| 0 |
Subject: How stats can deceive . . .
Posted by: jumpball, Co-Guru Jr
- Sustainer [480332121] Sat, Sep 29, 20:52
Consider the stats of the following 2 pitchers:
PITCHER #1
Starts: 31
Innings: 209 1/3
ERA: 3.48
Hits Allowed: 196
Homers Allowed: 18
Walks Allowed: 64
Strikeouts: 201
PITCHER #2
Starts: 31
Inns: 211 2/3
ERA: 3.27
Hits Allowed: 188
Homers Allowed: 18
Walks Allowed: 40
Strikeouts: 200
Looks like 2 pitchers having very good years.
Now, suppose I was to tell you that Pitcher #1 will likely win the Cy Young award and Pitcher #2 will not get any first place votes and may not even finish in the top 5. Would you say I was crazy?
If you're going to tell me that one pitches for a winning team and the other for a losing team, that would be a nice try, so I'll tell you now that they pitch for the same team!
You can probably figure out the 2 pitchers, but I'll post the reason for the apparent discrepency later.
|
| 1 | F Gump
ID: 578382615
Sat, Sep 29, 21:09
|
You left out one very important and VERY RELEVANT stat - which is WINS.
Wins are important. They are why the game is played. The pitcher who "finds a way to win" will always get more recognition than the one who consistently "barely loses."
Pitching "just good enough to win" is a world apart from "pitching just good enough to almost win." Giving up runs that "don't matter", vs giving up runs that "lose the game", can in the long run look similar on ERA and BA against, but they are far different.
|
| 2 | butt
ID: 248162921
Sat, Sep 29, 21:16
|
Maybe one pitcher is 20-2 and the other is 16-11, eh?
|
| 3 | jumpball, Co-Guru Jr
Sustainer
ID: 480332121
Sat, Sep 29, 21:26
|
The stat I left out is run support.
Pitcher #1 (yes, it's Clemens) is getting 6.3 runs per game from his teammates; one of the highest numbers for all starting pitchers.
Pitcher #2 (Mussina) is getting 4.1 runs per game support; one of the lowest numbers. That's 50% more runs each game for Clemens.
I think the run support has a lot more to do with the disparity in the respective W-L records than F Gump's reply.
You can't win a game if your teammates don't score.
Oh and I forgot another set of stats . . . Mussina has 3 complete games and 2 shutouts; Clemens has zero of those.
|
| 4 | biliruben
Sustainer
ID: 231045110
Sat, Sep 29, 21:32
|
F Gump - so two years ago when when Mussina went 18-7, he knew how to win, but after that he forgot? No, two years ago he had 6.82 runs of support. Last year? 3.71 runs and he goes 11-15.
This year? 4.67 runs. Clemens this year? 6.71 runs!
Every year Mussina's era was very good - in the mid 3s.
Clemens better share that award with the yankee offense.
|
| 5 | F Gump
ID: 578382615
Sat, Sep 29, 21:47
|
biliruben
Argue it any way you want. But the name of the game is to "win". And the awards are greatly skewed towards those who win. And, in my opinion, rightly so.
|
| 6 | biliruben
Sustainer
ID: 231045110
Sat, Sep 29, 21:59
|
The team's goal is to win. The pitcher's goal is to put them in position to win. How can Mussina somehow "will" the offense to hit better?
Maybe not buy Tino shots the night before? Maybe DO buy him shots!
|
| 7 | JKaye
Sustainer
ID: 4711592917
Sat, Sep 29, 22:03
|
biliruben is on the money. Of course wins are important--for teams. Wins for individuals are pointless and should not even be kept as a relevant stat. If pitchers get wins, why shouldn't hitters? If I here anyone else say Clemens has had an amazing season, while Mussina has been a dissapointment for the Yankees, I will lose my mind. Mussina is the best pitcher on the team hands down.
|
| 8 | ChicagoTRS
Sustainer
ID: 117541522
Sat, Sep 29, 22:24
|
Another important stat in Clemens favor...look up disaster starts...where the pitcher gives up more earned runs than innings pitched...I think Mussina has 4 or 5 this year...Clemens has 0. Clemens has also lasted into at least the 5-6 innings every start this year.
