Sorta like that pro-Jim Rice/HOF dude…except with a deep hankerin’ for tailgators creole boulette po’boy.
The fact is that most HOF pitchers were truly great for a period of about ten years. Pitchers like Walter Johnson, Maddux, Spahn, Clemens and Seaver who had multiple outstanding seasons outside their peak decade are the exception, not the rule. It is clear that Ron Guidry had a peak decade that is comparable to the peak decades of many Hall of Famers - Bunning, Drysdale, Lemon, Wynn, Sutton, Gomez, Hunter, Jenkins, Ruffing and Roberts, among others. It is also clear that none of these pitchers did anything outside of their peak decade that materially added to their HOF resume.
* * * * * *
I would humbly submit that by any statistical measure Guidry’s HOF qualifications are the equal of Bunning’s, Dyrsdale’s, Lemon’s, Newhouser’s, Vance’s and Gomez’s. To the extent they won more games in their career it is because they pitched in the era of four-man rotations. I would also submit that Guidry’s HOF qualifications are the equal of Ruffing’s, Hunter’s, Sutton’s and Niekro’s. To the extent they won more games than Guidry they did so primarily because they had many more seasons where they were perhaps competent major league pitchers but not HOF quality pitchers.
There will no doubt be those who argue that many of these pitchers don’t meet their particular idea of HOFers. Hunter, Bunning and Drysdale are examples of more recent HOF inductees who are frequently characterized as marginal inductees. Vance, Newhouser, Coveleski, Pennock, Hoyt and Faber are just a few examples of other pitchers who have been deemed by many to be marginal HOFers. I think it is fair to say that Guidry’s HOF qualifications stack up pretty well against the qualifications of all the pitchers I’ve named in this paragraph. If one wants to argue nonetheless that Guidry doesn’t belong in the Hall then they are in effect arguing for a much smaller Hall of Fame and for HOF standards that are radically more restrictive than the standards that have been observed for the last 75 years.
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.
I would like to applaud JJ1986 for this post that works in so many ways, and from a Mets fan no less. Huzzah!
Because that can turn into a huuuuge HoF, pretty quickly.
Again: if Guidry, why not Gooden?
If Gooden, why not the next guy, and the next? Where do you stop?
I mean, he's not arguing Guidry is anywhere near being a top-tier HoFer, is he?
I honestly don't even know what the guy is arguing anymore.
EDIT: clarity
"Hey, I got my guy." -- Don Liddle, September 29, 1954
Bert Blyleven post-season - 5-1, 2.47 ERA
Ron Guidry post-season - 5-2, 3.02 ERA
Yeah, Blyleven couldn't handle the pressure.
Now please come up with a list of randomly selected starts by Blyleven where he pitched poorly against good teams between September 2 and September 27 in day games when he allowed 10 base runners that supports your point and finish up by calling me an insulting name.
I'm assuming that the reasoning behind it must have somehow been physical, that Guidry couldn't handle the strain and workload of being a starter, since he certainly hadn't failed as a starter in the low minors before the conversion.
Anyone have any info on the details?
Do you think it's easy to give up 8 hits, walk 7 batters, have your team make one error and give up a stolen base -- and still give up only one run? Even with Nettles playing at the top of his form?
Every single Dodger starter, and one pinch hitter, got on base, and only one crossed the plate.
Helluva game.
EDIT: Throw in 10 hits and 3 walks to the Yankees, and the official game time is still only 2:27. Yowza
What's weird about this line of argument is how unsupported it is by fact. Let's take one example, their career performances in "close and late" situations.
Stieb : 1168 AB, hitters hit .244/.313/.360 against him
Blyleven : 2129 AB, hitters hit .259/.317/.368 against him
Guidry : 1045 AB, hitters hit .264/.309/.418 against him
That's 17 points of OBP and 51 points of slugging percentage worse than Guidry performs overall.
Does that mean a whole lot? Probably not, not to me anyway. It does mean that Guidry's game-on-the-line performances were, over his whole career, substantially worse than Stieb or Blyleven and also substantially worse than his performance in non-close-and-late situations. While Stieb and Blyleven were approximately the same in both kinds of situations.
So any argument resting on the idea that Guidry was substantially better in such situations has an evidence gap : in fact when the game was on the line, Guidry was not just worse than either Stieb or Blyleven, he was substantially worse than your average serving of Louisiana Lightnin'.
Let's see some further evidence about clutch performance, please, before this argument goes any further...
No doubt. Tommy is playing you fishies like a violin.
I'm hoping that was directed at the original post (which for what it's worth I took as a bit tongue in cheek) and not me. I was just clarifying what he meant for you, not necessarily taking up the argument as a valid one.
I feel like I'm missing some history or something here. The Guidry from the Hall of Fame argument seems a reasonable one to take up. Though I don't buy it myself, he's in the discussion. How the discussion could become so heated is beyond me.
Niekro pitched in three playoff series, and also Game 163 in 1980 (which he won).
