By In Stuff

Yo Joe! Getting to the heart

Happy September 9. On this day 15 years ago, Cal Ripken hit into his 329th career double play, breaking Henry Aaron’s record. There’s a very good chance that, before he’s done, Albert Pujols will pass Ripken.

From Brilliant Reader Ryan:

Yo Joe! Is it your intention to work the phrase “gets to the heart of” into every blog post? I’m considering turning into a one-man drinking game.


I have no idea what you’re talking about.

From Brilliant Reader Chris:

Yo Joe! I’m OK with a lot of the newer stats but think WAR is dumb.
I’ve never seen an answer to these questions:
1) what or who exactly is a replacement player?
give me a name if possible, someone who represents that perfectly.
2) what would a team’s record be if it had 25 guys like that?


I think people (and I’ve fallen for this) get too caught up in the theoretical “Who is a replacement player?” question. Generally speaking a replacement player is called that because it is supposed to be the level of player a team can easily acquire as a replacement (for, let’s say, an injury) through the minor leagues or low-level trades or mid-season signings.

But I think people tend to use this explanation as a way to bash the concept. The CONCEPT — this gets to the heart of what I’m trying to say — is simply to find a stable level to use for comparison. There is a system called “Wins Above Average” (WAA) which compares players to average, and it works pretty well when talking about the Hall of Fame (where it matters how much better a player is than the average). But if you used only WAA, average players would score exactly 0 and average players are worth quite a lot in Major League baseball. 

For instance, this year Milwaukee’s Wily Peralta scored EXACTLY 0.0 in Wins Above Average. He also has thrown 171 innings, he has a 3.84 ERA, the Brewers are 16-12 in games he started. To say he’s worth nothing is ludicrous. At the same time, if you compare him against a baseline of 0, his numbers won’t look that much better than a replacement level player. As it stands now, he has a 1.4 WAR, which shows him to be of value but not a star, which is where he probably should be ranked.

As for a team of replacement players — again it’s theoretical but I’ve seen numbers ranging from 35 to 50 wins. Like WAR, don’t like WAR, like WAR but have problems with it, whatever you think, it’s pretty clear that a team of replacement players would be awful.

From Brilliant Reader Arthur

Yo Joe! I love the game but also cringe at the prospect / memories of 4 hour extra innings epics.
How about each team has to subtract a fielder for every inning over
nine. So they play the 10th inning with 8 men on the field, the 11th
with 7, etc.


I like it — your thought gets to the heart of something I’ve long wanted in more sports: The penalty box theme. It’s my favorite part of hockey, the best penalty going in sports. You committed a penalty. You have to leave. Your team has to play shorthanded because of you. It’s the child-raising timeout long before there was a timeout, and I’m all for getting it into baseball, football, basketball, everything.

I think if football would have a penalty box for players, the game would be more interesting. An offensive lineman holds, fine, he has to sit out the rest of the series. Well maybe “interesting” is not the word I’m looking for here.

From Brilliant Reader D:

Here is a question that I’ve pondered for the last 5 or so years:

Would Derek Jeter be considered an all time great if he played his entire career in a small market like Pittsburgh, Kansas City, or Milwaukee?
My opinion is Jeter would be comparable to Barry Larkin – a Hall of Famer and local legend, but not nearly the immortal Jeter is made out to be.
I know this is probably sacrilegious to most, but I figure the inventor of Jeterate would know the answer.


Your question gets to the heart of this thought I’ve had about Jeter all year — that there has never been a player in baseball who was both so overrated and so underrated. 

I know a lot of people will chalk this up to New York — but I think it might be something else: Longevity. Jeter has just been in our baseball consciousness for so long. There is little doubt in my mind that, pound for pound, Barry Larkin was as good a player as Derek Jeter. I believe Alan Trammell was too. But Jeter got 3,000-plus more plate appearances than Larkin or Trammell — that’s six full seasons more.

So, yes, I feel sure he has been overrated by people who want to give him max grades on all the intangible things that can’t be measured reliably. People kept giving him Gold Gloves, and people will call him one of the world’s great leaders, and people will talk endlessly about the effects he had on the Yankees and baseball with his mere presence. This is Jeterating, and it’s at high pitch now that we’re at the end of his career.

