Taking Out Jason Motte

Last night, Jason Motte gave up two bloop singles and took the loss, but likely even more frustrating for the Cardinals not-closer closer, he had to watch the runs score from the dugout. After Ian Kinsler dumped one into no man’s land and then Elvis Andrus punched one into shallow right-center field, LaRussa was essentially faced with four decisions:

A. Leave Motte in to face Josh Hamilton, hoping that a non-healthy Hamilton would be overpowered by his fastball, and then have Motte match up against right-handed batters Adrian Beltre and Michael Young.

B. Intentionally walk Hamilton to load the bases, set up the force at home and a potential double play, and give Motte three consecutive right-handed batters to try and retire – Young, Beltre, and Nelson Cruz.

C. Replace Motte with Arthur Rhodes, get the left-on-left match-up against Hamilton, and then go to the bullpen again for another RHP (in this case, Lance Lynn) to go after Young and Beltre.

D. Bring in Rhodes to face Hamilton, but hide Motte somewhere in the field for that one batter so that he could return to face Young and Beltre.

LaRussa chose option C, and of course, it ended up not working out very well for St. Louis. Was there a better option that would have been more likely to help the Cardinals keep the lead, or at least not head to the bottom of the 9th down by a run?

Option A has some pretty obvious problems, as Motte/Hamilton isn’t a match-up that is likely to end with a strikeout – the result that St. Louis really needed with runners at second and third and no outs. Motte’s career K% against LHBs is 20.4%, while Hamilton’s career K% versus RHBs is just 16.5%. Even with his groin injury, he’s maintained his ability to get wood on the ball, and against an RHP, contact was a pretty likely outcome. Neither Motte nor Hamilton have strong GB tendencies, and Hamilton would clearly go up to the plate looking to hit the ball to the outfield.

However, Option C – the one that actually went down – had many of the same issues. While Hamilton’s strikeout rate against LHPs jumps to 22.1%, Rhodes K% against LHBs this year was just 16.1%. His career numbers are much better, but he’s not the same pitcher he was a few years ago, and Hamilton had hit an outfield fly against him the night before. Perhaps even more frightening should have been Rhodes FB% against LHBs this year, which was a staggering 55.8% – his prior ability to get lefties to hit the ball on the ground hasn’t showed up much in 2011.

Even considering the platoon advantage, it doesn’t seem obviously clear that the Cardinals were more likely to strikeout Hamilton with Rhodes on the mound, nor were they less likely to have him hit the ball in the air. Beyond just the fact that he’s left-handed, in fact, it doesn’t seem that bringing Rhodes in to face Hamilton actually increased the likelihood of any of the outcomes the Cardinals should have been trying to maximize in that situation.

If having Rhodes face Hamilton didn’t incur an obvious significant benefit over having Motte face him, then Option C seems demonstrably inferior to Option A, because having Motte face Hamilton means that he gets to stay in the game to face Young and Beltre. In order to justify the downgrade in the right-on-right match-ups that were essentially inevitable and also likely to be of extreme importance, you’d have to get a pretty massive bump in expectation of striking Hamilton out or getting him to hit the ball on the ground, and I don’t see that bringing Rhodes in did either of those things.

So, that leaves options A, B, and D. Option B has a lot of positive benefits – platoon advantages with Motte versus three RHBs, and any ground ball at an infielder likely keeps the run from scoring and might even get you a double play. However, it comes with the cost of putting an additional runner on base, and then an extra base hit could get that runner home and potentially expand the lead even further.

However, Hamilton’s potential third run has pretty significant diminishing returns due to the late nature of the game. By the time Andrus scored the go ahead run, the Cardinals win expectancy had already dipped to 18.3% – it would have fallen to just 7.8% had the Cardinals finished the inning with a third run scored, but Andrus scoring lowered the Cardinals odds from 32.5% – stranding him was a lot more important than stranding Hamilton.

Additionally, WPA doesn’t account for the fact that the Cardinals were going to send up the bottom of their batting order in the 9th inning, and asking Nick Punto, Skip Schumaker, and Rafael Furcal to generate offense against Neftali Feliz was likely to end in failure. While scoring two runs off Feliz was less likely than scoring one, they really needed to end up in a situation where they didn’t have to score any, and they could only do that by stranding Andrus on base.