I do agree that wins/losses for a pitcher is way overrated and has everything to do with run support.
If it was true that there are pitchers who "find a way to win" you would expect them to do that consistently year after year. Doesn't happen, a few years ago Mussina had 2-3 years in a row where he got great run support and won a lot of games. Did all of a sudden he forget how to find a way to win this year? after doing it in the past. Clemens a few years back had a couple bad seasons win/loss wise basically because of run support. Did all of a sudden Clemens learn how to win again? Rob Neyer had some good articles on the relevent facts about a month ago. It is all about run support.
Someone should look up the Neyer article has a lot of the data both ways for these two pitchers.
|
| 9 | F Gump
ID: 578382615
Sat, Sep 29, 22:28
|
biliruben
The problem with stats is that they are based on averages. And run support is one of those stats that sometimes tells the whole story, but other times is misleading. Is it misleading with Mussina? I don't know.
The problem with Mussina's record is the losses, 11 of them. Did "lack of run support" cause him to lose these following 7 - or was it crappy pitching? The "run support" was low - but in 6 of the 7 at least it was the pitching that would seem to be the culprit, and Mussina was the pitcher.
IP -4, runs allowed -6, support -3
IP -6.1, R -5, S - 3
IP -6, R -4, S - 3
IP -5, R -6, S - 3
IP -2, R -8, S - 3
IP -4, R -8, S - 2
IP -6, R -3, S - 1
In those 7 games, you can use stats to say "Mussina is such a hard luck pitcher, he is only getting 2 1/2 runs of support per game!" But other than the last game (which is borderline, a 4.50 ERA for the game is not great but could be acceptable) he DESERVED the losses. The "only 2 1/2 runs" makes him sound like a hard luck pitcher, rather than a crummy pitcher, in those 7 games.
The downside to this sort of argument, either way, is that to give an ACCURATE analysis, you have to examine every game of every pitcher, and every situation in every game, and assign credit or blame. The game, which already has tons of stats, then ends up mired in minutae and endless analysis that only the sabermetricians can unsort, and even THEY won't always agree on whether the pitcher did well enough or not.
Though I like stats, I think that to throw wins out the window would be a mistake. In the long run, the game is about winning - and pitchers are a crucial part of that equation, and should be judged accordingly.
|
| 10 | ChicagoTRS
Sustainer
ID: 117541522
Sat, Sep 29, 22:36
|
Neyer Article:
TUESDAY, SEPTEMBER 4
Here's an actual e-mail message that I received last Friday night, a few minutes after Roger Clemens beat the Red Sox 3-1:
You my friend must be an idiot, how do they let people like write sport. You must be another angry Boston Redsox fan. Just wait un till the end of the season then you can eat your words, you armature.
Leaving me to wonder, yet again, how someone capable of writing something like this figured out how to use a computer (and if anyone knows how my words would qualify me as an "armature" -- the rotating part of a dynamo -- please let me know).
Here's another one:
I guess Clemens wins without run support when he has to. A real gem tonite.
1 run, 0 BB and 10 SO. I suppose you will have something negative, but, believe me you are in the minority among good baseball analysts.
This sort of correspondence does represent a minority of what I receive (at least most days), as I'm blessed with an incredibly intelligent audience. And to be sure, a few intelligent readers suggested that while they did understand that Mike Mussina is a great pitcher, and that run support does play a part in a pitcher's record ... still, isn't there something to the notion that some pitchers just "know how to win"? And/or that pitchers "pitch to the score," resulting in win-loss records that don't necessarily correspond to their ERA?
Taking the second of those, while it's true that pitchers occasionally alter their style depending on the score -- that is, a pitcher with a big lead might throw more strikes -- there's little or no evidence to suggest that this has a big effect on their ERA. Why not? Three reasons. One, there just aren't that many blowouts. Two, while throwing more strikes may result in more hits, it also results in fewer walks. And three, pitchers don't like to give up runs, even if they've got a 16-1 lead. They just don't.
