WAR Leader and Cy Young Winner in the Same Season
Posted by Neil Paine on November 19, 2010
As Andy noted yesterday, King Felix won the AL Cy Young after also leading the AL in Wins Above Replacement. I was wondering how often pitchers have pulled off this double feat of winning the CYA and the WAR crown in the same season, so I called up our year-by-year WAR leaders and cross-referenced that list with the Cy Young winners file at Baseball-Databank.org. Here were the results:
| Year | Lg | Pitcher | Tm | WAR | Year | Lg | Pitcher | Tm | WAR | |
|---|---|---|---|---|---|---|---|---|---|---|
| 2010 | AL | Felix Hernandez | SEA | 6.0 | 1992 | NL | Greg Maddux | CHC | 8.4 | |
| 2009 | NL | Tim Lincecum | SFG | 6.3 | 1991 | NL | Tom Glavine | ATL | 7.4 | |
| 2009 | AL | Zack Greinke | KCR | 9.0 | 1991 | AL | Roger Clemens | BOS | 7.5 | |
| 2008 | NL | Tim Lincecum | SFG | 6.9 | 1989 | AL | Bret Saberhagen | KCR | 8.6 | |
| 2008 | AL | Cliff Lee | CLE | 7.3 | 1988 | NL | Orel Hershiser | LAD | 7.3 | |
| 2007 | NL | Jake Peavy | SDP | 6.2 | 1987 | AL | Roger Clemens | BOS | 8.4 | |
| 2007 | AL | CC Sabathia | CLE | 6.8 | 1986 | NL | Mike Scott | HOU | 7.8 | |
| 2006 | NL | Brandon Webb | ARI | 5.8 | 1985 | NL | Dwight Gooden | NYM | 11.7 | |
| 2006 | AL | Johan Santana | MIN | 7.0 | 1985 | AL | Bret Saberhagen | KCR | 6.7 | |
| 2004 | AL | Johan Santana | MIN | 7.4 | 1983 | NL | John Denny | PHI | 7.1 | |
| 2003 | AL | Roy Halladay | TOR | 7.5 | 1981 | NL | Fernando Valenzuela | LAD | 5.4 | |
| 2002 | NL | Randy Johnson | ARI | 8.8 | 1980 | NL | Steve Carlton | PHI | 9.4 | |
| 2001 | NL | Randy Johnson | ARI | 8.4 | 1978 | AL | Ron Guidry | NYY | 8.5 | |
| 2000 | NL | Randy Johnson | ARI | 7.6 | 1975 | NL | Tom Seaver | NYM | 7.7 | |
| 2000 | AL | Pedro Martinez | BOS | 10.1 | 1973 | NL | Tom Seaver | NYM | 9.5 | |
| 1999 | NL | Randy Johnson | ARI | 8.2 | 1972 | NL | Steve Carlton | PHI | 12.2 | |
| 1999 | AL | Pedro Martinez | BOS | 8.4 | 1972 | AL | Gaylord Perry | CLE | 10.5 | |
| 1998 | AL | Roger Clemens | TOR | 7.5 | 1971 | NL | Fergie Jenkins | CHC | 9.2 | |
| 1997 | NL | Pedro Martinez | MON | 8.2 | 1970 | NL | Bob Gibson | STL | 8.7 | |
| 1997 | AL | Roger Clemens | TOR | 10.3 | 1969 | AL | Denny McLain | DET | 7.5 | |
| 1996 | AL | Pat Hentgen | TOR | 8.4 | 1968 | NL | Bob Gibson | STL | 11.9 | |
| 1995 | NL | Greg Maddux | ATL | 8.8 | 1966 | NL | Sandy Koufax | LAD | 10.8 | |
| 1995 | AL | Randy Johnson | SEA | 7.8 | 1964 | AL | Dean Chance | LAA | 8.9 | |
| 1994 | NL | Greg Maddux | ATL | 7.8 | 1963 | NL | Sandy Koufax | LAD | 10.8 | |
| 1994 | AL | David Cone | KCR | 6.1 |
Of the 100 Cy Young Awards that have been awarded, 49 went to the pitcher who also led the league in WAR, so it's not really that rare of an accomplishment. Still, being one of the 31 players to pull off the feat puts you in some pretty good company:
November 19th, 2010 at 8:20 am
What's interesting in looking at the chart is that the voters have definitely gotten better over the years. In the first 28 years, there was only one year (1972) in which both winners led the league in WAR. Starting in 1991, it's happened quite frequently (10 times).
November 19th, 2010 at 8:59 am
Is anyone else having trouble going to Bob Gibson's page?
November 19th, 2010 at 9:06 am
You're right, Brett. That Bob Gibson link must be broken. I tried a search for "the Bob Gibson" and it came up right away, though, if that helps.
November 19th, 2010 at 9:22 am
Thanks for the tip Dr.!
November 19th, 2010 at 9:37 am
Looking only at starting pitchers, the WAR rate rises to 54% (49 of 91 winners). Of course, the Wins leaders still takes home the hardware almost 2/3 of the time -- 60 of 91. But the voters have been moving away from rubber-stamping the Wins leader in recent years, which is a positive sign!
November 19th, 2010 at 9:43 am
My favorite Felix stat is 8 tough losses and zero cheap wins. Typically you can negate a tough loss or two, but to go 8 the wrong way is really amazing. This means we can generally assume with an average offense he would've been assumed to go 21-4. Pretty sure it would've been unanimous if that had been the case.
November 19th, 2010 at 9:47 am
We want to be careful to not always assume that someone other than the top-WAR guy getting the Cy means it was "wrong". There are years where there are two or more guys closely vying for the WAR lead and the margin-of-error makes a choice of any of them legitimate. We might want to look at guys within half a win of the lead or something. That might give us a better idea and how often the Cy Young goes to a pitcher with a legitimate case for being considered the best in the league.
November 19th, 2010 at 9:47 am
Question: Felix Hernandez had a WAR of 6. His record was 13-12. Does this mean that a "AAAA" replacement pitcher would be expected to have gone 7-18? If so, does this not seem really low to anyone?
CC went 21-7 w a WAR of 5.4 (I'll round down to 5). Does this mean a AAAA pitcher the Yankees called up would be expected to go 16-12? This seems a bit off, unless I am missing something or misinterpreting WAR. Please help.
