This post is for voting and discussion in the 143rd round of balloting for the Circle of Greats (COG). This is the last of three rounds of balloting adding to the list of candidates eligible to receive your votes those players born in 1980. Rules and lists are after the jump.
The new group of 1980-born players, in order to join the eligible list, must, as usual, have played at least 10 seasons in the major leagues or generated at least 20 Wins Above Replacement (“WAR”, as calculated by baseball-reference.com, and for this purpose meaning 20 total WAR for everyday players and 20 pitching WAR for pitchers). This group of 1980-born candidates, comprising those with N-Z surnames, joins the eligible holdovers from previous rounds to comprise the full list of players eligible to appear on your ballots.
In addition to voting for COG election among players on the main ballot, there will be also be voting for elevation to the main ballot among players on the secondary ballot. For the main ballot election, voters must select three and only three eligible players, and list them in ranked order. The first player listed on each ballot receives three points, the second player listed receives two points, and the third listed receives one point. The one player accumulating the most points from all ballots cast in the round is inducted into the Circle of Greats. For the secondary ballot election, voters may select up to three eligible players, with the one player appearing on the most ballots cast elevated to the main ballot for the next COG election round. In the case of ties, a runoff election round will be held for COG election, while a tie-breaking process will be followed to determine the secondary ballot winner.
Players who fail to win either ballot but appear on half or more of the ballots that are cast win four added future rounds of ballot eligibility. Players who appear on 25% or more of the ballots cast, but less than 50%, earn two added future rounds of ballot eligibility. One additional round of eligibility is earned by any player who appears on at least 10% of the ballots cast or, for the main ballot only, any player finishing in the top 9 (including ties) in ballot appearances. Holdover candidates on the main ballot who exhaust their eligibility will drop to the secondary ballot for the next COG election round, as will first time main ballot candidates who attract one or more votes but do not earn additional main ballot eligibility. Secondary ballot candidates who exhaust their eligibility will drop from that ballot, but will become eligible for possible reinstatement in a future Redemption round election.
All voting for this round closes at 11:59 PM EDT Sunday, March 9th, while changes to previously cast ballots are allowed until 11:59 PM EST Friday, March 7th.
If you’d like to follow the vote tally, and/or check to make sure I’ve recorded your vote correctly, you can see my ballot-counting spreadsheet for this round here: COG 1980 Part 3 Vote Tally. I’ll be updating the spreadsheet periodically with the latest votes. Initially, there is a row in the spreadsheet for every voter who has cast a ballot in any of the past rounds, but new voters are entirely welcome — new voters will be added to the spreadsheet as their ballots are submitted. Also in the spreadsheet is a column for each of the holdover candidates; additional player columns from the new born-in-1980 group will be added to the spreadsheet as votes are cast for them.
Choose your three players, for both the main and secondary ballots, from the lists below of eligible players. The current holdovers are listed in order of the number of future rounds (including this one) through which they are assured eligibility, and alphabetically when the future eligibility number is the same. The 1980 birth-year players are listed below in order of the number of seasons each played in the majors, and alphabetically among players with the same number of seasons played.
Holdovers:
| MAIN BALLOT | ELIGIBILITY | SECONDARY BALLOT | ELIGIBILITY |
|---|---|---|---|
| Dick Allen | 10 rounds | Billy Williams | 6 rounds |
| Vladimir Guerrero | 7 rounds | Bobby Abreu | 4 rounds |
| Ted Lyons | 6 rounds | Ken Boyer | 4 rounds |
| David Ortiz | 4 rounds | Don Sutton | 4 rounds |
| Willie Randolph | 4 rounds | Andre Dawson | 3 rounds |
| Stan Coveleski | 3 rounds | Don Drysdale | 3 rounds |
| Todd Helton | 2 rounds | Monte Irvin | 3 rounds |
| Johan Santana | 2 rounds | Reggie Smith | 3 rounds |
| Luis Tiant | 2 rounds | Chase Utley | 2 rounds |
| Richie Ashburn | this round ONLY | ||
| Andruw Jones | this round ONLY | ||
| Gary Sheffield | this round ONLY | ||
| Ted Simmons | this round ONLY | ||
Everyday Players (born in 1980, ten or more seasons played in the major leagues or at least 20 WAR):
Albert Pujols
Mark Teixeira
Cody Ross
Nick Swisher
Shane Victorino
Laynce Nix
Skip Schumaker
Dan Uggla
Pitchers (born in 1980, ten or more seasons played in the major leagues or at least 20 WAR):
CC Sabathia
Pat Neshek
Jonathan Papelbon
C.J. Wilson
As is our custom, here are quiz questions for each of the new players on the ballot.
1. Albert Pujols recorded .995 OPS for his post-season career. Among retired players with 200+ post-season PA, who has the only higher career post-season OPS? (Carlos Beltran, 1.021 OPS)
2. CC Sabathia recorded 115 ERA+ aged 35-37 after managing only 83 ERA+ aged 32-34, a 32 point improvement that is among the largest by any pitcher with 400+ IP in both periods. Which pitcher recorded the largest such ERA+ improvement? (Lefty Grove, 49 points, from 125 to 174)
3. Mark Teixeira led his league in home runs, RBI and total bases playing for the 2009 Yankees. Which other two players had the same trifecta while playing in their first season for a new team and winning the World Series? (Frank Robinson 1966, Shohei Ohtani 2024)
4. Pat Neshek finished his career with twelve consecutive seasons (2007-19) with more appearances than innings pitched. Which pitcher recorded the longest run of such seasons to close out a career? (Joe Smith, all 15 seasons of his career)
5. Shane Victorino compiled 31.5 WAR for his career, the most for non-pitchers born in Hawaii. Which pitcher leads all Hawaii-born players in career WAR? (Charlie Hough, 38.4 WAR)
6. Cody Ross was MVP of the 2010 NLCS in which he homered twice off of Roy Halladay in game 1. Which other Giant homered twice in a post-season game against a future HOF pitcher? (Will Clark, 1989 NLCS game 1, off of Greg Maddux)
7. Nick Swisher’s .165 BA for his post-season career is lowest among all players with 150+ career post-season PA. Which player active in 2024 currently holds the record for lowest career post-season OPS in 150+ post-season PA? (Martin Maldonado, .501 OPS)
8. Jonathan Papelbon recorded 17 consecutive scoreless appearances to begin his post-season career, tied with Trevor Rosenthal for the longest such streak when averaging 1+ IP per game. Among retired pitchers with zero runs allowed for their post-season careers, who recorded the most post-season games while averaging at least one IP per appearance? (Matt Herges, 10 games)
9. Laynce Nix ranked first in Total Zone Runs among NL left-fielders in 2009, despite playing only 558 innings (equivalent to 62 games) at the position that season. Which other player, like Nix, recorded career totals including 100+ games at each outfield position, and positive oWAR and dWAR, with at least three times as much of the latter as the former? (Endy Chavez)
10. Skip Schumaker played at both 2B and LF in 31 games in 2009, the most games in one season for that combination of positions. Which player recorded the most career games playing at both of those positions? (Ben Zobrist, 68 games)
11. C.J. Wilson snagged a W in his post-season debut, his only win in 10 post-season starts. Which other pitcher saw his post-season career begin the same way? (Gary Nolan)
12. Dan Uggla ranks first in career home runs among second basemen for the first season of a career, the first 3 seasons, first 4, first 5, first 6, first 7, and first 8. Which player bumped Uggla to second place for the first 2 seasons? (Gleyber Torres)
Pujols, Williams, Allen
Thanks Jeff.
However, if I am to count your vote, you must limit your picks to the players listed under the “Main Ballot” heading plus the born-in-1980 players.
The players listed under Secondary ballot are in a separate election. If you want to participate in that election, please choose up to 3 players listed under the heading “Secondary Ballot”.
Main: Pujols, Allen, Guerrero
Main Ballot: Pujols, Sheffield, Papelbon
Secondary Ballot: Smith, Utley, Drysdale
I don’t usually look at the quiz questions, but I happened to notice the Teixeira one. The answer is Frank Robinson with Baltimore in 1966.
Robinson is correct.
