Evaluating Pitcher Talent

Dave · August 29, 2006 at 7:45 am · Filed Under Mariners 

The discussion of what statistics are useful in evaluating a pitcher came up in the game thread, again, last night. This issue comes up quite a bit around here, since I use a lot of non-conventional numbers, and new readers often don’t know what they mean, where to find them, or why they should bother. So, last night, I decided to write something of a primer on why I like to use the statistics that I use, what their usefulness is, and why I don’t really care about things like ERA, WHIP, or batting average against.

All the stats referenced, by the way, can be found at the Hardball Times, and detailed game logs using these numbers can be found at Fangraphs, which are two of the most awesome sites out there right now.

The mainstream tools for evaluating a pitcher’s success and abilities are won-loss record and earned run average, with fantasy baseball players often add WHIP (walks+hits per inning pitched) to the discussion, since it’s one of their categories. These statistics attempt to sum up pitcher effectiveness in total, giving an overview of the totality of his performance with just a few numbers.

I, personally, think they fail in that regard. ERA and WHIP group together a large string of individual events made by multiple players, making it extremely tough to separate out the credit for the pitcher, hitter, or defense. WHIP and ERA tell you there is no difference in an inning where three batters drive the ball to the fence and end up with three long flyouts or an inning where a pitcher strikes out the side. Clearly, they’re drastically different, but WHIP and ERA fail to account for the actual contributions of the pitcher. So, if the goal is to actually find out how well a pitcher threw, why not look at a micro level, instead of a macro level? That’s what I prefer to do.

For instance, what are the possible events in an at-bat that can occur?

A pitch can be thrown for a ball.
A pitch can be thrown for a strike.
A pitch can be swung at and missed.
The ball can be hit on the ground.
The ball can be hit on a line.
The ball can be hit in the air.

On any given pitch, those are the options. There are a few sub-categories under those options (outfield fly or infield fly, bunt grounder or normal grounder, etc…), but we can sum up every possible outcome of each pitch with those six options. Those outcomes might lead to wildly different events, but we’ll get to that later.

Which of these six outcomes are positive for the pitcher? Called strike, swinging strike, and groundball.

Which of these six outcomes are positive for the hitter? Called ball, line drive, and flyball.

If we can effectively determine which pitchers maximize their value in the “good outcomes” and minimize their harm in the “bad outcomes”, we can get a pretty firm grasp on who has pitching talent and who does not. Thankfully, Dave Studeman wrote a fantastic article called “Whats A Batted Ball Worth” in the 2006 Hardball Times Annual, and it includes the following run value chart. This chart will give a context to those good and bad outcome categories:

Line Drive: .356 – in other words, an average line drive is worth 35% of one run.
HBP: .342
Non-Intentional Walk: .315
Intentional Walk: .176
Outfield Fly: .035
Groundball: -.101
Bunts: -.103
Infield Fly: -.243
Strikeout: -.287

These run values were taken from real life play-by-play data, so this is an actual representation of events, not some theoretic formula. As you can see, a hit-by-pitch is a better event for the offense than a walk, even though they both simply put the batter on first base. Why? Because a hit-by-pitch is pretty much random, and can occur both at times when it is a critical situation and times when it isn’t. A walk, conversely, is far more likely to put a runner on first base in a run scoring situation, lowering it’s run value compared to the HBP.

As you can see, the difference between an outfield fly and a groundball isn’t huge, but its real, and it adds up over the course of the season. This is why, all things equal, a groundball pitcher is better than a flyball pitcher. All things are almost never equal, and flyball pitchers tend to have higher strikeout rates than groundball pitchers, but the theoretical best pitcher alive would be a groundball pitcher, not a flyball pitcher.

Also, bunting = bad.

So, now that we have some understanding of the possible outcomes and their relative value, instead of using statistics like ERA or WHIP that leave out critical information, our best bet is to try to quantify the six potential outcomes, and the events that result from those outcomes as best as we can.

BB% (Walks per Total Batters Faced) does a nice job evaluating how often a pitcher throws the ball in the strike zone. The average walkrate is 8% for a major league pitcher, though the DH makes the AL a higher walk league than the NL. Anything under 5% is tremendous, and anything over 11% is a problem. The Hardball Times publishes BB% and K% in a slightly different manner, calling it BB/G or K/G to make it scale more like the per nine innings numbers people are used to seeing. BB/G (and BB%, its derivitive) is more effective than BB/9 because it accounts for the actual amount of batters faced rather than using a proxy like innings pitched. It’s just more accurate.

K% (Strikeouts per Total Batters Faced) does a decent job evaluating how often a pitcher induces swings and misses or called strikes. 16% is league average, with 20% being terrific and 12% being a problem.

GB% (Groundballs per Balls In Play) does a very good job of telling us how often a pitcher induces a groundball. 42% is league average, and anything over 50% is terrific, with the best sinkerball pitchers posting rates in the 60-65% range, while anything below 35% can be a problem if its not offset with a high strikeout rate.

LD% (Line Drives per Balls In Play) does a very good job of telling us how often a pitcher gives up line drives. 20% is league average, 17% is good, and 23% is a serious problem. Because of the way line drives have been scored by Baseball Info Solutions the past couple of years, this number is hard to use for year to year analysis, and right now, it’s not a very effective tool. We don’t use it very often.

FB% (Flyballs per Balls In Play) does a very good job of telling us how often a pitcher gives up flyballs that leave the infield, and is basically the corollary to GB%. 36% is league average, while 32% is good and 40% could be a problem.

So we have five statistics that cover each of the six possible outcomes pretty effectively. Not perfect, but they do a credible job. They aren’t park adjusted (and yes, parks have an effect on things you might not expect, such as walk rates, strikeout rates, and groundball rates), but they’re pretty close for the majority of cases.

Thanks to the work of guys like Voros McCracken, Tom Tippett, Keith Woolner, and Dave Studeman, we also now know that the result of a particular ball in play is also not very consistent, and is due more to the actions of the hitter than the pitcher. So, when evaluating pitcher’s talent, we need to adjust for outlier type performances on converting outs on balls in play. If a pitcher has a lot of flyballs that are being caught on the warning track, or groundballs that are going right to infielders, that’s not likely to continue, and we shouldn’t assume that it will.

Not all balls in play are created equal, however, and so when we’re adjusting for outs on balls in play, we need to make sure we’re adjusting back to the type of ball in play the pitcher is giving up, since we’ve noted that they certainly do have control over their groundball or flyball tendencies.

An outfield fly becomes an out 77.7% of the time. A groundball becomes an out 74.8% of the time. A line drive becomes an out only 26.4% of the time, which is why it’s the worst possible outcome for a pitcher. An infield fly becomes an out 98.8% of the time. Because of this, flyball pitchers will post more outs on balls in play than groundball pitchers, and it won’t be a fluke. However, the non-outs that flyball pitchers give up are more harmful, and thus, the quality of the hits against flyball pitchers outweighs the relative lack of quantity. This is shown in the run value chart above, where an average groundball is a positive event for the pitcher and the outfield flyball is not.

Infield flies are automatic outs, essentially, so it’s best to separate them from outfield flies for analysis like this. Since evidence has shown that pitchers don’t have a strong year to year control over their infield fly percentage, however, when evaluating true talent levels, it’s best to assume something like a normal infield fly percentage for a pitcher, rather than the one he’s posting at the moment.

Two other big factors that we’ve identified that can have a great effect on run scoring are home run rates and stranding runners. In general, flyball pitchers give up more home runs than groundball pitchers, which is why a groundball is a positive event for the pitcher and a flyball is not.