29 post-season innings. 0 earned runs. That's Z-E-R-O.
Guidry never even pitched ONE post-season shutout.
Regular season? LATE regular season?
Guidry won 38 games in Sept/Oct.
Niekro won 40. FOUR ZERO. And that was on significantly worse teams.
Late in the game? Innings 7-9, Guidry coughed up a .700 OPS.
Niekro, just a .688. SIX EIGHT EIGHT. TWELVE points better.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
I never knew that, either. My guess is that Tommy is from either Fairfield or New Haven County. There are Yankee fans north and east of there, but those two places have the highest concentration f them.
I'm guessing this is accidentally the angriest thing GGC has ever posted.
Name-calling helps.
What are the betting lines - Tommy exits over 209 and 212, or ignores them like all the other damning evidence and tosses out more blather?
I think he's done, but will wait in mild amusement to see if he continues the charade.
He's the kind of guy who merits the "if only he wanted to do good, instead of evil" line.
Trolls bring nothing to the table, but this guy has some legit points amid the psychosis.
Is there an internet moniker for that mixed bag?
WGLP
Yeah, that's real damning evidence.
True, Guidry never pitched a post-season shutout. He's just the only guy in baseball history to make four or more World Series starts and go at least 7 innings and surrender 2 runs or less in every start. What a choker.
Yes, Niekro won 40 games in September. He also lost 40. Guidry was 38-16 in September. In Septembers where the Yankees were within five games of first at any point in September he was 26-4. Obviously Niekro was the better September pitcher.
Hmm. Todd Helton has a higher OPS than than Stan Musial, Mickey Mantle, Willie Mays, Frank Thomas, Joe Dimaggio, Hank Aaron and Frank Robinson. I guess that makes Helton a better hitter than those guys! Ya know, we have stats now that adjust for the fact that, say, Helton played in the greatest hitting park in the greatest hitting era of all time, and it says that Helton isn't anywhere near as good a hitter as any of those guys. The stat is called OPS+. Guidry's OPS+ Against in innings 7-9 was 92. Niekro's was 100.
A complete idiot like Fred Lynn can show up at any online forum. What's weird is that he trots out this crap and everybody else in this forum hails The Devastating Evidence! OPS rules, there's no need to adjust for park or era, and so Fred Lynn and his buddies tell me that Todd Helton is a better hitter than Ty Cobb and Mickey Mantle and Hank Aaron. Niekro won 40 games in September, and that means he's better than Guidry no matter how many games Niekro lost in September.
The funny thing is all you guys fancy yourselves real sabremetricians, dazzling statistical analysts, and you've got to more ignorant of baseball statistics and fundamental statistical analysis than the average 2nd grader. Seriously, Howie says that's real "damning evidence", because he believes that a 40-40 record is better than a 38-16 record. He believes that a 100 OPS+ Against is better than a 92 OPS+ Against.
I seriously doubt most of you could add 2 and 2.
He tells us that Guidry was worse than Blyleven in close games, citing those late and close numbers. And yet Guidry's record in one-run games was 21-13. Blyleven's was 48-63. Johan Santana's late and close numbers are very similar to Guidry's. He obviously can't win the close ones, and his .660+ winning percentage must be a function of winning just blowouts.
Late and close stats don't tell you much, certainly far less than the high leverage statistics, which are defined by the impact the situation has on likely game outcome. Guidry's high leverage stats are very good, Blyleven's very poor. The high leverage stats correlate very closely with a pitcher's record in close games. Late and close does not. Schilling also has poor late and close numbers...and a nearly .600 career winning percentage. Seaver also had a great record in one-run games. His late and close numbers aren't impressive (a tOPS of 100) but his leverage stats show that he improved by 6% in high leverage situations. Again, good high leverage performance correlates with winning close games. Late and close numbers do not.
Statistics are very dangerous things in the hands of people like JLAC. They require analysis. JLAC appears incapable of analysis.
Late and close doesn't measure clutch. It's a narrowly defined and rather arbitrary statistic. It doesn't include a situation in the sixth inning of a tie game with the bases loaded, but it does include a situation where you're ahead by one run in the seventh and no one is on base.
The Clutch statistic at Fangraphs defines clutch situations purely by win probability analysis. It is continuous metric, not arbitrary like late and close. It shows that Guidry has one of the highest Clutch indexes of pitchers in recent decades, a +4.09. It also shows that Blyleven had a -3.88, and that doesn't even include some of his worst performances in high leverage situations, which occurred in the early '70s and aren't yet included in Fangraphs stats. Stieb was also a poor clutch performer but not as bad as Blyleven. Stieb had a -1.52.
It is not coincidental that Blyleven and Stieb significantly underperformed their Pythagorean projections in their prime, and it is not coincidence that Guidry significantly overperformed his Pythagorean projections. The clutch statistic measures precisely performance in situations that have a high impact on game outcomes.
Don't take this the wrong way, JLAC, but you have to be the most statistically ignorant baseball fan I've ever encountered.
