M's #3-6 Starters Have Been Snakebit
Saunders, Maurer, Harang, Beavan all deserve better


You can talk all you want about Dustin Ackley, Jesus Montero, Justin Smoak and Jeff Clement.  But the fact remains:  the Mariners' team offense is 97.  That ain't the problem.

Further:  the Mariners have 5 guys hitting the ball real well ...

  • All the old guys that Zduriencik just brought in -- the 3 OF's and Morales
  • Seager

And the other 4* guys, they just now swapped out.  Meaning that --- > the M's have had a solid offense, and they just upgraded it.  So, fine.


Has the back of the rotation been THAT bad?  It has, but the funny thing is, it shouldn't have been:

Joe Saunders 5.20 4.58 1.6
Aaron Harang 5.82 4.20 4.3
Brandon Maurer 6.93 4.84 1.9
Blake Beavan 8.27 4.71 2.8

Scan down the ERA column; that's what the Mariners have coughed up on the actual green grass of the AL's* baseball diamonds.  Let's round it to integers, what say:

  • 5 ERA - Saunders
  • 6 ERA - Harang
  • 7 ERA - Maurer
  • 8 ERA - Beavan

I suppose you know what happens when you have starting pitchers giving up 6, 7, and 8 ERA's.

But the funny thing is -- scan down the xFIP columns, and ask "what if all these pitchers had ERA's comfortably in the 4's?"

For those who just joined us, or for those who have little interest in fine-grain baseball stats, xFIP tells you what a pitcher's ERA would have been if he had average defense, average balls-in-play-for-hits, and average HR's-per-flyball.


The $64,000 Question:  ERA-xFIP, luck or bad pitchability?

It's pretty blinkin' tough to run an ERA that is a run larger than your xFIP, and here the Mariners are with ALL of their #3-6 starters throwing one crap-out roll after another.  I mean, Blake Beavan has allowed +8 runs over his xFIP expecation, and that's in only 2 starts!

Here are the runs that each SP has allowed in excess of his xFIP xPectation:

  • +5 = Saunders
  • +8 = Beavan
  • +11 = Maurer
  • +8 = Harang

Those four SP's alone have "unlucked" into +32 runs beyond their K/BB/flyball outcomes.  And Jeremy Bonderman, oh by the way, allowed +4 runs beyond his xFIP in his first start.

The Mariners have currently scored +210 runs, and allowed -253.  That projects to a 24-34 record; they are one game better than that.

Had their #3-7 starters had ERA's matching their xFIP's, they'd have scored +210 runs and allowed -217, for a 28-28 to 27-29 expectation.


Funny thing is, xFIP presumes league-average defense.  The Mariners' DER is #5 in the AL and #12 in baseball; if you assume the SP's are snakebit, the defense is a lot better than that.

Had the M's starters been pitching in average luck, they'd probably be .500 or a little better.


Jeremy Bonderman is the poster boy for terrible pitchability, for EARNING an ERA that is a full run worse than his K/BB/HR's should predict.

Are the other four guys like Bonderman this year?  What do you guys think?  Have you seen a blizzard of pitches that you knew were coming as they left the pitchers' hands?  Or a landslide of middle-middle sliders?

I haven't.

To the extent that you fly by your instruments, not your instincts, to that same extent you would reason "let Saunders, Maurer, and Harang pitch.  They're due a little bit of good luck."

I dunno.  Mighta started on Monday.  Saunders whupped the Sox pretty easy.




Clearly those guys haven't been shafted by poor defense. I don't remember them being dinked and blooped to death. Homers, however, tell a different story.Saunders gives up 1.3/9 innings. OK, there, not bad. But Harang is at 1.7, Maurer 1.8, Beavan 2.2 and (in one start) Bonderman 5.8.
How many of the homers these guys give up are rocket shots and how may just get over the fence? I suspect they've been hit pretty hard on their mistakes, considering none of them are overpowering guys.
BTW, I want Maurer back in the rotation.

M's Watcher's picture

Since last season, Beavan had been bad-mouthed on the local forums. However, if he could have merely replicated last season, most here should have been pleased. But he showed up in ST with messed-up, "improved", downward plane mechanics and has not looked like the solid back-end starter he had been. Whose idea was that? Somebody took a former tall, hard-thrower and tried to make him Doug Fister. Picture trying to make RJ into Jamie Moyer. I hope Blake has some salvage value.
We just need Harang and Safeco Joe to pitch to their career averages, not the 5-8 ERAs we've seen. If they are not done yet with their careers, they should come around. Eat innings and buy time for the kids to mature.

blissedj's picture

I know this article covers pitching but can't help but be struck by the 97 for the M's offense. I don't know the stat well enough. Is it like xFip where it takes into account certain rate stats but is not necessarily a measure of results on the field? So we are basically average, yet I look in the Runs Scored Column and there we sit tied for 14 of 15 teams. Is the 97 score park adjusted and the tweaks to Safeco dimensions have made no difference? Struggling to see how we rank almost last in a very important category, actual runs scored on the field of play yet another stat says "hey, nothing to see here, league average". I don't even know the name of the stat that says we are "97", if you could provide it I'll read up on it myself and hopefully it will make more sense because my eyes and the scoreboard still see a team struggling to score enough runs.