As for the "knowing how to win" argument, let me present you with a pair of pitchers ...
ERA W-L
Pitcher A 2.76 8-16
Pitcher B 3.20 16-10
Now, it's obvious that Pitcher A, while certainly effective at times, just doesn't know how to win. And Pitcher B, on the other hand, while perhaps not quite as talented as Pitcher A, really knows how to win.
The problem here is that Pitcher A and Pitcher B are both Nolan Ryan, just two years apart. Now, while it's true that Ryan won 20 games only twice, does anyone really believe that he went 8-16 with a 2.76 ERA in 1987 because "he didn't know how to win"? And that he somehow remembered how to win two years later, when he went 16-10 with a 3.20 ERA?
Here are two sections of Mike Mussina's career:
ERA W-L WinPct
1991-1999 3.50 136-66 .673
2000-2001 3.61 25-26 .490
At the conclusion of the 1999 season, Mussina's .673 career winning percentage ranked sixth on the all-time list, and fourth among pitchers whose careers began after 1900. The only three post-1900 hurlers ahead of Mussina? Pedro Martinez, Whitey Ford and Lefty Grove.
Did Mussina win more than two-thirds of his decisions because he knew how to win? No, he won more than two-thirds of his decisions because he pitched brilliantly and was gifted with good run support from his teammates.
Has Mussina won fewer than half his decisions the past two seasons because he doesn't know how to win? Of course he hasn't; that notion is simply preposterous. He hasn't won even half his decisions because he's received lousy run support: 3.71 runs per nine innings last season, 4.30 runs this year ... the two lowest figures since his rookie season, 10 years ago. Now, is there really anyone out there who doesn't believe there's some correlation between Mussina's run support over the years and his W-L records?
Yes, I know that I'm preaching to the choir; most of you understand the importance of run support, and most of you understand that Mike Mussina has pitched just as well as Roger Clemens this season.
Or almost as well. If were voting for the Cy Young and had to pick between Clemens and Mussina, I would pick Clemens because Mussina has pitched poorly in more games than Clemens has. As I noted last week, Mussina has four starts this season in which his runs allowed exceeded his innings pitched, leaving the Yankees with little chance to win; indeed, the Yankees (and Mussina) lost all four of those games. Clemens, on the other hand, doesn't have any "disaster starts." So it might be argued that Clemens has, in each of his starts, given his club at least a reasonable chance to win. And yes, he deserves a bit of extra credit for that, enough to push him past Mussina on our hypothetical Cy Young ballot.
And enough to push him past Freddy Garcia and Tim Hudson? Maybe, because of the ballparks in which Garcia and Hudson do half their pitching. Enough to push him past Mark Buehrle, he of the league-leading 3.07 ERA? Maybe, because Buerhle is just 12-7. My point last week, and one of my points today, is that when it comes to the Cy Young in the American League there are still a lot of maybes. I don't believe in anointing an award winner in late August if the anointee hasn't done enough to deserve it.
Finally, one might reasonably ask, "Why do people continue to believe that won-lost records can be explained by 'knowing how (or how not) to win' "? Well, they believe it for the same reason that people believe a lot of untrue things; somewhat paradoxically, because of personal experience.
I'm sure that the great majority of you have played sports at some level. Haven't you all played poorly in a game of some sort, but felt OK about your performance because your team won anyway? You were able to focus on the positive aspect of your performance, because after all your team did win.
Well, the great majority of sports "wisdom" comes from professional athletes or ex-professional athletes, and in this respect they're little different than you or I; their primary focus is on winning. So when an ex-player tells you that winning is all that matters, he's merely expressing a heart-felt opinion from his own perspective. The problem, at least in this instance, is that it's a limited perspective. Just as players tend to value personality over performance in a teammate, they tend to value victories over effectiveness in a pitcher. It's all about perspective; as my friend Eddie Epstein often tells me, asking a player to evaluate baseball statistics is like asking a tree to evaluate a forest.
