Game 131, Angels at Mariners
Julio Mateo is out for the season with a broken left-hand. The Mariners have purchased the contract of Jon Huber from Tacoma, adding him to both the 25 man and 40 man rosters. He has had a very good year for the Rainiers, but has marginal, Mateo like stuff. He’s not part of the future here, but the M’s are just swapping out one replacement level reliever for another.
Escobar vs Hernandez, 7:05 pm. Feliz Dia de Felix!
I’m surprised no one has mentioned this in the local media, but the Mariners have lost 20 consecutive games to AL West opponents. If the M’s are going to break that streak, they’re going to have to do so against a quality pitcher and a line-up that is significantly better than it was earlier in the year. With the addition of Howie Kendrick and the realization that Juan Rivera is an everyday player, the Angels offense has improved tremendously since the last time the Angels rolled into town. They’re no longer Vlad and the eight dwarves.
Escobar, meanwhile, is a lot like Current Felix – inconsistent command leads to bouts of wildness, but high strikeout rate keeps runners from scoring and he’s generally effective. Either one has the ability to toss a shutout tonight, though with the M’s running out a line-up including Doyle hitting second, the M’s aren’t nearly as likely to go down helplessly as they were before his arrival.
This could be a fun game. The M’s just got to put an end to Boston’s season over the weekend, and now they have a chance to do that to the Angels as well.
Great line-up 1-8:
#9: Bloomquist, SS
Seriously, Felix is pitching, Mr. Ground Ball, and you give Betancourt a night off? Just insane.
Comments
277 Responses to “Game 131, Angels at Mariners”
Leave a Reply
You must be logged in to post a comment.
Dave, great points. What you are saying is that each stat is a piece of a larger puzzle. ERA is a small piece that actually fits in a bigger piece…say FIP. K/9 is another piece, etc. If you look at all the pieces you start to see the puzzle take its true form.
With my comments, I’m just trying to evaluate how much better one stat is over the other. Its too easy to say, ERA and WHIP suck (not that you are saying that, but you get my point). I’m trying to see how inferior they really are.
By the way, thanks for responding to my posts. You may get tired of answering questions and comments like mine all the time, but your time and thoughts in relation to my comments are appreciated.
OK, who farted in this thread?
Or, to frame this debate in a different way, getting away from FIP/xFIP vs ERA/WHIP. (and yes, I’m writing this so I don’t have to keep writing it every time this comes up)
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.
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.
ERA and WHIP group together a large string of individual events made by multiple players, making it extremely tough to parcel out the credit to the pitcher. WHIP and ERA tell you there is no difference in an inning where three batters whack the crap out of the ball 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 contribution of the pitcher.
So, instead of using statistics that leave out critical information, our best bet is to try to quantify the six potential 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.
K% (Strikeouts per Total Batters Faced) does a decent job evaluating how often a pitcher induces swings and misses or called strikes.
GB% (Groundballs per Balls In Play) does a very good job of telling us how often a pitcher induces a groundball.
LD% (Line Drives per Balls In Play) does a very good job of telling us how often a pitcher gives up line drives.
FB% (Flyballs per Balls In Play) does a very good job of telling us how often a pitcher gives up flyballs.
So we have five statistics that cover each of the six possible outcomes pretty effectively. Not perfect, but pretty good. Now, using these statistics, we’ve noticed that line drive percentage seems to be fairly random from year to year, and we currently believe that the hitter has more to do with whether a ball in play becomes a line drive than the pitcher does. Since we’re trying to evaluate the pitcher, we don’t use LD% very often for major league pitchers.
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. So, when evaluating pitchers talent, we need to adjust for outliar 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.
The other two 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 predictive, 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.
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 get an out 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. This is something of a reflection of skill.
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 how a pitcher is performing. There’s nothing that ERA/WHIP tell you that those component statistics do not, but ERA and WHIP certainly leave a lot of the underlying information out.
In this age of wonderful information, there’s just no reason to use ERA and WHIP for serious analysis of a pitcher’s ability.
I’ll concede. Maybe you have all these stats at your finger tips, but for me when comparing pitchers, the most readily available stats you can find online are era and whip. I never said it was the end all be all for pitcher evaluation. It is a good barometer and easily accessible.
Go to the hardballtimes.com. The site loads faster than ESPN, MLB.com, or whatever site you’re using to get ERA/WHIP, and these “fancy stats” are available for every pitcher in baseball for the last three years. You’ll love it.
Dave…I now believe. This post should be linked when this comes up in a week. Very easy to follow the logic in this post.