November 19th, 2010 at 10:05 am
@ 8
While team WAR correlates well with team wins I don't think you can just subtract a pitcher's WAR from their win total to determine what a replacement player's results would be, especially since you are referencing two pitchers that go deeper into games (not only due to the quality of their pitching, but due to having more stamina). I can't really say if a replacement pitcher would have a .280 winning percentage with Seattle (I'd think they could manage slightly higher) or a .570 winning percentage with New York (I'd expect slightly lower, perhaps not even .500 given the offenses he would've faced). Based on no information at all I'd say 5-15 for a replacement player pitching for Seattle and 12-11 for one pitching for New York.
November 19th, 2010 at 10:40 am
Just to clarify/elaborate on my post... I think it came off far more critical of the presentation here than I intended. My point was only that looking deeper into the data might reveal more about the tendencies of the voters and whether or not they are hitting better than .500 and/or getting better or worse. I don't know how to get at that idea without a lot more legwork, so I realize my musings are more than mere suggestions.
November 19th, 2010 at 10:44 am
Actually, 2005 was the last time the WAR leader DIDN'T win the CYA. It looks like the WAR proponents should have been happy the last 5 years, instead of regarding the F.Hernandez selection as "finally" being a SABR win.
And it (WAR leader winning CYA) would have continued though 2010 if Jimenez had not collapsed down the stretch while Halladay was ridiculous in September (5-0 in 5 starts) leading the Phils to the playoffs.
November 19th, 2010 at 11:09 am
# 1. Ed.
HILARIOUS!!!
You look at the trend and say; "aha, the voters are finally getting it".
Others might look at the trend and say that it points to the near uselessness of WAR to correctly gauge the value of players several decades back.
In much the same way that a time machine cannot go back to a time before it's invention, I think WAR really strains to compare players of the decades before it's invention.
November 19th, 2010 at 11:18 am
I was surprised to see that Koufax wasn't the WAR leader in '65 so I looked up the pitching leaders for that year. Koufax finished 3rd in WAR, pretty well behind Marichal and 0.1 behind Bunning. I could kind of understand Marichal based on park factors, but Bunning? Koufax blew him away in virtually every category. I know the old stadium in Philly was a hitters park (as are the two that succeeded it), but by that much? Can someone who really understands this stuff please explain?
November 19th, 2010 at 11:19 am
@7,10
I agree with you. Its hard to quibble about a close-WAR choice like Jim Palmer in 1975. Or perhaps even a choice like Maddux over Rijo in 1993. Its easy to see that Jose Rijo was underrated that year but "robbed" of a CYA? I don't know about that. (the main difference in that particular case looks to be UER)
Much easier to pick on the Lamarr Hoyt and Pete Vuckovich selections.
November 19th, 2010 at 11:41 am
@# 7 "We want to be careful to not always assume that someone other than the top-WAR guy getting the Cy means it was "wrong".
@# 11"Actually, 2005 was the last time the WAR leader DIDN'T win the CYA. It looks like the WAR proponents should have been happy the last 5 years, instead of regarding the F.Hernandez selection as "finally" being a SABR win".
Congrats to the sabermetricians for their win, which have been far more common than many have been willing to admit. As to the two blogs above, a difference of .6 isn't much, and so I still would have voted for CC or Price given the win and pitch environment differences (pressure vs. zero pressure). Feliz wasn't a bad pick since I see significant value in the WAR, but as far as Cy Young, the WAR-propenents have made significant headway. Lets hope the MVP award doesn't get too slanted towards the WAR.
I also think King Felix not winning last year likely gave him a bit more of a bump this year.
November 19th, 2010 at 11:41 am
Good list Neil,
I don't think Felix was a slam dunk case. I think you can make a strong case for Lee, Weaver, Liriano, and Verlander as well.
It's interesting that Felix's 6 WAR is the second lowest among any WAR/Cy Young winner and only 1 of 4 out of the 49 WAR/Cy Young winners to have a WAR below 6.5, Webb '06, Peavy '07 & Lincecum '09 were the other three. (Cone '94 & Fernando '81 were strike years).
You could also make a case that Webb deserved the Cy Young in '07.
November 19th, 2010 at 11:44 am
Ferguson Jenkins's WAR in 1971 was 9.2, not 12.2 (that number got copied by mistake from Carlton's the next year). Tom Seaver's WAR in 1971 was also 9.2, by the way, which points up the problem with using WAR to say who the best pitcher was. It's a cumulative stat, not a qualitative one--it measures how much, not how well. In this case, Seaver actually had much better year. Jenkins matched him in WAR because he pitched more innings. (And won the Cy Young because he won more games, but that's another issue.)
November 19th, 2010 at 11:51 am
Thanks Morten, I fixed Jenkins' WAR. Also, in response to cumulative vs. qualitative, WAR tries to combine the two by recognizing that the additional innings Jenkins pitched added real value in and of themselves... Every inning he pitched was an inning a lesser pitcher didn't have to.
November 19th, 2010 at 12:00 pm
I have the same questions as number 8 above.
Plus more.
I really don't understand this WAR stat for pitchers. I know that it includes a component for defensive support. But does being on a bad offensive team help a pitcher's WAR? Does being on a good offensive team hurt?
Here's an example of something I noticed a few days ago.
In 1985, Ron Guidry won the Cy Young with a pitching record of 22-6. He lead the league in wins. He was first in win percentage, 7th in ERA, 2nd in walks and hits per inning, and in the top 10 in a number of pitching categories, including 9th in ERA Plus. But at 4.2, his WAR was not even in the top 10 for pitchers.
Compare Guidry to Dave Stieb that same year. Stieb had an amazing year other than his 14-13 win loss record. Second in pitchers' WAR with a 6.5, led the league in ERA and ERA Plus; top 10 in many categories.
Here's what looks REALLY strange to me.
Stieb's team, Toronto, was 99-62. His win-loss percentage was .518 His team's win loss percentage was .615. Toronto scored 759 runs (4.7 per game) that year and finished in first place. Other pitchers on Toronto had much higher winning percentages, including Dennis Lamp (in relief) at 11-0 and Jimmy Key at 14-7. The other three pitchers who started over 20 games had over a .600 win loss ration, as did almost the entire pitching staff. Stieb led his team in LOSSES.
Guidry's team, NYY, was 97-64. They scored 839 runs. Guidry's win-loss was .786 compared to the team's .602. If Guidry won at the same ratio as his team, he would have been 16-12 instead of 22-6.
How is it that Guidry was only worth 4.2 WAR, but Stieb was worth 6.5? Toronto had a pretty average pitcher, Jim Clancy with an ERA Plus of about 113. His WAR was 1.8. His win loss ration was .600, just a tad under the team's.
I don't see how this makes sense.
November 19th, 2010 at 12:02 pm
P.S.: Guidry and Stieb pitched almost the identical amount of innings. 259 to 265, I think; so don't blame that.
November 19th, 2010 at 12:03 pm
@13 Argman,
I saw that the other day and I was thinking the same thing about the 1965 season.
I think basically it comes down to Total Zone and how it values defense.
Koufax's Dodgers are seen by Total Zone to be the best defensive team in the National League for 1965 with a +58 RDEF as a team. And when you think about all the mystique that surrounds Koufax one thing that's never brought up is the caliber of the Dodgers defense during those years: Willie Davis, Wills, Lefebvre, Fairly, and Parker.
Marichal's Giants are seen by Total Zone to be slightly below average with a -5 RDEF.
And Jim Bunning's Phillies are seen as the second worst defensive team in the majors with a (-40) RDEF.
It's interesting to note that Koufax's era+ of 160 is THIRD that year behind Marichal 169 and Vern Law 163.
I don't know if FIP numbers exist for 1965 but it would be interesting to see how Marichal & Koufax compare.
November 19th, 2010 at 12:05 pm
@19 Joseph,
Guidry didn't win the Cy Young in 1985, Bret Saberhagen won it.
November 19th, 2010 at 12:06 pm
@19,
I don't think Guidry won the CYA in 1985. Saberhagen maybe? Guidry only won in 1978.
November 19th, 2010 at 12:09 pm
HA HA, you guys are right. Damn my narrow nap top screen.
My question is still valid, however.
November 19th, 2010 at 12:11 pm
Argman, the major difference between Koufax and Bunning '65 appears to be defensive support. WAR sees the Dodgers as the best defense in the NL, and the Phillies close to the worst. A "gut-check" sort of supports that: PHI had Dr. Strangeglove and Dick Allen at the infield corners, and I only recognize Johnny Callison as having a good defensive rep. LA had Wes Parker, Maury Wills, and Willie Davis with good reps, maybe others too.
November 19th, 2010 at 12:11 pm
@18
Neil, Jenkins pitched about 40 more innings and allowed about 40 more runs. I'm not sure how that's adding value.
November 19th, 2010 at 12:12 pm
Too slow....
November 19th, 2010 at 12:17 pm
Johnny, John-Q -
Thanks for the explanation. I had dug a little deeper and was wondering if that was part of the reason. But that begs another question - does the WAR stat take into consideration Koufax's incredible strikeout numbers (especially in '65) when factoring in the team defense? What I'm trying to say is with the number of guys he K'd, how much did those superb fielders help him?
(As you may have guessed, I'm a Koufax partisan - my first baseball hero, and still my favorite.)
November 19th, 2010 at 12:20 pm
Re: #26 - But you also have to think about the runs that would have been allowed by a replacement-level pitcher in front of the same defense, in the same park, in the same innings... A replacement pitcher in Seaver's innings would have allowed only 143 runs that year; a replacement in Jenkins' innings would have allowed 199. So even though he allowed 53 more runs than Seaver, he added the same amount of value relative to what the replacement would have allowed.
November 19th, 2010 at 12:22 pm
Re: #28 - WAR only allocates credit away from the pitcher to his fielders on balls in play, so Koufax gets full credit for striking someone out.
November 19th, 2010 at 12:25 pm
@30 - thanks. so the rationale behind Bunning being ranked even with Koufax is based on his having to overcome shabby defense versus Koufax's benefitting from great defense? It seems like an awful lot of weight given to that, based on what I'd consider the fallibility of fielding statistics, especially "historic" ones. Am I wrong about that?
November 19th, 2010 at 12:42 pm
Well, that and the fact that Koufax benefited from an extreme pitcher's park (Bunning's PPF was 98, but Koufax's was 92!). Despite Koufax pitching 44 more innings, WAR sees a replacement pitcher in Bunning's situation actually allowing more runs (161) than the replacement would allow in Koufax's circumstances (158), simply due to park, defense, and strength of opposing hitters. That they allowed basically the same number of runs explains why they have essentially the same WAR.
November 19th, 2010 at 12:52 pm
@31,32
There's an 18 run gap between the two defensive contibutions (+11 Koufax, -7 Bunning).
I think it Argman might be right that its hard to quantify these things, but there's no denying the large gap between the two defenses. Willie Davis had a great year in CF and Maury Willis and Jim Lefebvre both had great years up the middle for the Dodgers. On the flip side, the Phillie had a bunch of stone-handed guys that the moved from position to position to try and minimize their impact. Taylor, Allen, Rojas, Stuart, etc.
When two pitchers pitch ~300 innings, the difference in defenses will add up. Whether its 18 runs, I don't know.
November 19th, 2010 at 12:52 pm
@19 Joseph,
I think run support per game is the biggest factor in the Steib, Guidry comparison.
Guidry received 5.8 runs per game in '85 which was ranked 3/55 in the American League that year. Steib on the other had received 4.5 runs per game which ranked 32/55 in the American League that year.
The Yankees 1985 bullpen is kind of forgotten but it was excellent that year:
Righetti-146era+
Shirly-153era+
Fisher-170era+
Bordi-126era+
Allen-148era+
So that bullpen definitely helped Guidry hold on to a number of wins. Guidry also had a great year when you consider he was 2nd in the league in k/bb, 1st bb/9, 2nd Whip, 9th in era+. It seems like Guidry's Achilles Heels was his HR/9 which ranked 26/42 and his K/9 22/42.
Steib also had a great bullpen maybe a notch below the Yankee version.
Steib's 1985 is kind of bizarre and just points out the arbitrary nature of W/L records.
Steib had a 173 era+ and a 2.48 era which led the league. It's just very odd that he only won 14 games that year.
I think what Steib's season comes down to was his big Achilles Heel: BB/9. Steib ranked 29/42 in BB/9 which probably accounted for his relatively low win total. Toronto had an excellent defense that year, the best in the majors by a large margin so Steib probably would have been better off putting more balls in play rather than giving up all those walks.
If you look by comparison Key and Alexander had excellent BB/9 in 1985. Key was 4th/42 and Alexander was 8th/42 so they let the Toronto defense do a lot of the work and it showed in their W/L totals.
