PTI on ... the 2001 M's Pitching

Q.  The 2001 Mariners did not have a terrific pitching staff. They had a league average pitching staff…below average rotation, well above bullpen.

A.  This is key, because it goes to the broader question of:  what happens when you have a sensational DER?, because the 2001 M's had a DER that was, quote, "eerily better than the rest of the league" as Baseball Prospectus put it in their 2002 annual.

I'm a fan of DNRA+ and always like to know what it says.  But all defensive (and defense-oriented) metrics will have their fliers, and this is one of them.  Any metric that accuses the 2001 Mariner pitchers of being "league average" is blowing a circuit in that loop.  :- ) Consider:

.............................

1.   There are three things that a pitcher can do, that are not affected by whether it is Jason Ellison, CF playing LF or whether it is Adam Dunn playing LF.  Those three things are:  strike three, ball four, and home run.   Neither Dunn, nor Ellison, can do ANYTHING whatsover about those three pitching results.

........................

2.  When you examine a pitcher according to those three pitching results, then you are getting an UNTHEORETICAL Hubble-telescope view into the murky nebula of whether a pitcher is winning his matchups.

UZR is theoretical.  DNRA+ is theoretical.  xERA is theoretical.  Strike three is something that happened on the field.

........................

3.  The 2001 Seattle Mariners pitchers, as a group, struck out 6.46 men per nine innings; this gave them the #5 total in the league, #5 out of 14 teams.  Four pitching staffs fanned more hitters; nine staffs fanned fewer.  This is not league-average.  It is not "below average."  This is above average.  

#5 in strikeouts is an important indication (one indication) that they were good.  Good within the 60 feet, 6 inches we are talking about, not good because of anything Carlos Guillen was doing at shortstop.

.........................

4.   The 2001 Mariners pitchers, as a group, walked a mere 2.86 men per game; this gave them the #3-best total in the league.  Two pitching staffs walked more batters; 11 staffs walked more.  This is not league-average.  This is excellent pitching.  

That 2.86 is not the walk rate for the relievers.  This is the walk rate for all 14 pitchers used by Lou Piniella that season.

When a pitcher is walking very few batters, that does not suggest to the objective mind that he is a below average pitcher. It is an indication, one indication, that he is a good pitcher.

..........................

5.  When a pitcher is good at BOTH striking out lots of hitters AND walking very few, it is very rare that he is an average pitcher, and these instances always occur when he gives up lots of home runs (he grooves the pitches on 3-1 so as to never walk anybody).

............................

6.  The 2001 Mariners pitchers, as a group, gave up ranking #5th in the league.  This means that the Mariners were good at ALL THREE "pure" pitching outcomes:  strikeouts, walks, AND home runs.

...............................................

7.  If it is possible for a pitcher to be above average at K's, BB's, and HR's, and yet be mediocre, I have no idea how that would work.

Do you think you could find a pitcher who was good at K's, good at BB's, and good at HR's, over a period of 1,465 innings -- and who was perfectly average?

Supposing you could.  How many could I find who were not average?

What are the odds that the 2001 Seattle Mariners' pitchers were good at K's, good at BB's, good at preventing HR's, and yet were average?  You guessed it.  The odds approach zero.

The 2001 M's had three stars in the rotation, two solid guys, and one questionable guy.  Then they had a mind-blowing bullpen.  IMHO, those guys, the pitchers, were the biggest cause of the excellent ERA.

Cheers,

Dr D

........................

image:  http://www.fhcrc.org/about/pubs/center_news/2004/jan22/Moyer_Jamie_pitch_vert.jpg

Comments

1
Sandy - Raleigh's picture

Was the 2001 pitching staff above average? No question. But, by definition, HALF of all baseball staffs are above average.
But, let's look at the TTO teams that WERE better than the Ms in 2001 -- and then examine the numbers, and see what makes sense for explaining their #1 ranking in fewest runs allowed:
rk -- team --- Ks -- runs (rank)
1) Yankees 1266 - 713 (3)
2) Red Sox 1259 - 745 (5)
3) Indians - 1218 - 821 (9)
4) Oakland - 1117 - 645 (2)
5) Mariners - 1051 - 627 (1)
The Yankees and Red Sox not only struck out more than the Ms -- they each fanned OVER 200 more batters than the Ms.
rk -- team -- BBs -- runs allowed
1) -- Oakland - 440 - 645 (2)
2) -- Twinkies - 445 - 766 (7)
3) -- Mariners - 465 - 627 (1)
3) -- Yankees - 465 - 713 (3)
9) -- Boston --- 544 - 745 (5)
Oakland was 25 walks better, but the Yankees were exactly tied with the Ms in fewest walks allowed. Boston trailed the Ms by 80 walks.
rk -- team -- HRs -- runs allowed
1) -- Boston - 146 -- 745 (5)
2) -- Indians - 148 -- 821 (9)
3) -- Oakland - 153 - 645 (2)
4) -- Yankees - 158 - 713 (4)
5) -- Seattle -- 160 - 627 (1)
Gee, the Yankees beat out Seattle by two HRs, tied in walks, and struck out 200 more batters. And they "only" allowed 86 more runs -- roughly 1/2 run PER GAME for the season. How is this possible?
rk -- team --- Hits -- runs allowed
1) -- Seattle -- 1293 - 627 (1)
2) -- Oakland - 1384 - 645 (2)
3) -- Boston - 1412 -- 745 (5)
4) -- Yankees - 1429 - 713 (3)
5) -- Anaheim - 1452 - 730 (4)
Remember how the Yankees and Boston both struck out an extra 200 batters? Well, Seattle allowed more than 100 FEWER hits, despite getting 200 extra balls in play.
Notice, also, that hits allowed shows the CLEARLY strongest correlation to runs allowed. The top 5 teams in hits allowed are ALSO the top 5 in runs allowed, with just minor shuffling. There aren't any stellar hits allowed teams finishing 9th in runs allowed.
Mind you - given "equal" defenses, the higher K team will end up better off. If you're trying to judge which PITCHING staff is better, then yes, you look at K, HR, BB. The Yankees *CLEARLY* had the best pitching staff in the AL in 2001. They beat or tied Seattle in EVERY metric -- and they DESTROYED them in strikeouts. And they finished 86 runs inferior to the team with the historically good defense. Note: This is NOT the same as historically good PITCHING.
By DER, in 2001:
1) Seattle - .727
2) Oakland - .702
3) Twinkies - .700
4) ChiSox -- .700
5) Anaheim - .699
9) Yankees - .683
10) Boston - .683
"Can" a stellar strikeout rate overcome a poor defense? Absolutely. The Yankees and Boston struck out an extra 200 guys, and managed to ONLY finish 80-100 runs behind Seattle.
Anaheim wasn't good at ANY of the pitching Metrics, yet finished 4th in runs allowed:
10th in Ks (947); 7th in HR (168); 7th in walks (525) -- yet 4th in runs allowed -- why? Because they were 5th in hits allowed.
Hits allowed *IS* the single largest factor for determining runs allowed. It OVERWHELMS the other categories in most cases. But, hits allowed is not PREDICTABLE. And since it isn't predictable, it is by and large waved away by many of the number crunchers.
Anyone claiming the 2001 staff was "mediocre" is speaking in hyperbole. They were above average. But, they weren't "stellar". They weren't "incredible". But, that defense behind them. THAT was a Hall of Fame defense.