It would be nice if in the future these stats meant something to the average fan. Like when someone says .300 hitter, it means something to everyone who watches baseball consistently. Maybe in 10 years, GB% will be stated and we will all nod our heads in understanding.
Any chance post #253 can be put on the front of the next USS Mariner t-shirt?
#233– he’s been wearing the brace since he came back in … was it May?
#241– I believe that was actually a mocking paraphrase of the interesting logic of the ESPN broadcast team covering the As/Sox game
While waiting for the Pirate All-Time webgems on BBTN, I did get to watch Kruk lay out the Phillies’ horrible play around Moyer in that 3rd inning this morning. When they came, the Pirates “top three plays” are a Jack Wilson DP, Dave Parker’s throw in the AS game, and Giles reaching over the wall to bring back a HR. Huh?
Boy, now I can’t wait to see what they deem the top three plays for the Mariners on wednesday.
Dave…I now believe. This post should be linked when this comes up in a week. Very easy to follow the logic in this post.
That’s the plan. The whole goal in writing that comment was to have something I could link to the next time this comes up.
It would be nice if in the future these stats meant something to the average fan. Like when someone says .300 hitter, it means something to everyone who watches baseball consistently. Maybe in 10 years, GB% will be stated and we will all nod our heads in understanding.
Yea, I realize the lack of a baseline can be a little confusing. So, let me throw a few at you, and try to figure out a way to get them into that comment:
BB% – 8% is about average. The AL is a little higher than the NL, thanks to the DH. Anything below 5% is spectacular.
K% – 16% is about average. This time, the NL is a little higher, but its still due to the DH. Anything over 20% is very good.
GB% – 42% is about average. Anything over 50% is very good. The most extreme guys are in the low-to-mid-60% range.
HR/FB% – 11% is the league average, but the more I look into this, the more I believe we need an average for relievers and an average for starters. I don’t have those baselines yet, but the number for relievers will be lower (probably 9% or so) and higher for starters (12%, maybe 13%).
LOB% – 70% is average. This one is a little tricky, because small swings in LOB% can produce big swings in ERA. A 67% LOB% and a 73% LOB% for a pitcher with the same peripherals is probably half a run of ERA, even though its “only” a 3% swing. Anything over 75% or under 65% is probably not a sustainable skill for a major league quality starting pitcher.
And, while I’m still here, I should note that most of this applies to starting pitchers. Relievers are a very weird, very different breed, and they represent a group of very polarized talents. We can still use these statistics to value relievers, but their jobs are different than starters, and the baselines are necessarily different. Since stranding runners is a key requirement for a reliever, LOB% is going to make up a disproportionate amount of their value.
The best way to approach evaluating relief pitching is to admit that they’re notoriously fickle and proceed with caution.
Any chance post #253 can be put on the front of the next USS Mariner t-shirt?
It would probably need to be on a dress instead.
the USS Mariner BBQ Apron…
This was an awesome game to attend. I’ve had great luck in both of my Felix starts this year–today’s game and the game he out-dueled Liriano. When Good Felix is in full effect, good times will abound in Safeco.
Seriously, Felix is pitching, Mr. Ground Ball, and you give Betancourt a night off? Just insane.
I had a similar thought at the beginning of the game. Can we put an asterisk on this start that notes Willie was at short? 🙂
#257: #241– I believe that was actually a mocking paraphrase of the interesting logic of the ESPN broadcast team covering the As/Sox game
Ah, that makes more sense. I went out for dinner and missed a big chunk of the games. (I was flipping back and forth before I went out.)
And it’s hard to imagine that the news for the Red Sox can get worse, but David Ortiz is scratched from the series and going back to Boston due to a recurrence of heart trouble.
Re 253.
Beat the shit out of my post. I’m copy / pasta’ing that thing to a .txt file just in case. Hell, maybe a .doc to add some cool borders and fonts. Maybe a few wingdings just to spice up the thing. You said you wrote that so you wont have to again, so pointing out that you should put that in the FAQ, or as an automatic obligatory link at the beginning of every game thread, is unecessary.
I want to ask though, whether we should ever take LOB% seriously. Much in the same way that HR/FB% is skewed against groundball pitchers (since groundball pitchers tend to have a higher percentage of their flyballs go for homeruns, simply because they throw fewer flyballs) LOB% could be skewed heavily one way or the other. Say you allow fewer runners, but more of them score. The LOB% would naturally be higher then someone who allows a lot of baserunners and more runs to score, but their LOB% looks better because of a larger set sample.
I’d also like to go around and test the theory of “getting ahead” in the count. Announcer guys like to say that those pitchers who throw strike one on the first pitch are better. But I’m wondering if there’s any corrolation between consistent first pitch strike throwers and whether they match up with what we observe with the 6 outcomes.
#258: The best way to approach evaluating relief pitching is to admit that they’re notoriously fickle and proceed with caution.
How much of that variance is due to the low IP numbers for relievers? It seems dangerous to divide by small numbers and make rate projections when you are dealing with small denominators.
242: If you’re trying to figure out what a pitcher will do in the future, however, ERA just isn’t very useful. The year to year correlation is quite low, something like .3, if I remember correctly.
246: Dave, I agree with your logic. One question, what is the year to year correlation between FIP and xFIP year to year?
It’s not super high. I can’t find the studies at the moment, but I think it was .52 for FIP and .56 for xFIP or something.
Good discussion and explanation, the one thing that’s missing is the data to support the statements that certain components (GB%) are relatively repeatable by a pitcher whereas others (HR/FB%) are not. E.g. what are the correlations of those components from year to year?
No I’m not actually asking for the data or the correlations, just point us in the direction of who did the studies and where to find them. I’m very familiar with the work that Voros and Tom Tippett did on the low correlations of BABIP from year to year, but I’m not familiar with the studies of the correlations of ERA, FIP, GB%, etc. from year to year. For awhile people liked to talk about the “Three True Outcomes” and look at BB%, HR%, and K% but what you’re describing goes farther, if I’m understanding correctly HR% is not as useful a measure as FB% … and HR/FB% is largely a matter of luck. Those are the results that I’m not familiar with.
#258: That’s the plan. The whole goal in writing that comment was to have something I could link to the next time this comes up.
This kind of analysis is easy to lose at the bottom of 200+ comments in a game thread. I wonder if you’d consider breaking it out into its own post. I expect you’d get a nice discussion going that would educate a lot of readers.
And finally, for anyone who was wondering while reading my long winded rambling above, here are the run values as calculated from 2002-2005 by Dave Studeman in the Hardball Times Annual for each event that can occur in an at-bat. Lumping things into “Good Outcomes” and “Bad Outcomes” makes the explanation simple, but also hides the difference between a great outcome (a strikeout) and a slightly good outcome (a groundball). Not all good or bad outcomes are created equal, and a run value chart helps show the difference in event quality:
Line Drive: .356 – basically, a 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 correlates pretty well with “struggling pitcher”, and so more struggles are likely to follow.
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 be a groundball pitcher, not a flyball pitcher.
Also, bunting = bad.
Hope this is informative.
Dave, you are on a roll. Do I get any credit for beating this out of you? I’m like the moses to your burning bush. Or more likely, I was the straw that finally broke the camel’s back.
Following your logic, I would expect LOB% to be more variable because it combines multiple outcomes from multiple at bats. Just like a pitcher who strikes out the side with the bases loaded is different from a pitcher who “induces” three consecutive batters to line out.
LOB% seems more like WHIP and ERA than GB% and K%. It shows what occured in the past, but doesn’t seem as robust in identifying future performance.
I want to ask though, whether we should ever take LOB% seriously. Much in the same way that HR/FB% is skewed against groundball pitchers (since groundball pitchers tend to have a higher percentage of their flyballs go for homeruns, simply because they throw fewer flyballs) LOB% could be skewed heavily one way or the other. Say you allow fewer runners, but more of them score. The LOB% would naturally be higher then someone who allows a lot of baserunners and more runs to score, but their LOB% looks better because of a larger set sample.
Absolutely we should take it seriously. It has a significant impact on run scoring. If you wanted to explain to someone why Felix was struggling, his 66.9% LOB% this year would be a great place to start. When he’s put guys on, they’ve scored. That’s been a problem.
Is it predictive? Kind of, but not really. Not any more so than establishing an expected LOB% based on a pitcher’s other component rates, anyways.
How much of that variance is due to the low IP numbers for relievers? It seems dangerous to divide by small numbers and make rate projections when you are dealing with small denominators.
That’s a huge chunk of it, no doubt. Reporters love to quote things like Julio Mateo’s 2.42 ERA in his last 21 outings, but that covers a whopping 22 innings. If he had given up 10 runs instead of 6 during that time, his ERA would be nearly 4.00. Using rate stats in small samples can make the differences look much larger than they really are.
Good discussion and explanation, the one thing that’s missing is the data to support the statements that certain components (GB%) are relatively repeatable by a pitcher whereas others (HR/FB%) are not. E.g. what are the correlations of those components from year to year?
