Josh Hamilton and the (Slow) Demise of the “Stretch Drive”?

We’re also really pleased to announce that Dayn Perry has come on board, and will now be contributing to FanGraphs and NotGraphs. His initial post over there gives some background on why you should be excited about him joining our team.

At first blush, it seems there’s nothing particularly notable about Josh Hamilton‘s winning the American League MVP last week. He’s a popular ballplayer with a compelling personal narrative, and he had a fine season for a winning team. That’s a prototype that appeals to your garden-variety BBWAA voter. Insofar as the numbers are concerned, there wasn’t much to separate one serious AL MVP candidate from another, but Hamilton is a perfectly defensible choice and perhaps, depending upon your criteria, the best choice. None of this, however, is what struck me about his selection.

What I found interesting is that Hamilton barely played in September. Received wisdom tells us that September is the most clarifying, mettle-testing month of all — the intoxicating power of “the stretch drive” and all that. Hamilton, though, appeared in just five regular season games in September and October, the majority of which occurred after his Rangers had already clinched the belt and the title. In fact, no MVP in the era of the modern schedule (i.e., since 1961, when the regular-season schedule expanded to 162 games) has logged fewer September/October plate appearances than Hamilton did this season (obviously, pitchers who won the MVP and both MVPs of the strike-truncated 1994 season are exempted). The numbers …

Rank MVP Sept./Oct. PAs
1 Josh Hamilton, 2010 (AL) 18
2 Reggie Jackson, 1973 (AL) 63
3 Kirk Gibson, 1988 (NL) 80
4 Elston Howard, 1963 (AL) 82
5 Joe Morgan, 1975 (NL) 84
T-6 Mickey Mantle, 1962 (AL) 88
T-6 Larry Walker, 1997 (NL) 88
8 George Brett, 1980 (AL) 89
9 Frank Robinson, 1961 (NL) 90
T-10 Orlando Cepeda, 1967 (NL) 92
T-10 Barry Larkin, 1995 (NL) 92

It’s one thing to give the MVP to someone who, like Reggie in ’73 and Gibson in ’88, missed some time because of injury, but it’s another thing to give it someone like Hamilton, who logged just 18 plate appearances.

MVP standards will always shift like the dunes. One year, contention matters. The next year, pitchers aren’t allowed to win. The next year, it’s an RBI contest. The year after that, it’s a premium defender over the “left side of the spectrum” slugger. And so on and on. Likewise, some years the September numbers are given primacy (witness the arguments marshaled in favor of Ryan Howard in 2006), and other years they aren’t. But never have they been so thoroughly ignored as this. Part of it is that the Rangers had a robust lead in the West at the time of Hamilton’s injury, and as such their September wasn’t so critical. Still, though, 18 plate appearances?

In the past, some players who played regularly but struggled in September — even in the throes of a tight race — have been awarded the MVP. But this can be owing to, say, defensive contributions, misplaced notions of value based on stats like RBI and batting average, or the elevation of things like presence, leadership, and other variants of je ne sais quoi. In other words, mainstream voters can explain away a candidate who posts a bad WAR for the stretch drive. They haven’t, until now, dismissed almost total absence.

Lest it seem otherwise, I should say I think this is a good thing. Hindsight tells us that, for a team missing the postseason by a single game, the 11-4 loss on April 11 was no less damaging than the 2-1 defeat on September 30. It doesn’t seem that way at the time, certainly, but “heat of the moment” perceptions are often too clever by half.

None of this is to suggest that Hamilton’s triumph implies a shift as tectonic as what’s happened to the Cy Young balloting in recent seasons. But maybe it’s the start of something.




Print This Post



Handsome Dayn Perry can be found making love to the reader at CBSSports.com's Eye on Baseball. He is available for all your Twitter needs.


44 Responses to “Josh Hamilton and the (Slow) Demise of the “Stretch Drive”?”

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

    Good stuff. Looking forward to more.