Walking Hamilton and having Motte face the three right-handed bats almost certainly gives you the best chance at stranding Andrus. You get to keep your best reliever on the mound, get him the platoon advantage in every at-bat, set up a force at home, and give yourself a chance at a double play on a well struck grounder. You might not be able to strand Kinsler without a lot of luck or top notch pitching, but Motte has performed well enough that he should have been trusted to keep Andrus from scoring, even against three good RHBs.

Option D would have perhaps been the most interesting to watch from a fan’s perspective, as the decision on where to put Motte would have been fascinating. The usual answer is left field, but when you’re facing a guy who is desperately trying to hit the ball in the air to set up a play at the plate, you don’t really want inferior outfield defense at that moment. Even though Motte’s a converted catcher and has a great arm, you probably don’t want to ask him to catch a pop fly or make an accurate throw to the plate if Hamilton was able to aim the ball to left.

The best option might have actually been third base, as left-handed batters rarely hit ground balls to the left side – Hamilton’s career GB% to the left side is just 18.7% – but then you’re essentially inviting Ron Washington to execute a squeeze play. Could Hamilton have gotten a bunt down far enough up the 3B line to make Motte field it? Could Motte field a ground ball from that angle and make an accurate throw home? It would have been fantastic theatre, but I can’t say that I would have had the stones to pull that move off in the World Series either.

To me, the best option was almost certainly putting Hamilton on and leaving Motte in to face the following trio of right-handed bats. The cost of the third baserunner is outweighed by all the positives that come from keeping Motte on the mound, and it would have given the Cardinals their best chance to enter the bottom of the ninth in a tie game instead of needing a run to keep playing.

Playing the match-ups can be a good idea, but taking out a good RHP for a mediocre LHP – especially with right-handed bats as far as the eye can see set to follow – seems like overmanaging at it’s worst.

Print This Post

Dave is a co-founder of USSMariner.com and contributes to the Wall Street Journal.

67 Responses to “Taking Out Jason Motte”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Aaron (UK) says:

    I like option D! When was the last time anyone did that in a game, let alone a playoff game? However overall agree that B is better.

    Vote -1 Vote +1

  2. Rich Mahogany says:

    I didn’t even realize that D was possible.

    Vote -1 Vote +1

  3. Rice Cube says:

    Re: option D, Lou Piniella did this at least once with the Cubs. I know Charlie Manuel was forced to use Roy Oswalt in an extra-inning game in LF. Then I think the Astros did it at least once with Wesley Wright this season.

    Vote -1 Vote +1

  4. TheGrandslamwich says:

    I agree that B is the best option (and during the game I was shouting at the TV when Motte was pulled), but after checking Lynn’s splits vs. RHB’s the move is slightly more understandable.

    Vote -1 Vote +1

  5. Mark says:

    If anyone has the stones to do option D, it would be TLR. One other problem with that: he had already used so much of his bench, I think he only had one guy left. To sacrifice him just to get one platoon advantage would be a tad decadent, especially if you are essentially banking on at least one run scoring and likely having to play extra frames. Maybe you go for it on the road, but not at home.

    Vote -1 Vote +1

  6. Brett says:

    Dave, to get around Motte fielding a squeeze, could TLR have temporarily moved Pujols over and put Motte at first? As an athletic pitcher, I don’t feel like he would subtract a ton of defense (though I may be underselling Pujols’ defense/positional adjustment), and would likely have a clear angle to the plate, should he end up fielding a ground ball. I understand that this downgrades defense at both positions, but Hamilton would certainly try to hit one in the air, as Pujols would have likely reduced the probability of a squeeze.

    Vote -1 Vote +1

  7. Patricio says:

    Option D – Why not make Motte the catcher? He has experience there, and it’s not like he has to worry about controlling the running game for that at bat. That would’ve been awesome!

    Vote -1 Vote +1

    • joseph says:

      Collision at the plate – do you really want Motte there over Yadier? (Plus, Molina was due to lead off the inning, so you don’t want him out of the game.)

      Vote -1 Vote +1

  8. Nestor Zucchini says:

    The Cardinals were actually asking Yadier Molina, Punto, Skip, and Furcal to generate offense. For whatever that’s worth.

    Vote -1 Vote +1

    • Jason B says:

      Speaking of Molina – was anyone else flabbergasted that they brought in Gerald Laird to pinch-run for Yadi? Under the guise that he’s faster?? Really?!? Laird looks like he runs a 7.5 forty (best-case, like if running toward a Dunkin Donuts).

      Vote -1 Vote +1

  9. Mark says:

    Why not bring in dotel instead of Lynn? Also does walking Hamilton force the use if a pr for him?

