Paxton 12-3, 2.70
and try to stay in shape can'cha Jamie

.

Preamble:  Paxton is 12-3 and has probably 10 starts left.  Think he can go 18-5 despite missing a month?  What is Paxton's shot at 20 wins, maybe four percent?

Bob Dutton pointed out that Sale and Paxton are now tied for the league ERA lead, 2.700006 to 2.700002, to the seventh decimal place or something.  Check me on that; just noodling from memory.  Remarkable coincidence in the numbers considering they've got such different innings pitched and fractions of innings pitched.

Anyway it's 2.70 to 2.70 but Paxton is zooming by Sale like a shot on the outside rail.  :- )

....

Bill James with another interesting article this week on pitcher wins.  It's behind the paywall, if you can call $3 a month a "paywall," so hopefully he doesn't mind a small excerpt.  (He's told us in the past that he doesn't; so far so good.) 

We don't have to litigate this yet again :- ) but just as a reminder:  of course there is a lot of static noise in the W stat.  Basically you're asking the starting pitcher to give up less runs than --- > a random number.  LOL.

But "a lot of noise" is not the same thing as "contains no information at all."  James Paxton has 7 pitcher Wins in 7 starts, and we all know that 7-W stat didn't pop into space-time like opposing particles at the beginning of the universe.*  We just watched James Paxton rip off 7 wins because ... well, because he went out and won 7 games.

Giving up less runs than a random number, it's probably better than giving up more runs than a random number.

....

James acknowledges, with gusto, that pitcher Wins have their limitations, though he (and Dr. D) counters Brian Kenny's position that the stat should be outlawed.*  As James lists the luckiest W-L pitchers ever, he does give this little observation while plowing through 100 years of data.  He lists a bunch of SP's who went 18-9 when they deserved 15-14, and then rubs his chin thoughtfully:

.

I realize that I am sort of implicitly making a pro-won-and-lost argument here, in that many of these discrepancies—which are the largest discrepancies in the history of the team—are not that remarkable.   Bryn Smith went 18-5 when he deserved to go 14-11—that’s the best you’ve got?   That’s the limit of good luck?   Doesn’t that show that the Won-Lost records ARE a pretty good indicator of how the pitcher has pitched, with the exception of an extreme case like Larry Christenson or Braden Looper?  

I’m not making that argument; I’m just noting that it is implicit in the data.   

.

Jamie Moyer, by the way, had an "extremely lucky" career W-L record.  Jamie won 269 and lost 209 but James calculates that his 'deserved' record was 247-226.  Or was it?  James runs through the luckiest W-L seasons -- not careers -- for each of the 30 franchises but he pulls a full stop when he gets to the Seattle Mariners.  There is a puzzling syndrome on the Mariners' "lucky W-L season" leaderboard:

.

 The Mariners, I’m going to vary the form to make a point you will see immediately:

First

Last

Year

W

L

WPct

Des Wins

Des Loss

Luck

Paul

Abbott

2001

17

4

.810

10

10

13.1

Jamie

Moyer

2001

20

6

.769

14

10

9.8

Jamie

Moyer

1997

17

5

.773

13

9

7.8

Jamie

Moyer

2003

21

7

.750

16

9

7.6

Felix

Hernandez

2015

18

9

.667

13

11

6.4

 

              Jamie Moyer had the second-, third-, and fourth-luckiest seasons in the history of the Mariners.   There is an argument that goes "If a pitcher does something year after year after year, doesn’t that prove that it isn’t luck?"  

              Well. . .no.   Maybe; I don’t know.   It’s not year after year after year; it’s three years in a seven-year span.   The Mariners had a loaded offense in those days, with Griffey and A-Rod and Olerud and Ichiro and Buhner and Edgar and Mike Cameron.    Moyer was a good pitcher, pitching for a tremendous team.   If pitchers in general were able to sustain luck, that would prove that it wasn’t luck.   If one pitcher is sort-of able to sustain good luck, I’m not sure that proves anything. 

.

My first QUESTION would be whether Moyer had a real ability to throw consistent Quality Starts with few shutouts -- and therefore tended to allow a consistent 2-3 runs when his offense was getting 4-5.

:: shrug :: Anyway.  If you don't like to even refer to the W stat, you can move on with Dr. D's blessing.  But I like the stat, used in context.  I also like hitters who get 100 RBI and pitchers who go 12-3.

Enjoy

Dr D

Comments

3

Seems like Felix must lead all pitchers in MLB history in unlucky losses/no decisions, don't it?

You would think that it would be a relatively simple matter to go back through the game logs to see if Moyer had an unusual distribution of runs allowed...one with less kurtosis and a smaller standard deviation.