We’ve seen very little evidence that major league pitchers have significant control over how often their flyballs go over the wall, so occassionally you’ll see a wild swing in performance that is not indicative of a players true talent level, simply because a pitcher is having more or less flyballs go over the wall than should be expected. Felix Hernandez in April and May of this year was a great example of a guy who allowed a lot of home runs per flyball, and that rate has steadily dropped as the season wore on. The average major league pitcher gives up home runs in about 11-12% of his outfield flies – significant variation from that is probably not an indicator of talent for a major league quality pitcher.

Stranding runners is also a big key, and a bit of a different animal. Naturally, good pitchers will strand more runners than bad pitchers. Since they’re good pitchers, they’re more likely to create an out in any situation, including with men on base, than if they weren’t a good pitcher. While the league average Left on Base Percentage is 70%, the bad pitchers often live in the low-to-mid-60% range, and the good pitchers live in the mid-to-high-70% range.

However, it’s not uncommon for bad pitchers to have flukily high strand rates that significantly lower than ERAs, and vice versa. Jarrod Washburn’s 2005 ERA was almost completely due to his high strand rate, as he posted the highest LOB% in the American League. That hasn’t held true in 2006, and we’ve seen his ERA rise a full run because of it. So, when you find a pitcher who is stranding runners at an unexpected rate when compared to his talent derived by BB%, K%, and GB%, it is prudent to expect that rate to regress back towards a more normal rate in the future.

So, looking at this breakdown, we see value in BB%, K%, GB%, HR/FB%, and LOB%. Those five statistics will tell you almost everything you need to know about what goes into why a pitcher is performing like he is, and all these statistics are easily available at The Hardball Times. There’s nothing that ERA or WHIP will tell you that those component statistics do not, but ERA and WHIP certainly leave a lot of the underlying information out.

However, it is understandable that people want one number that sums up pitcher performance. If you really prefer to not look through the prism of BB/K/GB/HR-FB/LOB percentages, you can always use FIP, or Fielding Indpendent Pitching (which I often call Fielding Independent ERA, since its scaled to look like ERA), which gives you an expected ERA for a pitcher based on his walk, strikeout, and home run rates. FIP isn’t perfect, either – it assumes that HR/FB is indeed a skill, and it assumes that all pitchers are equal at stranding runners, neither of which are true, but it’s better than ERA for summing up a pitcher’s total contributions to run prevention.

If you want to get really crazy, you can even use Expected FIP, or xFIP, which substitutes the league average home run per fly ball rate for the pitcher’s actual home run rate, giving a more accurate picture of how we’ll expect a pitcher to perform going forward as his HR/FB rate regresses towards the mean.

As I said, both FIP and xFIP have flaws, especially when it comes to evaluating relief pitchers, but if you’re insistent on using one number to sum up a pitcher’s contribution to run prevention, those would be your best bet.

In this age of wonderful information, there’s just no reason to use ERA and WHIP for serious analysis of a pitcher’s ability. We have better tools at our disposal. We’re doing ourselves an injustice if we continue to lean on inferior information.

Comments

91 Responses to “Evaluating Pitcher Talent”

  1. DoesntCompute on August 29th, 2006 7:58 am

    Very nice pitching analysis primer Dave. Has this been linked to the FAQ yet?

  2. strong silence on August 29th, 2006 8:09 am

    You should demand that future posters take and pass a test based on this essay.

    ERA and WHIP group together a large string of individual events made by multiple players, making it extremely tough to separate out the credit for the pitcher, hitter, or defense.

    Very well said.

  3. JI on August 29th, 2006 8:14 am

    That’s one hell of a post.

  4. pdb on August 29th, 2006 8:14 am

    Thanks for not only writing this up, but for also including what are normal, low, and high rates for each stat you write about. I find that’s usually the missing piece in a great many explanations of statistical analysis – most people don’t say “well, for stat X, a rate of Y is league average, Q is phenomenal, and L, not so hot”. It really helps the not-quite-numbers-savvy of us out here to make sense of what’s being explained.

    (and yes, parks have an effect on things you might not expect, such as walk rates, strikeout rates, and groundball rates),

    Just out of curiosity, how? If it’s a hugely detailed explanation, don’t sweat it here, but I’d be interested in knowing at some point.

    And, I have to be a pedant for a second – it’s *independent*, not *independant*. Doesn’t detract from the analysis, though, great job and I appreciate it…

  5. Dave on August 29th, 2006 8:20 am

    Just out of curiosity, how? If it’s a hugely detailed explanation, don’t sweat it here, but I’d be interested in knowing at some point.

    We don’t know for sure, but we have some decent guesses.

    Altitude – strikeout rates are consistently lower in Colorado, probably due to the thin air’s effect on breaking balls.

    Sightlines – If the hitters backdrop isn’t very effective, players can have a hard time picking up the ball out of the pitchers hands, and this will lead to lower BB and higher K rates.

    Mental Adjustments – I have no doubt that Safeco Field got into Mike Cameron’s head. He could hit the ball as hard as he possibly could, and it’d still be an out. That messed with his approach, and he just couldn’t figure out how to reinvent his swing to succeed at home. This is probably a bigger factor than we expect. I would imagine there are many players or pitchers who make adjustments in their swing or approach to account for a ballparks effects.

    Colorado and Florida have shown sustained significant effect on strikeout rates. Cleveland has showed a sustained positive effect on groundball rates. Safeco Field has induced more flyballs than expected. The differences are there, even if we don’t have concrete, quantifiable reasosns why.

  6. msb on August 29th, 2006 8:25 am

    #2– I’d fail. As I read, I can feel my brain cells pulling down the shades and shutting off the lights, not unlike their reaction to the introduction of mathematics into their little lives.

  7. The Ancient Mariner on August 29th, 2006 8:30 am

    Curiosity question. BB and K are taken as a percentage of total batters faced–why not do the same with GB, FB, and LD, rather than taking them as a percentage of a subset (balls in play)? I would think it might be more helpful to take them all as percentages of the same global set; that way, you could see at a glance the percentages for each outcome per plate appearance. At least, it would seem to me to be useful to be able to look at a pitcher (let’s call him Felix) and say, for example, that 15% of the plate appearances against him end in a walk, 20% in a strikeout, 40% end in a ground ball, 10% end in a line drive, and 15% end in a fly ball.

  8. robbbbbb on August 29th, 2006 8:31 am

    Great post, Dave. Thanks for the summary. I’d come to some of these conclusions by reading this site (among others) over the last few years, but having it laid out in one place like that is a great summation of the basic concepts.

    Are you willing to do a flipside post on hitters? One major point would be: “Grounders are, in general, bad for hitters. However there are specific hitters (think Ichiro, or to a lesser extent, YuBet) who benefit from above-average groundball rates.”

  9. Dave on August 29th, 2006 8:36 am

    Curiosity question. BB and K are taken as a percentage of total batters faced–why not do the same with GB, FB, and LD, rather than taking them as a percentage of a subset (balls in play)?

    That’s how I’d prefer to do it as well. The guys at THT don’t agree with me. So, since this is supposed to be a primer to help the average fan abandon WHIP and ERA, I didn’t want to encourage them to go to fangraphs, copy the game log into excel, and then create their own formulas. I’m the only one nerdy enough to do that.

    Are you willing to do a flipside post on hitters? One major point would be: “Grounders are, in general, bad for hitters. However there are specific hitters (think Ichiro, or to a lesser extent, YuBet) who benefit from above-average groundball rates.”