    Vote -1 Vote +1

  10. Telo says:

    Perfect article, great analysis. D would’ve been really awesome, but B was most definitely the right decision.

    Vote -1 Vote +1

  11. Ben says:

    Remember, this is LaRussa we’re talking about here. He’s the *inventer* of the LOOGY. It’s not surprising to see him go with the kinds of strategies that got him here.

    Vote -1 Vote +1

  12. Erik says:

    I think you put Motte at shortstop. That way you leave Freese in the game for offense and you have a 3B playing 3B in the event of a squeeze. I don’t have data to back this up, but I’d think that Hamilton hits few gb to SS, and that with the infield in, the likelihood of a real SS making a play is reduced, so you’re not sacrificing much (the likelihood of an opportunity is small).

    The catcher idea is a fun one, but good luck trying to convince Rhodes to throw down in the zone with a man on third and a closer behind the plate!

    Vote -1 Vote +1

  13. Ryan says:

    Nice article, thanks. Herzog used to put Worrell in the outfield to bring in a lefty to get an out.

    One other thing about this analysis: it assumes a pretty big difference between Motte and Lynn vs. RH. It’s hard to say exactly what the difference is between the two because Lynn hasn’t pitched much. He does actually have a higher K rate than Motte, though. His FIP isn’t too far off of Motte’s either. So, if Rhodes was a good option against Hamilton (and, here, I agree with you, not La Russa), I think the move makes sense.

    I doubt the moves are hugely different in terms of expected runs. Ultimately, the Cardinals need to score more than 1 run.

    Vote -1 Vote +1

  14. CircleChange11 says:

    I think in ANY situation where there are runners at 2nd and 3rd with no one out, you’re in deep poop. In the discussion, it should be accepted that we’re discussing a “Houdini Situation”. Some of the discussion has implied that there were good chances for StL to someone get out of the situation without losing the lead.

    Walking Hamilton to load the bases with one out is the situation that puts the most pressure on the pitcher. Not only can he not bounce a pitch, but now he can’t walk a batter, hit a batter, etc … he basically has to challenge 3 consecutive good hitting RHBs. He’s got the fastball to do it, but the numbers say that cards give up ~2 runs in that situation.

    Arthur Rhodes entered the game with 1st base open, meaning that you can nibble Hamilton a bit. He threw a belt high slider in the middle of the plate. Most likely Rhodes wasn’t aiming for that spot.

    The relievers came in and got all 3 of the batters they faced out. That’s pretty exceptional given the situation.

    The bad part is that runs scored on two of those plays.

    Again, we’re acting like the cards had some good chances at getting out of that inning without losing the lead. They did not.

    If Motte fails, then everyone says TLR should have brought in the lefty, and on and on.

    Vote -1 Vote +1

    • Xeifrank says:

      I agree it was a “pick your poison” situation. Likely not enough of a difference between any of the top three scenarios to roost anyone.

      Vote -1 Vote +1

    • Brad Johnson says:

      We’re also dismissing a potentially important component entirely.

      Having given up two hits, could TLR or Dave Duncan have decided/seen something to suggest that Motte “wasn’t on.” Remember that TLR doesn’t necessarily trust Motte much. There’s been a fairly strong and obvious case to be made that Motte should have been the Cardinals closer since sometime in mid-2010 yet TLR has been utterly resistant to the idea until recently.

      Vote -1 Vote +1

      • Brad Johnson says:

        I say that as someone who only watches 10-20 Cards games a year (aside from the postseason). Motte’s stuff always looks untouchable to me but I could just be catching his good days.

        Vote -1 Vote +1

    • Gregory H says:

      I agree with your assessment of the game and LaRussa’s quandry.

      Like Dave Cameron wrote, I would have stayed with Mott, walked Hamilton to load the bases, and then hope that Mott fanned the next two hitters. Mott was far more likely to record Ks than Rhodes or Lynn, and striking out the next two hitters was almost necessary to avoid any runs crossing home. One of the cardinal rules of baseball is never to intentionally walk the potential winning run. In this case, the potential winning run was already on 2nd base. Since the infield had to play in anyway, having a force-out at home is the lesser of two evils, and at least with Hamilton out of the way, Mott faces only RH batters. But LaRussa had a Hobson’s choice, and he went with the platoon split matchup. I can’t fault him for doing so. If Rhodes had executed better pitches, and with the way Hamilton had been swinging the bat against lefties, the Cardinals may have gotten out of the jam, and LaRussa would have looked like a genius.