I see that there is a second answer to the question, so I’ll rephrase the question and let you find the other player.
Answer to question 12 is Gleyber Torres who had 62 HR in his first 2 seasons as compared to Ugglas’s 58.
#1 has to be Carlos Ivan Beltran, who, BTW, did not hit well with the aid of Astros’ trash cans
It has to be.
Sadly, though, I suspect Beltran’s post-season career will be remembered more for his appearance on this list.
IIRC, “tough pitch”….he just froze
#6 is Pablo Sandoval who crushed 3 in the opening game of a WS – two off Verlander
Ah yes, BUT Verlander isn’t in the Hall … yet.
There’s another answer, involving a pitcher already in the HOF.
Yes, but , since they’re all active when they surrender the home runs, aren’t they all future possible Hall of Famers?
Including, future Hall of Famer Red Faber who, while active in the 1917 WS, was tagged twice by one-time Federal League superstar Benny Kauff?
Sorry, Will Clark off Greg Maddux, 1988 NLCS. Forget the Benny Kauff rant as he homered twice in a 1917 WS game but only one was off Red Faber
Clark is correct.
Kauff’s second homer was off of Dave Danforth, the lefty with a balk move to first (that was seldom called) and vast assortment of trick pitches, apparently aided by doctoring the ball. Danforth cratered in 1919 due to arm trouble and thus avoided the Black Sox scandal when he was sold to Columbus late in the season (Danforth sued the White Sox for a WS share and was eventually awarded a 1/4 share). Regaining his form, Danforth stayed in the minors for two years when major league clubs (the Yankees, notably) wouldn’t meet Columbus’s terms for acquiring him. When he made it back to the majors with the Browns in 1922 (the Browns offered up eleven players to Columbus, several of whom were not then with the Browns but were named in the event that St. Louis should acquire them in the future), it was post-Ray Chapman and Danforth became a lightning rod, constantly accused of still doctoring the ball. In late July, he entered an extra-inning game in relief, buzzed a Yankee hitter with some high cheese, and was immediately ejected for throwing a doctored ball and suspended for 10 games. Under pressure from all sides, Browns GM Bob Quinn demoted Danforth to the minors, where he helped Tulsa win the Western League championship, while the Browns finished a game back of the Yankees.
I knew nothing about Danforth, not even his name. Quite a remarkable career, both in the Majors and in the Minors. His unusually long SABR biography is good reading.
Doug: For question 2 I found 2 players with a larger improvement than Sabathia, Lefty Grove with an improvement of 49 and Justin Verlander with 41. Next in line is Sad Sam Jones with 31.
Looks like I screwed up the question (I think I meant to add “with ERA+ under 100 in the prior period”), as Sad Sam Jones was the answer I intended.
Good sleuthing.
To lay some groundwork for discussion, I’m posting ranked B-R stats for WAR/162G for candidates, separating the two ballots and position players from pitchers (total WAR in brackets). I’ve included Albert and CC from the 1980 class (on the Primary Ballot), as well as Papelbon, since he has a vote already. (Papelbon’s WAR figures aren’t informative, though, since relief pitcher rates are always quite low. For context, Rivera’s rate of 3.4 [56.3] could be viewed as a contemporary gold standard, and perhaps Wagner’s rate of 2.2 [27.7] as HoF threshold.)
Primary Ballot, position players
Allen 5.4 [58.7]
Pujols 5.3 [101.4]
Randolph 4.9 [65.9]
Ashburn 4.7 [64.2]
Andruw 4.6 [62.7]
Guerrero 4.5 [59.5]
Helton 4.5 [61.8]
Sheffield 3.8 [60.5]
Ortiz 3.7 [55.3]
Simmons 3.3 [50.4]
Primary Ballot, pitchers
Coveleski 5.5 [62.0]
Santana 5.4 [51.7]
Tiant 4.2 [66.1]
Lyons 4.2 [70.6]
Sabathia 3.7 [62.3]
Papelbon 2.3 [23.3]
Secondary Ballot, position players
Utley 5.4 [64.5]
Smith 5.3 [64.5]
Boyer 5.0 [62.8]
Irvin 5.0 [32.1]
Williams 4.1 [63.6]
Abreu 4.0 [60.2]
Dawson 4.0 [64.9]
Secondary Ballot, pitchers
Drysdale 4.2 [67.1]
Sutton 3.0 [66.7]
So, Pujols is on the same WAR rate footing as Allen, Utley, Smith, Coveleski and Santana?
Technically, yes, but again we have to decide whether to overlook Pujols’s lengthy decline phase, the same question I posed when we were discussing Simmons in the same vein earlier. Pujols was the beneficiary of a(nother) dubious Angels contract that was sure to be shelling out top dollar for Albert long after his prime. Saving face was the reason the Angels held onto him as long as they did, but since everybody knew it, it wasn’t really saving face at all. Instead, it was actually rather comical that the Angels regrettably came to the realization that it really didn’t make sense for Albert to be using up a roster spot, a judgment that was only coincidentally made in the final year of Pujols’ contract.
If the team and its fans were the primary concern, the Angels would have parted with Pujols a lot earlier and with a lot less theatrics. But, maybe it was all for the best that it happened the way it did, as everyone then got to bid farewell to Pujols the right way, in his phoenix-like “last hurrah” final season.
I don’t think we should overlook anything, Doug, and since Pujols’s CoG case is overwhelming it doesn’t affect the outcome anyway. WAR/162G captures one way of understanding the implications of Pujols’s long decline. The counter-weight is simply to stay aware of Pujols’s peak value, and the same would be true for Simmons (who doesn’t really fare well on either count when it comes to CoG norms).
I do think we should not treat Pujols as having no agency over ending his decline. No one forced him to play out his contract and there are plenty of examples of players who for one reason or another decided to forego their salaries to avoid playing beyond their productive years. (Dave McNally was once famous for having done so in a situation analogous to Pujols’s.) Pujols had already earned $20m by the time his value cratered — it’s not as though he and his family were in need. The trade-off was his choice. (I probably would have made the same one!)
Here are the CoG candidates ranked by their JAWS 7-yr. (non-consecutive) peaks, with Albert on top by a huge margin:
Primary Ballot, position players
Pujols 61.7
Helton 46.6
Andruw 46.4
Allen 45.9
Ashburn 44.5
Guerrero 41.2
Sheffield 38.0
Randolph 36.3
Ortiz 35.2
Simmons 34.8
Primary Ballot, pitchers
Coveleski 51.2
Santana 45.0
Tiant 44.1
Lyons 40.8
Sabathia 39.4
Papelbon 19.5
Secondary Ballot, position players
Utley 49.3
Boyer 46.2
Dawson 42.7
Abreu 41.6
Williams 41.3
Smith 38.6
Irvin 25.1
Secondary Ballot, pitchers
Drysdale 44.7
Sutton 33.9
Correction: “Pujols had already earned $20m” should have been “Pujols had already earned $200m.”
Reasonable arguments, Bob.
As you say, Pujols almost certainly didn’t keep playing because he needed the money. Rather, it was something he enjoyed doing and no doubt he still felt he was making a positive contribution, especially when he was healthy. Other players may retire sooner because of physical ailments, time away from family, or because they don’t want to perform at a level significantly below their prior peak. Everyone is different, so it’s not a knock on anyone to make the decision that makes the most sense to the player.
In Pujols case, when you take into account intangibles like mentoring teammates, or being an example to young players about preparation, training, conducting yourself on and off the field, or whatever, then no doubt he was making a positive contribution to the team in addition to whatever he did on the field. I doubt you’d find any Angel who would say otherwise.
I’m sure he added value through those intangibles, Doug, and that he didn’t want to retire. I’m also sure fans paid to see him. But Pujols was receiving a salary of $26-30m each year during his non-productive seasons (that would be ~$35m/yr now), playing at replacement level and blocking new talent. His underperformance contributed to his teammates having no experience of even a .500 season, and I can’t imagine that more fans would not have paid to see a contender.