November 19th, 2010 at 12:55 pm
Argman, you're not wrong. WAR is just a starting point. Its benefit is in pointing out that Bunning's season was also extremely good, and possibly just as impressive as Koufax's when you account for all the stuff which doesn't show up in the raw numbers. So it gets us to take a closer look at his season, and we can come to our own conclusions. The fielding numbers are rough. We can be pretty sure that LA's defense was better than PHI's, but do we "know" that Koufax's fielders helped him exactly 18 runs more than Bunning's? No, we can never be certain of that.
November 19th, 2010 at 12:55 pm
I think WAR for pitchers in the present day is kind of absurd.
When Felix played for crappy Seattle teams early in his career he was stuck with them, due to arbitration, etc.
But nowadays, pitchers get to pick who they pitch for. Hernandez signed a long term deal with a crappy team. Why should stats seek to restore/remedy his crappy decision?
November 19th, 2010 at 1:06 pm
@19 Joseph,
I bet if you go back and look at Steib's 1985 season, there are probably a bunch of really odd games that Steib pitched great and either lost of had a no decision. It's really next to impossible to have a 173era+ on a 99 win team with a great defense and only win 14 games. It has to be one of the toughest luck seasons a pitcher has had in the last 40-50 years.
I went back and checked a few game in April of 1985 and Steib lost a game to Royals 2-0, he got a no decision against the Royals leaving the game at 2-1 and he got a no-decision after leaving a game UP 7-2 against the Orioles.
November 19th, 2010 at 1:10 pm
Barkfart, your question makes no sense. I can't tell if you're trolling or just showing more of your anti-sabermetric feelings.
I'm sure Felix doesn't care one iota what his WAR is, so how is his "crappy" decision being affected? WAR doesn't care what team he is on. It is not seeking to restore anything. It is seeking to portray how well he pitched. It seems quite obvious Felix is an excellent pitcher who had an excellent season, and WAR reflects that. If you think his W-L record better reflects how well he pitched (i.e., average), then you are being deliberately obtuse.
November 19th, 2010 at 1:16 pm
@35 Twisto,
I tend to agree with your comments. I think WAR works better for position players than pitchers. I think there are too many variables in what a pitcher actually does and what he's responsible for to isolate it into one value like WAR.
Do you know if there are FIP numbers for the 1965 season? And if so how do Koufax, Marichal, Bunning rank?
November 19th, 2010 at 1:18 pm
For modern day pitchers I tend to add the BRWAR with the Fangraphs WAR and divide by two. I think the truth in the value of a pitcher is somewhere in the middle.
November 19th, 2010 at 1:35 pm
Barkfart,
It's not completely terrible to pitch in Seattle. It's not like pitching for the 1970's Cubs where you pitched in a great hitter's park with a bad defense. Granted the Mariners' offense is terrible, but remember Safeco is one of the best pitcher's parks in baseball plus the Mariners defense was great this year. Cliff Lee had a 8-3 record with a 2.34 era before he got traded so you can win and pitch well in Seattle. Actually I think Cliff Lee would have won the Cy Young had he stayed in Seattle.
November 19th, 2010 at 2:20 pm
John Q, the basic FIP is pretty easy to calculate. It's (13*HR + 3*(BB+HBP-IBB) - 2*K) / IP + a constant to make it match league ERA (for the '65 NL that would be +2.76). So:
Koufax 2.13 (actual ERA 2.04)
Marichal 2.79 (2.13)
Bunning 2.61 (2.60)
So if you went by that, it would appear Marichal got a big boost from his defense and Koufax and Bunning were relatively unaffected.
(No guarantees on accuracy of my calculations.)
November 19th, 2010 at 3:00 pm
Thanks Johnny Twisto, good stuff.
I don't understand adding the "constant" part to the innings pitched in the equation. Why 2.76 for 1965? The league NL league average era for 1965 was 3.54, why or how did you use 2.76 as a constant?
Also, why multiply 13 times the HR total and 3 times the (bb+hbp-ibb) and 2 times the K total? is this to make it jive with a baseball ERA?
November 19th, 2010 at 3:45 pm
Adding the constant is just to make the FIP look like a regular ERA. Doing the calculation of the HR, BB, and K per IP for the league as a whole, you get something like 0.8, so you add the constant to make it equal the league ERA of 3.54.
The coefficients are to weigh each event by its effect on run-scoring. This article seems to explain it pretty well: http://www.3-dbaseball.net/2009/10/evaluating-pitchers-with-fip-part-i.html
November 19th, 2010 at 5:19 pm
Johnny Twisto,
I think I'm getting it.
So do you take all the league totals and just plug them into the FIP formula to get your constant?
In this case the NL in 1965 had:
HR-1318
BB-4730
HPB-404
IBB-596
K-9649
Innings Pitched-14643
So I plug in the numbers following the FIP Formula: (13 x HR) + 3 x (BB+HBP-IBB) - (2 x K's) to get my numerator:
(13 x 1318) + 3 x (4730 + 404 - 596) - (2 x 9649)=11450 (Numerator)
Then I divide 11450/14643 (Innings Pitched-Denominator) and I get .78
Then I subtract .78 from 3.54(NL league average for 1965) and I get 2.76 which is my Constant. Yes?
November 19th, 2010 at 5:36 pm
Correct.
If you are too lazy to run numbers for the league, or don't have the numbers handy, you can just add on a default constant to get you in the right ballpark. 3.20 gets bandied about -- that's probably better for higher-scoring leagues, since that's what were in when the formula was derived. 1965 was a lower-scoring time, so I guess 2.76 is at the low end of the constants one would use. So I'd guess adding around 3.00 works for a more neutral-scoring league. (I've not played around with FIP much myself, so this is just sort of an educated guess on the changing constants.)
November 19th, 2010 at 5:37 pm
I think what got lost in the NL Cy Young is how good a season Josh Johnson had. He had a 2.41 FiP which was the best in the major leagues. He had a 183 era+ which was second in the majors he had a 6.4 WAR which was 3rd overall in the majors. The only thing he didn't have, were the innings pitched.
November 19th, 2010 at 5:46 pm
@46 Johnny Twisto,
Once FIP is explained it's really not that hard, it's just arithmetic with 7 or 8 more steps than a pitcher's ERA. I think part of the problem is that baseball writers don't take the time explain the statistic and what it means.
A slight problem with just relying on Fip like ERA is that it's a "Rate" stat so yeah Josh Johnson had the lowest Fip but he also only pitched about 183 innings.