blissedj's picture

It was just team OPS+ I guess. Sandy made a post about it regarding the Astros. Familiar with that, brain freeze:) Interesting that Houston is a 90, M's 97 yet Houston outscores us by 19 runs so far. Sorry for the clutter, feel free to delete if you desire :)


In all fairness, it is likely that the Ms offense is skewing a point or two high and the ERA a point or two low, because the park effects equation is based on PAST results.
That means it is going to take a little while for the new park effects to get measured and the adjustment to be properly included in the final + result.
That said ... the results from the guys who played in last years park and this year's park by and large don't indicate that the change is "drastic".
Seager is hitting a little better ... but that isn't unusual for young players and his current pace is to finish around 20 HRs.
Smoak is getting more HRs.
Saunders, after a quick start, is slumping is like Smoak is on a slower pace of HRs.
Clearly, Ryan, Ackley and Montero didn't get any major boost from the park reconfig.
That said, the current bbref park factors for the Ms THIS YEAR are: 93 / 94
For the most recent 3 year period ... 90 / 91
So, maybe the "correct" number is closer to 95 OPS+ and 93 ERA+. The simple truth is that teams like Oakland and Seattle are going to have lower run totals (both sides) than teams like Boston and Texas on a consistent basis. And in those rare occasions when a team like Boston or Texas is allowing the fewest runs, (Texas is currently at 188 earned runs allowed, 3.61 runs per game, best in the AL), it is ASTOUNDING. But, that's where ERA+ and OPS+ help show exceptional from pedestrian behavior. Texas' ERA+ of 133 is off the charts good. And they will probably give back a chunk of that when the warm weather jet stream hits Arlington later this year.
The trick is to not only look at one side of the park effect.
Seattle is major pitchers park. Therefore, offensive numbers are suppressed. So, an .800 OPS is actually more impressive for a Mariner than an .800 OPS for a Ranger.
Problem is ... when the park is helping, some people want to ignore that edge. Felix is a great pitcher. No question. But, by ERA he ranks 4th among active hurlers, while by ERA+ he ranks 7th. Oswalt, Santana and Halladay all jump past Felix after adjusting for home parks. That's actually a very reasonable adjustment. Understanding that Seattle suppresses offense, while Houston has been a hitter heaven ever since they left the Astrodome makes perfect sense that one guy is getting help and the other is pitching "up hill".
(I just wonder how many people are aware Adam Wainwright is actually 3rd on both the ERA and ERA+ lists among active pitchers - only trailing Marino and Kershaw ... but I digress).

blissedj's picture

Very well explained! One more thing I am curious about when calculating park factor for hitting in Safeco, are M's batting statistics weighted as 50% of the data, or are the M's hitters weighted equally with the other 14 teams in the American League? It would seem if you have had a decade worth of lousy home team hitters making up 50% of the park factor it would skew the results terribly and just be a reflection of whether the home team had great or poor hitters. If M's hitting results only account for 1/15 of the park factor data that seems much more logical. I read a bit on baseball reference about how they calculate park effect but was not clear on my first look at the information how they weigh the performance of the home team.


The important part to remember is that the calculation compares Seattle hitters (home) vs. Seattle Hitters (away).
In that respect, it is (mostly) irrelevant whether your hitters are good or bad. It is how much better or worse they hit at home versus on the road.
Say Seattle hitters OPS .750 at home and .800 on the road. The calculation is looking at the SKEW (that 50 point swing), not whether they are good or bad.
Same goes with pitchers. If Seattle pitchers accumulate a 3.50 team ERA at home and go 4.00 on the road, the skew is what matters.
So, the visiting teams are composited, meaning they are 50% of the rating ... but of course every opposing hitter is facing a Seattle pitcher for the calculation (and vice versa).
The one mitigating factor in all this is if a team manages to construct itself to take particularly good advantage of its park (or constructed horribly to be disadvantaged), then talent *could* skew the park effect. But that is part of the reason they calculate both a pitcher and hitter park effect as two numbers. That can potentially reveal any outlier results where ONLY your offense or defense is seriously impacted by the park.
While I don't know the entire calculation, the basic gist of it looks at the home/road skews. The "official" park adjustments are calculated using 3 years of data, as it is felt single season values are not as reliable (prone to random fluctuations). That's what I call the weather effect -- if one year the wind happens to be blowing in for 60 games at Wrigley and out for the other 21 ... and then the next year it is 40/40 ... that's going to impact the park effect. The multi-year method attempts to factor out the random fluctuations to reveal the "static" impact. (Of course, when you move the fences around that kind of throws the whole 3 year calculation out the window. In that regard, it will be 2015 before the Safeco park effects will become "vested".

blissedj's picture

Really appreciate you taking the time to explain this in more detail. Just looking over some formulas wasn't doing it for me :)

Add comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p><br>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.


  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.