out 200+ or be Derek Lowe. So from a SLG angle, you'd rather your pitchers put the ball in play.

Again, DIPS' lesson is not "balls out-of-play good, balls-in-play bad." DIPS' lesson is "strikeouts great, balls-in-play good, walks bad, HR very bad." Successful low-K pitchers like Glavine, Moyer, Tommy John, etc. are not DIPS anomalies.

To be a good pitcher, you can either K a lot (Ryan, Wood) or keep your walks & HR rates down (and thereby give up lots of balls in play). To be a great pitcher, do all three. The trick, obviously, for the Glavine types is giving up BIP without giving up a substantial number of HR.

Of course, OPS isn't the be-all and end-all. I'd be interested if someone wants to run the numbers through XR or LWTS or RC or what have you. But given that, on BIP, your average hitter against the Reds was Mark Grudzielanek (301/364/665), I'm willing to bet that an all BIP pitcher would look pretty good under any measurement.
   7. Dan 'The Boy' Werr Posted: October 09, 2002 at 12:54 AM (#606645)
Nice post, Walt. It's a good reminder that a guy like Derek Lowe can be good from a DIPS standpoint, too.

Gagne had better true outcome numbers than Dotel this year. Though you could argue that Dotel's 15 extra IP make up for that. I think you could easily make a case either way. Personally, I had Gagne 9th and originally had Dotel 10th, but I dropped him for Schmidt. A couple guys who probably deserved more recognition than they got were Nen and Isringhausen (no HR allowed in over 65 IP!).
   8. Mike Emeigh Posted: October 09, 2002 at 12:54 AM (#606646)
Again, DIPS' lesson is not "balls out-of-play good, balls-in-play bad." DIPS' lesson is "strikeouts great, balls-in-play good, walks bad, HR very bad."

This is better stated as "strikeouts great, fly balls and ground good, walks bad, HR and line drives very bad."

Walt is not entirely correct that allowing more balls to be put into play is a good thing. As a general rule, the more balls that are put into play, the more balls that are put into play that are hit *hard* - and the harder that a ball is hit, the less likely it is that the ball will be converted into an out. It's only a good idea to allow more balls in play if you can be sure that the extra balls aren't going to be hit very hard.

BA/SLG on BIP, 1998-2000 (from my licensed PBP database):

Ground balls: .231/.253
Fly balls: .114/.194
Line drives: .730/.942

I should add that because fly balls are more often converted into outs than are ground balls, fly ball pitchers do have a DIPS advantage over ground ball pitchers in general; if you take a group of extreme fly ballers (the Millwoods and Miltons) and compare them to a group of extreme ground ballers (the Madduxes and Lowes), the fly ballers as a group will have a lower BA/BIP, although the SLG will be fairly close because of the EBH off the fly ballers (the vast majority of GBH are singles).

-- MWE
   9. Walt Davis Posted: October 10, 2002 at 12:54 AM (#606653)
Walt is not entirely correct that allowing more balls to be put into play is a good thing. As a general rule, the more balls that are put into play, the more balls that are put into play that are hit *hard*

But so what? That the more BIP, the more that are hit hard is a truism. DIPS finding is that pitchers don't have control (or very little) over the results of BIP. This would seem to require that regardless of how many BIP there are, the _proportion_ that are hit hard is the same. Therefore the expected OBP, SLG, and OPS on BIP would be the same.

Or are you saying that the more BIP, the higher the proportion of hard hit BIP? If indeed more BIP leads to a larger proportion of hard hit BIP, then pitchers do affect the results of BIP. Which reopens the DIPS can of worms.

And even if so, just how bad are the results of those additional BIP? Even very good pitchers have OPNs against in the range of 800-1000, and it would be hard to do worse than that on BIP. If your average Cinci pitcher is at 292/372/664 on BIP, they'd have to be hit hard enough to get up to somewhere around 340/460/800 for it to stop paying off.

One argument I definitely am open to is the we really need to move HR's over to the "other side" of the equation. Taking HR's out of BIP makes sense when we're trying to assess the quality of a pitcher independent of their defense. But when we're trying to assess whether pitchers should concentrate their efforts on batters not making contact vs. hitting the ball, and especially if we're moving into how hard the ball is hit, it makes intuitive sense to me to include HR's in that part of the discussion. But I'm also not sure it makes a lot of difference since I think all it would tell us is what we already know -- pitchers with high BIP rates and that give up lots of HR's are in trouble...though whether they're in more trouble than low BIP pitchers who give up lots of HR's is up in the air.