2
Taro's picture

Good post Sandy.
The 2001 M's pitching staff was above-average. Average SP, great bullpen. The D was historic.
The 2003 D was possibly even more impressive with guys like Ryan Franklin running ERAs in the mid 3s.

3

How exactly do we know that it was defense that was historic (what about 2002?) and not that their luck was historic? Is there randomness anywhere considered here?
I happen to believe that defensive metrics are trying to measure something that they don't have enough data to do. Think of all the unaccounted for variables:
1. Speed of batted ball
2. Angle of batted ball
3. Spin of batted ball
4. Fielder positioning
5. Fielder energy level (road trip? extra innings the night before? Cleveland snowstorm makeup games?)
6. Length of grass
7. Wind speed
8. Wind gust speed
9. Timing of wind gust versus ball height
10. Favorable or unfavorable wind (do we assume they cancel out? why?)
11. Pujols has a cold the night you play them and does not have his 'usual' batted ball trajectory
12. Umpire strike zones and which ones cause teams to have to swing at more marginal pitches
I'm sure there are many more - point being: IMHO, using UZR and other measurements for predictive value seems like fool's gold.
While I admire organizations for looking into all aspects of the game, this is one where the trained eye has to be in play.
That's my .01
-R

4
Sandy - Raleigh's picture

Well, one can assume that EVERY stat skew from a single season is luck.
Maris' 61 HRs -- clearly luck.
Sandy Koufax 382 Ks in 1965 -- clearly luck, (his next best was only 317).
Ichiro's 262 hits -- obviously luck.
And, of course, 116 wins -- plainly luck.
Does anyone see a problem with this logic?
Do we UNDERSTAND defense? No, we don't.
But, Seattle DID have:
a .727 DER in 2001
a .704 DER in 2002
a .721 DER in 2003
Were they lucky in BOTH 2001 and 2003?
Or were they just unlucky in 2002?
"We cannot explain it" is not the same as "it didn't happen."
When the same club has two of the highest DER's of the modern era in a 3-year span, I'm inclined to NOT readily attempt to classify the feat as a fluke.

5

Sandy, my point is not that they *were* lucky. They probably weren't. I think the naked eye could tell us that they had skills on that side of the ball. My point is, that if they were lucky, how would you *know*. The variable list is clearly one that you could make a logical assumption should equal out over time. But over how much time? Are we reverse-engineering the metrics and then assuming random factors equal out? Suppose they did not...again, how would you know?
I think it's great that the sabermetric community is attempting to quantify this stuff. Perhaps they're on to something. I just have a hard time believing we can isolate the variables enough to get more than a pretty good correlation. And even then, if there is proof of causation.
I'm not trying to dismiss this in the "it's just to complex, let's not even try" vain. In fact, I would surprised if the ability to isolate some of those variables doesn't improve, with the increasing capability of today and tomorrow's technology. Especially with regard to some of the first on my list, like batted ball speed, trajectory, and positioning. There would be a lot of data to mine, but I can at least visualize some of that happening.

6
Taro's picture

Padna makes a good point.
We're going to have access to some of that data very soon.. It'll be interesting to see what blind spots exist.

7

Doc...you do know that the Mariners' home park plays havoc even with defense independent statistics, right?
Safeco's HR factor, for example is off the charts compared to the rest of the AL...only Oakland is close. The 2001 Mariners were well below average at preventing longballs once you account for park.
They were also only average-solid with their K rate and that was seriously inflated by their good bullpen.
I see a team that allowed a lot of HRs (relative to park), had an average K rate, and an above average (to the good side) walk rate. That's, by definition, an average pitching staff. Well above average pen...slightly below average rotation.

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.

shout_filter

  • 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.