Respect, Measured

Perhaps the greatest right-handed hitter in the history of Major League Baseball, Albert Pujols inspires fear into the hearts of opposing pitchers and managers alike. In the past 10 years, he has led baseball in intentional walks, and not by a small margin either — he has 72 more intentional walks than second place (Vladimir Guerrero). This is not an insignificant sum. Because the value of an intentional walk is about a 10th of a run, Pujols’ 251 career intentional walks are equivalent to about 2.5 wins. That means on average, Pujols gains about 2.3 runs per season purely out of managerial fear.

This postseason has fared no differently for the Cardinal’s legendary first baseman; he’s already gotten  seven intentional walks. But this tells us nothing about how pitchers approach Pujols — intentional walks don’t contain information about instances where pitchers simply pitch around him. To gain a greater grasp of the Pujols fear dynamic, we should visually examine how pitchers approach him:

This graph shows the difference between two different distributions. The first being the pitches thrown to Pujols from 2008 to 2011. The second is a random sample of 15,000 pitches thrown to right-handed batters in 2011. Comparing the two sets of pitch locations allows us to see where the league is pitching to Pujols, with the context already accounted for. Blue locations indicate areas where Pujols was pitched to less than an average right-handed batter, and red indicates locations where Pujols was pitched to more than right-handed batters. The graph is from the catcher’s perspective, so the left side of the graph corresponds to the inside of the plate for right-handed batters and the right side of the graph corresponds to the outside portion of the plate to right-handed batters. The dotted box represents the strike zone.

As you can see, pitchers work very hard to avoid the middle of the plate — and pitches that are up. Instead, they throw more pitches to Pujols that are down and down-and-away. Neither of these results is surprising, but it’s always nice to have visual analysis confirm what we would intuitively expect. Of course, there are other methods to quantitatively assess pitchers’ approach to Pujols.

Zone % — more complicated than it seems

A simple way — at least ostensibly — to look at how the league pitches Pujols — compared to the rest of the league — is zone %. This is the percentage of pitches that are within the strike zone. But how do we define the zone? Are we talking about the rulebook strike zone, or the strike zone that umpires really call? Are these two zones even different? These questions introduce some subjectivity into zone %.

According to Mike Fast’s research, the rulebook strike zone and the called strike zone are not interchangeable. In the interest of accuracy, I’ll calculate zone % using these updated strike-zone definitions and PITCHf/x data. I find that the average zone % to right handed batters in the sample is 50.9 % and that the zone % to Pujols, across the past four years, is 46.7%.

But this does not consider the fact that Pujols is pitched to in a different distribution of counts than the league-average batter. This is significant because pitchers don’t throw the same proportion of pitches within the zone for every count:

 

There’s generally an inverse relationship between the polarity of the count (from the pitcher’s perspective) and the amount of pitches thrown in the zone. In other words, when pitchers are ahead they throw more balls, and when they are behind they throw more strikes. This means that the distribution of counts that Pujols bats in might have a significant effect on his zone %. In fact, if Pujols is in more batter-favorable counts than average, then his zone % is actually going to be inflated because pitchers throw more strikes when they are behind, versus in other situations.

To account for this, I made the distribution of counts the same for both Pujols and the sample. Once we do this, we find that the new, count-adjusted zone % for Pujols is 46.1% and that the adjusted average is 52.3%.  This difference of 6.2% is larger than the initial estimate of 4.2%, albeit by a small amount. The zone % method does not give us the level of granularity of the visual method used earlier, but it does offer us one nice, overall number.

Despite pitchers’ greatest efforts to avoid giving Pujols anything to hit, he still finds ways to mash and get on base at a rate that few others can match. At this point, it’s very unlikely — if not inconceivable — that pitchers will ever develop a strategy to deal with him.

References and Resources

*PITCHf/x data from MLBAM via Darrel Zimmerman’s pbp2 database and scripts by Joseph Adler/Mike Fast/Darrel Zimmerman

*Research by Mike Fast on the actual called strike zone

*Tango’s linear weights




Print This Post



33 Responses to “Respect, Measured”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Sean says:

    Forgot about Bonds?

    Vote -1 Vote +1

    • Adam W says:

      Bonds was a lefty.

      Vote -1 Vote +1

    • Anon21 says:

      But he did forget about Aaron. And Mays. Basically, you can’t sensibly place Pujols in the all-time pantheon until you see his decline phase.