    Eventually, sure, but it’s not as necessary for hitters. OPS correlates to runs scored at .97. We don’t have any tool that evaluates pitcher effectiveness as well as OPS (or, its better versions, OPS+, EqA, or VORP) does for hitters. Since we have a summation tool for offense that is so effective, the need for breaking them down by ball-in-play type isn’t as necessary.

    It can still be interesting, however, and I may do something like that during the offseason. To note, though, groundballs aren’t as bad for hitters as you may think. Miguel Tejada, Derek Jeter, and Carl Crawford are all extreme groundball hitters.

  10. arbeck on August 29th, 2006 8:39 am

    Just out of curiosity, how? If it’s a hugely detailed explanation, don’t sweat it here, but I’d be interested in knowing at some point.

    I can’t give a total be all end all explanation, but I can give you some examples.

    One of the reasons Coors field has been such an offensive park is that the thin air causes less movement on pitches. When a pitch moves less, you are going to have fewer swings and misses and a lower SO%. Safeco originally had a batters eye that players hated. Certain parks have better or worse batters eyes, different amounts of shadow, etc. All of that is going to have an effect on the SO%.

    GB% is a little harder to quantify in my head, but if i were to guess, I’d say the same affects from above would tend to decrease the GB%. When the players see the ball better they hit the ball harder which would increase LD% and FB%. Also, certain parks tend to favor ground balls. A left handed hitter travelling to the Metrodome maybe more likely to try to hit the ball on the ground and scoot it through on the turf then try to deal with the outfield and the way it supresses left handed sock.

    BB% is going to be similar to SO%. Batters see the ball better they take more borderline pitches and get fooled less. They hit the ball harder, the pitcher throws more pitches. This increases fatigue which also increases BB%.

  11. arbeck on August 29th, 2006 8:40 am

    Damn, dave is quicker than me.

  12. gwangung on August 29th, 2006 8:41 am

    As you can see, a hit-by-pitch is a better event for the offense than a walk, even though they both simply put the batter on first base. Why? Because a hit-by-pitch correlates pretty well with “struggling pitcher”, and so more struggles are likely to follow.

    Something to remember…correlation is not causation. What we see here are not necessarily explanations or causes of events, but they’re big clues.

    And as Dave points out, they’re not park adjusted, so there’s still more analysis to be done.

    Mental Adjustments – I have no doubt that Safeco Field got into Mike Cameron’s head. He could hit the ball as hard as he possibly could, and it’d still be an out. That messed with his approach, and he just couldn’t figure out how to reinvent his swing to succeed at home. This is probably a bigger factor than we expect. I would imagine there are many players or pitchers who make adjustments in their swing or approach to account for a ballparks effects.

    Another salient point to remember.

  13. DCMariner on August 29th, 2006 8:41 am

    It’s posts like these that make me feel like I am stealing from this website…

  14. pdb on August 29th, 2006 8:45 am

    Thanks all for the clarifications. I feel smarter already.

  15. AQ on August 29th, 2006 8:46 am

    #13 – I hear that. I used to pray at the altar of ERA and WHIP until I started to regularly visit USSM about 2 years ago. This post will help me explain to others about why I believe that WHIP and ERA are not valuable indicators of pitching prowess. I knew that they weren’t good tools to use, but I couldn’t fully articulate why that was.

    Thanks Dave!

  16. Ralph Malph on August 29th, 2006 9:23 am

    Also, as far as park adjustments, don’t forget the amount of foul territory, which I think is a bigger factor than most people realize.

  17. Steve Nelson on August 29th, 2006 9:36 am

    #16: lso, as far as park adjustments, don’t forget the amount of foul territory, which I think is a bigger factor than most people realize.

    Foul territory also favors fly ball pitchers. Since a foul flyball ends an AB, that is part of the reason why a greater proportion of flyballs result in outs.

    I’ve sometimes considered that foul ball fly outs should be considered differently from balls in play. They’re not really a ball in play, because if the fielder fails to make a play, the batter does not reach base.

    If a pitcher has a skill of inducing weak fly balls that can produce outs without the possibility of a runner reaching base, I think that skill should be noted. In that respect, foul ball fly outs have seemed to me more akin to strikeouts than to balls in play.

  18. Dan W on August 29th, 2006 9:45 am

    Awesome stuff Dave. I have been completely cured of looking at ERA as a valuable indicator of a pitcher’s true ability.

    Would it be fair to say, however, that ERA IS a valid indicator of past ACTUAL performance, factoring in all variables including those he does not control?

  19. Dan W on August 29th, 2006 9:52 am

    I have to add that the timing of this discussion has been perfect, given Felix’s living, breathing demonstration of these principles last night. Living here in Halo-ville has not been much fun from a baseball standpoint lately, except for the perverse enjoyment of wathing obnoxious Redsox fans gloat over Angel misery last week.

  20. Dave on August 29th, 2006 9:53 am

    Would it be fair to say, however, that ERA IS a valid indicator of past ACTUAL performance, factoring in all variables including those he does not control?

    Maybe, if you have any faith in official scorers (I don’t), and you really don’t care about park effects or team defense. ERA does a decent job of telling you how well the the team prevented their opponents from scoring when that pitcher was on the mound, but it doesn’t really help you figure out how to spread the credit around, and it doesn’t adjust for context – mainly, the park being played in.

    So, even in that limited value, it still has significant flaws.

  21. Dan W on August 29th, 2006 9:53 am

    watching, not wathing

  22. arbeck on August 29th, 2006 9:54 am

    Steve Nelson,

    The problem is that foul ball flyouts are almost all the equivalent of infield pop out. If pitchers don’t have much skill in inducing infield pop ups, then why would they have skill in inducing foul ball fly outs? I agree that it should be taken into consideration when calculating park effects, but I don’t think it has much to do with pitcher ability.

  23. The Ancient Mariner on August 29th, 2006 9:55 am

    I’m not Dave, obviously, but from where I sit, I’d think it would be fair to say that ERA is a partially-valid descriptor of past actual results; after all, variables beyond a pitcher’s control are not a part of his performance, but rather elements which interact with his performance to affect the results of that performance. Also, ERA can only be partially valid given that it excludes runs which the scorer decides are unearned — these are also a part of said pitcher’s results, and they are a part for which he likely bears some responsibility even if the scorer labels them “unearned runs.”

  24. theberle on August 29th, 2006 10:03 am

    Dave: OPS correlates to runs scored at .97. We don’t have any tool that evaluates pitcher effectiveness as well as OPS.

    Dave, this article is awesome. One quibble about the above comment though: I’m certain ERA also correlates pretty well to runs allowed.

    One key to understanding why ERA is not a great tool for evaluating talent is its (weak) predictive value. The year to year correlation is much lower for a stat like ERA than it is for xFIP. This is why OPS is a better tool than ERA. OPS also has good predictive value. While ERA does correlate from year to year (good pitchers will stay good, most of the time), xFIP does a much much better job.

    And, yes, ERA and WHIP (and for that matter GWRBI for hitters) shows what a player has actually done this season, and I have no problems awarding the MVP based on stats like GWRBI’s or the Cy Young for Wins, since that measures the results of what a player actually did (although, there are team influences that make it much harder for a player from a losing team to win). However, it’s when evaluating a player for future performance (like when you’re offering a contract), when you have to use advanced statistics that have more predictive value, rather than “clutchiness” or “ability to win.” This is where the advanced metrics have their greatest value.

  25. bigpoppa01 on August 29th, 2006 10:06 am

    Dave,
    Great information…So based on these numbers whom do you think the M’s should go after in the offseason (via free agency or trade)? Play GM for a moment if you will.

  26. theberle on August 29th, 2006 10:06 am

    T.A.M. You beat me to part of my point. You also make a good point about unearned runs. One amazing note that still surprises me is a pitchers RA (total runs allowed) actually correlates better from year to year than their ERA.