    Vote -1 Vote +1

  2. Dirk Diggler says:

    Mauer missed all of April in ’09 and he won the MVP. Hamilton helped the Rangers lock up the division and then missed a month. I don’t see why which month you miss matters.

    Vote -1 Vote +1

    • Alireza says:

      April was a short month in 2009 because of the WBC

      Vote -1 Vote +1

    • chuckb says:

      In 2006, Ryan Howard won the MVP largely on the basis of his exceptionally strong September. Voters barely noticed Albert Pujols’s exceptionally strong early months. The conventional “stretch drive” meme has value to many MVP voters. Indeed many in St. Louis still point to the last month of 2009 in comparing Adam Wainwright’s candidacy for the Cy Young to Tim Lincecum’s. It probably shouldn’t matter much at all, but it often does.

      Vote -1 Vote +1

  3. Doug says:

    It matters because guys like Ryan Howard won the MVP based almost entirely around a very good September, where he almost certainly would not have won it if he’d had the same hot month in April and played the rest of the season the way he did April-August of that year.

    Vote -1 Vote +1

    • CircleChange11 says:

      Is that reality or your perception?

      I ask because, IMO, no one wins an award based “almost entirely” on 1 month. Even the best possible month say Sosa’s 20+ HR June) is probably only worth 1/3rd of the season’s production.

      You have to be talking about 2006.

      Howard had 58 HRs and 149 RBIs. Furthermore, he hit .313, OPS’d over 1.000 and had an OBP of .425 (108 BB’s).

      That’s not just a good MVP season, that’s an AWESOME MVP season. At the very least, it’s far more than just a hot september. Let me put it THIS way, he put up a ~6 WAR season as an average to poor defensive 1B.

      Another aspect is that Pujols missed a month due to an oblique injury (in a year when the cards were loaded with injuries).

      I prefer Pujols as an MVP that year, but scknowledge that missing an entire month is significant, especially considering Howard’s year AND that he played 159 games.

      Vote -1 Vote +1

    • fredsbank says:

      troy tulowitzki had arguably the second best september in history and only finished fifth, and a distant fifth at that.

      Vote -1 Vote +1

  4. Pat says:

    In both the Mauer and Hamilton cases, they were still the “obvious” best pick.

    Vote -1 Vote +1

    • Tom B says:

      Except Mauer wasn’t. At all.

      Vote -1 Vote +1

      • Patrick says:

        You are hilarious.

        Please, enlighten us – Are we to pick Zobrist, much of whose value came from tiny sample size defensive numbers? Or do we have another suggestion?

        Vote -1 Vote +1

  5. John says:

    I understand why BA isn’t as important as OBP, but why does everyone dismiss it completely? As far as I’m concerned, you can’t drive a runner in from second base with a walk, no matter what.

    Vote -1 Vote +1

    • oh dear says:

      OBP is not just walks. Everything included in BA is included in OBP and more, so it is a more inclusive stat.

      Vote -1 Vote +1

      • oh dear says:

        and to further my point, OBP is about not creating outs. Sure you can drive a runner in from second with a hit, but if your BA is .300 and your OBP is .300, you have a .700 chance of making an out in that situation. A player with a higher OBP would have a smaller chance of making an out with a man on second base (way more important than actually driving him in).

        Vote -1 Vote +1

      • John says:

        I do have to say thanks for not being condescending. I understand that. And I also do believe OBP is more important than BA, I just don’t like how many people now completely discount the stat. To be honest, I’d rather have a .303/.356/whatever than a .248/.375/whatever. This might be the case with everyone (maybe the averages are a little TOO far apart), I don’t know. What do you think, oh dear?

        Vote -1 Vote +1

      • oh dear says:

        it would be really hard to have that kind of OBP with such a low average.

        When reading up on more advanced thinking in baseball, you will come across “scarcity”. The most scarce thing in baseball is outs. You want to do everything possible to avoid making them. The natural effect of sending batters to the plate that do not create outs is runs.

        Whether they do it by walks or hits is (i think) largely irrelevant. Obviously there will be hitters that can be more effective than just being out or on-base by hitting for power. In turn you would look towards their SLG rather than their batting average.