I don't know how to throw a slider or turn a double play, and never will. What I do know is that when a pitcher's record doesn't match his other numbers, it's rarely because he doesn't know how to win. Most of the time, it's because his teammates didn't score any runs for him.
|
| 11 | F Gump
ID: 578382615
Sun, Sep 30, 01:18
|
Neyer's article is interesting, but he uses a straw man to knock down as his opponent. He sets up "knowing how to win" as the alternate crucial component of win/loss (as against his "run support"). Thus, if he can demonstrate "knowing how to win" is bogus, then he can say "I disproved that, so my alternate is proved."
However, it is all bogus. So Ryan did not WIN in one year, and won in another, big deal. It was not (nor has ever been postulated) that the component is KNOWLEDGE or even DESIRE (another straw man), but rather EXECUTION in the crucial situations. So because he was able to win more, in one year, with worse stats than in another year, neither proves or disproves anything.
Run support can be a legit factor. But as Neyer points out, in Mussina's case the losses came from not giving his team a chance - and thus, interestingly enough, the W-L records of Clemens and Mussina reflect somewhat accurately the value to their team this year in their starts - in fact, MORE accurately than ERA etc: and thus it would seem that Wins DOES have a valid role here, after all.
|
| 12 | steve houpt
ID: 32851214
Sun, Sep 30, 01:47
|
Let's look at both pitchers by (ERA) earned runs allowed each game. Less than 4.50 and 4.50 and above.
Less than 4.50 ERA games
Clemens (20G) 16-0, 2.06 ERA (NYY 19-1) (Runs 111-47)
Mussina (21G) 14-4, 1.36 ERA (NYY 15-6) (Runs 75-35)
14-4 in 21 starts with 1.36 ERA. Tough.
4.50 ERA and above games
Clemens (11G) 4-2, 6.33 ERA (NYY 8-3) (Runs 78-55)
Mussina (11G) 2-7, 8.89 ERA (NYY 3-8) (Runs 62-84)
Quite a feat to go 4-2 in the 11 games you have a 6.33 ERA.
Mussina has 3 games he has given up 6 or more runs, Clemens none. I give Mussina those three losses.
But Mussina is 16-8 in 29 other starts with 2.48 ERA, Clemens 20-2 in 31 starts with 3.48 ERA.
Sure looks like run support has 'something' to do with it.
|
| 13 | F Gump
ID: 578382615
Sun, Sep 30, 03:33
|
steve houpt
You can massage the numbers 100 different ways and support any conclusion, and yours is as valid as any ... my opinion is that Mussina was not quite as good of a pitcher as Clemens this year. Though Clemens had better run support, the difference in records is also due to just flat out more consistent pitching for Clemens - so to attribute the whole thing to "run support" which some do is going to be deceptive.
After all is said and done, the W-L difference somehow DOES accurately pick up that difference in this situation that does not seem to be shown in the stats based on averages. I think Ws are an important stat - since the pennant is not decided on year-long averages but rather on daily snapshots, the daily Ws can tell some things more accurately than the year long averages do, in evaluating performance.
Does run support impact wins? Yep, it can. But it can also be a deceptive stat (as all stats can be) and incredibly misused, in my opinion. As in my post above (9), Mussina had 7 losses that totalled about 2 1/2 average runs of support for him, yet he was NOT a hard-luck pitcher in any of those 7 losses. In his other games he is 16-4, and maybe you want to say he SHOULD have gone 18-2 or 17-3 instead of 16-4 in those games; but it can ALSO be argued that his record is about where it SHOULD be, since every once in a while you can just get outpitched, it would seem. Those numbers (18-2 or 17-3 in his "good outings") still put Mussina with 9 or 10 losses - rarely Cy Young caliber.
|
| 14 | Skidazl
Sustainer
ID: 506141811
Sun, Sep 30, 05:12
|
here some stats, and he probably won't get CY young either...