    Again, I don’t find it fair to evaluate a pitcher’s past performance based on their unearned runs (no matter how random official scorers are). But, if I’m a GM (or I just play one on blogs), before offering a contract, I’m going to look at the stats that correlate from year to year over the stats that change more wildly from year to year.

  27. Brian Rust on August 29th, 2006 10:09 am

    Great work, Dave. I’ll have to come back to this tonight (after work)!!!

    A couple of questions:

    How do I get to the “detailed game logs” at fangraphs? I can find the graphs and the box-score-like summary, but not a log.

    Can you share with us the definition of “LD” as a statistic? Or more precisely, how does B.I.S. distinguish LDs from FBs and GBs? If any analysis depends on dividing a velocity/trajectory/rotation continuum into three discrete categories, we need to understand what the dividing lines are.

    Finally, I think it’s not quite accurate to say “Safeco Field induces more fly balls than expected.” In reality it’s probably more accurate to say pitchers evaluate the park’s characteristics, and are more willing to induce (or allow) fly balls than they would be in, say, Arlington. In other words, pitchers may be changing their approach when on the mound at Safeco. It has serious implications for an evaluation of pitching based on the factors you so eloquently and succinctly prescribe above.

    Thanks, and keep up the good work.

  28. Brian Rust on August 29th, 2006 10:18 am

    More, relevant to the discussion ensuing while posting:

    If official scorers are too subjective in dividing outcomes into the discrete categories H and E, isn’t the segmentation of the GB/LD/FB continuum into discrete categories similarly flawed?

    Granted, it tells us more about the ball in play, but the subjectiveness will make it difficult to accurately tease out statistically significant conclusions about marginal differences.

  29. Dave on August 29th, 2006 10:19 am

    How do I get to the “detailed game logs” at fangraphs? I can find the graphs and the box-score-like summary, but not a log.

    Use the player search function, click on stats, then on game log.

    Can you share with us the definition of “LD” as a statistic? Or more precisely, how does B.I.S. distinguish LDs from FBs and GBs? If any analysis depends on dividing a velocity/trajectory/rotation continuum into three discrete categories, we need to understand what the dividing lines are.

    BIS scores every game from video, and it is up to the scorer to determine whether it was a a ground ball, a line drive, a flyball, or newly added this year, a “fliner” (one of those line drive fly balls that is hard to categorize). They do their best to get consistency across all scorers, but since its still a human judgment, there is some margin for error. I’d say they’re probably at least 97-98% accurate, though, so we can use the statistics for most of our purposes without fear.

    Finally, I think it’s not quite accurate to say “Safeco Field induces more fly balls than expected.” In reality it’s probably more accurate to say pitchers evaluate the park’s characteristics, and are more willing to induce (or allow) fly balls than they would be in, say, Arlington. In other words, pitchers may be changing their approach when on the mound at Safeco. It has serious implications for an evaluation of pitching based on the factors you so eloquently and succinctly prescribe above.

    This is just really semantics, but if Safeco’s features are convincing the pitchers to pitch differently (and I agree, that is whats happening), then isn’t Safeco ultimately responsible for causing the change in pitcher approach that leads to the increase in flyballs?

  30. TheEmrys on August 29th, 2006 10:26 am

    “A pitch can be thrown for a ball.
    A pitch can be thrown for a strike.
    A pitch can be swung at and missed.
    The ball can be hit on the ground.
    The ball can be hit on a line.
    The ball can be hit in the air.”

    I would like to see this modified. Actally, just the third supposition. I know it is unquantifiable with how things stand now, but a pitch can be a ball or a strike, and yet still swung on and missed. By dilineating between the two, I think it gives a good basis for “batter error.”

    I guess by the same token, a pitch could be a ball and be hit as well. At that point, there are 11 options.

    But technologically, we can are approaching the point to where it is possible to achieve this.

  31. JMB on August 29th, 2006 10:33 am

    If anyone cares, I’ve added this post to both the FAQ and the Features section, so you can find it later on (or point others to it).

  32. giuseppe on August 29th, 2006 10:38 am

    “…significant variation from that is probably not an indicator of talent for a major league quality pitcher.”

    This is why much of Dave’s illuminating post does not apply to many Mariner’s pitchers in recent history. [end sarcasm]

    Seriously great post. Well written, clear, concise.

    When is the U.S.S. Mariner annual fund drive?

    How much is the information you receive here worth to you? How many of us come here every day, multiple times? Contribute what you feel you can.

  33. scraps on August 29th, 2006 10:41 am

    This is terrific. Thank you, Dave.

    As you can see, a hit-by-pitch is a better event for the offense than a walk, even though they both simply put the batter on first base. Why? Because a hit-by-pitch correlates pretty well with “struggling pitcher”, and so more struggles are likely to follow.

    I think it’s also because even an unintentional walk is often due to cautious situational pitching, while a hit-by-pitch is more random, and therefore more likely to occur in a damaging sitation overall than even an unintentional walk.

    Maybe that’s just putting the same thing a different way.

  34. gwangung on August 29th, 2006 10:45 am

    I think it’s also because even an unintentional walk is often due to cautious situational pitching, while a hit-by-pitch is more random, and therefore more likely to occur in a damaging sitation overall than even an unintentional walk.

    Or maybe it’s because there’s more variation introduced by inconsistent calls from the umpire, or varying calls from different umpires (at any rate, you’re measuring contributions from more than one person). Can’t say that about HBPs.

  35. firova on August 29th, 2006 10:50 am

    Maybe this could be forwarded to the announcers and beat writers around here. It actually explains what we are seeing and can expect, rather than the guesswork that passes for commentary. I don’t think Fairly had a clue what Felix was up to last night.

    A copy for Bill Bavasi and one for Mike Hargrove too?

  36. Evan on August 29th, 2006 10:53 am

    I’m actually surprised that a HBP is a more negative outcome than an unintentional walk. A HBP requires one bad pitch. Sometimes not even that bad when facing guys like Melvin Mora or Reed Johnson. A walk requires at least four pitches, at least some of them thrown from behind in the count. It think it’s necessarily a higher stress event, and should have a bigger negative impact on the pitcher.

    That it doesn’t surprises me.

  37. The Ancient Mariner on August 29th, 2006 10:54 am

    Re #30: Myrddin, it seems to me that’s really only a consideration on the batter’s side; from the pitcher’s point of view, what matters is that he induced a swing-and-miss. Similarly, if a pitcher throws a pitch out of the strike zone and the batter makes contact, it’s the fact of the contact that matters.

  38. Nintendo Marios on August 29th, 2006 10:55 am

    13 – So true.

    It is like going to a preschool and having Dave read from Hardball Times every day for story time. And we can have beers at snack time.

    I tend to give Patient Teacher Dave Chef’s voice. “…and that’s why, children, ERA is not your friend when evaluating pitching talent.”

    I haven’t got a voice for Pissed Moderator Dave but it isn’t Issac Hayes.

  39. strong silence on August 29th, 2006 10:56 am

    Clearly, they’re drastically different, but WHIP and ERA fail to account for the actual contributions of the pitcher

    It is fair to use ERA in an evaluation of how a pitcher actually performed. (ERA is not an entirely useful predictive tool.) It is not fair to use ERA as the only measure of pitcher performance.

    So if Verlander ends the season with a 1.00 unadjusted-ERA advantage over Felix I’m confident in saying that Verlander actually pitched better in 2006.