November 19th, 2010 at 11:33 pm
Thanks for all your great in depth answers, gents, especially John Q and Johnny Twisto.
I think WAR does not make sense for pitchers and does not make sense to judge whether a Cy Young award winner is worthy based on WAR.
In 1970 Darold Knowles won 2 games. His pitching WAR was 2.9.
Somehow this doesn't make sense to me.
Does WAR account for situations like this:
A guy who pitched great games during the season, maybe 5 or 6innings, maybe allows two runs and his team got 2 runs, didn't get run support--he gets a no decision. At least his team gets a chance to rally and get the win.
November 20th, 2010 at 12:19 am
Joseph, I see not all of your questions from #19 were answered, as well as your new one. No, WAR does not account for offensive support. It doesn't account for decisions or for individual games where a pitcher kept his team in the game. Pitching WAR is actually pretty simple. Look at how many runs the pitcher gave up during the season, compared to what a replacement-level pitcher would be estimated to give up in the same number of IP. Adjust it for the park. Adjust it based on defensive support. That's it. That's runs above replacement, and it just gets converted to wins. It does not consider in any way whether the pitcher was credited with a win, a loss, pitched a shutout, had Babe Ruth as a teammate. Just total runs allowed and adjusted by defense.
(Also, relievers have an adjustment based on their leverage -- Mariano Rivera gets a boost because he pitches in more important situations than average.)
November 20th, 2010 at 12:27 am
As for Knowles, he gave up very few runs in 120 IP as the Senators' fireman. W-L records usually don't mean much for relievers. In his case, managing to lose 14 games despite giving up so few runs, I would assume he was not very good in the closest games, and indeed that is the case. He got hit much harder in tie or 1-run games, and in high-leverage situations, than in less important situations. WAR is not that context-dependent, so it makes no adjustment for when he allowed his runs.
November 20th, 2010 at 12:53 am
@49 Joseph,
You're welcome. Steib's career is very interesting in the way his W/L record doesn't match up with how good a pitcher he was. He was probably the best pitcher in the AL from 1981-1985 yet he never won a Cy Young award. That's very rare for a pitcher to be the best pitcher in a league for 5 years and not win at least 1 Cy Young.
As far as Knowles goes he was a relief pitcher so like Twisto said he pitched 119 innings. He also had a 2.04 era with a 174era+. ERA's are a little bit tricky with relief pitchers during this time period because of the way they would be used. Relief pitchers in the early 70's would pitch 1-3 innings and often come in during the middle of an inning.
Say the starter gives up a double and is taken out of a game for a relief pitcher. The relief pitcher gives up a single, the runner scores and then the relief pitcher gets 3 outs. That run is credited to the starting pitcher not the reliever.
Knowles 2-14 season is rather odd. I noticed he gave up 27 earned runs and 36 runs. That means that there were 9 errors made while he pitched. The 9 errors might account for the some of the losses and his relatively low era.
November 20th, 2010 at 1:04 am
I noticed he gave up 27 earned runs and 36 runs. That means that there were 9 errors made while he pitched.
Not necessarily. An error can result in multiple UER. (And it could result in 0 UER.) But 25% of runs being unearned is a lot. In one game he gave up 4 runs, all unearned, as the result of one error. He didn't pitch well, giving up 4 hits in that inning, but it looks like the error could have been an inning-ending DP which would have resulted in 0 runs. This was one of his 11 blown saves and 14 losses.
It is odd how Stieb's W-L records were consistently worse than they should have been. Sort of a Blyleven light. I know there has been some research indicating Blyleven probably didn't win as many games as he "should" have, based on his run support and ERA. I don't know if that's true for Stieb, or if he really just had bad luck with run and bullpen support year after year.
November 20th, 2010 at 2:20 am
Thanks for more great comments Twisto and Q.
You know, when it comes to pitchers and whether they won as many games as they were "supposed" to, I think it would be impossible to measure all the factors. Maybe pitcher A doesn't get run support because he follows pitcher B, who won a lot of games, and whenever that happened, the teams two best hitters would stay up half the night partying with Pitcher B to celebrate the win.
And somehow it doesn't feel right to have the Cy Young Winner have a 13-12 record.
November 20th, 2010 at 5:03 am
You can well argue that King Felix had more pressure than someone like CC, considering what run support they got & were likely to continue to get, how difficult it was to win, & where they were in their contracts.
A guy could lose every game & have by far the best season ever by a pitcher (if he had enough IP0. Of course this is unlikely, but reflects how W-L is NO good measure of pitching performance, with the many other powerful variables. Randy Johnson deserved the Cy young the year he went 14-16, though Clemens W-L was superb/dwarfed him. Ryan was 5th in WAR with an 8-16 record. But he was 9th in IP. That means he was at least quite close to the best per IP.
November 20th, 2010 at 7:58 am
Johnny Twisto and Don Q
not sure if this thread is still alive,
No, I'm not trolling- if I was doing that I'd launch into an anti-Blyleven rant.
I'm not ant-Sabremetric as a whole, but I do chafe at WAR's attempt to make up for the hard-luck Charlie Brown players and adjust their stats to their rightful place. We can all go back and wonder what Gaylord Perry would have done on a better team. It's fun. Maybe WAR can be helpful.
But pitchers today have their choice of teams they play on- especially when they're in their prime. Many pitchers choose money to play on a weak team. I don't know if that's helpful.
November 20th, 2010 at 12:15 pm
I simply cannot believe all of this statistical nonsense, in addition to the flagrant ignorance of stating that Ron Guidry won the Cy Young award in 1985! C'mon guys, like Stengel, we all know who he is right?, said, "you could look it up!" So now baseball is going to give post season awards based on statistical nonsense that I freely admit I don't understand? OK, call me ignorant, I've been going to baseball games since 1970, so I'll say I've seen a few things, and know a few things about this game. How about this: There is no way Felix Hernandez was the best pitcher in the American League in 2010. Why, he won 13 games! He lost 12! I don't care what team he played for, what his MSRP is, or dividing his VAT by his ERA, and comparing it to the total runs scored by his team in months with an R in it. Who can follow this stuff? I look around me when I'm at a game, and I have to say, if I asked ten of the people sitting around me in any given season to explain the WAR to me, they'd probably ask, which war am I talking about: Iraq, Vietnam, Korea, War of the Roses? This is just more proof that sportswriters are simply ITTG, which means Irrevlevant To The Game! The next thing you know, we won't be deciding who wins the World Series based on a best-of-seven on the field format, we'll be doing it virtually, so that some moron can argue that the Red Sox actually won the World Series in 2010 based on a bunch of made-up esoteric, irrelevant, complicated, arbitrary, and convoluted statistics! And the ITTG sportswriters will be in complete agreement and try to tell us why it should be this way. Having ranted all this, Felix Hernandez is a great pitcher, I'd love to have him on my team, I saw him pitch against the Yankees this year, one of those 3 vital wins that make him so much better than Sabathia, Price and Lester, and he was brilliant, Yanks couldn't touch him. That doesn't make him an award-winner. Sorry.