5

I need to get my backside in gear and begin again with the database...but at that point I will have five immediate projects of interest:

1) Determine whether earned run median is a better predictor than ERA for future performance (related: are there pitchers with statistically significantly different RA distributions...different from the average pitcher when adjusted for the median skill of the pitcher)

2) Experiment with ways to deal with rare event statistics. We judge players' entire seasons based on statistics that frequently are heavily influenced by rare events that shouldn't be shown as simple averages.

3) Study pitchability by the run value of each pitch vs. the run values of the fifty most similar pitches (velocity, horizontal and vertical movement, location, starting count). I believe this will capture the skill of pitch sequencing and situational awareness.

4) Determine whether the Mariners are actually worse against meatball pitchers than they should be given their overall offensive skill and the skill of the pitcher...in so doing...see if there exists any team that has a problem beating bad pitching or is better than usual against good pitching.

5) Continue work on league quality as measured by the spread in various team performance metrics and/or the change in performance of players year or year, with age differences factored out as much as possible.

6

i like twenty game winners. What does a twenty game win season show? I suppose that you went out there and did your job. Wilbur Wood supposedly cheated in his 20 game seasons with that knuckler that gave him lots of extra starts with no wear and tear. But that is really doing your job.

Speaking as a Mariner, Jamie holds special significance.

We were deprived of Randy johnson in those "lucky" years. But we had a two time 20 game winner for our consolation.

Foe me, I would get SO FRUSTRATED watching a Tommy John or Frank Tanana come into the Jingdome and beat us, make us look foolish throwing that left handed slop. Drove me crazy. So, it was a real cathartic pleasure for me to watch other team's offenses go wacko trying to hit Jamie.  Those quotes from Billy Beane of absolute frustration when facing Jamie were truly delicious. Heh heh heh. Jamie was my own personal "We have Jamie and you can't have him" With a "deal with it, suckers" added in.

I don't think either Frank or Tommy won 20 games after they lost their heat and resurrected their careers. Jamie is also 39th (!) lifetime in career strikeouts. Quite a feat for a pitcher like him Who never had a early career of blazing fastballs and actually became Jamie Moyer late in the game.  Yet he would get completely overlooked as the great pitcher he was EXCEPT FOR the two 20 game seasons. They demand a place in the discussion. Because Jamie had two 20 game seasons, 2400 K's, 269 wins, he earns a place in the Cooperstown discussions. I'm not saying he deserves induction. But he wouldn't look bad up there. It cerainly separates him from the rest of the pack.

7

He probably doesn't qualify objectively speaking - he's got 56/100 on the HOF Monitor scale.  But I could see an argument for him based on the idea that he had 230-some wins after age 30.  The archetype of an overachiever.  I could go along with a plaque for him based on that, like electing the game's best defensive 2B or a player who was freakish on saves or something.  And it's not like Moyer's 269 wins are anything to sneeze at.  300 wins is rubber-stamp HOF.

One of my roto friends, Mikey Jay, knew Moyer and says he and his wife are super cool, well-adjusted people even by non-celebrity standards.  Seattle was lucky to have them on so many levels.

Superb comment Rick!  :- )

8
Arne's picture

Moyer I remember having at least a few starts each season that were ugly, as in allowing 3 or 4 homers and 6+ runs in 4 innings. Retrosheet shows him allowing 8+ runs 11 times as an M. He didn't throw many shutouts to offset those ERA inflators,

9

I detest wins. I've hated them since 2010, when Geoff Baker had to go on a prolonged crusade to convince the crusty East Coast writers that Felix was better than CC Sabathia (21-7) and David Price (19-6), even though he had like 50% more WAR than they did. I think the Brian Kenny war to end all Wins is not against the stat itself, but against its horrible misusage by lazy, Sabr illiterate writers and HOF voters who decide awards based on W/L record far more often than they should. Those awards matter: they form the tapestry of the game. It’s not at all egalitarian or fair, and it cheapens the game’s history. In that sense, I'm behind him 100%.

As for the utility of the stat itself, it's obviously not useless. It's just very crude. When presented with a 3/4 or full season W/L record, you can usually infer what kind of year the pitcher had. No one goes 20-6 without deserving it, and no one goes 13-12 when they're hands-down the best pitcher in the league... oops, bad example. See, that's kind of the problem. It's a stat that suggests things, but you'll always have to consult other stats to back those up. I never, never, never trust that a pitcher's record is representative of their performance. If I see a guy has gone 12-3, I promptly look up his b-ref page and check the component stats to see if that means anything.

Let me put it another way: we all know that before blindly believing ERA, we should check FIP. Ditto BA and BABIP. ERA and BA are very useful indicators, and they are usually pretty well deserved. However, to be sure you have a proper read on the player you have to check the underlying modifiers. The problem with Ws is that to verify them, you need to check ERA. And to check ERA, you need to check FIP. So why bother with Wins at all? They're far enough removed from the truth that they aren't a particularly helpful indicator of it. In that sense they're a very, very small waste of time. And a small waste of space in a Root Sports graphic.