  40. Dr. Milos PHD on August 29th, 2006 10:56 am

    A couple of early observations and questions as I attempt to put some of these new–to me at least–findings to practice. It’s too bad HBT doesn’t list FB% with their GB% and LD%. Fangraph is a great tool to see all three in comparison to a pitchers tendencies, though.

    Questions: and I apologize if these are a repeat of past threads.

    Does a LOB% account only for runners that were allowed on-base by the pitcher or does it include inherited runners? This would be important when evelauating a reliever’s performance.

    I know you warned against using FIP as an end all be all number, and I’m trying to keep that in mind, put when looking at a pitchers FIP and xFIP, where does the league averages fall?

    Lastly, this is not a question, but last night you gave a list of pitchers that were “underrated” due to potentially higher era and whip numbers. While I didn’t particularly agree with the list you provided, I see where you were coming from–Bonderman has been a stud. In my mind that is still reflected in his overall, or “traditional” numbers–era/whip.

    Here is a list I compiled based solely on this years numbers using FIP, xFIP, GB/FB/LD%, K/G, BB/G. (I know you prefer K%, BB% but I didn’t have time to construct my own data in excel…yet)

    AL: Joe Blanton; Jake Westbrook, phenominal GB%, low LD% and BB/G, top 15 in AL FIP and xFIP, not the biggest strikout guy though(my number one underrated pitcher based on era and whip numbers)

    NL: Andy Pettitte; Tim Hudson; Zach Duke; and to a lesser extent Matt Cain, who as a flyball pitcher has increased his % of groundballs while reducing his flyballs and having a microscopic LD% and a very high K/GG ratio. If he cuts down on his BB, he is a great young pitcher going forward(not a huge secret by any means).

    Anyway, potentially boring post, just having fun with my new toy. Thanks Dave.

  41. Steve Nelson on August 29th, 2006 10:57 am

    #35: A copy for Bill Bavasi and one for Mike Hargrove too?

    As I’ve commented about Hargrove in other circumstances, you can lead a man to data but you can’t make him think.

  42. Dave on August 29th, 2006 11:15 am

    It is fair to use ERA in an evaluation of how a pitcher actually performed. (ERA is not an entirely useful predictive tool.) It is not fair to use ERA as the only measure of pitcher performance.

    So if Verlander ends the season with a 1.00 unadjusted-ERA advantage over Felix I’m confident in saying that Verlander actually pitched better in 2006.

    If you don’t care about defense or park effects, sure. Did you know the Tigers defense is, by far, the best in baseball this year? It’s not even close. And Comerica is notoriously pitcher friendly. Obviously, Safeco is too, so I’d agree with your conclusion that ’06 Verlander > ’06 Felix.

    But I don’t understand why you’re willing to overlook the glaring park/defense adjustments that ERA doesn’t even attempt to make.

    Does a LOB% account only for runners that were allowed on-base by the pitcher or does it include inherited runners? This would be important when evelauating a reliever’s performance.

    LOB% is actually a formula, not a real-world number of runners left standing when an inning ends. The formula can be found in THT’s statistics glossary. Baseball Prospectus has actual, real life inherited runners stranded numbers for evaluating relievers. LOB% is better for starters than it is for relievers.

    I know you warned against using FIP as an end all be all number, and I’m trying to keep that in mind, put when looking at a pitchers FIP and xFIP, where does the league averages fall?

    Because FIP and xFIP are scaled to match ERA, there’s a different multiplier put into the formula every year to make it match up to league average ERA. So, a league average FIP and a league average ERA should be the same in any given year – the difference will be in the spread. The best pitcher will have a higher FIP than ERA, and the worst pitcher will have a lower FIP than ERA, because a lot of what goes into “great years” are things that aren’t actually repeatable skills.

    Just as a baseline, though, a FIP of 4.00 is pretty solid, while an xFIP of 4.30 is a rough guess for a decent performance for a starter.

    AL: Joe Blanton; Jake Westbrook, phenominal GB%, low LD% and BB/G, top 15 in AL FIP and xFIP, not the biggest strikout guy though(my number one underrated pitcher based on era and whip numbers)

    I’m not sure what you’re using to say Blanton is underrated. His success this year is due entirely to one thing – a ridiculously low 6.0% HR/FB rate, which is absolutely not sustainable. His xFIP is 4.97 – thats not good.

    Westbrook has pitched about as well as you could expect – no big outlier performances there. His LOB% and HR/FB% are both a little better than you might expect, but he’s got the lowest infield fly percentage in the AL, and everything pretty much balances out. His 4.19 xFIP is a pretty close match to his 4.27 ERA.

    Pettitte’s a good candidate, yes, as is Hudson. Duke, I’m not so sure about – he’s not stranding runners, so that will probably improve next year, but his peripherals aren’t very good, and he was WAY over his head last year. I wouldn’t say Cain is underrated either – the strikeouts give hope for the future, but his command is bad, and until it improves, he’s a mediocre pitcher.

  43. Anthony on August 29th, 2006 11:24 am

    Dave, would you happen to know exactly how FIP is scaled to the league ERA? I know the basic FIP formula simply adds 3.2. But let’s say after adding 3.2, you get a lgFIP of 4.30 v. an actual lgERA of 4.55. Do they simply add 3.45 instead of 3.2, or do they multiply the 4.30 by 1.058? That would make a difference–admittedly small–for individual pitchers.

  44. arbeck on August 29th, 2006 11:28 am

    I think if you stop and think about it a HBP should be worse for the pitcher than a walk. A walk requires the pitcher to throw 4 balls. The HBP only takes one. So with a runner on first and second it only takes one mistake to load the bases versus four mistakes. An average major league pitcher can throw a strike when he absolutely has to most of the time, so walks in situations that are super critical are probably less likely than HBP. It also stands to reason that ALOT of walks are semi-intentional. For example the pitcher would rather face Carl Everrett with the bases loaded then Richie Sexson with runners on 1st and 3rd, so he pitches around Sexson by not giving him a good pitch and takes his chances with Everrett. The HBP is almost never intentional.

    Also, there are certain players who I believe have a skill of being hit by pitch. Craig Biggio and Jason Kendall are willing to “step into one” to get on base. It can probably be argued that the times they do this are when the affect would be most beneficial to the offense.

    Plus, there is a mental aspect to consider. I remember from my pitching days that hitting a batter usually shook me up a little. I tended to groove a few after that. Doing that in the big leagues is a recipe for disaster.

  45. Anthony on August 29th, 2006 11:34 am

    I know that preventing HR/F isn’t a repeatable skill for pitchers, and I think Roger Clemens and maybe John Lackey are the only ones to ever consistently limit that figure, but has anyone looked at Mariano Rivera’s THT page? His HR/F the last three years are 6.6%, 3.7% & 5.7%. I’d love to know what those numbers are prior to 2004.

  46. Dave on August 29th, 2006 11:41 am

    HR/F is totally different for relievers than starters. Almost every elite reliever beats the 11% benchmark. xFIP isn’t useful for relievers. FIP is preferable, but even that is off, because it regressese to a BABIP that is higher than what the average reliever gives up.

  47. Dave on August 29th, 2006 11:43 am

    Dave, would you happen to know exactly how FIP is scaled to the league ERA? I know the basic FIP formula simply adds 3.2. But let’s say after adding 3.2, you get a lgFIP of 4.30 v. an actual lgERA of 4.55. Do they simply add 3.45 instead of 3.2, or do they multiply the 4.30 by 1.058? That would make a difference–admittedly small–for individual pitchers.

    I’m not sure – FIP is a Tangotiger invention, and he hangs out around here a bunch, so maybe he’ll offer up an answer.

  48. Dr. Milos PHD on August 29th, 2006 12:02 pm

    Upon closer inspection, Blanton might have been a stretch. There weren’t as many good examples in the AL, so I reached. Millwood might have been a slightly better example, though he too is not stranding runners very effectively. I’ll have to do some more reading on LOB%, as well.

    Being in my infintile state of THT pitching stats, I am still learning how to process all these for a complete picture. What I was attempting to look at though were pitchers that were having off or bad years by “convential” standards, ie era. Blanton and Duke both have pretty horrendous era and whip numbers. And I thought, that they were pitching better than those numbers indicated. In the case of Duke, he is producing a great GB%, and has a low HR/F rate. Maybe underrated isn’t the right term.

    Duke was actually overrated coming into the year and has been a dissapointment in many eyes. But as a young pitcher, his numbers indicate there is still a quality major leaguer going forward, there.

  49. Dave on August 29th, 2006 12:06 pm

    Low HR/F isn’t a sign that a guy is underrated – its generally a good sign that he’s probably overrated, since that’s not reallys something they can control very well.

    If you want, you can sort THT’s stats by FIP-ERA – that will give you a list of guys whose ERA is a lot lower than their Fielding Independant ERA would suggest, or vice versa, and will help identify guys who are probably going to regress to the mean.

  50. joser on August 29th, 2006 12:12 pm

    The other thing about HBP vs the unintentional walk is that, as someone said, you have the umpire coming into it. For pitchers who aren’t the power/high-K guys, the strike zone variation can have an impact. Consider Jamie Moyer: on nights when he gets the strike zone he wants, he gets a lot of batters out by nibbling on the corners — or induces them to swing at pitches they otherwise would allow go by. If he doesn’t get the strike zone he wants, he has a lot of balls called (and he then as to go more into the heart of the zone, where batters eat him alive). But those balls aren’t an indication of loss of control the way a HBP would be.

  51. DarkKnight1680 on August 29th, 2006 12:36 pm

    Dave – Can you quantify the problem with Adam Kennedy this year? He’s second in the AL with a LD% of 27.3 but his BA is in teh .260s and his OPS in under .700. You’d think someone hitting the ball on a line essentially once a game would have better numbers with the average value of a line drive so high. Yes, I know Adam has no power, but it still seems strange for a full-season sample.

  52. DarkKnight1680 on August 29th, 2006 12:49 pm

    And another one quickly while I think of it…is there any possibility that Woods can learn to cut down on the number of walks he issues? His only real noteworthy season was in Arkansas (AA) in 2004, when he Kd 60 and walked just 19 in 90 innings. He moved to AAA the same year, pitched 83 more innings, again struck out 60 against better hitters, but walked a ridiculous 42 batters. Does he just have no control or can he learn to target better?

  53. Mat on August 29th, 2006 12:51 pm

    If you don’t care about defense or park effects, sure. Did you know the Tigers defense is, by far, the best in baseball this year? It’s not even close. And Comerica is notoriously pitcher friendly.

    Okay, I admit this is a fairly stupid nit to pick in this particular discussion, but I don’t think that the Tigers’ defense is by far the best in baseball this year. They have, by far, the best defensive efficiency. However, looking at defensive efficency totals so far this season, I don’t think it’s a coincidence that four of the five top teams play in notorious pitcher’s parks with large outfields. (As for the fifth, St. Louis, I’m unsure about how the park factor has played out this year in the new Busch Stadium.) After adjusting for park factors, I’d guess that Detroit comes back towards the pack a fair bit.

    The reason this is a stupid point to belabor is that for the purposes of singling out pitching performance we want to adjust for defense and park factors, and unadjusted defensive efficiency is perfectly legit for that purpose. But in judging defenses, I think park effects aren’t emphasized enough, so I decided to mention it anyway.

  54. Dave on August 29th, 2006 1:05 pm

    It’s not defensive efficiency that makes the Tigers so good – they’re making the plays that no one else does, and its not the outfield. If you go to the THT’s Team pages, you’ll see a column for +/- fielding by air and ground. Basically, this shows how many extra plays a team is making on groundballs and flyballs than would be expected base on ball in play types.

    The Tigers are +65 on the ground. +65! The next best team in the AL is Toronto at +16. Brandon Inge and Placido Polanco are both tremendous, elite defensive infielders. Pudge is still very good at getting out on balls in front of the plate. Shelton and Guillen are average defenders, and they’re the wink links. Even the Tigers reserve infielders, Omar Infante and Neifi Perez, can pick it.

    The Tigers defense isn’t a park effect illusion. Their infield defense is something else.

  55. Dave on August 29th, 2006 1:08 pm

    Dave – Can you quantify the problem with Adam Kennedy this year? He’s second in the AL with a LD% of 27.3 but his BA is in teh .260s and his OPS in under .700. You’d think someone hitting the ball on a line essentially once a game would have better numbers with the average value of a line drive so high. Yes, I know Adam has no power, but it still seems strange for a full-season sample.

    J.C. Bradbury developed Predicted OPS (called PrOPS), which is basically an expected BA/OBP/SLG line based on ball in play types. Kennedy’s PrOPS is .728, 40 points higher than his actual line, and his Projected BA is .291. So, you’re right, we’d expect a normal hitter with that kind of LD% to post better numbers.

    However, PrOPS overrates slow players, and Kennedy’s injury problems have cost him a lot of his speed, and it assumes that all line drives are created equal. Considering Kennedies total lack of power, that’s probably not a good assumption. More likely, his line drives are easier to catch than, say, Albert Pujols’ line drives.

    Kennedy may be experiencing a bit of bad luck, but even if he is, his lack of power still makes him a marginal major league hitter.

  56. ConorGlassey on August 29th, 2006 1:15 pm

    Dave – any idea which pitchers get the most swing-and-misses? And, do these pitchers generally match up with the scouts’ view of who has the best stuff?

  57. Dave on August 29th, 2006 1:20 pm

    Dave – any idea which pitchers get the most swing-and-misses? And, do these pitchers generally match up with the scouts’ view of who has the best stuff?

    No, this is something I really want to get my hands on, because I have a lot of theories about called strikes versus swinging strikes. But that data just isn’t publically avialable right now in any kind of sample that would lend itself to a thorough analysis.

  58. JI on August 29th, 2006 2:17 pm

    How random is BABIP for hitters? For example, Jason Giambi is posting the similar power/walks number as his glory years in Oakland, but he’s only hitting .260 or so, and he is near the bottom in BABIP. Where as Scott Podsednik has a good BABIP but a similar batting average. Is the reason that Giambi can get away with a lower BABIP because of his power, or is it random?

  59. Dave on August 29th, 2006 2:28 pm

    BABIP is much less random for hitters than it is for pitchers, but it’s still subject to fluctuation. A lot of power hitters are going to post low BABIPs because when they hit the ball well, it leaves the yard, so those get removed from the denominator as its not a ball in play.

    I personally don’t see BABIP adding a lot of information to our knowledge about hitters. It’s useful for pitchers, but not really for offensive evaluations.

  60. sidroo on August 29th, 2006 2:41 pm

    This is off the top of my head, but as poster #28 mentions, isn’t there a potential problem with how balls in play are classifed as one of three discrete event types?

    It seems a fair number of struck balls sit on a category cusp – a scorching one-hopper to the third baseman or a soft drive that falls in front of an outfielder. A differnce of three or four judged one way or another over six innings might amount to a significant difference, it seems.

    And biases in what various observers consider to be line drive/ground ball/fly ball would tend to be consistent and magnify differences between pitchers, since each observer would tend to follow one particular pitching staff.

    Maybe all of this is dealt with on the sites mentioned.

  61. Dave on August 29th, 2006 2:43 pm

    BIS does a good job recording the data. There’s no reason to be worried about the integrity of the information.

  62. Ralph Malph on August 29th, 2006 2:46 pm

    60 – sidroo raises a question I’ve wondered about. Who is collecting this data? If it’s a stringer who does it at each ballpark, you could get subtle “park effects” on these numbers based on, for instance, whether the Safeco field stringer tends to classify hard hit balls through the infield as line drives more often than an observer at another park. Or whatever.

  63. JI on August 29th, 2006 2:46 pm

    it leaves the yard, so those get removed from the denominator as its not a ball in play.

    Yeah that would make it pretty useless for hitters. I just saw it listed as a sortable for hitters, so… thanks for taking the question though. This beats the hell out of an ESPN chat.

  64. Dave on August 29th, 2006 2:58 pm

    60 – sidroo raises a question I’ve wondered about. Who is collecting this data? If it’s a stringer who does it at each ballpark, you could get subtle “park effects” on these numbers based on, for instance, whether the Safeco field stringer tends to classify hard hit balls through the infield as line drives more often than an observer at another park. Or whatever.

    It isn’t a stringer. Baseball Info Solutions trains their employees to ensure accuracy, and this is what they do for a business. They sell this info to major league teams – accuracy is their currency.

    The data is good.

  65. MickeyZ on August 29th, 2006 3:11 pm

    When they give a run value for bunting, is that bunting for a hit? Or bunting to advance a runner?

  66. Bodhizefa on August 29th, 2006 3:12 pm

    Fantastic post, Dave. I’ve given all my baseball buds the link to check it out, and it will be extremely useful in any discussion I ever have with anyone online about baseball.

    Well done.

  67. Dave on August 29th, 2006 3:14 pm

    When they give a run value for bunting, is that bunting for a hit? Or bunting to advance a runner?

    Both. It’s the act of squaring around and intentionally trying to hit the ball 20 feet.

  68. Evan on August 29th, 2006 3:29 pm

    Bunting is also necessarily skill-related. Some guys are really good bunters and do help create runs by bunting. But that’s not what these are designed to measure. League-wide, bunting is bad.

  69. Bender on August 29th, 2006 3:29 pm

    The real question I have after reading this is how in the world can major league teams still be ignorant to this kind of analysis?

    It seems like the Mariners are willfully ignorant of this kind of in depth thought and it really hurts the team. Why don’t they invest in some statisticians?

    What will it take to get them to open their eyes to more reliable ways of judging talent? Surely its not just an old boys network of ‘baseball guys’ who are sitting in some wood-paneled room smoking cigars and talking about how they should trade for someone because, “the kid’s got heart, damnit!” Right?

  70. Dan W on August 29th, 2006 3:40 pm

    I got a tour of the White Sox corporate offices at Comiskey (aka US Cellular) a few years back. It happened to be on a draft day early evening, and it really was a smoke-filled room (as in literally billowing into the hallway) of scouts and old baseball guys.

  71. Dan W on August 29th, 2006 3:48 pm

    During the same tour, they also went out of their way to discuss a little of their baseball-related technology. One system (I think it had a name, and is in use by most if not all major league teams) stored years of indexed video, such that if you wanted to see all Maglio Ordonez at-bats against Jamie Moyer during night games at home (for example), you could.

  72. studes on August 29th, 2006 3:56 pm

    Nice job, David. You really should try writing for those THT guys.

    Regarding BIS, they not only have stringers at ballparks, but they have every game doublechecked by QA guys. Each game is essentially scored twice, as I understand it. That said, the difference between a line drive and flyball is ephemeral to me.

    For FIP, we calculate basic FIP (without the add-on) and then calculate what the league-specific add-on is based on actual league ERA, then use that for all pitchers in that league.

  73. Steve T on August 29th, 2006 4:39 pm

    Best USSM thread ever?

  74. joser on August 29th, 2006 4:54 pm

    Surely its not just an old boys network of ‘baseball guys’ who are sitting in some wood-paneled room smoking cigars and talking about how they should trade for someone because, “the kid’s got heart, damnit!” Right?

    Well, and sometimes that works. But the teams that avail themselves of all information — scouts and stats both, as Dave has always suggested — will tend to separate themselves from those that don’t. Unfortunately the old boy network is strong, and so is tradition, and “baseball men” tend to surround themselves with other baseball men… so it may take a decade or two as they die off. Baseball is probably the worst sport in this regard for various reasons — not least because it doesn’t have a huge and highly-visible college analog where maverick ideas have a chance to be tried and succeed. Any short term failures by the “whiz kids” — and there will be some, of course — tend to mask the benefits of the approach (just ask Beane about the playoffs, or DePodesta about, well, anything) so it takes all the longer. It’s also worth noting that this kind of data is much harder to come by for minor league players, and of debatable worth when it is available, so that limits its applicability to trades and free-agent acquisitions (though of course that’s where most of your budget goes).

  75. joser on August 29th, 2006 5:08 pm

    Best USSM thread ever?

    I can’t think of a better one. But it starts with the post, and Dave’s work here is among the top two or three articles I’ve ever seen on the site (and among the top baseball-related articles I’ve ever read on the web).

    On a related note: I expect this article is going to get linked from all over, and lots of people will be telling others to read it, so it will probably become one of the most-read pages on the site. If ever there was a page on USSM that should have either some google adwords and/or a (more) prominent link to buy USSM swag, this is it. Articles of this quality improve my enjoyment of the game, and that’s a rare and special thing — and worth a lot. USSM deserves support in general, but it’s articles like this that make me hope they get more revenue so they can keep the thing going as more than a sheer act of love. In a better world Dave would get paid a lot to write like this, but I fear in this world that will only happen in a way that forces the rest of us pay to read him.

    And it would be nice if USSM had the resources so its server didn’t fall over the next time, say, Willie Bloomquist is DFA’d.

  76. Dave on August 29th, 2006 5:17 pm

    The real question I have after reading this is how in the world can major league teams still be ignorant to this kind of analysis?

    There’s not a major league team that doesn’t have someone on staff who understands these concepts.

    It seems like the Mariners are willfully ignorant of this kind of in depth thought and it really hurts the team. Why don’t they invest in some statisticians?

    Looks can be deceiving. The M’s have Mat Olkin on payroll, and Mat certainly knows more about baseball than I do. None of this would even cause him to blink an eye.

    What will it take to get them to open their eyes to more reliable ways of judging talent? Surely its not just an old boys network of ‘baseball guys’ who are sitting in some wood-paneled room smoking cigars and talking about how they should trade for someone because, “the kid’s got heart, damnit!” Right?

    No, it’s not. The Mariners, especially, are nothing like that.

  77. Dave on August 29th, 2006 5:18 pm

    Nice job, David. You really should try writing for those THT guys.

    I find that they have too many guys named Dave as it is.

    Best USSM thread ever?

    I don’t know that this can compete with traffic in Idaho.

  78. Bender on August 29th, 2006 5:37 pm

    Looks can be deceiving. The M’s have Mat Olkin on payroll, and Mat certainly knows more about baseball than I do. None of this would even cause him to blink an eye.

    But do they listen to him? I mean the Washburn signing seems like they pretty much paid for ERA, not any of the better indicators, or am I missing something?

  79. Dave on August 29th, 2006 5:41 pm

    But do they listen to him? I mean the Washburn signing seems like they pretty much paid for ERA, not any of the better indicators, or am I missing something?

    They listen to him. They may not act on what he says at all times, and there are certainly other voices in the front office, but that doesn’t mean the M’s are “willfully ignorant” of this kind of analysis.