November 20th, 2010 at 12:47 pm
B+
November 20th, 2010 at 1:36 pm
Ha ha...
Too much is made of the impact of the modern metrics. He led the league in both ERA and IP. Those two stats have been around forever.
Yes, the modern metrics tell a much more complete story, and I like that, but the 19th century metrics have him as the best "run preventer" in the league... hands down.
November 20th, 2010 at 2:16 pm
@57, Iron Horse, I tend to agree with you. I think there are things that we can't practically measure that makes a winner a winner (I don't like to call them intangibles, because, if there were a practical way to measure them, we might find some kind of statistical correlation).
And I don't think a guy who is 13-12 is a good Cy Young Award winner, unless maybe he also saved 40 or 50 games the same season. He might be a great pitcher by statistical standards, but he's not a "winner" in the sense that he puts everything in his life together properly to make sure he wins.
@ 59, John Q:
I was looking at Dave Stieb some more, and there is one thing that stands out to me as very unusual: He hit 114 batters between 1980 and 1990.
With pitchers with over 110ERA plus and 100 decisions in that time period, the next closest guy was Charlie Hough with 91.
And after that David Welch with only 62.
HBP isn't affected by fielding, right? And if the hit batter scores, the pitcher isn't charged with an earned run, is he?
Guys like Johnson, Clemens, Martinez, and even Kevin Brown, had similar numbers of HBP, innings pitched, and so forth, in the 1990's, but also struck out a lot more batters than Stieb did in the 80's. They also had higher ERA Plus and higher winning percentages than Stieb.
Does WAR or ERA Plus take into account HBP?
If not, then I don't see it as a true FIP stat.
November 20th, 2010 at 2:40 pm
Joseph Says:
HBP isn't affected by fielding, right? And if the hit batter scores, the pitcher isn't charged with an earned run, is he?
-------------
Runs that result from a hit batter are earned. Fielding metrics won't be affected by HBP. Both WAR and ERA+ account for HBP in that sometimes those runners end up scoring. HBP is a really easy factor to account for. You can pretty much group them in with walks almost all the time. In my opinion, you're barking up the wrong tree here.
And no one every claimed that ERA+ was "fielding independent". It doesn't count unearned runs but there's more to defense than that.
Sean said he was considering having a series of "how WAR is calculated" articles in the offseason. If he ends up doing that, maybe it would clear more things up.
November 20th, 2010 at 3:09 pm
Iron Horse. You admit you do not understand the new stats, though people explain them in detail on these threads & in links. I would not use the word "ignorant" if you do not know something, since the word has connotations of someone being vicious &/or racist or otherwise immature or anti-social. But when you start throwing around words like moron that qualifies.
Especially when you must know no-one would assign victories based upon team productivity. Your post does not make sense. If you freely admit you cannot fathom the reasoning behind the stats, & you give zero substantive arguments against them, you are merely engaging in mean spirited mockery.
Joseph, in many ways on this thread especially you have been shown how no other pitcher in the AL would have been ABLE to put together an even 13-12 record if you put him on Seattle in '10. take everything they have done & import it, also using the traditional metrics of ERA & IP. They absolutely could not have done as well, let alone better. It is merely that you cannot comprehend how the best pitcher can have a mediocre record. But overwhelming evidence shows that you are conflating many factors that a pitcher cannot control, like Seattle's historically low run production, with what a pitcher CAN control.
Barkfart, even when a guy's team does not collapse or do unexpectedly poorly, no rational measure of who is the best or who deserves individual awards should be based upon how good your team is. Arguably you can give a little bump to a guy who helps clinch a pennant, at least if he plays especially strongly down the stretch. But to disqualify a man due either being loyal to a team or going for money? Most go for money, but the better teams tend to give it to him. But I do not care if he is selfish or altruistic in choosing a bad team. We should not punish for things related to team performance, nor reward someone due to choosing to be on a 'winner'. They get plenty of intrinsic rewards & recognition already, we do not need to "cheat" for them by giving them awards & glory beyond what they deserve.
November 20th, 2010 at 3:17 pm
While WAR attempts to neutralize certain numbers by factoring in fielding and park effects, I wonder if it also counts 'non-forced running errors' against a pitcher.
Obviously an error essentially creates a 4 out inning. But a runner getting thrown out at the plate on a bone-head choice, picked off, falling down, essentially reduces the pitchers work by a third. In other words, it is an out he has not earned. I understand that over a long enough stretch, things like teams running themselves out of innings, tend to even out, but I still wonder if there is a stat for "unearned" outs. And if there is... does it factor into WAR??
November 20th, 2010 at 3:23 pm
2004, AL. Randy Johnson was clearly the best pitcher in the league. Not only was nobody within a run of him in WAR, but he lead the league in Ks, REA +, WHIP, & H/9.
He was 16-14. Clemens easily had the best W-L record, 18-3, but was definitely not as good. WAR has him about right at 5th best in the league. Run support is the biggest part of this, & he did not throw as many IP as Johnson & others. All measures of his raw #s shows him not the best or near so. That is forgetting that he had a releatively rocky September.
Another year he likely should have won the Cy Young. But on no planet anywhere is it rational to think that he would have had the best record, or as good as several others, had his performances been transplanted to the very best hurler's teams.
What he did to be 18-3 that year? If you put him on Arizona that year he would have had a LOSING record. But can those who ASSUME contrary to all evidence that those who do well in W-L did something mystical to merit that exact record consider it objectively?
If not, the-becoming extinct paradigm that is held is akin to, say, Creationism. Essentially a religious belief held in opposition to all that reason tells us, due to an emotional attachment automatically ingrained by early conditioning.