Now, if there's a positive correlation between BIP rates and HR rates, we might be onto something.
   10. Mike Emeigh Posted: October 10, 2002 at 12:54 AM (#606661)
Or are you saying that the more BIP, the higher the proportion of hard hit BIP?

Yes, that is generally true, from what I can tell.

If indeed more BIP leads to a larger proportion of hard hit BIP, then pitchers do affect the results of BIP.

I don't think it's unreasonable to suggest that pitchers do have an effect, but that the magnitude of that effect is smaller than the effect that either the hitters or the fielders have. It's not true that there is *no* year-to-year correlation in BA/BIP for pitchers; the correlation is not statistically significant, but it is weakly positive. Furthermore, as I noted earlier, fly ball pitchers in general tend to have lower BA/BIP than do ground ball pitchers. And as Voros has noted, knuckleballers tend to have lower BA/BIP than do other pitchers. Those conclusions aren't consistent with a finding of *no pitcher effect*.

The hitters are an important, often overlooked factor in this discussion. The hitter is primarily responsible in most cases for the location of the BIP, and in my opinion is also primarily responsible for the force with which the ball is hit. If that is the case, the pitcher has to either (a) prevent the hitter from making contact in the first place, or (b) have something else in his arsenal which reduces the likelihood that the hitter can make *solid* contact (such as a knuckleball) - because when the hitter belts a line drive, he's posting an OPS of 1.674.

-- MWE
   11. Mom makes botox doctors furious Posted: October 10, 2002 at 12:54 AM (#606665)
Hey Joe

(kinda sounds like some old song i've heard before.. hmm..)

On Greg Maddux, you wrote: "He is 36 years old, so those numbers can be expected to start slipping."

And the big question is: why? What backs up this conclusion? Are there other pitchers sharing a similar history who then wound down? If Maddux can, as you suggest (and I for one certainly hope he can) pitch those 220 innings, is there some evidence to suggest his expected performance will be less than has been the last few years?

While he pitched fewer innings this year, his numbers suggest that he pitched better than any season since 1998. Why a decline now?

Thanks..



   12. tangotiger Posted: October 10, 2002 at 12:54 AM (#606669)
If indeed more BIP leads to a larger proportion of hard hit BIP, then pitchers do affect the results of BIP.

Sounds fishy to me. If you take all games pitched by Randy, Curt, Greg, and the other star pitchers, and break them down into "games with lotsa BIP", and "games with normal BIP", and "game with little BIP", I would be surprised if the % of hard hit balls is different in each of these groups.

I suspect that *bad pitchers* probably give up lotsa BIP, and by virtue of being bad, also give up lotsa hard hit balls. I don't see the causal relationship between BIP and % hard hit balls. The causal agent is the level of pitcher.

Mike, can you elaborate on your findings, and perhaps run a controlled study where you have the same population of pitchers (and PA) in all groups?
   13. Walt Davis Posted: October 10, 2002 at 12:54 AM (#606670)
ME: If indeed more BIP leads to a larger proportion of hard hit BIP, then pitchers do affect the results of BIP.

I don't think it's unreasonable to suggest that pitchers do have an effect, but that the magnitude of that effect is smaller than the effect that either the hitters or the fielders have. It's not true that there is *no* year-to-year correlation in BA/BIP for pitchers; the correlation is not statistically significant, but it is weakly positive. Furthermore, as I noted earlier, fly ball pitchers in general tend to have lower BA/BIP than do ground ball pitchers. And as Voros has noted, knuckleballers tend to have lower BA/BIP than do other pitchers. Those conclusions aren't consistent with a finding of *no pitcher effect*.

Yes, some small effect seems to be the case. But is that small effect large enough to overcome the 150ish point difference between the OPS on BIP and the OPS on BnIP even for very good pitchers. If you strike out 200, walk 80 (a 2.5 K/bb ratio) and give up 20 HR's (about average), you've got an OBN of 333, an SLN of 364, and an OPN of 697. (For a real-life example, Barry Zito's OPN was 845 this year) Even that is a smidgen higher than the 664 OPI of Cinci pitchers. And remember, we're talking about Hamilton, Dessens, Moehler, Rijo, etc. -- if bad BIP pitchers get hit harder, this ought to be one of the teams to show it.

Let's take Detroit, who struck out fewer batters than any other team in the league (by a wide margin) and, by all accounts, is an awful defensive team. If any team is getting killed by hard hit balls and poor defense on non-hard hit balls, it should be the Tigers. Their OBI was 302, SLI was 400, giving them an OPI of 702. I'd wager there weren't more than about 10 ML starters with an OPN below 702. More importantly, the outcome of the average Detroit BIP was no worse than the outcome of a BnIP for the 200k, 80bb, 20hr pitcher above.