On the other hand, I feel sure he’s been underrated by people who miss the sheer relentlessness of his career. He played 150-plus games at shortstop THIRTEEN times — more than any player in baseball history including Cal Ripken. On top of that, he had one season with 149 games and another with 148. He cracked 200 hits eight times. He hit .300 or better 12 times. These aren’t advanced stats, but I’m not trying to make a point about how GOOD a player Jeter was but rather how PRESENT a player Jeter was. It’s one thing to say Larkin was as good a player as Jeter. It’s another to say that Jeter was as good as Larkin with about 25% more career.

Which finally gets to the heart of your question — would Jeter be as beloved and famous and all that if he played somewhere besides New York? I say yes, assuming as he played on a good team and got some time in the postseason. Cal Ripken is considered an all-time great. George Brett is considered an all-time great. Joe Morgan is considered an all-time great. Robin Yount is considered an all-time great. Roberto Clemente is considered an all-time great. All of these players had good careers, they had long careers, they had postseason moments, and they were in smaller markets.  Sure, the pinstripes have helped Jeter’s Q Rating, but it really has been the Yankees winning — and Jeter’s constant presence in October — that has most helped his legacy.

The way I see it is like this: The Young Nomar Garciaparra was a better player than Derek Jeter. The shortstop version of Alex Rodriguez was better than Derek Jeter too. But Jeter lasted. Nomah and A-Rod, for various reasons, did not. That’s what separates Jeter, and that would be true no matter where he played.

You can email me with comments, questions, or whatever.

Print Friendly

41 Responses to Yo Joe! Getting to the heart

  1. Carl says:

    GIDP is included in both versions of WAR, and is “extra deducted” for making an extra out. That’s good as an extra out is created, and guys who don’t hustle, or don’t have speed, everything else equal, get rewarded for only making one out rather than two. However, is not grounding into a DP, just like the RBI, more of a “team event” and matter of circumstances? After all, to ground into a twin killing, one needs a man on first and certain teams get more players on first than others. Also, certain batting order positoins will come up with a man on first than other positions. Is it really proper to include GIDP as solely the batter’s fault?

    • Carl says:

      Yo Joe!, Sorry meant to start that last question with a Yo Joe!

    • It’s like any counting stat. More ABs equates to higher numbers across the board….or, as you’ve noted, more ABs with a runner on first equates to higher DP numbers. If you want to equalize, you’d need to know the DP rate per “opportunity”, I.e. Pct of DPs per AB with a runner on 2nd. You’d probably have to factor out odd DP situations where a runner creates a DP with poor base running …. But those should be small numbers. Mainly DP rate with a runner on first should cover it.

    • invitro says:

      bWAR does what you want it to do with GIDP, I think. The following is from :

      “Rdp, Grounded into Double Play Runs

      GIDP avoidance is something that batters can have a significant impact on. Slow right-handed power hitters (like say Jim Rice) will ground into a ton of double plays, while someone like Carl Crawford won’t.

      Our run vaule for non-SO outs takes into account the league average # of DP’s incurred by these number of outs, so everyone is assumed there to ground into a certain number of DP’s, but we know that this skill can allow some players to beat out the back end of the DP when other players wouldn’t.

      GIDP opportunities are any infield ground ball with a runner on first, less than two outs and at least one out is recorded on the play. The play must not be scored a hit as well.

      The difference in runs scored between a double play and a double play avoided is on average .44 runs, but it can vary by the run environment of the league. The league GIDP rate is then calculated and:

      R_gidp = .44 × ( GIDP_OPPS_player * GIDP_RATE_lg – GIDP_player)

      2011 Best: 4.7 runs by Johnny Damon, 2011 Worst: -4.8 runs by Albert Pujols.”

      • Carl says:

        Hi Invitro,

        Awesome research and explanation of the GIDP adjustment that bWAR does. Thank you very much. Joe had always counted the GIDP in the Trout vs Cabrera articles he does and I wrongly thought it was included in the WAR calc.

        • invitro says:

          Carl: thanks, but I just found the blurb, I didn’t write it :). I may be misunderstanding you, but the GIDP adjustment -is- included in WAR. This just explains how it is calculated.