I can’t see any way that he did not take wins away from his team for years at the end of his career, so I don’t think the fact that his WAR/162G is lower than Allen or Utley is just a technical issue. It’s a consequence of sustained impact on game outcomes over a five-season stretch. I liked and admired Albert as a player and will happily help vote him into the Circle. I was very distressed as I saw him lower his career averages, year after year. Like you, I was delighted to see his farewell season bring back a glimmer of his greatness.
Thanks for coming up with all the quiz questions, Doug.
For #5, I believe there are two Hawaiian-born players who exceed Shane Victorino’s WAR total.
Oops. You’re right again.
I’ll rephrase the question, and then you can provide the answer.
Charlie Hough leads all Hawaiian-born players with 38.4 WAR (over 25 seasons).
Sid Fernandez had 31.4 pWAR, which would be just shy of Victorino’s 31.5 career WAR, but his 1.3 bWAR lift him into second place with 32.7 total.
Question #4 answer: Joe Smith with 15 seasons.
In front of everyone’s nose:
Ohtani in 2024 matches Robinson and Teixeira to complete the answer to question 3.
You’re on a roll, nsb.
Vote:
Andruw Jones
Willie Randolph
Ted Lyons
The one drawback of having an inner circle dude like Pijols on the ballot is that we usually only get a week to talk about him.
So I’m voting strategically in the hope that big Albert is still around when we do this again next year.
Secondary Vote:
Drysdale
Irvin
Utley
For question #10 I came up with Ben Zobrist who had 59 such games.
Stathead has Zobrist with 68 such games, but he’s the guy. That total of 68 is made up of 59 games playing only 2B and LF, and 9 games playing those positions plus one more. So, you’re both right.
Main vote:
Ortiz
Lyons
Coveleski
Secondary vote:
Williams
Abreu
Boyer
Last round, in response to Voomo’s posts, I began to notice some unusual elements in comparing Richie Ashburn and Andruw Jones. Andruw’s CoG case rests primarily on his reputation as the GOAT among centerfielders, based on fielding excellence. The B-R standard of Total Zone runs places Andruw, at 230, way ahead of the competition in CF (Willie Mays a distant second at 176), and behind only Brooks Robinson for all positions. On top of that, of course, Andruw hit over 400 HR. Compare Ashburn – Rtot 49 and only 29 career HR – and Richie seems like a minor leaguer by comparison. I want to outline a case for viewing this comparison as dead wrong. Although I’m an advocate for weighing fielding heavily, I think there appears to be a major weakness in Total Zone figures applied to Ashburn’s era, at least insofar as we can discern some basis for Rtot in raw data available to us.
To start out with some general points: Although Andruw and Ashburn had MLB careers of generally comparable length (17 and 15 years), Andruw only had 93% as many OF innings and 86% as many CF innings as Ashburn. So it’s probably better not to make OF/CF fielding comparisons on a career total basis. Second, we need to bear in mind (I think Paul raised this point) that given the rise in strikeouts, fewer balls were put in play during Andruw’s career, lowering his number of fielding opportunities – on the other hand, seasons provided 5% more games than in Ashburn’s time. (During Ashburn’s time with them, Phillies pitchers were on the whole slightly below average in SO, while Andruw’s Braves began as a high-K team and fell towards the bottom of the league.) Third, there was a universal improvement in fielding percentages between Ashburn’s era and Andruw’s. Comparing 1955 and 2005 error rates, NL players at all positions made about 47% more errors in Ashburn’s day than in Andruw’s.
So let me start by noting a couple of figures that purely relate these two fielders to their contemporaries:
Seasons leading league in PO: Ashburn 9; Jones 6
Seasons leading league in some form of OF Range Factor: Ashburn 11; Jones 4
(The latter figure accommodates seasons where one of the two may have led in CF but not OF.)
Ashburn was a league-leading fielder in a critical category for about two-thirds of his seasons, Andruw for one-third. (In the less robust categories of Assists and DPs, Ashburn led in 4 and 3 seasons, Andruw in 3 and 2.)
As noted, we can assume CF had fewer fielding opportunities in Andruw’s day, so we can expect lower absolute numbers than we’d expect from Ashburn, but we also have to bear in mind that this mean CF was a more critical position for game success in Ashburn’s time, and in all eras until recent ones. With this in mind, consider Ashburn’s all-time top-25 leaderboard profile in these critical season-leading categories (Andruw is there for comparison, but it’s mostly Ashburn’s prominence that matters):
Most CF putouts, season (all-time top-25)
Ashburn ##1, 4, 6, 7, 8, 13, 17
Jones #9t
Most OF putouts, season (all-time top-25)
Ashburn ##2, 3, 6, 8, 9, 10, 19
Jones #11
Range Factor/9, season (since 1948)
Ashburn ##4, 17, 20, 21, 22
Jones (top mark: #101)
Range Factor/G, season (since 1948)
Ashburn ##2, 7t, 7t, 11, 12, 15
Jones (top mark: #48)
I haven’t found other leaderboards where any player is as historically dominant as Ashburn is here. For a contemporary comparison, Mays’s single-season bests for these four boards rank ##46, 57, 164, 80.
Ashburn’s low Rtot is, in part, due to the absence of the stat for the first five years of his career. During those years he led the NL in critical fielding categories every season but compiled exactly 0 Rtot. (Of course, when his career ends with a four-year decline the figures are available as debits without corresponding early-career credits.) But even once Rtot kicks in in 1953, the relation to contemporary Rtot figures – the ones Andruw is unparalleled in – is unclear. Ashburn’s career high Rtot is 20 (in 1957); Andruw’s is 36. Take a look at the underlying raw data – for Ashburn it’s complete, but for Andruw there is a ton of additional data that systems like Retrosheet and Statcast made available.
Ashburn 20 [PO 499; A 17; DP 6; E 7; Range above League ave.: CF 0.54; OF 1.05]
Jones 36 [PO 492; A 13; DP 1; E 10; Range above League ave.: CF 0.51; OF 0.88]
So far as I can see (which ain’t very far), Jones, who lags Ashburn in every category, must be gaining very sizable credits for factors beyond the categories available for Ashburn’s era, while Ashburn simply receives no credit for categories beyond data recovery.
It’s true that Ashburn committed more than twice the number of errors that Andruw did: 111 to 50. But if you normalize for equal innings and contemporary era fielding percent, the difference is not significant (68 vs. 50), about an error per year, and is easily balanced by Ashburn’s margins in Assists (182 vs. 125) and DPs (44 vs. 35), roughly four extra outs per season, after corrections.
And, to add a point I may have made before, while Andruw had fine power on offense, he didn’t have much else. Ashburn walked, sacrificed, stole bases, led the league in triples twice, and, in the end, scratches out an OPS+ identical to Andruw’s, and an oWAR figure of 58.3 vs. Andruw’s 39.8.
None of this takes anything away from Andruw. It leaves his splendid career untouched. But it seems to me highly questionable to assume that Andruw is actually the GOAT in CF rather than Ashburn. I’m anticipating counter-arguments and will be ready to change my perspective. But as things stand I couldn’t justify voting for Andruw over Ashburn. Ashburn seems to me by far the more valuable player on both defense and offense.
Bob:
Two things:
One, by looking at the obvious to question the abstruse about a player, Richie Ashburn, from the 1950s, with the aim of arguing that the abstruse fails to credit his talent at a high enough level, you are doing what I’ve tried to do over the years regarding Whitey Ford. Even to admit that there are “outliers”—the trash bin of unexpected results—is enough to raise doubts in any rational mind as to the absolute accuracy of a supposed perfect system. Another player from that era I’ve reinvestigted recently who fails the WAR test is Bob Lemon. In the 1994 Total Baseball tomb on my shelf, believe it or not, Lemon ranks sixth among live-ball era pitchers in the Total Pitcher Index (Ford is seventh). then mathematics happened, and Lemon’s current pWAR, especially as figured by Fangraphs, now barely rises above mediocrity.
Lemon, as is well known, came to pitching late, but from age 27 through age 35, a nine year stretch, he won 186 games, almost twenty-one a year on average, never falling below 17. How does WAR evaluate such consistency? The team’s manager and general manager look at the roster in the spring, compare it to the challenges facing the team in the season, see Lemon’s name, and think, well, we can assume we start with a plus 20 in the win column, a fact that WAR does not factor.