So let's review the evidence. Guidry has a historically high Clutch rating as measured by win probability analysis. Blyleven has a terrible clutch rating. Guidry was 21-13 in one-run games. Blyleven was 48-63. Guidry significantly improved his performance in tie games and games within one run; Blyleven did not. Guidry won far more games than his Pythagorean projection, indicating he made very good use of his run support. Blylven significantly underperformed his Pythagorean projections, indcating he made poor use of his run support.
JLAC looks at this evidence and concludes that late and close stats are more telling about who is clutch, and apparently believes that Schilling is a choker because his performance declined nearly 10% in late and close situations, much more than Blyleven's declined.
Ignorance seems too weak a word for JLAC's peculiar combination of perfect illogic and imperviousness to reason. Sayonara, JLAC. I'll leave you to worship at your Todd Helton shrine and curse the choking Schilling!
First of all, wins and losses for pitchers are pretty much useless. Pitchers prevent runs, and teams score runs. Wins are a team stat, not an individual stat.
Nevertheless, ss far as pitchers underperforming their Pythagorean projections, I am guessing that run support has a huge bearing. Funny how that isn't mentioned.
Ron Guidry had an adjusted run support of 109 for his career. 100 is average. Catfish Hunter had an adjusted run support of 112. Those are 2 pitchers whose career w/l records are far better than they deserved. Everybody knows this.
Blyleven is 96.4, and Stieb is at 96.6. I'm wondering how much better their w/l records would have been if they had received Guidry or Hunter levels of run support.
One thing is clear and irrefutable, and it is that pitchers who receive excellent run support have better w/l percentages. Most objective people would mention things like that.
But in 1979, Joe Niekro was better in the clutch than any single year of Guidry's career.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
Tommy you've got some compelling stuff, but for most of us the Guidry file just wasn't good enough for long enough, nor peaky enough ala Pedro/Koufax.
And now over a 100 posts on clutchiness...yes!
What's most troubling about Guidry is how he could let his team down when the they needed him most. In 1986 the Yankees lost to the hated Red Sox by 5.5 games, largely because Guidry completely fell apart that year, going a pitiful 9 and 12 with a bloated ERA around 4.00. If Guidry merely pitches like a decent #3 on a contender the Yankees make the postseason. Add to that that in his last good year, 1985, the Yankees were never going to catch the Jays and it's clear Guidry's career was as much mere flash as it was actual substance, the latter of which the Yankees desperately needed, and Guidry failed to supply, in the debacle that was 1986.
edit: worse than I thought. As of August 1, 1986 the Yankees trail by 4 games. Guidry promptly has his second worst month of the season (.268/.301/.523/825 against), at the end of which the Yanks are hopelessly out of it, down by 10. They do a little better in September but it's obviously far too late. It's only then that Guidry puts up a respectable month.
I think this tells us a lot about Tommy. It's not that he's not smart, it's that his reading comprehension skills are just horrendous. JLAC never said Guidry was terrible, what he said was;
But Tommy goes and takes it personally. Unlike Tommy who has called Stieb and Blyleven and others chokers or worse I think everyone here has acknowledged that Guidry was a fine pitcher who had a career to be proud of, just not one that warrants Hall of Fame induction. Tommy has tried to give a statistical reason why he belongs but people have refuted this to which Tommy has resorted to name-calling and insults.
If anyone's frustrated, it's probably the guy who devolves into throwing insults once he gets challenged. Nobody else here has sand in their vagina just because everybody with two brain cells to rub together rejects Guidry for the HOF.
(1) Guidry in spite of doing well in clutch WPA is not clearly better than Steib in overall WPA and WPA/LI, because Guidry outside of clutch was otherwise a worse pitcher. Since Steib pitched significantly more innings, he would be much better in WPA over replacement.
(2) Saberhagen, Tommy's whipping boy, is clearly better than Guidry in WPA and WPA/LI. Sabes also has a better peak.
I will add:
Tommy concedes the use of WPA to measure clutch. Accept Guidry deserves 4 extra wins for clutch. He starts with a neutralized record of 136-116. Credit him with his clutch-adjusted record of 140-112. HOF-worthy?
Keep it down you nit, I'm just in the middle of preparing my "Bobby Murcer for HOF" article.
Here's a list of pitchers with similar ERA+ and Dec to Gator (the bottom line is that none are in the HOF):
Rk Player ERA+ W L IP Dec From To Age1 Dave Stieb 122 176 137 2895.1 313 1979 1998 21-40
2 Eddie Rommel 121 171 119 2556.1 290 1920 1932 22-34
3 Kevin Appier 121 169 137 2595.1 306 1989 2004 21-36
4 Andy Messersmith 121 130 99 2230.1 229 1968 1979 22-33
5 Hippo Vaughn 120 178 137 2730.0 315 1908 1921 20-33
6 Ron Guidry 119 170 91 2392.0 261 1975 1988 24-37
7 Thornton Lee 119 117 124 2331.1 241 1933 1948 26-41
8 Nap Rucker 119 134 134 2375.1 268 1907 1916 22-31
9 Dean Chance 118 128 115 2147.1 243 1961 1971 20-30
10 Mike Garcia 117 142 97 2174.2 239 1948 1961 24-37
11 Firpo Marberry 116 148 88 2067.1 236 1923 1936 24-37
12 Virgil Trucks 116 177 135 2682.1 312 1941 1958 24-41
13 Eddie Lopat 116 166 112 2439.1 278 1944 1955 26-37
14 Curt Davis 116 158 131 2325.0 289 1934 1946 30-42
15 Steve Rogers 116 158 152 2837.2 310 1973 1985 23-35
You forgot postseason WINZZZZZZ!!