      • Mark Daniel says:

        That seems like a convoluted way of determining GDP effects. Unless I’m mistaken, they are looking only at ground balls with a man on 1st and less than 2 outs when an out is recorded. Basically, they are looking at the ability of a player to beat the throw to first when a force is made at 2nd. Why do they only look at this? Why not look at all plate appearances with men on 1st and less than 2 outs? For example, let’s say we have two players (A and B). Both players had 100 total PAs with men on 1st and less than 2 outs, and each player hit into 20 DPs.
        Let’s say player A hit into 40 GIDP_opps as indicated above by B-R, and player B hit into 30 GIDP_opps. Player A would be rated as better at avoiding the DP than player B, even though they both avoided the DP equally well, it’s just that player B hit fewer grounders.

        This would seem to punish sluggers because they are often put in spots where men get on base a lot, and the reason they are put in these spots is because they are sluggers. They might just hit a home run or a double or a triple. Is hitting a home run with a man on 1st and less than 2 outs not a way to avoid the DP?

        It sounds like they are overvaluing speed here. What am I missing?

  2. Dan W. says:

    Off topic but if Jayson Stark is writing about it …

    Would it offend the baseball gods too much if every play at home plate was deemed a force play? Simply put, if runner is not in a force and he approaches homeplate (to be defined) the simple act of catching the ball and touching home plate before the runner arrives is sufficient to make the runner out. Doing this would (a) eliminate the need to tag a runner and (b) eliminate the difficulty of calling a tag play and (c) eliminate the need for there to be any collisions at home plate.

    Concerning home plate, do the current rules only apply to the catcher or to any play at the plate? If a pitcher or other fielder ends up at home plate to make a play would they also need to “provide a path” to the base? In any case, my answer is to make all plays at the plate force plays. Do this and the chance the world series is decided on a technicality is eliminated.

    • Jaack says:

      Well then the technicality could be whether or not the player was approaching home plate. If you rule it as moving off of 3rd base at all, then players won’t be able to round third base to slow down. If you offer leeway in the definition of approaching third base, now you add an even less definable aspect that’s open to just as much umpire judgement calls.

      • buddaley says:

        Easy to solve. Place a line across the base path some feet away from home plate. It becomes a force play once the runner crosses that line. (That is not to say I agree with the suggestion.)

  3. Mikey says:

    I find the whole concept of the replacement player unnecessary and confusing to the average fan. Why not just measure each player in terms of wins generated? What does it really add to our understanding of player performance to say that a player contributed 2 wins compared to a hypothetical being rather than to say they contributed 4 wins? The whole replacement player concept makes the stat less useful because we can’t know with precision how many wins a replacement level team would earn but we do know how many wins each team earned in the real world and those totals help us verify the accuracy of the stat.

    • Dr. Doom says:

      Because you CAN’T compare to zero. What is zero? 0/600 on the year? Okay, fine. Are you suggesting that a player who has 600 PAs and manages only one hit the entire year has POSITIVE value for his team? I think every baseball fan ever knows that this is not true. We all have seen “black holes” in our favorite teams’ batting orders – the player who not only isn’t helping, but is in fact HURTING. The idea is to quantify at what point that happens. WAR attempts to do that.

      • Somehow BJ Upton has a +.3 WARE despite hitting .206/.284/.327. His defense has also been terrible. So if BJ Upton is slightly better than a replacement player…. As the worst starting offensive player in baseball, a replacement player really sucks. Does that help with the definition?

          • Kris says:

            BBRef’s WAR (stupidly, IMO) includes a position adjustment in their measure of oWAR. Both versions of WAR indicate that B.J. Upton is in fact worse than a replacement level player, while both also indicate that his defense is in the range of average, not terrible. Average defense in CF is worth a good bit, which is why B.J. Upton does not rate as one of the worst seasons ever.

    • invitro says:

      “What does it really add to our understanding of player performance to say that a player contributed 2 wins compared to a hypothetical being rather than to say they contributed 4 wins?”

      If I understand it correctly, the comparison to replacement is crucial to the main goal of WAR: to allow comparison of players who play different positions, and in different leagues and eras.

      Example: Suppose Freddie plays 1B and contributes 4 wins, calculated absolutely. Suppose Rafael plays SS and also contributes 4 wins. Are they equally valuable? Looks like it. But suppose also that a replacement level 1B contributes 0.5 win, while a replacement level SS contributes 2.5 wins. Now, Freddie is “clearly” more valuable than Rafael: you could give Rafael away, replace him with a “free” player, and lose only 1.5 win. Losing Freddie would cost you 3.5 wins.