Those Cleveland teams were good, of course, with Boudreau early and Rosen later on, Doby for the stretch, and Feller, Garcia, and Wynn joining Bob on the mound, but Lemon was the anchor for two pennant winners and five second place finishes, plus two other teams that won 89 and 93. He actually compares without blushing to Warren Spahn, given their differing strengths, in everything but WAR.
My second point is cautionary: If there are outliers in a system at the low end, then they undoubtedly exist up and down the rankings.
A final comment: Bob, by my reading this is the best post you’ve ever put up, and that is saying a lot.
Tomb is as good a tome for its TPI ratings, it seems.
Thanks for the kind words, nsb. I appreciate them.
I spent time last night going over Lemon’s stats, trying to understand the factors that produce his surprisingly low pWAR and thinking about the conceptual issues involved. Some are clear enough. WAR doesn’t measure reliability, so when you refer to the impact on a GM of Lemon’s prolonged steadiness we should understand that the point is legitimate and must be supplied by us outside WAR. Others concern the degree to which the value assumptions of WAR are ones we endorse. Lemon is negatively impacted in a major way by park factor and high quality defense (as is Spahn) — that’s built into the system and we can argue whether that is legitimate or not (or the degree to which it is). I think pWAR (and, I think, fWAR even more) favors power pitchers as intrinsically more valuable than finesse pitchers, and I think there are both clear reasons to agree and clear reasons to disagree. The main task for us is to understand what’s going on with the evaluation before we reduce all that complexity to a binary choice by casting a vote.
I started to do more analysis of Lemon’s record last night, but may not have time to return to that till tomorrow. If I still have my old copy of “Total Baseball” it’s in the attic in a box. I’m unable to find online a good description of the components of its Total Pitcher Index. It would be helpful if you could extract and share a description of how TPI worked so we can use that as a way to better understand why it yields results so different from WAR.
One contrast that does distinguish the Lemon case from Ashburn’s, so far as I’ve seen to this point, is that I think that in Ashburn’s case there is a specific defect within the WAR parameters — incommensurable data across eras for Rtot — that seems to produce results too counterintuitive to rely on, whereas in Lemon’s case the issue is that the parameters themselves may disadvantage some players in a way that requires adjustment external to the system.
nsb, I’ve been doing some work on Lemon, picking up on your passing comparison between Lemon and Spahn. I took as a base Lemon’s ten consecutive good years (1947-56), which comprised virtually his entire career: well over 90%. I took ten of Spahn’s solid years for comparison, 1953-1962: his initial ten years in Milwaukee.
There’s a lot of comparability in those blocks. Spahn has a few more innings (2729 vs. 2613), but they’re close. Their W-L records are very close too, as are their ERA+ figures. More to the point when it comes to the factors WAR uses, they have identical defense quality behind them and both pitched in pitcher parks. Here are some relevant figures for the two ten-year periods:
…………Spahn……..…..Lemon
W-L…..205-118 .635…..197-111 .640
ERA+…….123……………122
RA9def…..0.17……………0.17
PPFp………92……………..96
(Just to point out the obvious, from a GM’s standpoint, looking at the huge W-L records, these are closely equivalent pitchers in a way that WAR ignores.)
Lemon can be expected to get a slight bump because County Stadium seems to have been a more extreme pitcher’s park than Municipal Stadium in Cleveland. Nevertheless, B-R given Spahn a very large pWAR edge (49.9 vs. 36.9). (The gap in narrower for fWAR, but both are lower: 37.6 vs. 32.3 – not unexpected because the FIP gap is similar, 3.43 vs. 3.76.)
What explains WAR so clearly favoring Spahn. Spahn’s edge in IP probably is balanced by Lemon’s PPFp advantage; in ERA+ and defense they are virtually identical. When I looked at HR, Lemon had a clear advantage, with a 162G rate of 16 vs. Spahn’s 22. But the FIP differential seems to be a key, and the reason simply involves walks and strikeouts. While Spahn’s strikeout rate is just a bit over Lemon’s (K/9: 4.3 vs. 4.1), the BB profiles are very different:
………….. BB/162G……….BB/9……..WHIP…….K/BB
Spahn………65……………..2.3……….1.165……..1.90
Lemon……..104…………….3.8……….1.307……..1.07
That walk differential of 39 baserunners per season may be an important element in Lemon’s disadvantage with Spahn. I also found gaps in WPA and, particularly, RE24 (I’m not sure either is factored into WAR) – these are totals for ten seasons, not rates:
………………WPA……….RE24
Spahn………..33.2……….339.02
Lemon……….30.0……….258.77
Now, your point was that both Spahn and Lemon were disadvantaged by WAR, and I suspect that’s right, because I think WAR does privilege power pitches, and while Lemon and Spahn both led their leagues in Ks, neither ever rose above fourth in K/9, and were mostly not in the top ten. So there’s more work to do, comparing this pair to a contemporary more known for strikeout rates. I’m thinking of trying Billy Pierce vs. Lemon next, although there are some complications there, since Pierce wasn’t as consistent.
I suspect that the RE24 difference is also related to the walk difference. Even just one extra walk per game can add up to a sizable RE24 difference at the end of the season. For example, if Spahn and Lemon pitch identical games all season long, except that Lemon walks the first batter of each game and Spahn retires him, that lone difference amounts to 0.65 RE24 for the game. Times 35 starts becomes 22.75 for the season. Times ten seasons becomes 227.5.
It isn’t always remembered that the ten year stretch or so following WWII was the era of the Walk. A number of batters, following the lead of Eddie Stanky, focussed on drawing bases on balls. There were the three other Eddies, Lake, Joost, and Yost (In 1951 Stanky, Yost and, Joost finished 3nd, 4rd, and 5th behind Ted Williams, who was drawing upwards of 140 free passes in the years he was whole and not at war, and Ralph Kiner at about 130/yr).Ferris Fain was another Walker of renown. Earl Torgeson.
High profile pitchers, notably Tommy Byrne early, and, yes, Bob Lemon, plus Early Wynn and Allie Reynolds throughout the era, with Bob Turley coming along late, somehow managed to overcome the handicap of putting excesses of men on base to win in big numbers.
The culture of the era, in other words, took bases on balls as a fact of life to be dealt with, not as a sin on the part of pitchers and a virtue on that of batters. I have long suspected that the umpires of the era had an overstrict notion as to the boundaries of the strike zone, contributing to the situation.
In 1958 and 1959, for what it’s worth, the CY Young recipients led the league in BBs, although with modest totals in comparison to those racked up a few years earlier.
Addendum:
Checking league totals for the period, I find that the epidemic raged far more in the junior circuit, with up to a thousand more walks per year at the peak.
Continuing the discussion about Bob Lemon, nsb, here’s what I’ve come up with comparing Bob Lemon and Billy Pierce, which may help illustrate (a little more) how pWAR deviates from intuitions (and how fWAR deviates pretty wildly!). As with the comparison with Spahn, WAR is harder on Lemon, but this time the comparison may be more apt precisely because Lemon and Pierce were pitching in the same league and the “walking man” phenomenon you describe was focused there, as you noted. Lemon is in the Hall with total WAR of 48.2, only 35.5 from pWAR (32.2 fWAR), and a 162G pWAR rate of 3.2. Pierce had 53.2 WAR (52.6 fWAR) and a 162G rate of 3.6, but never got Hall notice, maxing out at 1.9% of the vote.
The figures below show Lemon’s ten great seasons against Pierce’s best consecutive eleven (adding for Pierce in order to get the IP comparable and let B-R do the calculations with its consecutive season total tool).