      When this was unfolding last night, I wondered if LaRussa intentionally walks a healthy Josh Hamilton, as opposed to going after a hitter who is obviously ailing and has looked awful at the plate in this Series.

      Vote -1 Vote +1

      • Xeifrank says:

        Yeah, there are plus and minuses all the way around. The intentional walk also forces a run across the plate on an unintentional walk to one of the next batters. Can’t imagine there is much of a change to the win expectancy either way here.

        Vote -1 Vote +1

  15. Scout Finch says:

    The Cards bullpen doesn’t really offer much, does it ?

    Vote -1 Vote +1

  16. Jeff C. says:

    It’s incredible that statheads are criticizing a manger for NOT intentionally walking a batter! It’s almost always the other way around. Also, you have to give LaRussa credit for being willing to pull his closer in the 9th inning. Most managers would have stuck with their closer good or bad, win or lose. The fact is, it was a no-win situation. The result wasn’t bad, it was just unfortunate for the Cardinals that the first sac fly also got Andrus over to third. Imagine if they had walked Hamilton and then Texas put 3+ runs on the board. Then everyone would be here criticizing LaRussa for the intentional walk. I think if you force the opponent to hit two straight sac flies to beat you, that’s the best you can hope for and you just have to tip your cap to them. LOTS of teams fail to get runners home from third with less than two out. Texas had to do it twice, and move a runner from second to third to boot.

    Vote -1 Vote +1

  17. Richie says:

    Walking the bases full with no outs is typically a horrible percentage move. As shown by what happens to batting average when the sacks are loaded. It goes up bigtime. Silly to put a sabermetric spin on the issue, and ignore that elephant.

    Vote -1 Vote +1

    • los says:

      Hitters don’t suddenly get better with the bases loaded. Major selection bias since crappy pitchers are more likely to load the bases than good pitchers.

      Vote -1 Vote +1

      • Richie says:

        Crappy pitchers often leave after they’ve loaded the bases. A la what (would have) happened here.

        You can check out the numbers. Hitters get immediately suddenly mucho better, as they now get much better pitches to hit. Walks go unsurprisingly down, BABIP up and slugging through the roof.

        Vote -1 Vote +1

      • williams .482 says:

        Bad pitchers ge replaced when they load the bases … with other bad pitchers (in many cases at least). and most likely in front of a poor defense the whole time.

        Vote -1 Vote +1

  18. Richie says:

    Taking out your best reliever and replacing him with your worst reliever just to get a platoon advantage against a hitter who doesn’t have a big platoon split? Stupid. Tony was just engaging in habit-driven behavior. As we all typically do under time constraints.

    Vote -1 Vote +1

    • todmod says:

      I agree with the sentiment (mainly that Rhodes is not that great of a lefty specialist to be worth the hassle), but Hamilton definitely has a solid platoon split.

      107 wRC vs lefties career, 146 vs righties.

      Vote -1 Vote +1

  19. Brandon says:

    Don’t trust Option D. I remember the Braves did that a few years ago, and Chris Resop ended up making an error in LF..

    Vote -1 Vote +1

  20. bowie says:

    I would have tried to get Hamilton to chase a bad pitch and if he didn’t chase let him walk.

    Vote -1 Vote +1

  21. joe says:

    Two bloop singles? I know it’s a minor point but do you really need to spin the details to make your point.

    The first hit was obviously a bloop – the Andrus hit was a liner to CF; it was hit hard enough that even a fast runner like Kinsler couldn’t score from 2nd in it.

    It’s a minor detail, but you seem to do this somewhat frequently… you couch/spin some of the details to fit your conclusion when it really isn’t necessary at all.

    Vote -1 Vote +1

    • johnefive says:

      neither hit was well hit at all

      Vote -1 Vote +1

      • J Walter Weatherman says:

        I really hope you don’t believe that. Two separate FanGraphs writers praise Andrus’ “liner” and PA. And despite his oddly-worded first sentence, Dave even seems to give proper acknowledgement to the second single by noting that it was “punched.” Kinsler’s single was by all means a bloop, but it is wholly inaccurate to lump Andrus’ hit into that bucket.

        Vote -1 Vote +1

  22. nate says:

    Option D is interesting, but you leave out a major factor: Jason Motte’s blurry vision (http://goo.gl/pWHeL). He claims that he can’t stand contact lenses, so he doesn’t wear them.

    Vote -1 Vote +1

  23. Rob says:

    It’s not about whether the cards had a good chance, its about whether their manager gave them the best chance. He did not do that.