        BA is just too limited, as it only shows one simple set of data. How many times you got a “hit” (which is a small scope of results from an at-bat) over your at-bats. Considering OBP and SLG (separately, OPS is pointless) will give you better chance of guessing which player might excel in different situations.

        I don’t agree with completely disregarding BA, but there is not much point in looking at it to determine anything other than “this guy got some hits”. When there are other stats that are more inclusive and less restricted they should be used.

        It obviously doesn’t stop at OBP or SLG… you go as far down the rabbit hole as you deem necessary to come away with a conclusion. Only considering BA would be like looking at the opening of the hole without looking inside it.

        Vote -1 Vote +1

      • oh dear says:

        so in your example, a runner on second base…

        with 2 outs, you would be better off having a hitter with a higher batting average (and other peripherals to match) as they are more likely to get the hit that will allow the runner to score.

        with less than 2 outs… you would likely be better off with the player that won’t end the inning faster.

        i’m not sure if the math works out in that scenario (things are funny like… you are more likely to score from 1st with no outs than from second with 1 out) but its a decent enough example.

        Vote -1 Vote +1

      • Paul says:

        can’t reply to the reply posted by oh dear regarding how difficult it is to have a lowish average and high OBP

        but an average of 250-260 and a obp of 360-370 (i.e. low AVE, with an excess of 100-120 points in OBP) is not that difficult to do, plently of sluggers (Dunn/Ortiz/Teix/Jose Bautista had that type of year) do this most years, it’s a typical Swisher/Dunn/Cust/last few years Thome type line – it’s not common, but there is a ‘type’ of player that sort of line fits

        It wouldn’t be a surprising year from guys like Uggla, Weeks who can talk a walk, but are average worries each year

        Vote -1 Vote +1

    • Christian says:

      Unless balls 3 and 4 are wild pitches.

      Vote -1 Vote +1

    • Travis says:

      The basic philosophy is that anything a batter does that does not result in an out, should be considered a success.

      Linear weights do show that singles are more valuable than walks. However, BA does not show the type of consistency between player-seasons that walks do. Many studies have been done that show a player’s BA is subject to luck much more than OBP. As a single datum, check out Ichiro’s year-by-year BA.

      So between the philosophical premise of “don’t make an out” and the understanding that BA can hide a lot of luck, it just doesn’t make much sense to use BA.

      In addition, BA conceals a lot of contributions of a hitter. At a basic level, a hitter does two things: doesn’t make outs, and gets extra base hits. BA doesn’t really convey that information better than either OBP or SLG, so we use the latter 2 as a cheap shortcut.

      Vote -1 Vote +1

  6. Mr Punch says:

    If you’d gone back one more year, to 1960, you’d have come to NL MVP Dick Groat, whose wrist was broken on Sept. 6.

    (BTW, Frank Robinson’s ’61 MVP award was in the NL.)

    Vote -1 Vote +1

  7. CircleChange11 says:

    Received wisdom tells us that September is the most clarifying, mettle-testing month of all–the intoxicating power of “the stretch drive”

    Tulo disagrees.

    Vote -1 Vote +1

    • oh dear says:

      Tulo hits at home in the second half, and that’s it. The rest of his season is too debilitating to keep him in serious MVP contention He’s never finished higher than 5th in MVP voting.

      Vote -1 Vote +1

      • CircleChange11 says:

        I was not advocating that Tulo be an MVP candidate.

        I was illustrating another example of the “stretch drive” not being important.

        Adam Wainwright lighting it up in 2009, while Lincecum faltered big time is yet another example.

        I think sometimes things are attributed to the media that are nothing more than a Strawman or saberist’s wishful thinking.

        Vote -1 Vote +1

      • oh dear says:

        I think we are in agreement.

        Trying to figure out “what the media is thinking” is pointless because they prove time and time again that they don’t, in fact, think.