starts - 33
innings - 232.2
ERA - 3.02
hits allowed - 193
HRs allowed - 16
walks - 68
K's - 156
18-6
barring w-l, looks like a pitcher having an even better year.....
|
| 15 | F Gump
ID: 578382615
Sun, Sep 30, 07:30
|
skidazl
Actually, Clemens gaudy W-L record makes him the front-runner, but Garcia does have a shot. His stats are all near the top, and he will get bonus consideration because he is a Mariner. If his Ks were HIGHER than Clemens, he might have been able to pull it off, since his ERA is better. (In my analysis, the votes usually come down to W-L, ERA, and K stats, with "team success" having some factor at times. Not sure if ANY of the other stats are factors that the voters weigh - don't ever seem to be, as I see it).
|
| 16 | ChicagoTRS
Sustainer
ID: 117541522
Sun, Sep 30, 08:51
|
fgump...Neyer not only used Ryan as proof of disproving the know how to win theory but used Mussina himself:
"Here are two sections of Mike Mussina's career:
ERA W-L WinPct
1991-1999 3.50 136-66 .673
2000-2001 3.61 25-26 .490
At the conclusion of the 1999 season, Mussina's .673 career winning percentage ranked sixth on the all-time list, and fourth among pitchers whose careers began after 1900. The only three post-1900 hurlers ahead of Mussina? Pedro Martinez, Whitey Ford and Lefty Grove.
Did Mussina win more than two-thirds of his decisions because he knew how to win? No, he won more than two-thirds of his decisions because he pitched brilliantly and was gifted with good run support from his teammates.
Has Mussina won fewer than half his decisions the past two seasons because he doesn't know how to win? Of course he hasn't; that notion is simply preposterous. He hasn't won even half his decisions because he's received lousy run support:
3.71 runs per nine innings last season, 4.30 runs this year ... the two lowest figures since his rookie season, 10 years ago. Now, is there really anyone out there who doesn't believe there's some correlation between Mussina's run support over the years and his W-L records?"
|
| 17 | steve houpt
ID: 32851214
Sun, Sep 30, 09:22
|
The numbers may not support Mussina having many more wins or less losses. His record might be right on for the 'way' he has pitched.
But I don't think the numbers Clemens has put up, 'support' a pitcher with only 2 losses. Twenty wins, sure. But he still has 'only' won 20 of 31 starts (.645). It is second best in the majors (Abbott #1 - 17 of 26). But others are not far behind. Morris 21 of 33.
If you zero in on the three games Clemens left trailing and two were he gave up 5 runs and left in 5-5 ties, he looks like a 20-7 pitcher. Give him the two games he left with the lead they blew, he's up to 22-7. Other two games he left with score tied 2-2. A little support and he's 24-7.
Take the 4 games where Mussina pitched 30 innings (7.5 per start) and gave up 2, 3, 3, 3 runs and Yankees scored a total of 4 runs and give him any kind of support and Mussina is a 20-7 pitcher. Give him two losses for the games they took him off the hook when he left trailing and he's 20-9. Other three games he left with score 0-0, 1-1 and 2-2. Give him a little support and he's 23-9.
I'll even chalk up the 5/17 game when Mussina left after 6 trailing 4-3 (one run unearned) to Mussina not knowing how to win. Yankees went on to lose 8-3.
So still a slight advantage Clemens the way I calculate on wins.
Here are some links to one of those sabremtrics. I'm kind to Clemens compared to this guy (Michael Wolverton).
About Support Neutral Wins and Losses
Each of the three numbers is calculated separately for each individual start, and then summed to get seasonal totals. Looking at a starter's performance game-by-game like this removes distortions that can be introduced by looking at cumulative run prevention (e.g., ERA or Thorn and Palmer's Adjusted Pitching Runs). In particular, the SN stats recognize that a pitcher can only cost his team a single game in a single start, so it puts a limit on how much a single bad outing (e.g., 2 IP, 11 R) can hurt his season/career value.
Support-Neutral W/L Report: September 29, 2001
Mussina 15.9 - 8.6
Clemens 13.1 - 8.8
Interresting.
Joe Mays 16.1 - 7.0
Here is last years stats just for a comparison. Not something Michael Wolverton just thought up to make Clemens look bad. Been doing this for years.
2000 SNWL stats
In depth description of SNWL
NOTE: I am not trying to take anything away from Clemens. There is a great deal of skill to winning. Team mates sense it. They feel like they are going to win when certain pitchers are on the mound. 20-2 is an amazing record. Considering the competition, I think 20-2 is the deciding factor in the Cy Young voting.
Enjoy these type of discussions.
|
| 18 | KrazyKoalaBears
ID: 266182910
Sun, Sep 30, 10:00
|
"Considering the competition, I think 20-2 is the deciding factor in the Cy Young voting." I think that, overall, is the deciding factor. First off, nobody has ever, ever, gone 20-1 no matter what type of run support or ERA they have. That in and of itself is absolutely amazing and worthy of the Cy Young Award. The fact that Clemens and Mussina get different amounts of run support lends credence to what source also alludes to when he says, "Teammates sense it. They feel like they are going to win when certain pitchers are on the mound." If this is not true, then why does Clemens get more run support from the same exact hitters? The fact of the matter is that Clemens has done something that has never been done before as a pitcher and he happens to have pretty good stats to back it up. Cy Young material? Definitely. A few worthy competitors that might have won the Cy Young in a different year with the same stats? Definitely.
|
| 19 | butt
ID: 54893010
Sun, Sep 30, 10:09
|
Dan Duquette, Executive VP and GM, Boston Red Sox, December 13, 1996:
"The Red Sox and our fans were fortunate to see Roger Clemens play in his prime and we had hoped to keep him in Boston during the twilight of his career."
What a moron.
|
| 20 | ChicagoTRS
Sustainer
ID: 4324316
Sun, Sep 30, 12:03
|
"then why does Clemens get more run support from the same exact hitters?"
One word can explain this "CHANCE" the 30 or so starts a starting pitcher gets is a relatively small sample and next year maybe Clemens will only get 4.2 run support and Mussina will get 6.2. Then what will people say? Mussina knows how to win and Clemens doesn't or the Yankee hitters felt like they were going to win when Mussina was on the mound. Run support is generally random over the careers of a pitcher...there are no pitchers who get consistently high run support every year...generally run support goes up and down...chance...thus pitchers win loss records tend to fluctuate even if there actual stats are identical.
|
| 21 | KrazyKoalaBears
ID: 266182910
Sun, Sep 30, 15:16
|
ChicagoTRS, if it's "chance", then why has Clemens not had less than 4.00 RS average since '93? And why has he only had less than 4.00 RS average once (3.66 in '93) since the stat was registered in '88? Also, why has he averaged more than 5.00 RS in his 3 years with NYY (4.94, 5.37, 6.71)? Orlando Hernandez? Never had less than 4.00 RS average in his 3 full years with NYY (7.34, 5.54, 4.42). And this year, an incomplete year, he's getting 3.67. Given a full year, it's likely he would be above 4.00 again. Andy Pettitte? Never less than 5.12 RS average in his 7 years with NYY. David Cone with NYY? Never less than 4.75 RS average. David Wells with NYY? 5.12 and 6.84. Also never less than 4.38 since '88. So how much "chance" is it really? Sure, there is variation, much like a batters batting average from year to year, but it doesn't mean it's completely "chance". The fact of the matter is that Mike Mussina never got less than 3.71 RS average in his 10 years with BAL. If you throw out his first two years and his last year with BAL, then he never had less than 4.91 RS average ('95). So why is he at 4.67 this year? I, personally, say it's a lot more than just "chance".
|
| 22 | F Gump
ID: 578382615
Sun, Sep 30, 15:38
|
CHICAGO TRS (post 16)
What relevance does "knowing how to win" have to do with the importance/relevance of "run support" stats? And, who cares whether such nonsense is examined by Neyer with respect to Ryan, or Mussina, or a million pitchers? As I said in post 11, no one other than Neyer (and you, I guess) put any credence in such a theory - primarily creating it in order to make it look silly, I guess. I have heard the idea stated, but my logical deduction is that what is really being praised is a pitcher's ability to "know and execute" what it takes to win a game - and even though the knowledge is there, it is the lack of EXECUTION that can vary.