    Vote -1 Vote +1

  24. MGL says:

    Wow. Without actually applying some NUMBERS to each of those options, it is impossible to know which one is correct (yields the highest WP for the Cards). Each of us can have an “opinion” on which one is correct, but without NUMBERS, I am afraid that opinion isn’t worth much.

    I’ve been doing these kinds of analyses for 25 years and I have no idea which one is correct. I suspect that D might be, but you/I would have to figure out how much the bad “D” in the field costs in WP as well as removing another player (decent chance for a tie game and extra innings). It is really complicated to figure all this out, but it can be done (approximated at least, so we have SOME idea as to which option might have been correct).

    I am completely agnostic as far as walking the bases loaded. Normally you never do that with 0 outs (other than perhaps in the bottom of the 9th in a tie game), but here, I don’t know. I don’t think (no, I KNOW) that Dave or anyone else knows without “running the numbers.”

    I’m also not sure why Dave is obsessed with the Cards trying to make Hamilton hit the ball on the ground. I have no idea whether that would be better or worse than a fly ball. Lots of fly balls don’t score the runner at third (short ones of course) and lots and lots of fly balls don’t move the runner to third. Same thing with ground balls. Some are base hits, some move both runner, etc. I don’t remember if the IF was playing in or not (probably not), but even if it were, I’m STILL not sure whether a fly ball or ground ball would be better. IOW, I’m not sure whether I want a GB or FB pitcher to pitch, everything else being equal.

    Also, I am very uncomfortable when an analyst gets to choose which sample he wants to present to support his point or his opinion. This year only? Last 2 years? 3 years? Career? Lately, as in last half season? You should not be allowed to do that, for obvious reasons (cherry picking your evidence makes your arguments intellectually dishonest, or misleading at best).

    For example, Dave said this:

    “While Hamilton’s strikeout rate against LHPs jumps to 22.1%, Rhodes K% against LHBs this year was just 16.1%. His career numbers are much better, but he’s not the same pitcher he was a few years ago, and Hamilton had hit an outfield fly against him the night before.”

    Yes, he is not the same pitchers, but if this year his K% was higher than his career numbers, Dave would obviously be quoting his career numbers (heck, I would to if I had the choice!). The analysts should NOT have the choice. He should always be quoting a projection which is some kind of weighted career average!

    And for the last part of that last sentence, about Hammy hitting a fly ball the night before, David should get immediately thrown into the MGL jail. I can’t believe he even said that in that context. Shame on you Dave!

    “…but then you’re essentially inviting Ron Washington to execute a squeeze play. ”

    I think there is about a zero chance of Hamilton squeezing just because Motte is playing third. That should not even be in the analysis unless you want to use it as a tie breaker in a dead heat.

    Finally, Davis didn’t even mention one of the most important part of the puzzle, and one that made LaRussa’s decision likely awful. Lynn is a terrible pitcher! I don’t care how he has done lately (has it been good). I an other projections experts have him as near replacement level! If you know that he is going to bring in Lynn, as opposed to say, Dotel, then it is a no-brainer not taking out Motte (or putting him in the field).

    So while David definitely brought up most of the relevant facts in order to determine which option was best, I don’t think that any of us is any closer to the answer….

    Vote -1 Vote +1

    • CircleChange11 says:

      I think there is about a zero chance of Hamilton squeezing just because Motte is playing third.


      If the 3B has to field a non-pop squeeze play, then the runner is safe.

      The only way you get the runner out on a squeeze is if the batters misses the bunt, pops it up, or bunts it right back to the picther who comes home quickly.

      No one would squeeze with 2nd and 3rd no outs. Seriously, 99% risk, 1% reward (numbers represent my opinion magnitude, not research). Anybody that would squeeze in that situation deserves to have it popped up to the pitcher and a triple play turned. Oh, the #3 hitter was up.

      The thing that invites a squeeze play is “having a left-handed pitcher on the mound” especially if he “goes from the windup”.

      Vote -1 Vote +1

  25. bothstillplaying says:

    Why use CAREER numbers for Motte and Hamilton, but switch to 2011 numbers (and just 33 IP at that) for Rhodes…..when someone uses a “55.8% FB% against LHBs this year” and LHBs had only 53 ABs against him all season, I become nervous…….talk about small sample size bias???…..while with TEX, Rhodes had 27 GB and 55 FB out of 105 BF, probably where most of that 2011 FB% came from…..since joining STL, before last night, Rhodes has had a more normal for him 15 GB and 13 FB from 41 BF, including 9 K & 4 BB…..and since Sept 9, the last 21 batters he’s faced have 1 H, 2 BB, 6 K, 8 GB & 5 FB……we well know the relevant time frames we use when looking at fantasy pitching stats, so why wouldn’t LaRussa place more weight on these current data points, especially when they are more consistent with Rhodes’ career stats, than the brief, but irrelevant, time spent early this year with TEX?