        Vote -1 Vote +1

      • oh dear says:

        but he is a great example of a stretch run player not being rewarded. ryan howard basically won an MVP on a september performance, so the mood of the media seems to be more determinate than expecting them to vote the same way twice.

        Vote -1 Vote +1

      • CircleChange11 says:

        In regards to Howard, the point was brought up earlier.

        Not only did he have a good september, but he also had a traditional line of …

        .313-58-149

        His slash stats are .313/.425/.659

        There aren’t too many years when that doesn’t win an MVP.

        Considering Pujols missed a month with an oblique, and he trailed Howard in HRs and RBIs (still important stats even if they are not the most revealing).

        Take out the name “Ryan Howard” and there’s little to dislike about the season as an MVP winner.

        37 IBB as well.

        Even Howard’s wOBA (.436) and Pujols’ (.448) aren’t likely enough of a disparity to overcome the additional GP, and other counting stats like HR and RBI.

        If one were to say that Howard’s 58 HRs were the reason he won, I’d agree … and add that’d be the case in almost every year.

        Howard’s August 2006 was .348-14-41. He also had 13-35 in May.

        His 2nd half was .355-30-78. But the award has never been about real month-to-month consistency.

        Vote -1 Vote +1

      • RoyaleWithCheese says:

        “Adam Wainwright lighting it up in 2009, while Lincecum faltered big time is yet another example.”

        Lincecum faltered? His FIP was 2.97 in September 2009. In August, it was 2.94.

        Vote -1 Vote +1

      • chuckb says:

        @ Royale —

        the traditional numbers — wins and losses and ERA tell a different story for the final month of ’09. That’s what many Cards’ fans point to as the injustice of Wainwright’s missing out on the ’09 Cy Young.

        Vote -1 Vote +1

      • CircleChange11 says:

        You’re making the assumption that FIP is IT for pitchers.

        To me, it’s not. I would never use FIP on its own to evaluate a pitcher’s performance, and the reasons why are obvious.

        Look at runs allowed over the last 2 months. No defense is that bad.

        Same deal with Cliff Lee’s horrible 7 start stretch THIS year. His FIP allowed him to zoom from 4th in fWAR to 1st, while he fell way down the rWAR list.

        IMHO, what you bring up is more of an issue about FIP than it is about how well/poorly Lincecum pitched.

        Again, please don’t use FIP in a manner as if it trumps everything else. It doesn’t.

        Vote -1 Vote +1

      • RoyaleWithCheese says:

        “Look at runs allowed over the last 2 months. No defense is that bad.”

        Fine. Lincecum allowed 25 runs over the last two months of ’09. 24 of them were earned. His ERA was 2.82. I still don’t see what you’re getting at.

        I’m aware that Wainwright supporters pointed to his 2-3 record in September, but that was the only argument they could make.

        Vote -1 Vote +1

  8. Jose Bautista's 2010 says:

    “If one were to say that Howard’s 58 HRs were the reason he won, I’d agree … and add that’d be the case in almost every year.”

    I like the way you think…

    Vote -1 Vote +1

    • CircleChange11 says:

      *grin*

      Big reason why I said “almost every year”.

      Cecil Fielder and mark McGwire would agree.

      Vote -1 Vote +1

  9. Dwight Schrute says:

    Well the guy that finished second, Miguel Cabrera didn’t do a whole lot in September either due to injury(he missed the last week or so and played poorly before that because of the injury) so that may have helped Hamilton’s case. Also Robinson Cano who finished 3rd had a rather lackluster September as well, posting a .650OPS.

    So I think if either of them 2 put up a great “stretch drive” it may have made the voting more interesting. As it stands Hamilton may have actually benefited in the voters eyes by not playing because his main competition for the award performed so poorly in September.

    Vote -1 Vote +1

  10. Scott says:

    I think the lack of a playoff challenge in the AL in September helped Hamilton. If say there had been a 3rd team running with the Rays and Yankees the whole month I think maybe someone else could be there.