  4. Ryan says:

    Well done, Joe. I’m starting to feel a little buzzed.

  5. To question 2 —

    1) what or who exactly is a replacement player?
    give me a name if possible, someone who represents that perfectly.

    On Tango’s blog, a reader in the comments suggested Dewayne Wise as a great example of a replacement level player. He had 13 year career for 7 teams, playing multiple seasons in both the majors and the minors, never reaching 225+ PA in a single season. Career WAR (per Fangraphs) is 0.5 WAR.

    2) what would a team’s record be if it had 25 guys like that?

    The new definition of WAR allocates 1,000 “wins” to all players across all teams. If there are 2,430 wins to be earned over the course of a season (30 teams x 162 games x 1/2 as winners), and 1,000 of those wins are “above replacement”, then that leaves 1,430 “replacement level” wins to be allocated to each team. Which means a team with 0 WAR would earn 47.7 wins, 114.3 losses, for a W% of 0.294

    2004 Arizona Diamondbacks – 51-111, -0.2 fWAR

  6. Yo Joe!

    You are also forgetting that Jeter has been untouchable as far as anything deemed a vice or crime. He doesn’t gamble, drink, do drugs, get DUI’s, hit women, etc. He’s respectful of his elders which is why the old time fans like him. He dates super models which is why the younger fans like him. He appeals to everyone. The only ones that really criticize him are the hardcore fans of other teams that don’t play the Yankees a lot.

  7. Scotty says:

    Dan W: I like your idea as a way to eliminate the only occasion where there is intentional hard and potentially injurious contact in an otherwise non-contact sport. One catch I see is for the potential play where a runner changes his mind at some point after rounding third base to go back to third to avoid the play at home and the catcher gets the ball and tags the plate. Seems there would be need for umpire discretion to make a ruling on intent, or depending on how far down the baseline towards home he was.

    And would there ever be another rundown between third and home any more, or is the runner out by the fielder tagging home?

  8. nobody78 says:

    Jeter’s postseason play has to be counted as part of his career, too. He’s played in 158 postseason games — that’s another full season – and put up excellent numbers, in line with his (excellent) career regular season numbers (.308 / .374 / .465 postseason; .310 /.378/.440 career).

    Sure, neither Larkin nor Trammell had the opportunity to play in so many games, because they played on lesser teams and because, for almost all of Trammell’s career and half of Larkin’s, there was no Wild Card… but of course, that full season’s worth of play is still a huge part of who Jeter is as a player. And it makes the differences in longevity that Joe talked about even more pronounced.

    • invitro says:

      Since you bring up Trammell… apparently it is not (yet?) common knowledge that he is one of the all-time greatest postseason players ever, with a sparking .992 OPS in 13 games.

      Jeter has of course also been postseason-excellent, with a .838 OPS. And so has Larkin, at .862!

      The three players have almost identical regular-season careers (by WAR, at least). If their postseasons are included, as they should be, Trammell comes out ahead. (By how much, I don’t know.)

  9. Mike says:

    Derek Jeter has 13 seasons of 150+ games.

    Only 9 of those 13 did he play SS 150+ games. For example, in 2012, he played DH 25 times.

    Just saying.

  10. […] writing on Martin Brodeur because Joe Posnanski answered a mailbag question about Derek Jeter today that reminded me of a topic I had been meaning to write about for some […]

  11. thoughtsandsox says:

    “I think if football would have a penalty box for players, the game would be more interesting. An offensive lineman holds, fine, he has to sit out the rest of the series.”

    The problem with this line of thinking is all the pass interference calls now. It would become a pass only league because there is no way for a team to have no pass interference calls for the course of a game nowadays. So a team just has to keep throwing the ball until all the DBs are in the penalty box.

  12. Thile says:

    Coincidentally, or ironically, whichever you may prefer.

    By, fWAR, Derek Jeter is a replacement level player 0.0 WAR (.260/.308/.311).

    (By baseball-reference, Jean Segura is a 0.0 WAR player at .235/.279/.313.)

    Also, I think as nobody alludes to above, Jeter playing in so many postseason games elevates his status through more visibility as well especially perhaps to the more casual base ball fan.

  13. Matthew says:

    FanGraphs had an excellent article covering Brad Mills’ trade from the Brewers to the Athletics for exactly $1. If that’s not illustrative of Mills being considered “replacement level”, I don’t know what is.