November 20th, 2010 at 4:57 pm
I tried figuring what Felix Hernandez's ERA would have to have been for him to go 21-7, like CC. I came up with 1.33. Maybe some of you smart boys can come up with something better, but I'll go with that for now. Obviously Felix needs to do a lot more work on getting his life together. And maybe some bionic parts.
November 20th, 2010 at 5:53 pm
I think there were valid points made all around.
@Iron Horse,
I think you make a valid point in all of these Saber-Stats being really confusing and as a side note I don't think the Saber community has done a very good job in explaining these stats. There is also a problem within a lot of Saber groups as far as having a group-think mentality and being very condescending and patronizing to people who question the authenticity of their conclusions.
I have to disagree with you that these saber stats take a mediocre pitcher and turn him into a Cy Young winner. It's not like the Saber Stats turn Oliver Perez into a Cy Young winner. Felix by all accounts was at least one of the top 5 pitchers in the AL it's just that W-L is a terrible way to judge a pitcher.
W-L Record is a holdover from the 19th century and boxing. A win or a loss is really something a Team does not an individual so why assign a win or a loss to a pitcher? Imagine if the NFL used W-L record with Quarterbacks? Joe Namath was 62-63 for his career. Dan Fouts was 86-84 for his career. Warren Moon was 102-101. All 3 are in the NFL Hall of Fame, in traditional baseball terms they would be looked upon as mediocre players. Meanwhile a guy like Jay Schroeder with his 61-38 record would be looked upon as a great player. No one in their right mind would rank Schroeder anywhere near Namath, so why should a baseball pitcher with a 61-38 record be considered much better than a pitcher with a 62-63 record?
There's 3 basic flaws with W-L record:
1- A pitcher's Run Support
2-A pitcher's Team Defense
3-Bullpen
The case in point I always use is Steve Trachsel's 2006. Trachsel was awful in 2006. He had a 4.97 era which was 37/41 qualifiers in the NL that year. He had a1.59 Whip which was 40th/41 qualifiers in the NL that year. He had a k/bb 1.01 which was 41/41 qualifiers that year. Put all that together and he had a 15-8 record. That's a .652 win% which made him tied for 6th/41 in the NL that year.
He won 15 games because the Mets had a very good bullpen in 2006, they had a good defensive team and most importantly he got amazing run support. 5.5 runs per start which ranked 6th in the NL that year.
November 20th, 2010 at 6:00 pm
As far as Felix's 13 wins, I really don't see this as a big deal when you put it into historical perspective. 7 pitchers won LESS than 13 games and won the Cy and 3 of them WON the MVP as well.
Mike Marshall won the Cy Young in 1974 with 15 wins
Sparky Lyle won the Cy Young in 1977 with 13 wins
Bruce Sutter had a 6-6 record in 1979 and won the Cy
Rollie Fingers won the Cy AND the MVP winning 6 games in '81
Willie Hernandez won the Cy AND the MVP winning 9 games in '84
Steve Bedrosian went 5-3 and won the Cy in 1987
Mark Davis went 4-3 and won the Cy in 1989
Dennis Eckersly won 7 games and the Cy AND the MVP in 1992
Greg Gagne went 2-3 and won the Cy in 2003.
November 20th, 2010 at 6:55 pm
@60 Joseph,
That is kind of interesting Steib's HBP total, he led the league 5 times. Steib's great strength was his H/9 which he finished in the top 3, 5 times in his career.
ERA takes into account HBP if the batter scores after he was hit by the pitch.
HPB isn't factored in his Whip or his k/bb rates so they look slightly better then if he walked someone instead of hitting them.
I don't think HBP is that big a factor because of the relatively small amount. At most he hit 15 batters in a year usually it was around 11, some other years 5-7. Remember this was also from a guy that was averaging 250 innings a year between 1980-1985, and that's including the strike shortened season of 1981.
Like I said previously Steib was the best pitcher in the AL from 1981-1985
and it's very odd that he didn't win at least One Cy Young award if not multiple awards. I can't think of another pitcher that was the best in his league for 5 consecutive years and did not win at least a Cy Young.
What it basically came down to with Steib was his terrible run support during that time period:
1980: 3.6 runs per game ranked 46/56
1981: 2.8 runs per game ranked 57/60
1982: 3.9 runs per game ranked 47/59
1983: 4.4 runs per game ranked 33/57
1984: 4.2 runs per game ranked 34/58
1985: 4.5 runs per game ranked 27/55
If he had any type of decent run support in '81, '82 or '84, he wins 3 Cy Young awards easily. They were won by 2 relief pitchers and a terrible selection in Pete Vukovich in 1982. Top 15 run support in 83 and 85 could have won him two more.
War ranks Steib as the best pitcher in the AL in '82, 83 & '84 and 2nd in '81 & 85 and 3rd in 1990.
November 20th, 2010 at 9:55 pm
Stieb
"I" before "E"... for him anyways.
November 20th, 2010 at 10:01 pm
@67-Every single one of those guys were relievers, not starters.
Listen, I think Felix Hernandez is a stud pitcher no question. However, 13 wins does not cut it period for being a Cy Young winner as a bonafied number 1 starter.
Wins ARE important. If wins were not an important statistic, you would not have the standings in the paper every day. What are the stat geeks going to do next, petition MLB to no longer keep win/loss records and have the top 2 teams with the best WAR for each player play in the series?
November 20th, 2010 at 10:46 pm
David RF, good catch my bad.
@70 JR,
Obviously those guys were relievers my point was they won less than 13 games and won the Cy Young and 3 of them won the MVP.
Obviously Wins are important in a TEAM concept. As far as assigning them to a Starting pitcher it makes little sense because of the arbitrary nature of the variables involved.
I'll use my Football analogy. You wouldn't put any importance on a W-L record of a Quarterback so why would you put ALL the importance of a W-L record on a Pitcher? It makes no sense. If anything a Quarterback has much more of an impact on a football game than a staring pitcher has on a baseball game and a Quarterback's W-L record isn't taken into consideration for his overall productivity.
Joe Namath had a 62-63 W-L record. If he was judged like major league pitchers are judged he would be considered a mediocre QB. Jay Schroeder had a 61-38 record. If Football QB's were judged like Baseball Pitchers then Schroeder would be in the NFL Hall of Fame and Namath wouldn't.
November 20th, 2010 at 10:56 pm
John Q: If you evaluated Namath by any of his stats, he wouldn't be in the hall of fame, either. He threw like 50 more interceptions than touch downs. His passer rating for his career was about 65, which is not very good at all.