      Vote -1 Vote +1

      • test says:

        Hence the “perhaps”. First word of the article. And Pujols certainly could end up being the greatest RH hitter ever. He is by several measures right now, just not career value. And don’t over-estimate the decline pahse either – most greats maintain a high level, and play less, so each season carries less weight.

        Vote -1 Vote +1

      • chuckb says:

        He qualified it with a “perhaps.” I would agree that Pujols isn’t there and likely won’t ever be there but he’s got to be in the conversation.

        Vote -1 Vote +1

      • adohaj says:

        decline phase hahahaha you assume much

        Vote -1 Vote +1

      • DCN says:

        Jimmie Foxx! Or, more recently, Frank Thomas.

        Pujols definitely could end up being in first place, but Thomas through the same age was pretty comparable. A little better OBP, although surprisingly far behind in homers.

        Vote -1 Vote +1

      • Anon21 says:

        adohaj: I assume nothing more than that he’s a human being, whose reaction times will decline as he gets older. That’s pretty damn parsimonious, since it’s a pattern demonstrated by every human being who ever got old.

        Vote -1 Vote +1

      • test says:

        You can’t leave Pujols out of the conversation of best ever just because he hasn’t retired yet. If a RH hitter was going to become the best ever, his career would look lot like Pujols so far.

        Yes, lots of guys have fallen off later (Foxx, Thomas among those who had a shot at this entirely imaginary title), and Pujols might (probably will?) too. But as of this point in his career, Pujols is better than Mays and Aaron had been, at the plate at least. Unless you prefer career shapes like Bonds – who doesn’t show up on the “as of age XX” leaderboards on baseball reference for OPS+ and batting wins until age 36 – this is what the all-time great careers look like.

        Vote -1 Vote +1

  2. JayT says:

    Even though he (was made to) quit five years ago, Bonds still has 88 more intentional walks over the last ten years then Pujols does.

    Vote -1 Vote +1

    • Yes, Bonds has more walks. Sorry, the numbers that I referenced were for qualified players only.

      Vote -1 Vote +1

      • Yirmiyahu says:

        Odd to use “qualified players” for a counting stat.

        Vote -1 Vote +1

      • Yirmiyahu says:

        BTW, how did you figure an intentional walk is worth 0.1 runs? According to Tango, it varies by player. He calculates the value of an intentional walk as being equal to a given player’s wOBA. In other words, the value of an IBB is equal to the player’s average plate appearance.

        In Pujols case, he’s produced 662.7 batting runs over 7433 PA’s. So each IBB would work out to 0.089 runs.

        I’m curious if you used this method and just rounded, or got the 0.1 number from elsewhere.

        Vote -1 Vote +1

      • Tanner Scheppers says:

        There is no need to build a dubious stat (most IBB among all players that have played in all of the past 10 seasons!) to fit the “Pujols is awesome and gets free passes” narrative.

        Vote -1 Vote +1

      • Yirmiyahu,

        The linear weight value of .1 for intentional walks is from the page linked to in my very last reference. I just used the average value.

        Vote -1 Vote +1

  3. Yirmiyahu says:

    Remember earlier this year, when people said that Pujols was finally human and/or was entering his decline phase? He had a .722 OPS as of May 29, and was pretty much ignored for the remainder of the regular season.

    Including the postseason, he’s hit .330/.400/.640/1.040 since May 30. 487 PA’s.

    Vote -1 Vote +1

  4. You’ve got apples and oranges problems here.

    “To account for this, I made the distribution of counts the same for both Pujols and the sample.”

    One of the reasons that Pujols has more favorably counts is that his strike zone is smaller than that of the average right handed hitter, ironically because of “umpire respect”. Counts are actual data, but pitch F/X is hypothetical data. There should be some sort of adjustment made. Your conclusion blends two different types of data, count (actual) and F/X (hypothetical) to come up with a result.

    Vote -1 Vote +1

  5. Matt says:

    That location graph is sweet, thanks!

    Vote -1 Vote +1

  6. Brad Johnson says:

    Small nitpick, 2-1 and 3-2 are also considered neutral counts, at least by pitching coaches.

    Vote -1 Vote +1

    • barkey Walker says:

      So… is this optimal? Where does Pujols hit his home runs?

      Vote -1 Vote +1

    • For purposes of simplicity, I defined neutral as balls = strikes. Batters perform better than average in both 2-1 and 3-2 counts.