On BnIP, Detroit had an OBN of 464, SLN 681, and OPN of 1145.

The hitters are an important, often overlooked factor in this discussion. The hitter is primarily responsible in most cases for the location of the BIP, and in my opinion is also primarily responsible for the force with which the ball is hit. If that is the case, the pitcher has to either (a) prevent the hitter from making contact in the first place, or (b) have something else in his arsenal which reduces the likelihood that the hitter can make *solid* contact (such as a knuckleball) - because when the hitter belts a line drive, he's posting an OPS of 1.674.

I agree the hitter is the forgotten man here, and that's why I think we should consider moving HR to the other side of the equation. But judging by Detroit's BIP #'s, apparently even their pitchers have come up with something in their arsenal that pretty well negates that 1674 OPS on line drives.

Now before I commit something I often chastise others for, let me point out that Detroit's decent overall numbers on BIP does not mean that there aren't individual pitchers on Detroit getting hammered on BIP. But it would mean that for each of their pitchers getting hammered, there's somebody else doing pretty good to balance them out.

Let me also ask about Comerica's park factors. I know it decreases HR and scoring overall, but I'd imagine it's a pretty good doubles and triples park and its size might increase singles too. So I would guess, if anything, that Comerica makes Detroit's BIP numbers look worse than they'd be in a neutral park. But why guess when somebody out there must have the answer.

I don't mean to pick on Mike nor am I trying to tear down DIPS. I'm just pointing out that it's wrong to think that DIPS leads to the conclusion that BIP are bad. On average, BIP outcomes are very favorable to the defense.

One thing we can definitely say with DIPS is the following:

1) if two pitchers face the same number of batters
2) and give up the same # of HR's and walks,
then
3) you want the pitcher who strikes out more batters.

Along with obvious corralaries like "if they K the same and give up the same # of HR, you want the one with fewer walks." Strikeouts are better than BIPs; walks and HR's are worse. That 1674 OPS on line drives looks mighty scary, but not as bad as the 5000 OPS on HRs or the 1000 OBP, undefined SLG on walks. :-)

So the part that's not so easy is when it's a choice between the guy who K's a lot, but also gives up lots of walks and/or a good number of HR vs. a guy who K's and walks fewer and/or gives up fewer HR. As I tried to demonstrate above, a Detroit pitcher who magically kept every ball in play would post an OPS against of about 702 (on average); a Detroit pitcher who had 200 k, 80 bb, and 20 hr would have an OPS against right around 700 (on average). There wouldn't seem to be a lot of difference between those two, but I know Detroit will have an easier time finding guys who keep the ball in play than ones that K 200 with 2.5 K/bb ratios.

And as noted earlier, I'm perfectly open to supported arguments that OPS gives a distorted picture here. Since the outcomes of BnIP are so extreme (K vs. bb/hr), perhaps OPS is the wrong way to go.

But until it's explained away, I continue to offer this as the Tewksbury paradox. Pitchers like John, Glavine, Moyer, etc. are not DIPS anomalies, they are successful because they limit BBs and HRs and let their defenses do the work for them...even if they also draw some small advantages from being GB or FB pitchers or, as shown in a recent primer piece on Glavine, exhibit some control over the direction of the BIP.

BIP are your friend. Just remember to duck when they rip one back through the box.
   14. tangotiger Posted: October 11, 2002 at 12:54 AM (#606680)
I don't know why we keep talking about Glavine as if he's bucking the trend.

Career $H, pitchers born since 1940, min 8000 BIP

Other than the obvious Charlie Hough being a knuckler, I don't see the "type" of pitcher having much impact here.
   15. Walt Davis Posted: October 11, 2002 at 12:54 AM (#606692)
Well, Glavine's not out of whack in terms of $H and I'm not sure anyone suggested he was. What I did mention was an article, I'm pretty sure posted here, that showed that Glavine did have control over the direction the ball was hit. This control _might_ explain his slight advantage in $H.

But I bring up Glavine because he's a very successful low K pitcher. Like Tommy John, who not only kept the ball in play a lot, but who had a higher $H than his teams. Why? Because putting the ball in play is actually a pretty good thing. Just look at those $H numbers for the pitchers or the teams. They're around 270-280. 72-73% of the time a BIP results in an out -- that's a pretty nice outcome -- and the majority of the hits will be singles.
   16. Mike Emeigh Posted: October 12, 2002 at 12:55 AM (#606707)
Well, Glavine's not out of whack in terms of $H and I'm not sure anyone suggested he was. What I did mention was an article, I'm pretty sure posted here, that showed that Glavine did have control over the direction the ball was hit. This control _might_ explain his slight advantage in $H.

The referenced article was one that I wrote back in August. The data shows that hitters take the ball the other way against Glavine significantly more often than they do against other pitchers, which suggests that Glavine pitches in such a way so that he might be able to influence the direction in which the ball is hit. There are very few differences of this type among pitchers; most pitchers do not differ significantly from league norms in terms of the location of balls in play.

But I bring up Glavine because he's a very successful low K pitcher. Like Tommy John, who not only kept the ball in play a lot, but who had a higher $H than his teams. Why? Because putting the ball in play is actually a pretty good thing.

I looked at my Gillette/Palmer play-by-play data for 2000, and identified groups of pitchers who (a) faced at least 300 hitters in that season and (b) had a ball-in-play percentage of one standard deviation or more above or below the league average percentage of balls in play. In the AL, the high-BIP group included pitchers who allowed BIP in 76% of more of the plate appearances against them, the low-BIP group included pitchers who allowed BIP in 66% or fewer. In the NL, the cutoff percentages were 75% or higher and 63% or fewer respectively. There were a total of 48 high-BIP pitchers across the two leagues, 33 low-BIP pitchers. I didn't combine performances across two teams, which led to Curt Schilling's 2000 performance numbers with Arizona being included in the high-BIP group, largely because Schilling fanned just 72 batters in 97 2/3 innings with the Snakes that year. I should also add that, since my database identifies IPHR, I included them as hits on balls in play; there were seven of them in 2000.

Anyway, here are the resulting performance numbers:

AL league BA/SLG on BIP: .300/.381
AL high-BIP BA/SLG on BIP: .305/.393
AL low-BIP BA/SLG on BIP: .286/.360

AL league OBP/SLG on BNIP: .456/.643
AL high-BIP OBP/SLG on NBIP: .458/.720
AL low-BIP OBP/SLG on NBIP: .379/.439

NL league BA/SLG on BIP: .294/.375
NL high-BIP BA/SLG on BIP: .293/.372
NL low-BIP BA/SLG on BIP: .285/.366

NL league OBP/SLG on NBIP: .442/.588
NL high-BIP OBP/SLG on BIP: .483/.821
NL low-BIP OBP/SLG on NBIP: .386/.365

The AL high-BIP pitchers posted an ERA of 4.66 against the league average of 4.92, while the low-BIP pitchers posted an ERA of 3.97. The NL high-BIP pitchers posted an ERA of 4.75 against the league average of 4.64, while the low-BIP pitchers posted an ERA of 3.71.

What's striking about this to me is that the low-BIP pitchers, as a group, were better than the league norm on the results from BIP, as well as significantly better on the results from NBIP. The high-BIP pitchers, on the other hand, didn't differ from the league norm much in terms of results from BIP, but allowed more HR as a percentage of NBIP with the result that their OPS on NBIP were quite a bit higher.

Looking at all of the pitchers for that year who faced 300 hitters ro more, I also noted that (just like with DIPS) the correlation between the % of BIP allowed to hitters and the OPS on BIP, while not significant, is still positive (r=+0.224 in the AL and +0.166 in the NL), as is the correlation between % BIP and overall OPS (r=+0.214 in the AL, +0.344 in the NL). That, also, is not consistent with a theory that a pitcher benefits by allowing more balls in play - in fact it suggests the opposite conclusion.

I go back to my earlier point. Pitchers who are successful when allowing large numbers of balls in play are (a) rare, and (b) able to do so because there is some other aspect of their game which allows them to avoid allowing a large number of hard-hit balls.

-- MWE
   17. tangotiger Posted: October 13, 2002 at 12:55 AM (#606714)
Mike, great stuff!

If possible can you continue your study based on my earlier suggestion
If you take all games pitched by Randy, Curt, Greg, and the other star pitchers, and break them down into "games with lotsa BIP", and "games with normal BIP", and "game with little BIP", I would be surprised if the % of hard hit balls is different in each of these groups.

I suspect that *bad pitchers* probably give up lotsa BIP, and by virtue of being bad, also give up lotsa hard hit balls. I don't see the causal relationship between BIP and % hard hit balls. The causal agent is the level of pitcher.

Mike, can you elaborate on your findings, and perhaps run a controlled study where you have the same population of pitchers (and PA) in all groups?


In your particular example them, let's only look at the low-BIP pitchers you've already identified, and break up their individual games into groups of "lotsa BIP", "avg BIP", "low BIP". Having the same pitchers in the groups will help in identifying the effect.

Thanks again..

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
1k5v3L
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.4241 seconds
66 querie(s) executed