Also, Wins are like 90% useless on a per-game basis. How many times does a guy "pitch well enough to win," and lose? Or vice versa? Constantly. The only reason I find them 10% useful is that they are an attempt to capture in-game strategy. If a SP gets spotted 5 runs in the first, and spends the rest of the game catching more of the plate than usual so he can go through seven innings, it's cool that Wins captures that and ERA just sees that he gave up an extra run. Still, that's a verrrry limited application, which implies that the use of the stat should be equally limited. If we ever reach a point where it’s only used to answer questions like “can a team win when Erasmo Ramirez goes 5.1 quality innings” or “was Jamie Moyer unusually good at pitching with a lead,” then I’ll support it. That’s a far cry from where we are now.

10

If your first exposure to Disneyland is a 2-hour line into the Dumbos, your feeling about the place will be 1 thing ... if your first day was watching your 3- and 4-year olds walking down Main Street, your feeling might be another thing.  :- )

That puts it in a way I can understand it Sherm, as you so often do.  If your generation started off by watching sportswriters insist that 20-6 was more important than pitching great, I get the "ugh."

....

Everybody HERE understands that 12 W, 3 L is a fun stat to enjoy AFTER you've established that James Paxton is pitching great in the first place.

11

For sure Doc. Timing is everything. Felix 2010 was the quintessential narrative to highlight what's wrong with wins. Kinda like how the 2000 election really should have made us scrap the electoral college, and Pearl Harbor demonstrated why isolationism doesn't work.

I've never, ever seen someone on SSI "misuse" wins in the way that irks me. Which isn't surprising at all, because this is not a community of lazy thinkers. Personally I love when you throw out lines like "that template can get you a long string of 13-9 seasons really easily." It's a true but not accurate description of a meme (as in memetics, not internet banality), and phrases like "20 game winner" and "2 and 8 in the first half" get ideas across quite effectively.

12

and two hours at the Dumbo line first time at D-land - yeah, I've been there. But...

Wins are a dusty, macho stat from the old days when your starters went mano-a-mano, and all that mattered was not whether you gave your bullpen a break, or pitched well enough to win, but if you limited your opponent's team to fewer runs than the other team's pitcher. It is a warrior stat, like a fighter pilot putting kills on his plane or a gunfighter putting notches on his belt.

You don't get moral victories, any more than the Lee did against Grant. Winning was all that mattered, and you better figure out how to do it, even if your fielders boot two balls and load the bases with no outs in a scoreless tie. It's still your job to find a way out.

13

Well when you put it like that Rick, suddenly Wins sound pretty hardcore...

In that case, maybe bullpens have ruined the symbolic significance of SP wins and losses. If we could guarantee that both SPs would go 7-9 innings, it would be cool to know how many scalps your braves had collected by the end of each season. I know that the true pitchers duels, like Felix used to have with every ace he faced, always did make who got the W at the end seem to matter for the individual pitchers. Good stuff!

I love the idea of using batter WPA for that exact reason: it gives guys credit for Killshots. Any gamer knows that kills are cool man, no matter how you get them.

14

Like the 1970's.  :- )

You'd be in August or September, and it would be the Big Red Machine against the Little Blue Bicycle, and Don Gullett against Andy Messersmith on NBC's Game of the Week, and either guy was looked at as a wussy if he failed to hand the ball to the closer much less if he had to come out after 6.

If Messersmith (or Sutton or Tommy John) locked down the Reds for a 4-1 win, the starting pitcher was the story.  Messersmith "won."  A lot of those "wins" WERE "hardcore," as you aptly put it.  The 1970's teams expected a whale of a lot from their staff aces.

Wins were still way overrated in the 1970's, as we now know, but in that era Ace-against-Ace a pitcher win did mean something a little different.  Nowadays you got seven guys in the 'pen throwing 95+ ...

15

Stuff like Holds, Saves, Wins, Losses...

Unlike a home run or an RBI, they are essentially arbitrary definitions on how a pitcher is credited with a number.

So what would happen if you redefined a pitching win or a save?

What if you made a starting win require 6 innings or even a quality start?

What would the rescaling do vis a vis the macho starters of the past?

16

They've actually been talking about that very thing at BJOL recently.  You could re-tune the stat for sure.  ... if any sabes outside BJOL and SSI were not actively campaigning against the very use of the stat, that is.

The M's had a "flash sale" of $6 right field blimp seats after Paxton's 6th win, and this a.m. had a followup $7 flash sale for his start against the Orioles next Tuesday.  By my calculations Cindy and I will be ordering $16 seats on Sept. 15 for his immolation of Cleveland the week after! 

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.