WPA is a useful statistic for determining a players contribution in the context of the game. It is not a measure of the quality of a player or pitcher, as the originator of the state, Tom Tango, has cautioned. It is purely dependant on game context. WPA/LI is also useful but suffers from a similar problem: it completely strips game context and leverage from the equation. Players who perform best in clutch situations are consequently shorted.
Here's a simple example of the limitations of WPA that even you guys should be able to understand. WPA is dependent on the average leverage index for the season. This is something entirely out of the players control. Take two seasons that are exactly equal in every respect. One player had an average leverage index of 1.1 and the other .9. Now assume that the second player, the one with the lower average leverage index, performed better in the clutch. Which player ends up with the higher WPA? The first player. His non-WPA stats are exactly the same as the second player, he played worse in the clutch, but he has a higher WPA because his average leverage index for the season was higher.
As Tango said, a useful statistic but not a measure of player quality.
As for neutralized or translated stats, the methodology behind it is complex and also, unfortunately, convoluted and dubious. This example, however, ought to make anyone question the methodology behind the neutralized stats at B-R.com.
The 1983 Yankees won 91 games and had a .562 winning percentage. They were a good hitting team but a very mediocre pitching team, posting a 99 ERA+. Guidry got precisely average run support from the Yankees, i.e., the same run support the average Yankee pitcher got that year. If Guidry posted a 99 ERA+ that year, one would expect he would have a .562 winning percentage - same run support as Yankee pitching generally, same ERA+, so he should have the same W-L record. Now consider that Guidry actually had a 113 ERA+. Consider further that he performed very well in high leverage situations, too. These considerations suggest that Guidry should have a significantly better winning percentage than .562, and in fact he went 21-9 that year. But what does the neutralized stats project for his record? 13-13.
13-13? The average Yankee pitcher got 4.8 runs/game that year, had a 99 ERA+, and had a .562 winning percentage. Guidry got 4.8 runs/game, too, had a much better ERA+, and the neutralized stats produce a 13-13 record?
Anyone who can't see that the neutralized stats are complete nonsense should take the opportunity to explain that 13-13 projection for Guidry in '83.
I'll be right back, I just need to go make a bowl of popcorn.
One of the things that neutralized stats are supposed to is take away the fact that Guidry was playing for a good-hitting 90-win team. Hence, neutralized stats. It puts him in the context of playing on an average-hitting, average-pitching, 81-win team, in an league-average stadium, playing 162 games in a league scoring 4.4 runs per game**, 90% of runs earned. A player is given 1 decision per 9 (adjusted) innings pitched, and the wins and losses are done pythagorean of his adjusted runs allowed versus league average ie 4.4 runs per game.
So Guidry loses his hitting support, and he also gives up a few more unearned runs by this method,. He also loses a couple of decisions, and any "clutchiness" over his Pythagorean. And so we'd expect him to wind up 13-13 (to the nearest integer). Yes, his ERA+ was a little over 100 so he's close to 14-12, but that's the nature of integers for you.
**The detailed explanation still says 4.63 runs per game but this has since been changed.
I'm actually (gasp) going to agree with Tommy on this - I have no idea how B-R's neutralized wins and losses are figured, and that particular result makes no sense. I have a little toy method of my own that I'm sure Tommy will yell at for not being granular enough (which isn't an unfair criticism, but it's just a toy method anyway) - it's based on runs allowed, innings pitched, and innings pitched per game. It gives Guidry a 17-11 record in '83 (actually 16.7-11.4). How you take a 113 ERA+ and only 4 unearned runs allowed in 250.1 innings and get a .500 winning percentage, I'm not sure.
Guidry in 1983 gave up 3.59 R/9 IP. The Yankees scored 4.8 runs/game (my spreadsheet shows Guidry actually got 4.75 R/G, but it's not enough to matter).
So a typical won-loss record for 4.8 runs scored versus 3.6 runs allowed is .636, by Pythag^2. Figure expected wins in 30 decisions as 30 * .636 = 19 wins. 30 - 19 wins = 11 Losses.
So Guidry's actual won-loss record was 2 wins better than expected, 21-9 vs 19-11. I'm guessing this isn't a very uncommon occurence.