………Lemon (’47-’56)….Pierce (’50-’60)
IP………..2613………………2579
IP/162……246……………….243
W-L…..197-111 .640……169-128 .569
ERA……..3.18………………3.10
ERA+…….122……………….126
RA9def…..0.17*……………..0.23
PPFp………96………………..98*
BB/162..…104………………..83*
BB/9………3.8………………..3.5*
WHIP……1.307………………1.238*
K/BB……..1.07……………….1.80**
FIP……….3.76……………….3.37**
WPA……..30.0.………………35.1*
RE24……258.77…………….288.64*
WAR……..36.9……………….46.5
fWAR…….32.3……………….46.1
WAR/162…3.5…………………4.4
I’ve asterisked the better figure in each underlying stat, and doubled those in the two related categories that I think explain the degree to which both types of WAR favor Pierce over Lemon. Using a more traditional approach, I think the Hall voters may have been noticing the factors that let Lemon compile a W-L record in his prime that was .050 above his very good team’s (~.590), while Pierce’s figure was .020 above a less outstanding team’s (~.550).
I have never thought of Billy Pierce as Hall quality and while I thought Lemon was short of that threshold, that’s because I’m a small Hall guy, not because I thought he didn’t meet the established standard. But I can understand to a degree the reasons pWAR gets where it does on these two, and I think that’s a valid way to start the discussion. (I think I also have some understanding of the reasons fWAR gets where it does, but since I think its methodology is fundamentally flawed – because it relies on excluding half the actual, existing game (the fielders), and assuming the other half is a valid index for the whole – I do not see fWAR as a valid way to start the discussion.) I am still troubled by the very low pWAR total, but part of that, at least, is simply due to Lemon’s short career. His rate is, essentially, the same as Sutton’s. (That’s another comparison I’d like to undertake, despite the difference in eras and pitching strengths.)
On element of the discussion that needs to follow on Lemon is, as you’ve noted, his astonishing, Spahn-like consistency of volume and quality over ten seasons (for Spahn, of course, it’s more like seventeen), something Pierce cannot match, though he comes reasonably close. (Apart from lower IP, Pierce had a serious off-year in 1954, the middle of his eleven-year stretch: interestingly, although he fell to 0.8 pWAR for reasons that seem obvious enough, fWAR sees it as a solid year, with a value of 3.2.)
Interesting that Lemon has the FIP edge despite more walks, fewer strikeouts and higher WHIP than Pierce. Lemon is helped by allowing 23% fewer HR, aided, no doubt, by the distant fences at Municipal Stadium (though Comiskey was hardly a homer haven, so that may be a wash). Pierce did allow fewer stolen bases (38 vs 59), but he should allow fewer being a southpaw, so probably not as significant a difference as it may appear.
The other thing to bear in mind is that Lemon got to pitch to the White Sox instead of the Indians, a not inconsequential advantage for those seasons, relative to Pierce. That said, the White Sox were hardly a pushover, winning 90+ games four times and finishing at least 8 games over .500 every one of Pierce’s seasons (in your comparison), save the first.
As to 1954, Pierce pitched 67 IP total against the league-leading Yankees and Indians, and only 45 IP combined against the second division Red Sox, Tigers and A’s. He held his own against the Indians, but struggled against the Yankees, Red Sox and Tigers. Owning the 100-loss Orioles probably didn’t help his WAR much, despite a 1.01 ERA in 44 IP.
These sorts of granular data would be really interesting to compile, Doug — I didn’t have the available time this round and I’m glad you went a few steps further. What I’d love to have a year to do is go through the game records to better understand the implications of the WPA and RE24 figures.
I think it’s actually Pierce who has the FIP edge, lower being better. I wonder about pitcher handedness and our assessments. Lefties generally have an edge in baseball because of being rarer and, in the case of batters, the extra step avoided. (Of course, they pay the price of being excluded from several high-value fielding positions.) But I don’t believe we generally penalize players for natural physical advantages. It’s all part of the arbitrary “talent” package each player gets. I think, though, that we do have a natural sympathy with players who find ways to compensate for disadvantages in their natural skill sets, which is why I appended a point about Ashburn’s fielding style in an earlier post above.
“… it’s actually Pierce who has the FIP edge“
Indeed he does. So, on second thought, Pierce’s lower FIP makes perfect sense, the extra homers notwithstanding.
Memo to self. Double-check what you think you see before heading down a rabbit hole.
Doug, I’ve written (and later misplaced) that memo many times since joining this site.
Andruw played in circumstances that did not favor being far above average in plays made, range factor per 9 innings. I checked every full year in Atlanta, (1997-2007), and the pitchers were generally average or above in strikeouts and consistently had a noticeably higher than average ratio of ground balls to fly balls. Nor was the park very big in center field. The only criticism of his rating that I’ve heard is that many of his catches may have been stolen. presumably mostly from his left and right fielders, but also from infielders, the second base umpire and a coupla beer vendors in the center field bleachers. There is a present day defensive rating system that has him noticeably lower, but I have not seen inside the “black box” enough to judge its output.
Shibe/Connie Mack had one of largest center fields of all time. I think that Total Zone takes this into account when estimating how many balls the CFer would be expected to catch, the main part of rtot, which is runs above average for each year. We don’t have GB/FB ratios from that era, but we do have goundouts/airouts which are much lower (in groundballs), than was the average; so very consistent with a very large number of flyballs available to be caught. In baseball analysis, as in life, it is very dangerous to assume that, all else is equal and we don’t have to check the context. I think that TZ does as good as we can do to place the numbers in context, and I’m not at all sure that it underrates Richie Ashburn.
KDS
This is a very strong argument, KDS. It seems to me to boil down to this: Ashburn had more room to field, a higher percentage of balls in play (because of lower strikeouts and CF HR), and a higher percentage of fly balls among balls in play.
In comparison with Jones I think these are all unquestionably true, and field size and ground ball ratio were factors I didn’t consider. (Just thinking of Robin Roberts vs. Greg Maddux gives a good picture of different GB/FB ratios.)
The first thing I did after seeing your post this morning was to look into field size — I hadn’t realized how large Shibe Park was; for Philly I guess I always focused on the small size of the Baker Bowl. CF was the corner of a rough square, with park dimensions in Ashburn’s time of:
Deep Left Center: 420 ft
Center Field: 447 ft
Deep Right Center: 405 ft
This is certainly deeper than most parks, including, of course, Turner Field. There were, however, parks with larger center fields and great outfielders, and Ashburn’s standing isn’t just in comparison to Andruw, but to the universe of center fields from the beginning of time (or shortly thereafter). For example, consider Willie Mays’s five seasons in the Polo Grounds (1951, 1954-57):
Left-center: 450 ft
Center field: 483 ft
Right-center: 449 ft
Ashburn led NL center fielders in Range Factor all those seasons, despite being four years older and having less room than Mays, who was among the top center fielders of all time. This is one path I’d use to try countering your argument. (There were other fields as large as Shibe Park, too. Forbes Field, for example, as well as Yankee Stadium in the AL — not to wave away your argument on this ground but just to indicate that its validity does have limits.)
I can’t easily respond to the GB/FB element of your argument (which is excellent), because I don’t know how to find team stats for those. If you could point me to them I’d appreciate it.
Having acknowledged the strength of your points and tried to push back partway on one of them, I’m going to point to one weakness I see. When you mention Andruw stealing put-outs from unsuspecting beer vendors and others in the outfield you are describing how, in the modest confines of Turner Field, Andruw was essentially expanding CF at cost to others who could have provided the Braves with the identical putouts. Ashburn, on the other hand, was (for the sake of argument) solely responsible for the expansive territory assigned him, which would have been too large to allow him the luxury of putout theft that did not create a net benefit to the team. I don’t really mean this as a knock on Andruw — I’m sure his teammates were content to give his superior skill room to work (not sure about the beer guys). But it does impact the force of the size-of-field argument with respect to the straight-up Ashburn/Jones comparison.
Finally, one interesting point about Ashburn. Compared to Andruw, Ashburn was a little guy and he had a comparatively weak arm. Yet he managed to produce a larger number of Assists and DPs (albeit this too could be partly related to park effects, pitching, etc.). His method for this was to play deep so that he would more often than normal be running in to make a catch, and he learned how to throw on the run, using the momentum of his motion to add speed to the throw. It was an unusual approach to center field. I mention this only to suggest that, among the intangibles (which I tend to think of as real factors that nsb will notice and WAR won’t), a contrast between Andruw and Ashburn may be that Andruw was making the most of his gifts while Ashburn was successfully compensating for a lack of them, at least in one respect. (That would place Ashburn alongside a player and sometime Phillie who he is in some other ways the opposite of: Pete Rose.)