      Vote -1 Vote +1

    • f says:

      How is a 2-1 count neutral? The batter has the complete advantage here. He looks for one pitch; if he doesn’t get it, he doesn’t swing. If the pitcher misses the strike zone, now it’s 3-1. If there’s a called strike, it’s only 2-2. Not sure why pitching coaches would think that. It’s a fastball count for the majority of pitchers out there.

      Vote -1 Vote +1

  7. adohaj says:

    I wonder if they started pitching down the middle and up on Pujols if he would stop hitting as good as he does. Well until he adjusts in a week.

    Vote -1 Vote +1

  8. jake says:

    If an IBB is worth .1 runs and 10 runs is worth 1 win, then it would have made since to IBB Barry Bonds every time he came up to the plate from 2001-2004.

    And the math is not even close. 47.8 WAR vs ~25 WAR (plus defense and baserunning added in, which wouldn’t influence the final number much).

    Vote -1 Vote +1

    • DCN says:

      How do you figure?

      Tango made a pretty good chart in 2003 on when it actually made sense to walk Bonds using WE data.

      Vote -1 Vote +1

      • aweb says:

        .1 runs *600 PA ->60runs, or about 6 WAR. I’m assuming that’s the rough estimate he did there. Something is wrong, since by that measure it makes sense to IBB a lot of players everytime. But I’m not sure what exactly…I assume IBB’s have a lower value than normal walks due to situational use? That is, they aren’t optimal pretty much ever (Bonds aside), but they are handed out at the least worst times?

        Vote -1 Vote +1

      • DCN says:

        Wins in the context of WAR don’t really equate to wins in the sense that he means walks are 0.1 runs (~0.01 wins). I’m sure you understand WAR – how many runs a player nets his team above the replacement (“AAA+”) player at that position.

        The linear weight 0.1 for IBBs, on the other hand, is a WPA-base stat. It’s how much the action helps or hurts the team’s expected run production, based on accumulated situational data. 0.1 is approximately the average WE for intentional walks across all situations where they’re issued.

        So you’re replacing the event, not the player.

        Of course, RE data isn’t tailored to particular players. So the value of an intentional walk over the replacement Pujols at bat is going to be less than 0.1.

        In fact, there still will be seasons where Pujols’ RE24 – run expectancy above average based on those tables – per at-bat will be high enough to merit intentional walks based on those linear weights, if you use the simple formula of RE24/PA. The average cost of pitching to Pujols in RE was higher than the average cost of an intentional walk.

        Even so, that only speaks for walking him in the situations where intentional walks were empirically issued. If you start walking him every at bat, you approach the RE value of non-intentional walk, because those walks are distributed across different situations. The RE value of a non-intentional walk is about 0.33, so if you walk Pujols every PA he starts putting up 200+ RE24 a season, which would make him easily the best player of all time. Most Bonds ever got to was 127.

        But you do bring up a good point – when you start replacing more plate appearances with IBBs, then the value of an IBB is going to change.

        And this brings me to another point – Weinstock, are you sure that he’s gaining 2.3 runs a season on those intentional walks? Because if you give him those plate appearances that the intentional walks replaced, he’s going to outperform the run expectancy because he’s Pujols. You’d have to subtract the expected increase in RE from Pujols batting from that figure, no?

        Vote -1 Vote +1

  9. cs3 says:

    So what this article is really saying is that Barry Bonds garnered more than 3 times as much respected as Albert Pujols…
    Since 2001:
    Bonds 368 IBB
    Pujols 251 IBB

    and the part that really matters:
    Bonds 9.4 PA/IBB
    Pujols 29.6 PA/IBB

    Amazing how much more feared (and how much better) Bonds was than ANY of his contemporaries, Pujols included

    Vote -1 Vote +1

    • DCN says:

      Intentional walks to Bonds, career: 688
      Intentional walks to #2 in history (Hank Aaron): 293

      Intentional walks to Bonds in 2004: 120
      Most intentional walks in a season by a non-Bonds player: 45 (McCovey).

      You have to figure some of the 2004 IBBs are spite walks rather than respect walks, but that’s still ridiculous.

      Part of it is his plate patience. You aren’t likely to get him to chase a bad pitch and get an out if you want to pitch around him, so you might as well save your arm, give him the free pass, and not risk making a mistake and getting punished.

      Vote -1 Vote +1

  10. eric_con says:

    Whoa, this was some fantastic work!

    Vote -1 Vote +1

  11. strtobill says:

    Pujols is NOT the leader in intentional walks over the past 10 years.
    Barry Bonds had 333 IBB from 2002 thru 2007, easily surpassing
    Pujols.

    Vote -1 Vote +1

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *