Top 20 Angels Pitcher Seasons

by elricsi on Nov 2, 2008 1:27 PM PST Comment 20 comments

I mentioned before about how win shares are great but they need to be adjusted for context (basically how many losses are generated too). So I've been entering the numbers in a database of every Angels pitcher season ever, and I want to share some preliminary results to see if I'm on the right track. So without further ado, the 20 best seasons for Angels pitchers:

Year Name, PWS, W-L, ERA, etc.

1. 64 Chance 31.7, 20-9, 1.65, 11 shoutouts, 4 saves

2. 76 Tanana 26.5, 19-10, 2.43, 23 CG

3. 89 Blyleven 22.2, 17-5, 2.73, 5 SHO

4. 69 Tatum 19.4, 7-2, 1.36, 22 saves

5. 04 Frankie 16.6, 4-1, 1.82, 12 saves

6. 64 Bob Lee 20.1, 6-5, 1.51, 19 saves

7. 90 Finley 22.6, 18-9, 2.40, 2 SHO

8. 73 Ryan 27.8, 21-16, 2.87, 26 CG

9. 06 Frankie 17.1, 2-3, 1.73, 47 saves

10. 75 Tanana 22.0, 16-9, 2.62, 5 SHO

11. 91 Harvey 17.8, 2-4, 1.60, 46 saves

12. 96 Percival 16.0, 0-2, 2.31, 36 saves

13. 65 Bob Lee 20.2, 9-7, 1.92, 23 saves

14. 86 Witt 22.5, 18-10, 2.84, 14 CG

15. 07 Lackey 21.5, 19-9, 3.01, 2 SHO

16. 08 Frankie 16.4, 2-3, 2.24, 62 saves

17. 91 Langston 20.1, 19-8, 3.00, 7 CG

18. 69 Messersmith 22.4, 16-11, 2.52, 10 CG, 2 saves

19. 85 Moore 19.9, 8-8, 1.92, 31 saves

20. 08 Santana 19.0, 16-7, 3.49. 1 SHO

This Fan-Post is authored by an independent fan. Tell us what you think and how you feel.

0 recs | Comment 20 comments

Story-email Email Printer Print

Comments

What's your exact adjustment for "losses generated"?

by mattwelch on Nov 2, 2008 1:52 PM PST reply actions 0 recs

Fibonacci

In Bill James HOF book, he came up with a stat called “Fibonacci Wins”, which can convert won-loss record to a single number. So I have been using Win Shares instead of wins, and that is the number. Formula is: Wins*(winning pct.) + wins – losses

P.S. I got your McCain book and it’s good stuff, I had been looking for something that is not all positive or negative, but tells the good and bad.

by elricsi on Nov 2, 2008 2:13 PM PST up reply actions 0 recs

I think that's problematic

James, IIRC, was trying to give extra credit for super-high WP% guys like Koufax & Pedro Martinez, as a quick way to recognize how they differ from the Milt Pappases of the world. That kind of thing is useful mostly as a quick eyeballer, not as a precise evaluation of a career, let alone a season. A season is small enough so that the variables that otherwise even out over time (run support, etc.) can be all over the map.

In Ryan’s case, he was piling up 300 IP a year for a team that really could not hit the baseball (until ‘78-79, when he wasn’t nearly as effective), and had a disastrous relief corps for almost his entire Angel career. Yes, if you start 39 games for a crap team and complete two-thirds of ‘em, you’re gonna rack up 16 losses. Put that kind of effort out front of a Scioscia team, and you’d be talking about more 26-10, less 22-16.

p.s. Thanks!

by mattwelch on Nov 2, 2008 2:40 PM PST up reply actions 0 recs

Ryan in '73...

…. pitched 327 innings (3rd in the league), threw 26 complete games (2nd), won 21 games (4th), with a 2.87 ERA (4th), threw 4 shutouts (4th) allowed 6.57 hits per 9 (2nd), 1.23 walks+hits per 9 (8th), and struck out 383 freaking batters (can’t touch this). All for a team that scored 3.56 runs per game in his starts — actually, 2.92 over 36 of his starts, then 11+ for three blowouts — and featured a truly brutal bullpen. He pitched 3 times on 2 days rest (2-0 with a 1.52 ERA), and pitched a two-inning save on zero games rest for good measure.

Bert Blyleven’s ‘89 and Chuck Finley’s ’90 are just not part of that conversation.

by mattwelch on Nov 2, 2008 2:11 PM PST reply actions 0 recs

Nolan

That will be the biggest issue, I’m sure.

He had 4 great seasons for the Angels, but lost 16 games each time. (Walking 200 in 2 of the 4).

That is a lot of losses, and needs to be accounted for somehow. Nolan had his unstoppable games, which are legend, but he also crapped the bed on many occasion, and people forget those. I am thinking of dividing losses by 2 on that last part of the formula.

by elricsi on Nov 2, 2008 2:16 PM PST up reply actions 0 recs

I think you're better off...

… doing something like applying league-average run support & distribution for every pitcher under consideration. Even then you can’t control much for relief support, but that’s comparatively marginal in effect.

Ryan had 11 games in ’73 where he received 1 or fewer runs in support, including five shutouts. He went 0-11 in those starts (with 6 complete games!); and was 21-5 when the team scored even 2 or more.

by mattwelch on Nov 2, 2008 2:47 PM PST up reply actions 0 recs

Agreed

Wins and losses should play a very little part of assessing any pitcher’s value, or worth, or whatever over a season. Ryan lost 16 games because the hitting was atrocious, not because he was. It’s pretty unfair to discount him for something he had no control over.

I feel the need, the need...for speed!

by Gorbachav5 on Nov 2, 2008 9:11 PM PST up reply actions 0 recs

agree!

those 26 CG’s alone are remarkable

by lilbobdog on Nov 3, 2008 11:54 AM PST up reply actions 0 recs

That was a sign of the times.

You weren’t a real pitcher if you couldn’t go the distance.

Angels fan since '67

by red floyd on Nov 3, 2008 1:35 PM PST up reply actions 0 recs

Well....

He had a little control over it at least. Also, I’m not really using pitcher wins, which I agree is a crap stat. I’m using pitching win shares (or which pitcher wins is but a small component).

Anyway, I’m playing around with estimating the number of losses the pitcher would have had with average offensive support. I’m curious what Nolan’s loss totals would have been then. Also, I think I’ll be ditching the Fibonacci thing, and simply deducting a fraction of losses.

While Ryan’s 73 certainly appears better than Finley’s 90, was it really better than Blyleven’s 89? The Angels were not the best then, and those numbers are superb for the time.

by elricsi on Nov 3, 2008 8:26 AM PST reply actions 0 recs

Win Shares itself is a perfectly good stat

Win Shares takes into consideration all that stuff, so no reason to reinvent the wheel, unless you just don’t like the metric.

As for Bly vs. Nolan, 80 innings is a whole damned lot. Just compare their ranks on the leaderboards — Nolan edges him out, probably, though it’s close — but Blyleven got very good run support, had a pretty good defense behind him, a terrific bullpen …. and just flat-out pitched 80 innings less. Give him those 80 innings, and it’s close; but until you do, it’s not.

by mattwelch on Nov 3, 2008 10:56 AM PST up reply actions 0 recs

We don't need no steenking loss shares

Win Shares is excellent true, but the zero point is not replacement level, it’s 52% worse than average. You can make the argument that replacement level for pitchers (including relievers) is likely in the range of 25-35 below average. Even the master, Bill James, is working on a formula for loss shares to give the complete picture. I mean CC and Ben Sheets both had 15.2 win shares for the Brewers this year, but CC’s work for them was far more great.

Also, if we only use straight win shares, I believe relievers are underrated. The best rankings are #16, 19, 21, 22, 29 for Angel’s seasons. 3 of those being relievers from the 60’s. But great closers have maybe 80% the win shares total, but with far fewer losses (and expected losses). I would agree my preliminary list above goes overboard with relief pitchers.

Anyway, I have some numbers now for projected losses in a run-neutral environment. This adjusts for league scoring, but not park effects. (Angels stadium has been historically around the middle, or actually a bit favorable to pitchers in the 60’s and 70’s)

Nolan Ryan
Year-WS-Losses-Expected Losses
72-23.6-16-12.7
73-27.8-16-13.2
74-20.7-16-16.1
75-11.6-12-11.2
76-16.7-18-16.4
77-22.1-16-14.0

P.S. My whole point for this is to reduce each season to a single number, then add the best 3 seasons together to have an objective list of the best Angels pitchers. The ranking will not be perfect, of course, but hopefully it will remind us of some forgotten excellence.

by elricsi on Nov 3, 2008 8:37 PM PST up reply actions 0 recs

Oops

I wish we could edit here. My numbers for Nolan’s expected losses are wrong.

From 72 to 77, they should be:

11.0, 12.6, 15.2, 10.5, 14.5, 11.6

by elricsi on Nov 3, 2008 9:29 PM PST up reply actions 0 recs

The C.C./Sheets conundrum is solved simply

By Win Shares percentage. Like the Hardball Times guys do.

But I would hesitate to rely on those too heavily for full seasons, since part of the point for starters is to chew up mass quantities of innings. That’s valuable in itself, and is ably reflected in … wait for it … Win Shares!

But yeah, the reliever/starter stuff is hard to feel comfortable with. I prefer to use WS in apples-to-apples situations, for safety.

by mattwelch on Nov 3, 2008 9:48 PM PST up reply actions 0 recs

Colon

How is Colon’s Cy Young Year at very least not better than Santana’s?

legen... wait for it... dary

by thebigA on Nov 3, 2008 10:57 AM PST reply actions 0 recs

Colon's Cy Young

was voter error. He had a good season, but it was not even close to being as good as Johan’s season that year. He was helped by great run support, which got him the great win totals, which got him the Cy Young.

Ervin had 50 more K’s, a better ERA+, and a better WHIP in only 3 less innings. Santana also had a shutout and both had two complete games. The only difference is Colon’s win total, which Santana largely had no control over.

I feel the need, the need...for speed!

by Gorbachav5 on Nov 3, 2008 12:44 PM PST up reply actions 0 recs