Sorry it took me so long to reply, one version got eaten by the interwebs, before completion.
Think about the Polo Grounds; if you add together the distances down the lines to the poles, you get just over 500 ft. I don’t know of any not short term MLB park that is significantly less than 600 ft. A 300 ft fly ball is an easy out in every park, but not if pulled down either line on that side of the Harlem River. I don’t think that this was unknown 70, or 100 years ago.
With that background, I think that it is quite likely that batters in the Polo Grounds were trying hard to pull fly balls down the line, perhaps to the point that there were considerably fewer fly balls to center field then would be expected. This may explain why Mays was less far above league average in fly balls caught then Ashburn was. I also think that it would be extremely un- likely to find good evidence to prove or disprove this theory.
KDS
Heck, KDS. Even I knew about it 70 years ago. I watched the ’54 Series with Dusty Rhodes’ famous first-game pinch-hit HR just into the RF seats (off Bob Lemon). It certainly contributed to the high HR totals Giants teams racked up.
But I think the jury’s out on whether visiting players made significant changes to reach the corner seats when they came to town (hence Vic Wertz hitting straight away and into Mays’s glove at a critical moment a couple of innings earlier).
I took one step to look into this and since the result points in a direction I like I’ll report it. I looked at the ’54 Giants, who tied for the league lead in HR (with Brooklyn) at 186. Then I looked at NL parks that season to see where the most HR were hit. Three parks stood out: Crosley Field (199), Ebbets Field (193), and the Polo Grounds (187). The Redlegs (sic) hit only 147 HR, so visitors at Crosley were likely having a field day, so to speak, when they came to play — the corner distances were ordinary but overall the fences were short. You’d expect from these data that Polo Grounds visitors had many fewer home runs, since many of the 187 HR there would likely have come from the powerful home team (actually counting would take too long!).
Why wouldn’t visitors have exploited those short porches? Perhaps because their aim had to be really good — the wall did not curve, it went straight out from the foul poles, so a fair ball off by even a few degrees would become a can of corn. (Ebbets Field’s RF was, in practical terms, more inviting, at 297′ with straightaway right only 318′.) When Yankee Stadium was built for Ruth, RF was only 258′ and the fence curved towards CF. Of course, the fact that visitors didn’t (based on super-limited data) hit a lot of HR in the Polo Grounds doesn’t mean they weren’t trying.
I’m going to vote early and see whether conversation will persuade me to make a change:
The CoG project is to do better than the BBWAA, and I don’t think that playing around with strategic voting when there’s an inner-circle CoG candidate fits the mission. After the first three votes (which would be 50% of the expected vote, if this round matches last round), Albert was tied with Jones, Ortiz, and Lyons. No way I’m going to put his election in jeopardy, even though I’m a long-time Lyons booster. Last round I was uneasy about successfully advocating for Minoso over a pretty unquestionable candidate, Rolen, so I put Rolen over Lyons, and it turned out that pushed Rolen in by a vote over Lyons — if I’d reversed them the outcome would have flipped. So I’d really like to make up for it by kicking Lyons into the lead over Albert, but it just wouldn’t be right.
My argument above about Ashburn has not only convinced me that he was superior to Jones, but the evidence that he is severely underrated by dWAR suggests he should actually be an 75-80 WAR player, though I wonder whether others will discover flaws in my argument that will force me to retract that. I see no one in the Class of 1981 I expect to vote for next year, and if the BBWAA gives us three more slots to work with, I expect to advocate for Ashburn, Lyons, and Coveleski in 2026 (assuming the world is still here, and I with it), though my expectations do change as arguments are made, and we’ll see who joins from the Secondary Ballot.
My Secondary Ballot picks are Irvin, Smith, and Boyer.
Among these three, Irvin is the only one whose name I’d include on a CoG Primary Ballot vote right now. I made a detailed argument for Irvin last Round, and it was so elegant that I convinced one voter he’s truly Circleworthy: me. It moved the needle for no one else (and the same fate may befall my Ashburn brief). But I’ll hold onto my advocacy until I’m either talked out of it or the terminal futility of my vote becomes obvious.
Bob,
Fret not, mon frere. You convinced me that Ashburn was better than Jones…..
For the quiz-minded, here’s another one involving CC Sabathia.
Besides Sabathia in 2009, which other two pitchers have led their league in Wins while playing in their first season for a new team and winning the World Series?
Sabathia, incidentally, is one of only 5 pitchers with 3500 IP, 3000 K’s, 250 Wins and a .600 W-L%. The other four are all over 4000 IP and 100 WAR. If we drop the thresholds to 3000 IP and 2500 K’s, only 7 more pitchers get included.
So, while the focus this round has been on Pujols, don’t forget about casting a vote in Sabathia’s direction; his candidacy definitely deserves your consideration.
One of the two is Curt Schilling, with 21 wins for the Red Sox in 2004.
If the question was phrased “first full season for a new team” (it isn’t), then the second could also be Schilling, with 22 wins for the Diamondbacks in 2001 who acquired him in a trade deadline week deal the previous July.
Schilling in 2004 is correct.
Not counting Schilling’s 2001 season, since he started with the D-Backs in 2000. Thus, he was not playing for a new team in 2001.
I did a check of pitchers leading their league in Wins and playing for more than one team in that league. Red Barrett in 1945 is the only such pitcher, and the team he was traded to (the Cardinals) were not WS winners. It’s possible that one (or more) of the other league leaders also played for another team in the other league in the same season, but I didn’t find such a player on a WS champion. So, probably need only consider pitchers who played only on a WS winner and only in the their first season with that team.
I knew Schilling did not fit the criteria in 2001, but I thought it worth mentioning how close he came to pulling off the feat twice.
The other pitcher was Jack Morris in 1992 with the Blue Jays. In ‘91, he was two wins behind teammate Scott Erickson (and Detroit’s Bill Gullickson) in his only season with the Twins, falling just short of winning a World Series and leading his league in wins his first year with a team in consecutive seasons.
That Blue Jay season with a 21-6 record is one of 184 in the modern era with 20+ wins and a .750 or higher W-L%. Morris’s 101 ERA+ that season is the second lowest of that group, besting only Lefty Gomez’s 97 ERA+ for another World Series winner, the 1932 Yankees.
In the pWAR leaderboards, CC is right there with David Cone.
Here is a discussion about Cone that happened on these pages 13 years ago:
http://www.highheatstats.com/2012/09/lets-talk-about-david-cone/
And there is this:
https://highheatstats.blogspot.com/2012/01/best-5-year-war-by-non-hof-pitchers.html?m=1
Who knew that I had a disqus profile? Pretty handy.
Voomo,
From the article:
“A little perspective on Kevin Brown‘s 1996-2000 peak of 34.6 WAR:
From what I can tell (or, at least be told by B-R), it’s actually 39 WAR for Johnson for 2000-2004 (5 seasons) and 35.7 WAR for Santana (2004-2008)? Halladay is at 34 WAR for his best 5 consecutive seasons ?
Has there been some sort of recalculation of WAR since the 2012 article? There isn’t much batting WAR for these guys to make up the difference?
We have several similar pitchers on the two ballots for this round. So, here are their stats side by side. Or, click here for a quick peek.
To my mind, Coveleski is a fair bit ahead of the others, and Sutton a fair bit behind. Not a lot of difference between the others, and especially not between Tiant, Sabathia and Drysdale, who, again to my mind, are a smidge ahead of Lyons.
And, their best 5 year WAR, total WAR from their best 5 WAR seasons, and number of 4 WAR seasons outside of the best 5 year span (in parentheses).
41.4, 41.4, 3 – Coveleski (1917-21)
30.4, 30.4, 1 – Sabathia (2007-11)
30.0, 32.1, 4 – Drysdale (1960-64)
28.2, 34.6, 4 – Tiant (1972-76)
23.8, 28.8, 6 – Lyons (1925-29)
22.6, 27.5, 2 – Sutton (1971-75)
The two WAR totals are for the same seasons for Coveleski and Sabathia. The others had high WAR seasons outside their best 5 year span, in particular Tiant, whose best 5 year span does not include his career best WAR total (8.5 in 1968), an anomaly mainly attributable to Tiant missing most of two prime seasons (age 29-30) to injury.