    Vote -1 Vote +1

  26. LarryTheBerry says:

    To me, pitching to Hamilton with Motte would have been the correct choice because of his injury, if he’s not hurt that changes everything. Of the choices Dave listed, I personally have no idea which is best for WE. Thus, while I disagree with the move TLR made, I don’t have as much an issue with it as I did with what I felt was his absolute worst move, which was to bring in Lynn instead of Dotel. I questioned it during the game and still can’t understand why Dotel was passed over.

    Also would have loved to see option D, that would be so awesome, even if it didn’t work.

    Vote -1 Vote +1

    • Geoff Buchan says:

      Dotel is an extreme fly ball pitcher, while Lynn gets more grounders than fly ball outs (at least so far in his tiny MLB sample). So if the FB out is more dangerous, that might make one tilt to Lynn. Also, Dotel had pitched in game 1, and might not have been as well rested.

      Lynn’s minor league numbers were actually pretty good in AA, and so-so in AAA. I suspect the projection systems may consider Lynn a starter, where he has to pace himself, and what he did in the minors, not a reliever. So sure, his MLB career sample size is tiny, but what numbers he has there are actually pretty good.

      And doesn’t research show the pitcher has an edge when facing a team for the first time? Think Matt Moore’s first ALDS start against Texas for an extreme example (albeit with a *much* more talented pitcher). So if anything, Lynn may be more likely to pitch better than his true talent level (and his MLB regular season stats may well reflect that).

      What you need most is a strikeout, and Lynn’s strikeout rate in 2011 was the same as Dotel’s. Mitchell Boggs and Jake Westbrook were also in the bullpen, but I’m guessing Westbrook might not have been able to get ready fast enough, and, although he’s an *extreme* GB pitcher (good, if fearing the SF is correct), he’s a poor strikeout pitcher. Boggs doesn’t strike out people as much as Dotel.

      It’s not obvious to me that Dotel would have been a clearly superior choice. You do want a strikeout badly, but an unknown Lynn may have been just as likely to get that as Dotel, while being more likely to get a ground ball.

      Vote -1 Vote +1

  27. MGL says:


    Don’t know if you read the post above yours, but an analyst should not be using any old time frame that he wants to support his thesis. He needs to use a time frame which is “generally accepted in the sabermetric community” which is usually some weighted career numbers. Using last 3 or 4 years is also acceptable if they are weighted. Even then (using accepted principles), there is, unfortunately, some wiggle room.

    Choosing to use one years worth of stats to support an argument is definitely not acceptable. Using career stats is a little lazy (it should be weighted), but I suppose that is OK, as long as the analyst’s argument does not rest on that evidence too greatly.

    And, as I said, using one PA for anything, even mentioning it, is a capital crime!

    Vote -1 Vote +1

  28. MGL says:

    I know we have all heard and read in the media (and from some analysts) that because Hammy is injured, the hard thrower would have been the best choice, but is there any evidence that that is true whatsoever. Is it even rational to think that? Don’t you think TLR knew that Hammy was injured and Motte throws hard?

    And yes, Dotel appears to be much the better pitcher against RHB (he has extreme platoon splits, probably because he is mostly a fastball, slider pitcher).

    In a thread on The Book Blog, someone mentioned that Lynn was a highly rated prospect by John Sickels (B+). I pointed out that he has terrible projections from ZIPS, Pecota, Oliver, and me. Not that any one of these (Sickels or any of the forecasters) has a high level of certainty…

    Vote -1 Vote +1

  29. bothstillplaying says:

    Michael (MGL)….while Dave switched from career to 2011 when the data suited him, my point was that Rhodes’ 2011 was an even super smaller sample size (33 IP) than most might expect……and to carry it further, bifurcate Rhodes’ 2011 season between TEX and STL and you get polar opposites…..then split the STL part into halves and Rhodes generates a stat line not unlike Motte’s 1 hit in the last 10 innings pitched, just for Tim McCarver & Joe Buck’s benefit of course…..but then Dave took it to the extreme and committed the capital offense in your eyes…..so the same data “proves” Dave’s case or vindicates TLR, just depends on how you slice it!