    It just means they value it less than we do.

  80. Steve T on August 29th, 2006 5:57 pm

    The reason I like this post so much is that it’s extremely clear. Dave isn’t always as easy to follow as this (though he’s far, far, lightyears far from the worst in this regard) — sometimes he blazes past the obvious into the world of acronyms too quickly, which is understandable. But once in a while it’s good to sit down and start from scratch, and painstakingly mark out exactly what you know, step by step, and what the conclusions are. This is valuable even for people who know it all already, or think they do. And for newcomers it’s invaluable. Bravo, Mr. Dave.

    The followup discussion has been top-notch, too.

    But none of this analysis can explain that little grin Felix tried but failed to supress in the eighth when I think it was Kennedy swung at a pitch that fell off the table.

  81. msb on August 29th, 2006 6:08 pm

    #71–One system (I think it had a name, and is in use by most if not all major league teams) stored years of indexed video, such that if you wanted to see all Maglio Ordonez at-bats against Jamie Moyer during night games at home (for example), you could.

    hey! the Mariners have one of those, it’s called Carl Hamilton :)

    actually it was this spring that Carl was the 1st recipient of the Professional Baseball Video Coordinators Association Award for Excellence. The award was also named the Carl Hamilton Award

  82. drjeff on August 29th, 2006 6:28 pm

    Bookmarked and forwarded to a number of friends. Thanks!

  83. MissingEdgar on August 30th, 2006 8:55 am

    Dave-

    [First time posting, so please be gentle if I inadvertently violate local etiquette.]

    I was wondering if “damage control” is a real pitching skill, whether it has been systematically assessed, and if it is comprehensibly explainable by the factors you’ve discussed. I assume that strikeouts and ground balls are good outcomes for a pitcher in trouble, but it seems to me that some pitchers are more than just lucky at getting these when most important. For years I’ve heard M’s broadcasters say that certain pitchers are “good at damage control” while others “can’t stop the bleeding.” Is the ability to “minimize the damage” a real pitching skill that carries over from season to season, or upon close examination does it regress to the mean?

    I’d think such a skill would be somehow evident in how the runs scored in scoring innings were distributed (fraction of runs scored in one run innings, fraction of runs scored in two run innings, etc.). I’d also guess that pitchers would get better at this as they matured, but would gradually regress as they lost their “stuff”. They’d learn what pitch to throw with experience and be able to make it, but eventually be unable to make quality pitches on demand.

    Obviously, I know my guesses don’t constitute data, but I don’t have any idea how to assess this myself. I’d appreciate your thoughts on the matter.

    Since this is my first time posting, I’d like to thank you for the many hours of enjoyment I’ve had at this site, and for helping me understand the game I love so much better.

  84. Dave on August 30th, 2006 4:36 pm

    I think it’s rare, but it exists in some extremely intelligent pitchers. Tom Glavine, for one, has consistently posted really low HR rates with runners on base, but his walk rate goes up, and with the bases empty, the walks go down and the homers go up. The evidence suggests that when Glavine gets in trouble, he nibbles and refuses to let hitters get a pitch they can clear the bases with. When he’s only risking one run, however, he attacks. As such, he’s managed to keep runners from scoring as often as we would expect a normal pitcher with his skill to do.

    But there aren’t many examples like Glavine. So, I’m going to say its possible, but it’s not something we’d expect to see in a pitcher.

  85. tangotiger on September 1st, 2006 1:05 pm

    Good summary Dave!

    ***

    For FIP, you adjust the constant to reflect the league. Since 1994, it’s been around 3.20. Pre-1994, it’s been 3.00. FIP is a quick equation, and therefore, try to keep it as simple as possible.

    ***

    “Why? Because a hit-by-pitch correlates pretty well with “struggling pitcher”, and so more struggles are likely to follow.”

    This part is not true. The fact is that a walk is given out more with 1b open, or 2 outs, than otherwise. That is, a walk is more non-random than a hit batter. If you look at each of the 24 base/out states (and given a large enough sample), the run values of the walk and HBP will be virtually the same, for each state. But, since the frequency of each state will be much different, this will account for the difference between a walk and a hit batter.

  86. john1313 on September 4th, 2006 7:55 pm

    Where do they show K% and BB% on the THT site and fangraphs….all i see is K/9 and BB/9……..not total batters faced.

  87. Evan on May 31st, 2007 1:51 pm

    Nate Silver has specifically complained that Glavine’s approach breaks PECOTA.

  88. teddyballgame on July 29th, 2007 9:30 am

    I’m skeptical. What are the actual results that prove the assertion? It seems to me that all of this misses the forest for the trees. Of course there’s a difference between a line drive out and fly ball out. WHIP and ERA don’t claim otherwise. They remain the most useful measure of trends and results across an entire season played out in all kinds of weather and with usable splits. I’ll bet that pitchers with a danger-zone high percentage of line drives and fly balls have high ERAs and high WHIPs, too. Those two numbers take less labor to derive, though, and I’ll stick with them until someone provides real data that shows me they don’t work.

  89. timc on July 29th, 2007 10:36 am

    I’m skeptical. What are the actual results that prove the assertion? It seems to me that all of this misses the forest for the trees.

    You could start with the fact that FIP is a better predictor of ERA than (past) ERA itself is. In other words, these component statistics have less year-to-year noise than ERA, which is saddled by dependence on luck/defense/opposition.

    Of course there’s a difference between a line drive out and fly ball out. WHIP and ERA don’t claim otherwise.

    Actually, WHIP and ERA do claim otherwise. They don’t distinguish between types of outs at all. For example, a pitcher who gets 3 strikeouts in an inning and a pitcher who induces 3 lineouts in an inning both have a WHIP of zero for that inning. The types of outs make no difference to the calculation of WHIP (or ERA).

    They remain the most useful measure of trends and results across an entire season played out in all kinds of weather and with usable splits.

    This is an assertion that you’ll need to provide some supporting evidence for.
    You should note that nobody refutes the idea that ERA and WHIP provide a picture of what has occurred. The issue is that what has occurred is coloured by the actions of the defense and by the quality of the opposing batters. In the context of evaluating pitchers, it makes sense to try to remove the influences of the defense/opposition, and that’s what the component statistics do.

    I’ll bet that pitchers with a danger-zone high percentage of line drives and fly balls have high ERAs and high WHIPs, too.

    Sure, and you can print out a document with an 80s-style dot matrix printer. But the picture is so much clearer if you use a modern printer.

    Those two numbers take less labor to derive, though

    There are several links to the Hardball Times in the post and comments. It’s not exactly a lot of labour to look up the numbers. And if you want to calculate them yourself from the raw numbers, it isn’t hard either: BB% = (# walks)/(# batters faced), and so on.

    and I’ll stick with them until someone provides real data that shows me they don’t work.

    The Jarrod Washburn example in the post is a good example of how ERA can break down (high strand rate in this case) but unfortunately I can’t find anything which shows the poor year-to-year correlation of ERA. Anyone?

  90. equaltojake on December 18th, 2007 4:27 pm

    Oh dear.
    I thought from his FIP that at least he wouldn’t be a really bad pitcher, but from what you’ve said it sounds like his chances of collapsing are actually pretty high. Which means that 4th year (if not the others) could turn out to be a real pain.
    Why is there never good news?
    I guess that at least Silva could be OK for a while if his command lasts, whereas Kuroda might be dreadful from the start.
    That silver linings looking thinner all the time though.

  91. Taylor H on November 11th, 2009 2:21 pm

    Yeah!

Leave a Reply

You must be logged in to post a comment.