Main:
Allen – RoY, MVP, best ML hitter from 1964 – 1974…
Ashburn – Bob’s posting convinced me that he’s the superior of Andruw Jones
Pujols – the only debate is whether he’s the 2nd or 3rd best 1B of all time
Secondary:
Williams
Sutton
Smith
I’m not sure of this one but I came up with Endy Chavez for question 9.
Chavez is correct.
After this round, I hope we can put some effort into a Second Chance ballot.
We have been doing this for a long time, and while I think we’ve done a great job, we have had some stacked ballots where borderline players have been overlooked.
Consider David Cone.
If he were on the ballot today, we would be arguing about his candidacy vs CC Sabathia.
However, when we did the 1963 year, he shared the space with Randy Johnson. And other holdovers included
Holdovers:
John Smoltz (eligible through 1958)
Tom Glavine (1960)
Mike Mussina (1960)
Curt Schilling (1960)
Craig Biggio (1961)
Larry Walker (1962)
Barry Larkin (1962)
Roberto Alomar (1963)
Kenny Lofton (1963)
Cone got one vote.
Yes, we will do that. Haven’t had a redemption round in quite some time.
Cone has an impressive WAR rate over 2 per 100 IP, in a group of just 46 pitchers in 1500 IP modern era careers, including Coveleski and Santana on the current ballot.
Once again I find myself in the position of needing to vote, but for a peculiar reason. To me it seems as if too many of the scant votership left have forgotten the purpose of the vote. Electing the best candidate to me ought to be the greatest priority, not to maneuver one candidate or another or a pair into prominence.
Pujols ought to be the primary choice on all ballots, simply because he dominates statistically by such a great margin. Leaving him off the ballot entirely smacks of injustice if not insult. Sorry to be so blunt.
Pujols
Lyons
Ashburn
Irvin
Smith
Dawson (but only because the ballot calls for three)
Concerning the discussion above of Bob Lemon, Pierce, etc:
Here are a pair of comparative charts from my delvings that might be of interest to your probe of the two, put in the context of two additional contemporary pitchers with better stats.
ERA+ and Six Other Stats by Innings Pitched
———————ERA+——PR———PW———RE24——WPA——-WPA/LI——REW
Whitey Ford—-133——.103———.0110——-.126———.0117——-.0101——-.0139
Warren Spahn-119——.071———.0075——-.098———.0113——-.0090——-.0109
Billy Pierce——119——.076———.0080——.089———.0103——-.0080——-.0096
Bob Lemon—-119——.067———.0069——-.085———.0099——-.0082——-.0092
Some additional comparisons:
———————WAR/IP——-fWAR/IP——-FIP———WHIP
Ford——————.0169——-.—0173—-——3.26———1.215
Spahn—————.0176———-..0143———-3.46———1.195
Pierce—————.0162————.0159———3.50———1.260
Lemon—————.0123———-.0113—-——3.81———1.337
Hope this helps to confuse the issue even more.
nsb, I want to put Ford to the side for now because I’m not up to speed on his record at the moment (I know Ford is your prime example of why pWAR is unreliable), and I think Spahn, based in the NL, is hard to compare in detail, given the relevant era and the shape of Spahn’s career.
I take your bottom line to be, essentially: Why does pWAR/IP award Lemon 24% less value than Pierce when the advanced stats you cite have him as much as 2.5% better and no more than 14% lower, and his ERA+ is identical? My initial answer is, ‘I have no idea.’
To get beyond that, one element has to be that ERA+, the WPA-based stats, and WAR are largely independent systems. As I understand them, ERA+ takes into account park factors; the WPA figures focus on situational contexts, and WAR includes elements of both these along with fielding runs (which works in Lemon’s favor) and strength of opposition. (I’m not interested in FIP because I do not think it can be applied holistically, as fWAR does; I’m interested in WHIP but it’s a traditional stat and different in kind.)
Let’s compare the key bottom line elements of WAR for Pierce and Lemon, starting with the simple stat of Runs Allowed per 9IP:
…………… RA9………RA9avg….RA9avg-RA9…….RAA/162G
Pierce……3.61……….4.24………….0.63…………………16
Lemon…..3.74……….4.21………….0.47…………………12
According to calculations that include strength of fielding, park effects, and strength of opposition in the context of the specific parks where these two pitched each game, Pierce’s slightly better RA/9 average is enhanced. He faced somewhat stronger opposition (he faced the Indians; Lemon faced the ChiSox) and did so overall in parks somewhat more favorable to hitters, which means the league average runs allowed would have predicted a higher rather than a lower RA9 than in Lemon’s case, although Lemon’s defensive support was slightly weaker. This resulted in Lemon saving runs compared to an average pitcher at a rate about 25% below Pierce’s rate (RA9avg-RA9 is my own figure, relying on arithmetic mastery). There’s a further situational factor applied to the final pWAR figure, pertaining only to relief appearances (gmLI), that works very slightly in Pierce’s favor too (I’m lost on the math of it) — it does not apply to starts, as the WPA series does, so the relevance here is low.
So I guess that now I feel I have an idea of why the magnitude of the pWAR differential between these pitchers is what it is. But it’s also true that the WPA series looks at the entire issue from a somewhat different angle and points to a somewhat different assessment. I don’t think it actually contradicts pWAR so much as it focuses with far more clarity on one important element and is somewhat orthogonal to pWAR (intersecting it only at one moment of relief appearances), though providing value within its range. The best approach would be to combine the WPA series within pWAR (and perhaps I’ve missed some wrinkle in pWAR that already does this), but the next best approach is simply to consider those figures in addition to WAR. And, of course, there are important elements that are completely outside both these systems that we need to factor in: consistency, seasonal workload, performance under pennant pressure (always for me, the Drysdale vs. Koufax factor), and so forth.
I still think pWAR and pWAR/162G are the most comprehensive and best starting points. But your scepticism is essential to our doing this exercise well. It’s really easy just to let those two figures end the discussion.
Bob:
I’ve never said that pWAR isn’t an important tool or that it isn’t the most important tool available in evaluating pitchers, but I can’t accept it as perfect or near perfect as the true measurement of what pitcher is the best in every circumstance. My skepticism goes in several directions.
Here’s the first. I’ll elaborate on the others at a later date.
When Tony Gwynn for the last place Padres in 1987 shows the highest batting WAR in the NL, I feel confident that the rating is just, since he finished first or second in the league in measurable Leaderboard stats, old and new, a dozen times. He led the league in Runs Created of a team that was desperate for runs. His .370 BA wasn’t an empty figure. He also finished in the top five in four other categories, and the top ten in two more.
When Hunter Greene for the next-to-last place Reds in 2024 shows the highest pWAR in the NL I can find almost no evidence that his performance isn’t a statistical anomaly. He appears in the top ten on significant positive Leaderboards in those six stats, Pitching Runs, pitching wins, RE24, etc., that I’ve become interested in, true, but his best showing there is 4th, and he appears on no other Leaderboards at all.
In other words, while singles hitter Gwynn was nevertheless arguably the best hitter in 1987, based on his up and down performance, despite that being the notorious year of the Home Run, the only argument for Greene being the best pitcher last year is one statistic, pWAR. And if you look back over the records of the years, you find again and again the same phenomenon, a pitcher whose pWAR is the highest recorded, or among the highest, with little to support that outcome on Leaderboards. Not coincidentally, though, these pitchers invariably plied their trade on teams ranging from poor to abysmal.
My deduction is that the way pWARais configured shows an unintended bias toward winners on losing teams.
nsb, I think the bias you detect in pWAR isn’t as unintended as you suggest. Pitchers on good teams tend to have traditional stats that are buoyed by the quality of the teams behind them, while pitchers on losing teams are weighted down. One of WAR’s primary purposes is precisely to reset the scales so that pitchers on poor teams who have unremarkable records have a chance to emerge in competition with pitchers with flashy stats on good teams. So here’s a looked at Greene’s record compared to the “runner-up” in 2024 NL pWAR, Chris Sale.