    Vote -1 Vote +1

  30. MGL says:

    Right, if you are allowed to split the samples up anyway you want, you can probably support just about any thesis from one end of the spectrum to the other. Which is why a standard must be used. As in all scientific fields, in sabermetrcis there is a generally acceptable standard in the industry – weighted career (or to simplify last 3 or 4 years).

    That is not an arbitrary method mind you. When we are trying to answer questions such as “who should be used in an upcoming situation,” we are essentially asking the question, “How do we expect so and so to perform at some time in the future, in most case as in this, the immediate future, as in the next PA or tomorrow?”

    To do that, again, the accepted standard in the industry, after years of very thorough research and analysis, is to use a “Marcel-like” projection for component rates, GB and FB frequency, platoon splits, etc. It is also accepted standard to ignore things like clutch, home/road splits (other than the normal one of course), day/night, pitcher/batter historical matchups, hot and cold streaks, etc. Not because we KNOW that these don’t exist but because we find, again, after years of thorough research, that even if they exist, they have little predictive value.

    So I implore all analysts, including David, who is a fine one, to use these standards when presenting a thesis. If time or other constraints exist, which I understand, then some semblance of these standards should be used, or some qualifications issued, rather than disingenuously using one year or other similarly small and/or misleading samples (such as un-weighted career) in order to support an argument.

    Vote -1 Vote +1

  31. Travis Touchdown says:

    I think option A is the best, with the caveat of telling Motte to try his best to avoid throwing a pitch in the strike zone.

    Hamilton is not a very patient hitter. He may have struck himself out if Motte throws pitches out of the zone

    And if he happens to hit the sac fly, well the game is only tied and Motte is still in there to face some RHP’s. Still not a disaster.

    And if he does work the walk, well then you are in the Option B state, so it still is an acceptable outcome.

    Vote -1 Vote +1

    • J Walter Weatherman says:

      Not sure I would call Hamilton K-prone, as you suggest. He may not be “patient” in that he has a very high K%, but he doesn’t walk very often for a player with his power. Compare to Joey Votto and Jose Bautista over the past 2-3 seasons (all near 18%). While the Joeys walk at a much higher rate, I wouldn’t feel too good about the chance of K for any of these hitters.

      Vote -1 Vote +1

  32. TK says:

    Any manager would do D every once Ina whole. In the world series? Not stoned. Boulders.

    Vote -1 Vote +1

  33. hk says:

    My problem with this article that the premise the author used in determining which option was best was which pitcher had the best chance to strike Hamilton out. While a strikeout is the goal, I think the manager also has to look at issues like which pitcher had the best chance to prevent Hamilton from getting a hit or prevent Hamilton from moving Andrus to third or, maybe more importantly, which series of pitchers gave STL the best chance to escape the inning with the game tied. If STL could have survived the 9th tied. To me, the fact that TLR went with pure platoon split advantages and his relievers got three consecutive outs justifies the decision to go to Rhodes to face Hamilton and a RHP to face Young and Beltre.

    As others have pointed out, if I am going to question TLR for his management of the 9th inning, I would question his use of Lynn over Dotel more than his use of Rhodes to face Hamilton.

    Vote -1 Vote +1

  34. jj3bagger says:

    Bobby Cox in ’08 had Chris Resop play LF for a batter, brought in lefty Royce Ring to pitch to a lefty (Adam LaRoche) and then brought him back into pitch to the next hitter. That was the first time it happened since ’93.


    Vote -1 Vote +1

  35. MGL says:

    A back of he envelope calculation:

    A pitcher fielding a position (very poorly) is a -50 runs per 150 games fielder. That is around .009 runs per PA. You could probably leverage that by putting him at a position that is not likely to get a batted ball. So, call that .005 runs loss.

    If you can have one reliever who is even .5 runs per 9 inning better than another reliever (say, Motte over Lynn), and that is not much (Motte could be 1 run per 9 better than Lynn), that would be a savings of .013 per PA, or around a .03 savings if Motte pitches to 2 or 3 batters rather than Lynn.

    So it definitely seems worth it to put your best pitcher in the field for one PA if that pitcher is even .5 runs per 9 better than the alternative and even if he pitches to only one more batter, let alone 2 or 3.

    Vote -1 Vote +1

    • CircleChange11 says:

      So it definitely seems worth it to put your best pitcher in the field for one PA if that pitcher is even .5 runs per 9 better than the alternative and even if he pitches to only one more batter, let alone 2 or 3.