    Vote -1 Vote +1

  11. Mike says:

    I wonder how many folks voted for Hamilton over Cabrera simply because Josh had a higher batting average (despite Cabrera have a higher OBP).

    Vote -1 Vote +1

  12. Adam says:

    I think you may be overlooking a possible other reason: by the time September rolled around, the Rangers had already run away and hid in the West, so there WAS no stretch drive in the common sense for Texas, so everybody had already tuned out. People were paying attention to Texas in June when they were shocking everybody atop the West, not in September, as they had no real competition down the stretch. Guys (right or wrong) get looked at down the stretch because they are locked in more dramatic races.

    Vote -1 Vote +1

  13. I wouldn’t count the lack of plate appearances in the last two months against him or the way people voted for the MVP.

    Injuries and stuff happens like this all of the time. If you asked him, I bet he would have wanted to play every day.

    As a manager I would have kept him out a few games against the Mariners and A’s.

    Vote -1 Vote +1

  14. CircleChange11 says:

    I think that “correlation” and “causation” are being confused in this thread.

    Playing on a division winner is a BIG factor in MVP, in general. naturally, performance down the stretch run will be associated with that.

    But, IMO, the examples being given in support of that idea, could also be in support of MANY ideas.

    For example, Ryan Howard could have won the MVP because he led the league in HRs, RBIs, and hit .313. He could also have won it because he had outstanding season stats. His May numbers were pretty much the same as his August numbers. His September numbers were pretty similar to June numbers.

    What’s the “stretch run” September? August? 2nd Half?

    IMO, there are assumptions being treated as realistic facts when little evidence shows causation.

    Vote -1 Vote +1

  15. CircleChange11 says:

    I think the more accurate idea would be “humans have short memories”. What happens at the end of the season is more “fresh” than what happens earlier in the season.

    It’s not so much that the “stretch run” counts for more, as it it is our brains remember the most recent stuff more vividly.

    People keep bringing up Ryan Howard’s 2006 as “proof”, but ignoring Lincecum’s 2009 or Tulo’s 2010. There’s probably numerous examples of good/bad Septembers winning the CYA or MVP.

    Ryan Howard di NOT hit 58 HRs and drive in 149 runs due to one good month. Let’s stop being silly. I know it’s Ryan Howard, and I know it’s “RBIs”, but still … let’s not lose our objectivism or act as if there’s never more than one deserving candidate each year.

    Vote -1 Vote +1

  16. Dwight Schrute says:

    “I think the more accurate idea would be “humans have short memories”

    I think that pretty much sums it up. For example ESPN.com asked a poll question a few weeks ago(it was a day or 2 after Mike Vick exploded on MNF) asking either “Who is the best QB in the NFL?” or “Who would you want as your QB on your team?” and it included your usual suspects like Peyton Manning, Brees, Brady and I think Aaron Rodgers along with Vick and Vick actually had the most votes. Now if you asked that question 2 or 3 weeks prior Vick probably wouldn’t have even bid on the ballot let alone winning it but since he had a few great games that were fresh in people’s minds he did.

    Vote -1 Vote +1

  17. CircleChange11 says:

    Ask that same poll question NOW, and it might not be either Vick nor Manning.

    Then ask it again in two weeks.

    You always know the answer to a poll question based on what news story preceeded it. *grin*

    Baseball is not as bad as football in this regard, because there’s a game every day. In the NFL there’s a game and then 6 days to talk about it, so the really good days and the really bad days get blown way out of proportion.

    In one day the Bears went from a fake 7-3 to the best team in the NFC. One. Game.

    Amazing how perceptions of Brian Urlacher change every few games. Here’s how you can tell if Brian Urlacher is good … Does he have good DT’s playing in front of him? If the answer is no, then he’ll be blocked by OL and FB all year and they’ll talk about how he’s over-rated. If the answer is yes, and he’s allowed to run sideline to sideline and use his best asset (speed), then he’ll make the pro-bowl. Same guy, different playing situations.

    Anyway, we were saying something about short memories?

    Vote -1 Vote +1