We can go game-by-game:
GUIDRY
1977, ALCS Game 2: spots Royals 1-0 lead through five, until Yankee bats bail him out with five runs in the 5th and 6th. Ends 6-2, Yankees.
1977, ALCS Game 5: spots Royals 2-0 and 3-1 lead before being yanked in the 3rd inning. Mike Torrez saves the Yankee season with 5-plus clutch innings, and the Yankee bats eventually come through. Ends 5-3, Yankees.
1978, Game 163: spots Red Sox 2-0 lead through six, until Yankee bats bail him out with four runs in the 7th. Guidry manages to bring the tying run to the plate in the bottom half, when he's wisely yanked for HOF relief pitcher Goose Gossage. Ends 5-4, Yankees.
1978, ALCS Game 4: spots Royals 1-0 lead, Yankee bats (homers by White and Chambliss) again bail him out. Ends 2-1, Yankees.
1978, WS Game 3: "The Nettles Game." Guidry gives up eight hits and seven walks, Nettles' glove and the Yankee bats save his bacon. Ends 5-1, Yankees.
1980, ALCS Game 1: Guidry finally gets a Game 1 start and immediately implodes, lasting just three innings in a 7-2 loss.
1981, ALDS Game 1: Guidry spots Brewers 2-0 lead through 4, Yankee bats bail him out with 4 in the 4th. Guidry tries to give it right back (2 hits and a walk in the bottom half), gets yanked, bullpen saves him. Ends 5-3, Yankees.
1981, ALDS Game 5: Guidry again spots Brewers 2-0 lead, Yankee bats again bail him out with 4 in the 4th, Guidry immediately pulled. Ends 7-2, Yankees.
1981, WS Game 1: Guidry "successful" when Yankees get him an early 5-0 lead; Guidry pulled when he walks two in the 8th to put the tying run on deck. Game ends 5-3, Dodgers.
1981, WS Game 5: with a chance to turn the Series' momentum, Guidry squanders a 1-0 lead when he surrenders back-to-back Dodger home runs. Dodgers win 2-1, take the Series at Yankee Stadium in the next game.
Guidry almost never goes more than seven innings, and doesn't win unless the Yankees get him at least five runs (he's just 1-1 otherwise).
Now, Joe Niekro.
1980, Game 163: Niekro throttles Dodgers, going the distance while surrendering just one late unearned run (and delivering two clutch sac bunts) as Astros win pennant, 7-1.
1980, NLCS Game 3: Niekro scatters 6 hits and 1 walk, shutting out the Phillies for 10 - TEN! - innings before 'Stros finally win it in the 11th.
1981, NLDS Game 2: Niekro delivers again, shutting out the Dodgers for 8 innings, before the Astros finally win it (again) in the 11th.
1987, WS Game 4: finally given a chance to pitch in the WS, Niekro delivers two more shutout innings, this time in relief. Again, his team gets him no runs.
Niekro goes deeper into big games, and gives up fewer runs, than Guidry -- and in 1981, the comparison is even more significant, because Niekro is more effective against the same team: the Dodgers.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
The method seems to work pretty well for hitters but is a complete disaster for pitchers. It purports to put a pitcher in a neutral ballpark (i.e., park factor of 100) in a neutral scoring era (i.e., 4.42 runs/game) and gives him average run support. However, it doesn't merely adjust the runs allowed and earned runs allowed by the pitcher, it recalculates them using a basic "runs created" formula that translates hits, BBs, and every other offensive event into runs. In concept, a pitcher who had a relatively good ERA+ but a mediocre or poor OPS+ Against would suffer as a result of this approach.
In my example I didn't assume league average run support for Guidry, I assumed average Yankee run support for 1983. Neutralized stats assume league average support. But if one assumes league average support for Guidry in 1983, then his record is 18-12 rather than 21-9, because the Yankees scored more runs than the average team that year. But that doesn't account for park factor, and Yankee stadium was a pitchers park in '83. But if you adjust for that Guidry's record is still 17-13 that year.
I initially assumed that Guidry's park adjusted ERA for '83, which was 13% better than league average, must not comport with the neutralized method of recreating runs allowed by using the "runs created" method. But the "runs created" method would generally correlate with OPS+ Against. Although Guidry's ERA+ may have been 113 that year, if his OPS+ was merely average then the "runs created" approach would penalize him. But Guidry's OPS+ Against that year wasn't average, it was 90, which comports with his ERA+ of 113.
A simple Pythagorean projection for a pitcher who received average run support and had an ERA+ of 113 results in a record of 17-13 for Guidry in '83. Again, this varies significantly from the neutralized projection, presumably because of a conceptual defect in the neutralized method of a calculating "runs created" and normalizing it park and league.
A guy who should really benefit from the neutralization method should be Curt Schilling, because his W-L record both underperformed his Pythagorean projection and isn't as good as his ERA+ stats. But the neutralization projection essentially generates the same career W-L record as his actual record.