November 21st, 2010 at 12:30 am
Joseph,
That's not true, you're only referring to his "passer rating."
In 5 different seasons Namath finished 1st or 2nd in passing yards. 7 different seasons he finished in the top 5 in T.D. passes. 7 times he finished in the top 4 in passing yds per game. 5 consecutive years he finished in the top 4 in completion %. Remember too that players of that time periods didn't have very long careers.
He was the AP & UPI player of the year for 1968, He was the Super Bowl MVP in Super Bowl lll, He was the AP player of the year for 1969.
When he retired in 1977, he was #10 all time in pass completions & passing yards and #15 in T.D.'s.
November 21st, 2010 at 12:35 am
Joseph,
I just picked Namath as one example you can use plenty of others.
Fran Tarkenton had a 124-109 record
Dan Fouts had a 86-84 record
Warren Moon had a 102-101 record
All three are in the NFL Hall of Fame and none of them would be in the Hall of Fame is they were judged like Baseball pitchers.
November 21st, 2010 at 9:24 am
Wins ARE important. If wins were not an important statistic, you would not have the standings in the paper every day.
Team wins are not the same thing as pitcher wins. Team wins are the ultimate point of the game for every player. Pitcher wins are determined by an arbitrary formula much of which does not track the pitcher's value toward producing the ultimate goal: team wins.
If a pitcher has 100-120 pitches coming out of the 5th/6th/7th inning of a tie game or down 1 run and is starting to slow down but still pitching ok -- do they come out? If you have good relievers, the team's chance to win the game is probably higher if they come out, but that pitchers chance to get a win is zero if they come out, and fairly significant if they stay in. The pitcher doesn't even have control over this decision, just input.
If the offense produces a ton of runs, all the pitcher has to do to get a W is not pitch badly. If the offense produces no runs, that pitcher has to completely dominate or pitch superbly and get some luck merely avoid a loss, and can't ever win. The pitcher has *no* control over this.
The whole point of WAR is to judge players based on the expected contribution to team wins of what they did. It's not perfect, but this is what it is trying to do, and if you can make a good statistical argument for a way in which it is wrong and how to make it better, the people who use it and quote it are going to care, and maybe start using your new improved statistic.
WPA judges players on the actual situational contribution to team wins. This adds in any "pitching to the score" or "clutch" factors. The problem is that it also adds a great deal of noise, for we know from analysis that there is very little difference in clutch-ness between the clutchest and least clutch players (at least for those least clutch players who might ever be considered for the hall). The difference is so tight, that it's very hard to produce a study that can verify that there is any such thing as clutch ability at all, and easy to produce one that suggests there is not.
"wins" for pitchers doesn't do either of those things. It uses an arbitrary way to assign value to pitchers that was invented before anybody had a good sense of how to statistically judge player value, and in the days before computers made complicated calculations fast and accurate.
This is what the anti-stat crowd doesn't seem get. Pitcher wins is still a derived stat. You aren't choose to avoid stats, you're just choosing one algorithm over another. And the one you are choosing is just a bad one for the purposes you want to use it for (deciding who is a better pitcher). And you even *know* this most of the time. When comparing two pitchers, one of whom went 21-7 with a 3.75 ERA on a good team, and another who went 18-10 with a 2.50 ERA on a bad team -- almost everyone will vote for guy number 2. It's just that some people freak out when that "win" stat gets too low, because you are anchored to the idea that pitchers get wins and losses and that matters.
I'm not saying you should just trust WAR. That would be as stupid as just trusting pitcher W-L.
November 21st, 2010 at 3:42 pm
Well, now I'm glad that Cy Young is the all time leader in WAR, since the award is named after him.
I'm pretty sure nobody would want an award named after the 2nd place guy, Roger Clemens.
And the third place guy--well, the "Johnson" award sounds like it should go to somebody for something you don't see on the diamond, but in the locker room.
November 21st, 2010 at 5:11 pm
LOL, Yeah, the "Johnson Award" would sound a little bit embarrassing.
November 21st, 2010 at 8:26 pm
My Avatar is not amused!
Well seriously, Walter Johnson should have an award named after him for sportsmanship. About the kindest & most gentle man you could imagine. And if you add in offensive contributions, his WAR is clearly #2 amongst hurlers. Career OPS + was 76, great for a pitcher. Also still holds the pitcher record with a .433 B.A. An injury from a batter ball ended his career at 39. Imagine if they had the medical care they had today.
November 25th, 2010 at 5:09 pm
"However, 13 wins does not cut it period for being a Cy Young winner as a bonafied number 1 starter."
In 15 of his 34 starts, the Mariners scored two runs or less. In four of them, they got shut out--including one in which he pitched a 2-hit complete game, but since one of those two hits was a solo homer... Why was that loss his fault? 90+% of the time, if you only give up one run in eight innings, your team is going to win.
His first game of the season, he left the game with the M's up 3-1 (but two runners on base with two out). A reliever came on and let both of those runners score (from first and second, no less) before he could close out the inning. Did Felix really deserve a no-decision for that one?
His eighth game of the season, he left the game after 7 with the M's up 5-1. They lost 6-5. He should've had the win for that one, but the bullpen blew it for him. His eleventh game of the season, he allowed one run through eight innings, but because the M's only scored one run of their own, he had to sit back and watch the bullpen blow another game (a walk-off grand slam in extra innings, no less)
His sixteenth game, he went nine innings and only gave up two runs. But because the M's only scored two of their own, he had to sit back and watch the bullpen lose it in extra innings. His eighteenth game, he went seven innings, leaving the M's up 4-2. They lost 6-4. His twentieth game, he went eight innings and only gave up three...but the M's only scored two, so he soaked the loss.
His twenty-first game, he had a 2-hit shutout through eight, but management had decided that they wanted to start being more conservative with his pitch count, so they didn't send him out in the ninth despite the fact that he hadn't cracked 100 pitches yet. He got a no-decision because the M's didn't score until the bottom of the ninth. His twenty-ninth game, he had a shutout through seven, but since the M's didn't score until after he was pulled, he got another no-decision.
So I'm seeing at least three games that his bullpen cost him victories, and another six that his offense cost him victories. Give him a league-average offense and a less incendiary bullpen, and he probably would've gotten at least 17 wins. (yes, my math makes it 22, but that's assuming everything had gone perfectly, which is a rotten assumption.)
And it's "bona fide." It's Latin for "in good faith."