Most of this work has been done by the guys at The Hardball Times. David Gassko and J.C. Bradbury have an article in the 2006 Annual dealing with these issues. You really should buy the book, but here are some year to year correlations for you:
K%: .77
BB%: .69
HR%: .28 (that’s not HR/FB, by the way – thats HR/TBF)
LD%: -.03 (like I said, pretty random)
GB%: .79
FB%: .73
The THT article has a ton of other great stuff, too, including more accurate baselines than I included above and correlations on a bunch of other statistics, plus the analysis that goes along with things like HR/FB rates and outs on balls in play.
if I’m understanding correctly HR% is not as useful a measure as FB% … and HR/FB% is largely a matter of luck.
Right – once you account for FB%, home run rate is almost totally random – .08 year to year correlation.
This kind of analysis is easy to lose at the bottom of 200+ comments in a game thread. I wonder if you’d consider breaking it out into its own post. I expect you’d get a nice discussion going that would educate a lot of readers.
Not a bad idea.
LOB% seems more like WHIP and ERA than GB% and K%. It shows what occured in the past, but doesn’t seem as robust in identifying future performance.
Bingo. It’s still good information to have, because if you’re looking at a guy with a low walk rate, a high strikeout rate, a low home run rate, and a high ERA, you’ll wonder what the heck is going on. That theoretical pitcher almost certainly has a ridiculously low LOB%.
Is it my imagination or does it seem like with each win that the possibility of Grover coming back INCREASES?
270,
I may be going out on a limb here, but I don’t think there’s a snowball’s chance in Hell Grover comes back next year. Bavasi will find a way to get rid of him. I just think that Bill’s had it with Grover’s mindset. I realize Bavasi said that he just provides the pieces and says nothing to the manager about using them, but when the manager never uses them at all, and you’re spending good money to have them there, it can get irritating. Hargrove’s misuse of his bench, bullpen, and other factors warrant his dismisal alone, but “wasting” the company’s money? Bavasi can put that through the FO.
Thanks for the hardballtimes link. I am sure it will provide hours of gazing at numbers for me. As it is very late on the east coast I haven’t gotten through all of your, now several, posts. Hopefully at the end of this I will come to a better understanding. But I do have to say, that all these things are a larger piece of a puzzle, as pointed out. But I don’t understand why GB% is better than FB%. A ground ball has just as much of a chance of finding a whole, as a fly ball does dropping in. Of course a grounder can never “fly” out of the park. And none of it really acconts for a pitchers ability to fool hitters and inducing outs.
Anyway, as I said it is very late and I’ve already stayed up later then I wanted. I will review the rest of the post and hopefully arise at a clearer understanding.
But I don’t understand why GB% is better than FB%. A ground ball has just as much of a chance of finding a whole, as a fly ball does dropping in. Of course a grounder can never “fly†out of the park. And none of it really acconts for a pitchers ability to fool hitters and inducing outs.
A groundball will be turned into an out approximately 61% of the time, while a flyball will be turned into an out approximately 78% of the time. The grounder actually has more of a chance of finding a hole than a flyball does of dropping in.
The issue, though, is that when a groundball “finds a hole”, its a single, which isn’t a huge deal. When a flyball “drops in”, its a double, a triple, or a home run, which is a huge deal. The run value of a home run is 1.45 runs, while the run value of a single is just .45 runs.
So, the groundball pitcher will give up more hits, but less impactful hits, than the flyball pitcher.
That’s why I said “all things being equal, the groundball pitcher is better.” But again, things are almost never equal, so I’m certainly not advocating acquiring a staff full of groundball pitchers and ignoring all the other skills.
RE 272: Ground balls actually go for hits more often than fly balls, but hardly ever for extra bases. Essentially, you’re trading a couple points of opposing OBP for a huge drop in slugging. Then you combine that with higher chances to get a DP ball when you need it, etc…
268: Most of this work has been done by the guys at The Hardball Times. David Gassko and J.C. Bradbury have an article in the 2006 Annual dealing with these issues. You really should buy the book, but here are some year to year correlations for you:
K%: .77
BB%: .69
HR%: .28 (that’s not HR/FB, by the way – thats HR/TBF)
LD%: -.03 (like I said, pretty random)
GB%: .79
FB%: .73
Excellent, thanks for both those answers (the cite and the correlations).
Wow. Nice posts Dave. Glad I checked in tonight.
I appreciate the developed series or posts here, Dave, which put your thinking and the relevant information in a compact context. While you’ve said these things before, these are the numbers _which really count_, and why they count. Numbers aren’t everything, but these numbers are among the best tools going. Knowledge+context+power.