How in world could you rate Greene’s year as more valuable that Sale’s spectacular comeback year!? Sale prevails on virtually every traditional stat and walked away with the Cy Young Award (Greene placed eighth with 2% of the vote shares). Greene wasn’t even a qualifier for rate stat leader boards.
…………….IP……W-L .Pct…….ERA….ERA+…..WHIP……H/9….BB/9….K/9
Greene….150….9-5 .643…….2.75…..160………1.018…..5.7….3.5…..10.1
Sale……..178…18-3 .857…….2.38…..174………1.013…..7.1….2.0…..11.4
No comparison overall (the bold figures are, indeed, black ink). But there is one anomaly: Greene’s exceptional H/9 figure, which almost erases the WHIP differential with Sale. Zack Wheeler led the league in H/9 with 6.3 — not Greene, because he fell 11.2 IP short of qualifying. (Sale was only 15.2 IP above it, so their workloads were actually not as far apart as the W-L record suggests.)
Greene’s low H/9 figure results in the following Batting-Against slash lines:
……………..BA……..OBP…….SLG…….OPS………BAbip
Greene…. .183….. .283…… .298……. .581……… .239
Sale……… .219…. .275…… .353 …… .629 ……… .297
Sale is great but Greene is really terrific. However — reality check! — Sale went 18-3 and Greene went 9-5. C’mon! The percentage gap is what it is — Sale would get my CYA vote. But for the Batting-Against figures the difference is still only 28 IP (actually 27.1). On this measure, Green strongly outperforms Sale.
WAR basically takes the basic figures and tries to measure how much friction each pitcher was working against, in terms of the specific strength of opposing teams faced each game, the quality of defensive support behind them, and the level of challenges to pitching stats presented by the specific parks they pitched in. Because both pitchers started only about 1/6 of team games, there would be more variability in the luck they’d draw in the mix of teams they faced than with pitchers on a regular 5-day rotation (not to mention the old 4-day rotation era guys). Here are the key stats:
…………..RA9opp……RA9def….PFF
Greene……4.72……….-0.19……106
Sale…..……4.50………+0.18……100
All these work pretty strongly in Greene’s favor. He faced teams with higher run production than was the case for Sale, his defensive support was far weaker (which seems to make his BAbip figure even more remarkable), and he drew significantly more hitter-friendly parks. 106 is a very high figure, while parks were neutral for Sale. Here’s how that plays out in the final calculation:
…………… RA9………RA9avg….RA9avg-RA9…….RAA/162G……pWAR
Greene……2.81……….5.40………….2.59…………………43…………6.3
Sale…..……2.43……….4.54………….2.21…………………42…………6.2
Again, there are more minor tweaks in pWAR. I’d guess these slightly favor Greene too, since his pWAR figure relative to Sale has to withstand his lower IP total. But WAR is doing its job by calculating micro-level stats like game-by-game park factors and opponent strength, and defensive contexts. Greene was overcoming strong headwinds, while Sale was speeding in advance of a tailwind.
Now, if the counter-intuitive element here is too much to overcome (and, really, you picked a great example, because it’s a huge challenge to credit pWAR here), you can explain that statistically, in part, by appealing to your WPA series, which significantly favors Sale — that, and (I hope) not peer pressure, is why I’d still vote the Cy Young to Sale. (I’m also a little warned off by Greene’s sole non-WAR black-ink category: 19 HBP.) But even if we plunk for Sale in the end, it won’t be because of the obvious case: it will be because fine-tooth combing of the stats first revealed Greene as an unexpected challenger and then demoted him slightly for less convincing (but still excellent) leverage performance.
All of our posts about this seem to come out in strong agreement about the appropriate way to look at pWAR, despite our different subjective approaches to using it. It’s a valid starting point, perhaps the best starting point, but inappropriate to rely on without careful cross-assessments using other lenses. In terms of Circle discussions, it’s sometimes as far as I’m willing to go in outlining an initial ranking of candidate lists because of the time investment in going further. But when we’re narrowed to an argument about just two or three candidates, it’s important to widen the scope to include other stats and non-statistical factors.
Bob, a virtuoso performance in unpacking pWAR. Bravo!
Thanks for the kind words, Doug. Not virtuoso, though — I’m just testing out basics. There are many formulas that are invisible and no way for us to assess, and I’m still in the shallows on the visible ones. And it’s all nsb’s doing, anyway. If he were not hammering at the implausibility of some pWAR ratings we’d all just be following along without effort beyond, perhaps, being fWAR curious.
My votes:
Main: Pujols, Coveleski, Tiant
Secondary: Drysdale, Utley, Smith
Main: Pujols, Sabathia, Ashburn
Secondary: Irvin, Boyer, Utley
Heading into the final hours of Round 3, these are the tabulations I have to this point. We seem to have eight ballots, matching the high point for this year, so the voting might be over, but who knows? – we could have a last minute newcomer or the return of a long lost friend of the site. It looks as though Albert has locked up his Circle induction, but a single late vote could change the Secondary Ballot outcome.
In the tabulations below, I’ve indicated the weighted votes for the Primary Ballot, and placed the total number of ballot mentions in brackets – that’s the number that will determine how many added rounds of eligibility candidates will earn. As things stand, Ashburn would gain four, Lyons and Covey two each, and those with a single vote would add one round, keeping their eligibility unchanged. Simmons, who has no extra years of eligibility left right now, is at risk of slipping onto the Secondary Ballot.
By my count, valid ballots that have been cast by: Doug, Jeff M, nsb, Paul E, Richard C, Scary Tuna, Voomo, and Bob E.
Primary Ballot
16 Pujols [6]
—— 50% or more
6 Ashburn [4]
—— 25%-49%
5 Lyons [3]
4 Coveleski [3]
—— 10%-24%
3 Allen [1], Ortiz [1], Jones [1]
2 Randolph [1], Sheffield [1], Sabathia [1]
1 Papelbon, Tiant
——
0 Guerrero, Helton, Santana, Simoons
Secondary Ballot
5 Smith
—— 50% or more
4 Irvin, Utley
—— 25%-49%
3 Boyer, Drysdale
2 Williams
—— 10%-24%
1 Abreu, Dawson, Sutton
For the 1980 – Part 3 election, I’m voting for:
Albert Pujols
CC Sabathia
Todd Helton
Other top candidates I considered highly (and/or will consider in future rounds):
-Randolph
-Guerrero
-Ortiz
-Tiant
-Allen
-Lyons
-Coveleski
-Sheffield
Thanks!
For the Secondary Ballot, I’m voting for:
-Don Sutton
-Andre Dawson
-Chase Utley
When I saw opal611’s vote I thought the Secondary Ballot result had changed, but on reflection I realized I’d been wrong to say one vote could change it — I’d forgotten the tiebreaker rule.
Here’s what I have for the outcome, with 9 total ballots:
Primary Ballot
19 Pujols [7]
—— 25%-49%
6 Ashburn [4]
5 Lyons [3]
4 Coveleski [3]
—— 10%-24%
4 Sabathia [2]
3 Allen [1], Ortiz [1], Jones [1]
2 Randolph [1], Sheffield [1],
1 Papelbon, Tiant, Helton
——
0 Guerrero, Santana, Simmons
Secondary Ballot
5 Smith
—— 50% or more
5 Utley
—— 25%-49%
4 Irvin
3 Boyer, Drysdale
—— 10%-24%
2 Williams, Dawson, Sutton
1 Abreu
Apologies if I missed your answers to the remaining quiz questions, which are:
#7 – Martin Maldonado has the lowest career post-season OPS (.501) in 150+ post-season PA.
#8 – Matt Herges recorded the most post-season games (10) among retired pitchers with as many (or more) IP as games and zero runs allowed in their post-season careers.
#11 – Gary Nolan started his post-season career the same way as C.J. Wilson, with a W in his post-season debut (an 8 hit, 4 walk shutout on only 110 pitches), and only L’s and ND’s for his next nine starts. Wilson’s post-season career ended there, but Nolan had one more game, snagging a second W for the Reds’ clinching win over the Yankees in the 1976 World Series.