When David Ortiz has Been Locked In

For a hot minute, used to be the story of the World Series was wacky finishes. More generally, it was overall wackiness, taking into account some defensive blunders. But then we were treated to a more or less clean and conventional Game 5, and now the clear story is David Ortiz, and how he’s presently un-get-outtable. I mean, I guess the real story is how the Red Sox are on the verge of another championship, but as far as players are concerned, Ortiz is the guy. He’s the main guy on the Red Sox, and he’s thought to be the main focus of the Cardinals.

In case you haven’t heard, so far Ortiz has had one of the most productive World Series of all time. He’s got 11 hits in 15 at-bats, and that doesn’t include a grand slam he had taken away by Carlos Beltran, which left him with a meager sacrifice fly. Always a presence, right now Ortiz feels like either a dream or a nightmare, depending on your loyalty. The sense is that he’s seeing the ball better than ever, and hitting the ball better than ever, and as a consequence, if you look around the Internet you’ll recognize the familiar debate about the nature and very existence of hot streaks. They say David Ortiz is locked in. It’s an easy thing to believe. It’s a more difficult thing to prove.

Nobody disagrees that David Ortiz has been hot, if you define hotness by the statistics. The statistics are white-hot. Ouch!! The disagreement is in the interpretation. Some people say that, going forward, Ortiz is just himself, and should be expected to hit like he’d normally hit. Other people say that Ortiz is just in the zone, and therefore at least his short-term expected numbers should be terrific. At the extreme, there are people who think the Cardinals should just put Ortiz on base instead of pitch to him. Pitch to Ortiz and you might give up a dinger. Put him on and you can take your chances with the next guy.

We’ve all been involved in this kind of conversation before. Many of us have observed it take place with people much smarter. The familiar faces in sabermetric research have attempted to identify meaningful hot streaks, and nothing’s really turned up. The conclusion, therefore, is that hot streaks are just blips, ending as quickly as they begin. Then there are players and coaches, and people who believe players and coaches, and that argument is that a hitter can most definitely get in the zone for a stretch. What player in baseball doesn’t believe in the existence and significance of a hot streak? Who are we to disagree with baseball players?

I don’t have anything really deep and meaningful to add. I just figured I’d go back through Ortiz’s own personal game logs. What might we find when we examine his history of hot streaks? What has he gone on to do next? It’s tricky, of course, to identify a hot streak in a number or two. But so far in this World Series, Ortiz has started five times, and he’s slugged at least .500 in all five games. I thought what I’d do is examine Ortiz’s history for streaks of 5+ games in which he slugged at least .500. It’s not perfect — it could never be perfect — but it’s something.

I was left with a total of 20 streaks, the first in 2000, the most recent coming this season (excluding the current World Series). I collected Ortiz’s numbers from his sixth start after the beginning of the streak. So, we have streaks of at least five games, and then we’re going to look at the numbers from the sixth game, regardless of whether or not the streak continued on. In all that means we’re dealing with a sample size of just 88 plate appearances, but, whatever, it’s better than nothing and we might as well see what they say.

Following streak: .315/.432/.589
Overall (weighted): .290/.382/.560

Just taking the numbers straight, we see improvements in all three, and especially in OBP. We don’t see a completely different hitter, and for a sense of the effect of the small sample size, one of the streaks led into the All-Star break. In the first game after the break, Ortiz homered, but that was several days later. Would he have still been in the zone? Should that game be eliminated? If you eliminate it, Ortiz’s slugging percentage following the streaks drops to .549. You always have to think about the error bars.

And there are some other considerations. All these streaks came in the regular season, and if Ortiz hit really well for a stretch, you figure that’s probably selective for inferior pitching staffs. Which means Ortiz could’ve been facing inferior staffs in start no. 6. You don’t want to dig too deep into any sample of 88 plate appearances, so I’ll quit it. The general point: when Ortiz has been locked in before, by the above criteria, he’s subsequently hit like a slightly better version of himself, and those numbers come with big error bars.

And this is the crux of the sabermetric argument: streaks aren’t predictive. Streaks exist, statistically, but if they reflect some change in true talent, that doesn’t show up. In this World Series, Ortiz has 11 hits on 30 swings. He had one grand slam taken away, but then he also turned one weak grounder into an infield single. During the regular season, Ortiz turned just under 5% of his swings into hits. The difference between a swing-hit and a swing-non-hit is extremely small, and Ortiz has missed or mis-hit a few pitches this past week. He’s just done it less often, but that doesn’t mean it’s all been because of Ortiz. There’s an element of chance involved, and there could be an element of the Cardinals just throwing him some more hittable pitches, mostly by accident.

Just thinking about it, it’s hard to explain why a hot streak might exist. Cold streaks are easier. A hitter might be hurt, or he might have his mechanics out of whack. A hitter can’t be more healthy than healthy, and he can’t have his mechanics any better than fine. Maybe, inexplicably, someone can just see the ball better for a little while, perhaps driven by confidence or clarity of mind, but this, like a good or bad mood, is presumably ephemeral. Something inexplicable can’t be counted on to repeat the next day, or two days later. And there’s nothing a hitter can do to make sure he’s better than he usually is. If there were, he’d always be better than he usually is, such that he’d never really be better than he usually is.

Two additional and opposing points: Ortiz hasn’t struck out since the second game of the ALCS. He’s gone 38 consecutive plate appearances without going down on strikes. What that suggests is that Ortiz has been locked in. On the other hand, in the ALCS, Ortiz had two hits in 22 at-bats. He had a .427 OPS. Instantly, he went from cold to hot, which should indicate to you something about the likelihood of going in the opposite direction.

And as one last somewhat throwaway point, the Tigers and Cardinals have both been starting all right-handers. Ortiz, then, has been facing more righties than usual, and Ortiz hits righties better than lefties, so we should expect him to have a good amount of success. That’s the thing about David Ortiz: hot streak or no hot streak, he’s still David Ortiz. He’s still one of the best hitters in baseball.

But if you’re Mike Matheny and the Cardinals, you have to see Ortiz as himself. You can’t see him as more than himself, because then you’re going to be going against the actual odds. There are cases, certainly, where Ortiz should just be walked. In all cases, Ortiz should be pitched carefully, and he should always be treated like he’s awesome. But the overwhelming likelihood is that Ortiz isn’t on some new and much higher level. It’s so easy to be biased by recent results. It’s so wrong to make decisions based on those results. It’s less risky from a PR perspective, but it isn’t less risky from a win-the-baseball-games perspective. David Ortiz has been awesome, and right now he’s the probable World Series MVP. He could win the damned thing tonight. But David Ortiz can be pitched to. All you need to do is make sure the pitches aren’t bad.




Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


38 Responses to “When David Ortiz has Been Locked In”

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

    Mike Matheny has managed like a scared child this World Series.

    His mantra of “trusting his players” has gone out the window as he’s gone away from guys he’s used and trusted all season long like Choate and Siegrist because of one plate attempt by each of them.

    He’s walking Ortiz in situations where it’s mind boggling, warming up his lefties and then being paralyzed that if he makes the move, he’ll take the blame, instead leaving his starter out there to suffer the consequences. It’s painful to watch.

    David Ortiz is a great player and he’s playing great, obviously, but Matheny is making Ortiz’s job that much easier.

    +7 Vote -1 Vote +1

    • Dave says:

      Exactly.

      Choate’s sole purpose on the Cards roster is to retire the toughest left handed batter(s) in the other team’s lineup. And Ortiz clearly fits that description. To have Choate warming up in the bullpen and then decide against bringing him in to face Ortiz is just strange.

      Vote -1 Vote +1

  2. nada says:

    I did a similar analysis (http://makenolittleplans.net/?p=231). Ortiz’s streak is impressive, but I found that performance in any given five game stretch is completely uncorrelated with performance in the sixth game.

    In any case, I wonder whether Matheny is in some sense making Ortiz locked in, by behaving so differently when he’s at bat… advising his starters to basically walk him, failing to use the correct relievers, etc. Even if Ortiz isn’t actually a better hitter now than he usually is, if the opposing team believes him to be, he may get better opportunities in his at-bats and so on.

    Vote -1 Vote +1

  3. GreggB says:

    Do hot and cold streaks exist (based on skill ebbing and flowing)? If they don’t exist in baseball, it would be the ONLY field of human endeavour where performance is perfectly consistent. Whatever your field of work, you KNOW that there are weeks where you are more productive or less productive. Ask Jay Leno if he doesn’t have weeks where he is “on” and weeks where he struggles. Ask anyone if their performance level varies. It is part of being a human being.

    So let’s look at Papi. Ortiz has been facing extraordinarily effective pitching, right-handers or not. The rest of his team — the team that led all of baseball in most hitting categories — is batting .151 in this series, and striking out at a historically high pace. Ortiz normally strikes out 15% of the time — which over 38 at bats against league-average pitching would have been five or six K’s. He has zero against extraordinary pitching. To maintain that this is simply a random grouping of events is asking us to deny everything we see with our eyes, measure with our logic and know about our own reactions.

    Vote -1 Vote +1

    • GilaMonster says:

      Hot and Cold streaks do exist, but hold no predictive because they could end at any moment. With an adequate sample size, things based on skill rarely change without a major factor.

      Imagine you are a good golfer that hits for par on a course. Chances are you end up between a few over and a few under evertime, but you’ll average par. You have a fluky day where you end up 10 under par. Chances are the next time you golf, you’ll end up closer to your average than that 10 under because you underlying skill hasn’t changed.

      Vote -1 Vote +1

      • NS says:

        You’re saying *clustering* exists, which is obvious to everyone. “Hot and cold streaks” implies something more meaningful than clustering.

        Vote -1 Vote +1

      • B N says:

        Baseball is a lot different than golf because it is directly competitive and strategic. There are a number of reasons why you would expect hot streaks (i.e., periods of performance that are of high quality, but unsustainable). Such as:

        1. Game-theoretic Hot Streaks: In game theory, your long-term performance is going to be bounded by the pitcher-hitter matchup and strategies. These are mixed strategies: if you know a guy is bad at hitting sliders, you throw him lots of sliders. But you don’t ALWAYS throw him sliders, because then he will sit on the slider. In the long term, you hit your best mix of sliders vs. fastballs and the batter reaches his best mix of sitting on sliders vs. looking for the fastball. The scouts and stat guys at a team will look for these patterns and distribute them.

        1.A. The information asymmetry hot streak. You get this when a guy is called up, takes notes on your players, but you don’t have notes on him. You can’t throw out an optimal strategy based on his weaknesses, but he can throw out a good one based on yours. This is a very likely reason for the “sophomore slump” and the acceleration of scouting and stats is why you’d expect players to hit that slump sooner rather than later in recent years. So your “sophomore slump” is not a slump, but your true talent. Your first season was a “hot streak” caused by teams lacking enough info about your strengths and weaknesses.

        1.B: The suboptimal strategy hot streak. Other teams know you. They know your tendencies and how you bat. You intentionally alter those tendencies for a short period to take advantage of the weaknesses in their strategies. This type of strategy would give you a hot start, followed by a fairly rapid drop later once teams figure out your new plan. At that point, you revert to your original strategy or you will actually be worse off (as your original strategy was optimal long-term). Assuming all games are equal, this doesn’t make much sense to do. But for games that are worth more, it can make a lot of sense to steer into your weaknesses for the element of surprise. Kind of like David Ortiz bunting for hits or stealing bases in some postseason games.

        Neither of these could explain the length of streak that Ortiz had in the WS, but they’re both perfectly adequate ways to explain certain hot streaks in general. Just because a mixed strategy is probablistic doesn’t mean that the probabilities cannot be altered, especially over short periods.

        2. Physiological/Mental Hot Streaks: These are streaks you get due to bodily factors. Most of these are pretty simple, but could explain mild bumps/declines in performance:

        2.A. Stress-related streak: People respond differently to stress and adrenaline. For some people, being on a postseason stage may not change their state much. For others, it may lead to a spike in adrenaline and/or stress during games. For a certain class of players, this may be harmful. Increased arousal reduces your ability to plan effectively and can harm motor movements. However, it may also be helpful. High levels of arousal can lead to faster reflexes, better visual detection of movement, faster processing for automatic responses, and bursts of strength. A playoff atmosphere will likely have beneficial or harmful effects on performance for certain players.

        2.B. Health-related streak. This is a simple one. Many players, especially older ones, deal with a multitude of nagging injuries at nearly all times. They will likely have certain periods where, by some stroke of chance, NONE of the injuries are bothering them. This may lead to increased performance. It may also lead to decreased performance, by causing them to overestimate their abilities (i.e., feeling great as a pitcher, so trying to blow the ball past batters rather than strategizing).

        3.C. Mental streak. Players have lives. Profound life events that impact a players’ emotional state could significantly impact performance. This may be due to a direct factor (e.g., distraction or non-distraction). It may also be due to an impact on lifestyle factors (i.e., you are in a better mood so you work out more, put in more BP time, read more scouting reports, and eat better).

        Ortiz historically hits better in the postseason, when most players hit worse. It may be that he is prone to hot streaks in high-pressure series due to how he responds to stress. Maybe he’s just not stressed enough during the regular season to perform optimally. Obviously, an OBP of 0.750 is not his true talent level. But when you take his OBP against facing lots of righties, then bump it up by a percentage, it makes such a performance more likely.

        It could just be that he is a random outlier on a distribution (i.e., “somebody” is bound to hit ~0.700 and it happened to be him). However, shouting about randomness as if it is gospel is meaningless. First, we know that baseball is not particularly “random.” It is chaotic, which means that it is a series of basically deterministic events (from a classical physics standpoint) that happen to result in noisy, largely unpredictable event chains. So any or all of these potential “hot streak” factors can be at play, with the knowledge that they only control a small fraction of the overall system and outcomes.

        Vote -1 Vote +1

        • B N says:

          I should also note that the potential (and downright inevitability) of game-theoretic streaks is just about mathematically provable. The only question about them is how long they can last.

          Physiological streaks are a hairier issue. Health-related ones are pretty hard to contest, and you’d expect them to last for days to months. You don’t say “Mike Sweeney is a 0.275 hitter.” You more accurately could say “Mike Sweeney is a 0.300 hitter when healthy, and a 0.250 hitter when not.” That is not a “random variation” in outcomes, but a random time-series of changing ability level.

          Stress-related ones are not known. You can show that stress can impact one outcome. But performance under long-term stress/higher adrenaline for something like baseball? I haven’t heard much on it. Likewise, mentally-related streaks seem likely, but I haven’t seen work on them. Worse, I would imagine that different players would react differently to stress and life events, so it’s not like you can just throw everyone’s data into a big lump and regress.

          Vote -1 Vote +1

        • Struck Out Swinging says:

          “Increased arousal reduces your ability to plan effectively and can harm motor movements.”

          That’s what my girlfriend always used to say! Talk about a cold streak…

          Vote -1 Vote +1

        • B N says:

          @SOS: I should qualify that it can harm fine motor movements especially. So your “sophomore slump” may last longer than you’d hope!

          Vote -1 Vote +1

        • Struck Out Swinging says:

          But I thought these streaks weren’t predictive!

          (/sobs)

          Vote -1 Vote +1

        • RC says:

          BN, thank you for saying much more eloquently what I’ve been trying to convince people of.

          Vote -1 Vote +1

    • Catoblepas says:

      That’s not what Jeff or anyone else making this sort of argument is saying. Ortiz is clearly not completely consistent, or else he would get approximately 35% of a hit every at-bat, and it would earn him exactly 56% of a base every time. There is clearly variation, but when that variation is all upside for a few consecutive days, it doesn’t mean the next day will continue in the same way. To use your metaphor, just because Jay Leno had four great shows in a row doesn’t mean the fifth will be good. It absolutely isn’t “random”, in the sense that there is a reason this is happening to David Ortiz and not me, but it is random in the sense that it isn’t predictive of future performance in the slightest.

      Vote -1 Vote +1

      • GreggB says:

        Catoblepas, what I disagree with Jeff about is that he seems to not accept that there are differences in what he at one point refers to as “true talent,” or what I call skill. I believe that, eliminating random and other explainable factors, Papi’s raw ability — skill or “true talent” — to see the ball and drive it has been at its peak the last six games, while Drew’s has been at low ebb.

        I DON’T disagree that it is difficult to use this to predict (although I think the major error is that this has been measured by the results of at-bats (which have a high degree of randomness, and thus a LOT of noise) instead of measuring strikeouts and speed-off-bat, which are the only things a batter can control.

        Vote -1 Vote +1

  4. Ruki Motomiya says:

    I think the problem I find here is this: “And this is the crux of the sabermetric argument: streaks aren’t predictive. Streaks exist, statistically, but if they reflect some change in true talent, that doesn’t show up.”

    Um…yeah? I mean, that’s kind of the definition of a streak: Something which is above the talent level (I understand what true talent level means, but players exceed their observed true talent all the time, since we have no extremely accurate way to predict true talent) that then comes down back to Earth, ergo it isn’t predictive for the rest of the year…

    I would say the question is this: Are streaks predictive of streaks or the length of streaks? Psychologically speaking, streaks probably exist in some way, because they have human brains and human brains vary how well they do a task based on many factors that we both do and do not understand and baseball players should follow this same thing. Not to mention other biology such as how the body varies over the year.

    The real problem is seperating it from everything else, because psychological aspects are a lot harder to tell, and I don’t mean “ooooh, you can’t predict clutch!” or some shit, I mean we don’t have a lot of information on what is going on through any one player’s mind at any time.

    I would say it is likely streaks exist in some form, it’s just that by definition a streak is not predictive of future performance, since it is a series of events of elevated (Or for cold streaks, reduced) performance compared to the norm.

    Vote -1 Vote +1

  5. ShishBoomBah says:

    There’s definitely a feel when your swing a golf club right and a collection of ill-feelings when you don’t. People spend lots of time trying to achieve that feel- mental tricks, muscle memory, etc. I don’t think the “feel” is something produced by a random clustering of good shots. But I don’t know what it is.

    Vote -1 Vote +1

    • cass says:

      I think you have the causality backwards. The feeling is produced by making good shots, not the other way around. So yeah, mostly clustering.

      It could be something that’s better mechanically, but then we could identify that on the video. And why would the mechanics then suddenly fall apart? Might make sense for an amateur, but less so for a pro.

      Vote -1 Vote +1

      • RC says:

        “The feeling is produced by making good shots, not the other way around. ”

        No, its not. There are days when you know you’re going to play well, and days when you know you’re not.

        Vote -1 Vote +1

  6. pft says:

    Part of it is in how the Cardinals have pitched to Ortiz. They never back him off the plate with a good fast ball up and in or hard slider at the ankles. Everything is away except when they miss and are in the middle. That’s the Yankees approach and Papi kills the Yankees.

    Plunk him in the ribs and save 3 pitches on a IBB.

    Vote -1 Vote +1

  7. pft says:

    I would say streaks are predictive. Regression always kicks in so the streak can safely be predicted to not continue forever.

    Vote -1 Vote +1

    • Brandon says:

      This is fangraphs. Do your research first–batters streaks are not predictive. In a sample taken from all “hot” players over multiple years, the players were found to have a wOba of roughly .3-.5 points higher in following games. That’s not very significant.

      Vote -1 Vote +1

      • Snowman says:

        I think your brain stopped paying attention after his first sentence there, because your reply doesn’t fit his second sentence at all.

        Vote -1 Vote +1

      • GreggB says:

        We don’t judge a rookie’s batting skill until he has at least several hundred at-bats. This is because we KNOW that dribblers can become hits, screaming line drives can be caught, etc. making small sample sizes based on the OUTCOME of an at bat virtually meaningless.

        Yet all the research I’ve seen about streaks is based on the outcome of at-bats — comparing the outcome of small sample sizes of at bats with even smaller sample sizes (usually one game!) of at bats. This is trying to predict the certainly-meaningless from the almost-meaningless. So of course there is little predictability.

        What needs to be studied is what CAN be measured in small sample sizes with at least some degree of validity. We need to largely ignore the outcome of where the ball goes and look at the batter’s reduction in swings and misses, extending the count by recognizing pitches, and of course driving the ball. That is what the batter sets out to do, and it is all he can control. And this WILL give more valid information in small sample sizes.

        If anyone has seen research defined in that way, please let me know. But unless and until that has been studied, I believe that there is a fundamental flaw in the existing studies that dismiss steaks and their predictability — because we have been measuring things that we already know are meaningless in small doses.

        Vote -1 Vote +1

        • Brandon says:

          Here’s the best article I know of for when sample sizes for different statistics begin to become meaningful.

          http://www.baseballprospectus.com/article.php?articleid=17659

          Vote -1 Vote +1

        • GreggB says:

          Thanks Brandon, this is very helpful. A quick summary of the article is that the sample size required to accurately measure the outcome of at-bats (average, OBP, HR rate, etc.) ranges between 170 and 910 plate appearances. Yet all the studies I’ve read that purport to dismiss streaks are compare, for example a 20 or 30 plate appearance sample with a five plate appearance sample — two largely meaningless statistics.

          Vote -1 Vote +1

  8. NS says:

    “Locked in” joins the ranks of worst baseball terms of all time. Its persistence is as peculiar and unnecessary as that of the term “eat” when people refer to a team paying a portion of a traded players contract.

    Vote -1 Vote +1

  9. PackBob says:

    A player can get “locked in” the same way as a player can get distracted and either probably effect how well a player performs at that time. But there is still everything else. A locked in player might not fare any better than normally if he’s facing good pitchers and things don’t go his way. Or he might hit three frozen-rope line drives directly at a fielder. A distracted player might miss the ball just enough to bloop a few hits over the infield.

    Players often talk about whether they are seeing the ball well or not. When they are, if things are also falling their way, they will probably be labeled as “locked in.”

    Vote -1 Vote +1

  10. Synovia says:

    “Something inexplicable can’t be counted on to repeat the next day, or two days later”

    This is the whole problem with this argument, the insistence that something has to be predictive to be real. We’ve been told for years that pitchers influencing their BABIP isn’t predictive, but oh, you start stripping away confounding factors, and look, it is.

    Also, the insistence that very incomplete studies prove anything.

    Vote -1 Vote +1

    • cass says:

      If the streak were based on something real, then we would expect it to be more likely to continue than not on a given day. If it is not predictive, than it is simply clustering and nothing more.

      It’s amazing how forcefully people hold onto their myths after they have been disproven. But it’s not like baseball is the only avenue of life where that’s true.

      Vote -1 Vote +1

      • pft says:

        The streak is real, its the interpretation that is the subject of discussion.

        The failure to understand that the lack of proof is not disproof is what stymies discussion and brings us back to the dark ages.

        We don’t know for sure why an individual has hot streaks or cold streaks. The one commonality seems to be in how the player sees the ball (at least when an injury is not involved), which is very subjective and does not lend itself to any kind of statistical test . There are all kinds of variables, quality of opponent, pitch type and location, travel, personal issues, health, BABIP luck, park, leverage, change of approach, PED use, etc.

        To say definitively that streaks are simply random noise is simply a strongly held belief not worth much more than religious dogma.

        The failure to predict the end of a streak is kind of like earthquakes that end streaks of seismic inactivity. We can’t predict the end of the seismic inactivity (yet), but earthquakes still exist. There are many other examples of phenomena that we can not predict accurately. A streak of good health can end abruptly with no warning following a clean bill of health from your Dr. The solar cycle has proven unpredictable in recent times. Like a batters hot streak and cold streaks, we know its coming and will end, we just can’t say for sure when.

        Comets and eclipses were unpredictable in older times and so thought to be signs from the Gods, since they lacked the understanding of what caused them. So too with batting streaks it seems, even in 2013, only “God” is called by another name, which is “random noise”.

        Vote -1 Vote +1

        • Snowman says:

          A lot of truth there, pft. Sabermetrics is far more cosmology than quantum mechanics. We have metrics that can predict things on the large scale, over large chunks of playing time, ignoring the fluctuations below that eventually cancel one another out. Our Uncertainty Principle covers far too many things, from human psychology to random luck fluctuations and many more in-between, to even attempt to think we know for certain what does and does not matter in the short term, down on that molecular short series level.

          Vote -1 Vote +1

      • B N says:

        “If the streak were based on something real, then we would expect it to be more likely to continue than not on a given day.”

        Not necessarily. Pretend you magically bump someone’s talent level up by 10% for 6 games. Now make sure that the toughest starter they face occurs on the 6th game. We know, by the original assumption, that the player is 10% better. We also know that their overall outcomes compared to their baseline will still be worse when they hit that tough starter. Unless you control for things like this, your noise is going to be vastly greater than your signal and wash it out entirely.

        Vote -1 Vote +1

      • RC says:

        “If the streak were based on something real, then we would expect it to be more likely to continue than not on a given day.”

        We would only expect it to be more likely to continue if whatever was causing the streak to continue did.

        IE, if the streak was a result of the player eating a more balanced breakfast, we would expect it to continue until the day he is rushed and stops at McDonalds.

        It would look random in the stats, but would not actually be random.

        Vote -1 Vote +1

    • Johnhavok says:

      You missed the point entirely. Streaks are real because after the fact, you can see them in the statistics.

      All anyone is trying to get across is that you cannot predict when a streak will end, or start, and that adjusting how you face that hitter based on a streak, and not the situation itself, is pure guesswork.

      It can be dumned down to the most basic streak of 2 hits in a row. Is it more or less likely that the player who has 2 hits in a row will get a hit the third time? The fact that he has two hits in the two previous at bats has little to no correlation to whether or not that player will get a hit in his third at bat, hence, there is no predictability on whether or not the streak will continue. As such, treating a player on a hot streak like he’s a better true talent hitter than his career indicates (ie, treating Ortiz like he’s Barry Bonds based on a hot streak) doesn’t make sense when you are trying to win ball games.

      That’s the only point anyone is trying to make. Nobody is arguing that streaks don’t exist.

      Vote -1 Vote +1

      • RC says:

        “All anyone is trying to get across is that you cannot predict when a streak will end, ”

        This is incomplete. We can not predict when a streak will begin/end using the current data that is being used in these studies. There’s no reason to believe that we won’t be able to isolate/remove more variables and do this in the future.

        “. Nobody is arguing that streaks don’t exist”

        Plenty of people on this thread, and the last one are arguing just that.

        Vote -1 Vote +1

  11. Bryan says:

    I don’t understand the narrow definition of “real” used in this posting: if it can’t be proven to exist using (incomplete) predictive models than it doesn’t exist in reality. That’s bad logic.

    Vote -1 Vote +1

    • RC says:

      Exactly.

      And its a HUGE problem with Sabermetrics; the insistence that if we can’t prove something using a hugely incomplete study, it doesn’t exist.

      I mean, advanced football statistics are only about 10 years old, and they’re adjusting for the quality of opponent on every single play. Baseball has been doing this for a lot longer, and we don’t even consider the pitchers that a batter faced. Half or our ‘advanced stats’ don’t even use park effect.

      We just assume that it all evens out, or the noise disappears, etc.

      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>