      In this case it would likely be LF with Hammy at the plate.

      The problem after that is that StL has (I think) Laird and Theriot on the bench with all RHBs coming to the plate.

      From an entertainment standpoint this option would be incredible to see.

      Vote -1 Vote +1

  36. CircleChange11 says:

    I think you have to start the scenario by asking “What is our goal?” or “How many runs allowed is acceptable?”, given that we still get to bat in the 9th. Are we trying to get out of this with the lead? Tied? Down by one?

    Here are the run expectancies for the situation: Runners on 2B & 3B, no out: 2.052 runs. Results aside from this are against the odds.

    The vast amount of data suggests on average the result at the end of the inning is TEX 2, StL 1, which is likely not favorable to StL given that Feliz is on the mound and the bottom of their lineup is coming up.

    StL has an above average reliever on the bump. TEX has 4 above average batters coming up (and a well above average batter after that). How does this affect the run expectancy? I don’t know.

    In these situations I view K% and IFF% as being equal outcomes. Both events get an out without the runners advancing.

    We also don’t know what % Hamilton’s injury changes things.


    OPTION A – Leave Motte in.

    Hamilton v. RHP: 110 tOPS+
    Hamilton v. LHP: 77 tOPS+

    Motte v. RHB: 69 tOPS+
    Motte v. LHB: 153 tOPS+

    I don’t think you allow Motte to pitch to Hamilton. If he does he has to K him or get him to pop up. Kinsler is speedy so a routine fly ball scores him, and Hammy is a pull hitter probably especially pull on the ground so a GB likely scores Kinsler and moves up Andrus, where runner on 3B and 1-out yields .983 runs.


    OPTION B – IBB Hamilton

    Run Expectancy:

    2B + 3B
    0 Outs: 2.052
    1 Out: 1.467
    2 Outs: 0.634

    1B + 2B + 3B
    0 Outs: 2.417
    1 Out: 1.650
    2 Outs: .815

    If you IBB Hamilton, you are putting the odds more heavily in TEX’s favor. I don’t like it. My main objection is that it gives the P no wiggle room, a walk is unacceptable.


    OPTION C – Bring in Rhodes.

    Hamilton v. RHP: 110 tOPS+
    Hamilton v. LHP: 77 tOPS+

    Rhodes v. RHB: 77 tOPS+
    Rhodes v. LHB: 111 tOPS+

    Rhodes is a better option than Motte, IMO. I think you send him in there with the instruction to keep the ball down and paint corners (yeah, easy to say). The worst case tolerable scenario is an unintentional walk if you can’t get him to chase.

    Rhodes came in a threw a pitch where I’m confident he did not want to, elevated slider in the hear of the plate. Basically he gave Hamilton a fly ball.


    OPTION D – Rhodes in the game, Motte in LF

    It’s interesting as it give StL the best platoon advantages, but anything hit to Motte is going to have to be right at him. He could reasonably make a blunder such as first step in on a flyball and botch it and 2 runs score and now a runner on 2B or 3B, still no outs.

    THEN, who comes into LF when Motte returns to the mound? Theriot? Laird?


    I think you pretty much do what TLR did. Bring in the LHP given Moote’s and Hamilton’s splits. If they get Hammy out without a run scoring, then you IBB to load em up and bring in Lynn looking to get a ground ball that gets you out of the inning with the win. It’s a rabbit our a hat situation for sure, but that’s what you’re looking to get with a 1-0 lead and 1-out.

    As it was, StL relievers retired all 3 batters they faced, which should be viewed as a big plus. StL was down by one with a fighting chance. If the relievers give up hits, the inning could reasonably ended being down by 2 or more runs.

    I’m still curious as to [1] how many situations like this end up with zero runs scored (or 1 run scored) and [2] how they did it. I wish I knew how to find that information?

    Vote -1 Vote +1

    • CircleChange11 says:

      Also, on a fly ball with Hammy at the plate and Kinsler on 3rd, the OF’s have to decide whether to try and throw him out at the plate, or try to keep Andrus from taking 3rd. They tried to get Andrus at 3rd, and even with a good throw he was safe.

      If Hamilton gets out, and you IBB Young, the run expectancy only increases from 1.467 to 1.650.

      That’s a reasonable trade off to me, as a manager, given that I’m now trying to get out of the inning with the win, and am really in no worse situation that I was before.

      With 2-outs, the run expectancies are either balanced between the O and D or shift to the D, if we only get one out instead of a DP.

      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>