The plainest demonstration of the defects inherent in the neutralization method is that on the whole the neutralized projection of pitcher's records should be evenly distributed, with the number of pitchers who go up or down in winning percentage, and the amount by which they go up or down, being essentially equal. But the neutralized stats penalize almost all pitchers. Very few see an increase in winning percentage, and almost all go down. Furthermore, those that go up in winning percentage go up very slightly, while many who go down see a precipitous drop in winning percentage.
The neutralization method should be absolutely agnostic in adjusting records up or down across the population of pitchers through history. It isn't - not even close. The method is conceptually flawed in various ways that are difficult to describe given the convoluted and dubious methodology described at B-R.com
Nope. From the link:
So if a pitcher is better at preventing earned runs than his component stats suggest, that skill is preserved through the translation.
Your method generates a sensible outcome, unlike the neutralization method. Guidry outperformed a Pythagorean projection, and your projection, for reasons I've previously discussed. He pitched very well in the clutch that year, which means that his leverage weighted runs allowed are far less than his nominal runs allowed, and he surrendered lots of runs late in games where he a huge leads and ended up winning the game by a huge amount anyway. All of that renders his ERA+ that year and his nominal runs allowed total very deceptive.
The neutralization method takes no account of these factors. But the real problem is that even a system that doesn't take account of these leverage issues should generate a record like the one you project for a pitcher with a 113 ERA+ and average run support. But the neutralization method comes up with a 13-13 record for a pitcher with a 113 ERA+ and an approximately 121 LevERA+. It makes no sense.
The odd thing is that the neutralization method seems to work pretty well for hitters. The methodology for pitchers, however, has some major conceptual defects.
Not in key games, he didn't.
And I've already talked about how he cost the Yankees the 1985 division title.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
Gibson gave up 4 runs in game one of the '64 Series and lost. He blew a 2 run lead in the bottom of the ninth in game two but is bailed out by Cardinal bats in extra innings. He gives up FIVE runs in game 7 but is bailed out by Cardinal bats.
Gibson blows a lead in game one of the '67 Series but is bailed out by Cardinal bats. He spots the Tigers a three run lead in game 7 and loses, 4-1.
Claude Osteen had a 0.86 ERA in 21 World Series innings.
Ergo, Claude Osteen is clutch and Gibson is a choker.
Rational people look at the results generated by their methodology and re-examine their methodology and processes. But Freddie Lynn just sticks with his methodology and never questions it even though it calls Gibson a choker and Osteen one of the great clutch pitchers, on the basis of nothing more than three games in the post-season.
I was trying to have a serious discussion with gay guy and Eric J and then Freddie weighs in with his usual nonsense. Who's the troll?
So Bill James does not have 2 brain cells to rub together? From his 2010 Gold Mine article on Season Scores:
Not that I necessarily agree with the Bearded Wizard. But I think he has at least 2 brain cells to rub together.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
Niekro was even better than Guidry in the clutch, against the same team.
You don't need any fancy stats or "normalization" to see that -- you just need to read the box-scores!
EDIT: Niekro was traded to the Yankees on September 15, 1985. AFTER that date, Guidry went 3-1. Do you think THAT was just a coincidence?
But that assumption only works if the runs created method is valid. But the neutralization method resulted in a 15% adjustment for Pedro's runs created even though the deviation of runs/game from the assumed scoring level was 20%. This is a significant difference in this context, and it skews the results. When OPS+ is poor relative to ERA+, this deviation is compounded because the proportional adjustment to ERA+ is undercounted as a result of ERA+ being small relative to the runs created figure.
Perhaps I didn't express myself well. I didn't mean to suggest that the neutralization method didn't ultimately translate actual earned runs allowed, I argued that it underadjusts where (i) the runs created methodology produces a result that varies from the actual difference in run environments and (ii) translates it in the context of a pitcher whose ERA+ is low relative to his OPS+.
Yeah, I read that Gold Mine article. The important thing to note here, and James repeatedly emphasizes it, is that he's not arguing that his Season Scores method is the best method for evaluating Hall of Fame qualifications, he's merely arguing that it produces results that very closely mirror actual HOF decisions. He wanted to account for why certain pitchers have historically been inducted even though they didn't compile huge career stats. James's method explains why the Hall has apparently inducted pitchers like Lemon, Vance, Gomez, Drysdale and Bunning - their careers may have been relatively short, but they were either among the best pitchers in their league for a substantial number of seasons, or were the very best pitcher in their league for a few seasons.
My argument is basically Bill's: if the HOF used its historical approach to inducting pitchers, they'd induct Guidry. But it appears that the Hall may have shifted in recent years to a focus on career stats and long careers.
7.15 ERA, 1.61 WHIP
(137.1 IP in 26 starts, 221 H/BB, 109 ER)
Question for the stat-crunchers: What would an expected W/L record be for a typical pitcher involved in 26 such starts?
(*also known as the "I want that 5-1 month, dammit" codicil)
You can't calculate it because projected W/L records are a function of runs scored, not runs scored while a starting pitcher is pitching.