  14. Marc says:

    A few thoughts have come to mind:

    1. I’ve always preferred Bill James’ ‘Win Shares’ over WAR – the basis is 0, but a player really has to not contribute to not count towards their wins. 3 Win Shares equals a win for the team, and it matches the actual record for the team. I’ve tried to add up these WAR figures for the players on a roster, and while they correlate to the record, they aren’t exact.

    2. My friends and I had a debate about which 3 people had the greatest “dating” resume, and we agreed Jeter was probably #1 – then I read about some side effects of that resume. I’m surprised more of it hasn’t been mentioned – the adoring press has no doubt helped keep it out of the news (one can easily search for it, don’t think I need to mention it in detail here)

    3. I’ve argued that before MLB puts a new rule into place, they should try it out in an independent league for a few years and see what effects it has, then tweak it accordingly. Without doing so, they wind up with major issues like we’re seeing this season.

    4. Doesn’t the rule in question state that the catcher can not block the plate without the ball, correct? If the catcher has the ball with the runner 45 feet away, how can the runner possibly be called safe?

    • dshorwich says:

      MLB has no power to implement rules in the independent leagues – that’s why they’re called “independent”, after all.

  15. Justin Zeth says:

    The question, “Who is a replacement level player?” has an easy and obvious answer: Willie Bloomquist. A few years ago Tom Tango wrote (in a classic Ha Ha Only Serious) that someday he’s going to change the name of WAR in The Book to WOW–Wins Over Willie. It’s actually true that WAR basically measures how much more value, at any position, a given player will give you in a year than Willie Bloomquist.

    It works wonderfully on every level–the entire reason Bloomquist has enjoyed such a long career in the major leagues is that he has always geninuely had the remarkable ability to play *every* position (except pitcher and catcher, of course) at pretty much exactly replacement level. It’s actually a very handy thing to have on your roster, because he can fill in *anywhere* without having to pull anyone up from the minors. He’s basically have a AAA roster by himself in that way. And in his entire career, only twice has he posted a WAR that wasn’t between -1 and 1.

    Wins Over Willie.

    • Thile says:

      I had not heard that before. Hilarious.

    • invitro says:

      I tried to find a player who was a better archetypical replacement than Blomqvist, but couldn’t, at least a player who played -all- positions.

      Here’s a replacement-player lineup for you. First year played is in parentheses. All of these guys played at least 10 seasons, had at least 3000 PA, and WARs very close to zero.

      C Bob Swift (1940)
      1B Larry Biittner (1970)
      2B Don Gutteridge (1936)
      SS Bobby Wine (1960)
      3B Sibby Sisti (1939)
      LF Jimmy Wasdell (1937)
      CF Roger Cedeno (1995)
      RF Randy Bush (1982)
      DH Lee Stevens (1990)
      PH Willie Bloomquist (2002)

      Swift is the champ of being replaceable: all of his 14 seasons have a WAR between -0.8 and 0.6!

  16. Brian says:

    “Brilliant” reader Chris calls WAR dumb, then admits he’s never seen answers to two questions that have been answered countless times by people who use/defend WAR. Something tells me Chris has never really cared about the answers to his questions; he just wants to dismiss WAR as being dumb.

    Am I right, Chris? Joe answered your questions. Is WAR still dumb?

    • Brian says:

      By the way, if you need the hypothetical replacement player in WAR to be represented by a specific player in order to take it seriously, then WAR was never meant for you.

      • Edward says:

        People understand the “replacement player” concept better if there are concrete examples. For your consideration, I present Don Kelly of the Tigers. He’s a super-utility guy, doesn’t start often, but is a consistent late-inning defensive replacement — basically your quintessential 25th guy on the roster. He’s also been consistently between 1.0 and -1.0 WAR every year (except for 2010, when he had 2.0 WAR).

    • John says:

      You beat me to this exact point. I was going to make the assumption that Brilliant Reader Chris doesn’t have the Google in his house. I was reminded of those young Earth creationists holding up signs asking “gotcha” questions to Bill Nye which had long ago been answered or were so far off base as to be meaningless.

  17. Edward says:

    A full season of replacement players – a) the 1962 Mets at 40-120, and b) the 2003 Tigers at 43-119.

Leave a Reply

Your email address will not be published. Required fields are marked *