Given that Guidry's run support for his 30 starts in September pennant races was 4.93, the projected W/L record for an average pitcher from those years would be approximately 16.5 - 13.5.
I find your data about the opposing starting pitchers to be odd, if not wrong, because the Yanks didn't give Guidry significantly better run support in those games than they usually gave him, and certainly 4.93 runs/game is not a figure that would begin to explain his 26-4 record. It's possible, though, that opposing pitchers pitched that poorly, but it's also irrelevant for purposes of projecting a W-L record.
Realistically, with a normal offense, that's probably close to a 3-15 level record.
It shows that Guidry outperformed his team on a normalized basis by 27% during his peak years. Seaver and Maddux are up at 37%. Gibson is at 28%. Schilling is at 27%. Palmer is at 20.4% (somewhat surprising). Bunning is at 19%. Drysdale is low man at 9.75%. Stieb is at 17%, which is actually pretty good.
His LevERA+ concept was interesting, and it certainly addresses the obvious defects with ERA+, but it still leaves the ERA+ stat somewhat limited in my view. This "performance relative to team", though, seems interesting. It seems to account for and normalize a lot of factors that effect a pitcher's record but have little to do with his performance.
These analyses are for peak periods running from nine years for most to 11 for Bunning (I think he did this because Bunning benefited by it, some of his best seasons having occurred at the beginning and end of that 11 year period).
It certainly exposes one big myth about Guidry: that his great W-L record was a function of run support and playing for great teams.
Hey, Tommy finally, finally, finally said what the #### his point is!
In James's term, this is the "if-one-then" argument: if those guys, why not this guy?
If these marginal candidates, why not this marginal candidate?
If Gomez, why not Guidry?
If Guidry, why not Gooden?
If Gooden, why not Key?
Etc.
No answer has yet been provided, but perhaps one will follow quite soon.
(my guess: the line should be right behind Ron Guidry)
Anyway, according to the Tommy-approved Fangraphs site, Guidry was terrible in the clutch in 1978, when he was supposedly the best pitcher in baseball. What's up with that?
Also, Guidry was 1-1 in key games* when the Yankees scored fewer than five runs. No fancy stats here, just looking at the box-scores. No myth, just fact. Is that Hall-worthy?
* Game 163 + post-season.
It is of course hopeless to engage with Davey NoPoint, but I didn't argue that leverage is irrelevant. The Clutch stat is based on WPA, but it is derived by subtracting WPA/LI (deleveraged WPA) from WPA and adjusting for the average Leverage Index for that season. WPA doesn't adjust for the average leverage index, and is in part a function of it. As Tango points out, it's not a good measurement for assessing a players performance to the extent that it benefits a player who just happened to play in a generally high leverage environment (regardless of how he performed in high and low situations) and hurts a player who played in a low leverage environment over the course of the season. Given that it's the objective of a team to produce low leverage situations, this is a counterproductive result.
The classic example is Dwight Gooden in 1985. Great season, great Clutch performance, but his record high WPA is skewed by the fact that his average leverage index was an extraordinary 1.15 in '85. It's largely happenstance that such a high average LI would occur, but it can also be produced by a pitcher who "pitches to the score", giving up more runs when he has runs to work with, thereby keeping the game closer and raising the average LI.
Guidry was a great clutch performer, as the Clutch stat shows, but it happens that all of his great seasons occurred when his average LI was extremely low.
1-1 in key games when the Yankees scored fewer than five runs = "great clutch"?
Come on. I'm just looking at the box scores, no fancy stats needed.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
Anyway, if Guidry, why not a bunch of other guys, too?
EDIT: I ask this because you did finally say what your point was, here.
EDIT EDIT: never mind.
Sloppy work, Fred. You missed some games.
Fourth game of the '77 ALCS. Guidry wins 4-2. That's not a key game?
Game 4 of the '78 ALCS. Guidry wins 2-1 to clinch the pennant. Not a key game?
Game 158 of the '85 season. Yanks need to win in order to stay alive going into a three-game series against division leading Blue Jays. Guidry wins, 3-0. Not a key game?
Sept. 15, 1978. Yanks are 1 1/2 games up on the Sox. Guidry pitches a two-hit shutout against Sox to win, 4-0. Not a key game?
Sept. 13, 1977. Yanks are 1 1/2 games up on the Sox. Guidry beats Sox 4-2. Not a key game?
I assume Freddie Lynn is a Sox fan. Sox fans thought those games were pretty big in '77 and '78. In fact, you btiches were ready to slit your wrists over those games.
I guess we know how much we can rely on Freddie's stats, don't we?
He did. He lead the league in starts that year. The pain only stopped because of the strike.
How the hell were the Twins only 7 games under .500 w/ that team!?!?
It was a key game -- in which Guidry did not appear, and which the Yankees won 6-4.
Game 4 of the '78 ALCS was Guidry's only win in a key game where they Yankees scored fewer than five runs. And he spotted the Royals the lead before two Yankee homers bailed him out late.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
Post 247 covered all this.
FTFY
He knows. He was there, and prefers his own interpretation.
Because he's trolling, and thus consistency of argument is a vice, not a virtue.
I have never seen somebody twist numbers as much as Tommy has in this thread - and that's saying something.
Hey, I tried.
And you didn't count it as a "key game." I guess "key games" are only the ones where Guidry didn't win.
Man, that's gotta be tough on your already dwindling credibility.
Your skepticism is odd and wrong. Those numbers (7.15 ERA, 1.61 WHIP) are the actual numbers the actual starters actually compiled while Guidry was actually winning his 26 actual starts. I'd have thought that someone who likes to lovingly deconstruct hand-selected close-and-late innings in which Dave Stieb walked the leadoff batter wouldn't have a problem with a simple "this is what the other guy did" average.
And I still wonder how many wins and losses the prototypical AL starter would have had in 26 September opportunities, if his 26 opponents lasted an average of 5.3 innings and compiled a 7.15 ERA. A small sample may provide a small clue: we know how Guidry's opponents did in nearly identical circumstances when their opponent-- Guidry-- had an 8.07 ERA and a 1.75 WHIP. They were 4-0. Whether or not those were four clutch wins I'll leave to finer minds than mine to decide.
I mean, I'm a Jays fan and I'm not even a big fan of Stieb. I always preferred Key.
I think Stieb was a better pitcher, but I just liked Key better.
Did anyone really like Stieb? He seemed like a bit of a dick...but he was still a damn fine pitcher.
Um...you were the one who referred to game four of the ALCS.
He meant Freddie caught his [Tommy's] mistake, and he [Tommy] was correcting himself.
See?
I'm an equal opportunity clarifier!
1978, WS Game 4: The "I Don't Give a ####, Dougie!" game.* Doug Rau spots the Yankees a three-run lead in the second. Guidry tries to give it back with a 2-run bomb by Davey Lopes (following a ground-rule double by pitcher Rick Rhoden). Game ends 4-2, as Dodger relievers sport a better line than the Yankee starter.
Basically, there's no evidence whatsoever that Guidry was better in the clutch than Joe Niekro.
Post 247 covered all this.
Anyway, you said your point was, "If the HOF used its historical approach to inducting pitchers, they'd induct Guidry. But it appears that the Hall may have shifted in recent years to a focus on career stats and long careers."
My question is, if Guidry, why not Gooden?
If Gooden, why not Key?
Etc.
I'm thinking you'd put the line right behind Guidry.
But - specifically - why Guidry, and not a bunch of other guys?
EDIT: I'm not asking whether he's better than somebody already in the Hall. I'm asking why he's better than everybody who's NOT in the Hall.
* those of you who haven't heard the recording of Lasorda on the mound should totally check it out.
I already gave you the calculation. And since Guidry was 26-0 in those games, it's academic, isn't it?
Koufax got knocked out in a huge game in late Sept. 1966. He didn't make it out of the 2nd inning in a the first game of the 1962 three-game playoff against the Giants, basically blowing the pennant. He got knocked out of three games in September during the '61 pennant race. His ERA in these five starts was almost 18.00.
I guess Koufax stunk.
You're a little slow, but even you should have understood that I didn't claim Guidry was infallible in a pennant race. He was just the closest thing to it.
Do you care to elaborate on your theory that Koufax stunk because he had a 17.00+ ERA in five September starts? I think he's generally considered a pretty good pitcher by knowledgeable people. I guess "knowledgeable" is the key here...
I guess there's a lot of difference between 5 starts and 6 or 9. So Koufax is still great, but Appier and Gooden get demerits.
edit: and of course mentions of Stieb choking in-season.
Ron Guidry was 26-0, there's no disputing it. Actual Guidry made the most of his 26 actual opportunities. I'm just wondering whether Bob Walk or Freddy Garcia or Kevin Millwood wouldn't have gone, say, 22-1. I understand that this speculation might be troubling or even hurtful to you.
YOU are the one that has been deciding a player's HoF worthiness based on small sample "pennant race" stats. For one pitcher, you'll point to one September where he stunk as evidence that he's not good, and then when someone points out someone else that stunk one September, you'll come up with some reason for why it isn't valid (or just ignore it).
You might be a little better at this trolling thing if you weren't so inconsistent, ridiculous and over the top.
Between his first and second retirements, Stieb indicated that he was frequently a colossal dick when he was active, and that he regretted not being better to his teammates.
In his first 2 starts in the playoffs, against the Astros, Gooden pitched 17 innings, gave up 1 run. He lost 1-0, on a Davis home run, then held the Astros scoreless for 10 innings, which the Mets won after he left the game. But somehow, because he did not win these games, he is unclutch? He also got NDs in 1988, when he started 2 games against the Dodgers.
In the stretch playoff drive of 1985, he did not allow a run in September. Zero. But Tommy knows that he was unclutch.
He did not pitch well against the Sox in the World Series.
You must be Registered and Logged In to post comments.
<< Back to main