Season long stats don't always tell the full story - and Mussina's high number of Ls for such a low ERA accurately portray the FACT that on a sizable number of days (in re to such a good ERA) he was too lousy.
It is not the "knowing" but rather the execution in doing what it takes to consistently win that Clemens has done better than Mussina this year. Run support has played a factor, but there is a LOT more to it than run support in the way Clemens has managed to produce more wins and fewer losses than Musina, as examined above. And thus, the W-L record is a true and relevant stat itself.
Or, as KKB so eloquently puts it, it really doesn't matter HOW Clemens did it, 20-2 is "in and of itself absolutely amazing and worthy of the Cy Young Award".
|
| 23 | KrazyKoalaBears
ID: 266182910
Sun, Sep 30, 15:47
|
So how is it that Clemens only got 1 run worth of support today and managed to avoid an "L"? I mean, he's only good because he has so much run support, right? ;)
|
| 24 | ChicagoTRS
Sustainer
ID: 117541522
Sun, Sep 30, 18:19
|
huh...I was wondering why Clemens gave up a run when he was only getting one run support today? I figured he would want the W.
"Season long stats don't always tell the full story - and Mussina's high number of Ls for such a low ERA accurately portray the FACT that on a sizable number of days (in re to such a good ERA) he was too lousy." --Wonder when Mussina lost this ability? He had the 6th best win/loss% in baseball history at one point in his career. Seems odd that he turned lousy in the last two years.
I am not saying Mussina should be the Cy Young winner...I do think Clemens should win the award (for other reasons than won/loss record)...what I am saying is win/loss record is a very overrated stat for pitchers and has a lot more to do with run support than the pitchers ability. It is a team stat. Very similar to RBIs, Runs scored, and clutch hitting. RBIs and Runs are very dependent on teammates and clutch hitting is generally a myth.
|
| 25 | F Gump
ID: 578382615
Sun, Sep 30, 22:23
|
CHICAGO TRS
The fact that Mussina has had an inordinate number of extra-crappy outings this year (for a pitcher of his ilk) is a fact. It is one factor why his losses are up, and as I detailed earlier his record reflects the fact that he had a bunch of excellent outings interspersed with some where he EARNED A LOSS. Accordingly, his record reflects that. That's just how it is.
You want me to explain WHY he was non-competitive so many times this year (in comparison to prior years)? I have no idea. But it is true, and his W-L record tells the results. And it is not by any means something you can pass off as "no run support" when you look at each game and how he pitched therein.
There is no way you can eliminate the element of chance from baseball in any one play, or in any one inning, or in any one game. Hits fall just out of the reach of fielders, balls are pitched a few inches from their intended spot and are hit rather than missed, errors are made and non-errors (where balls are misplayed) also occur. But when all is said and done, the sport demands that you set aside the whining and explaining and give credit to the winners who find a way to win, whether by skill or luck or some combo thereof. And thus, to recognize the pitcher for a W or a L, on the day they pitch, because they have such a vital role in that game, is quite proper in my opinion.
W-L record is not the ONLY criteria for evaluating pitching excellence. But it is one well-recognized and well-deserved item in the list of whether a pitcher might be pitching well or poorly in any given time interval. So when judging season-long awards, it certainly MUST be given a good bit of consideration IMO.
|
| 26 | Hawkes
ID: 2491315
Mon, Oct 01, 05:13
|
Clemens must be the all time master of "knowing how to win", or perhaps more accurately knowing how not to lose, as it would seem his ability in this regard stretches to when he has been removed from the game and is no longer pitching. Perhaps these are relevant stats that should be weighed against Clemens' 20-2 record, as opposed to Garcia's 18-6, when the Cy Young award is being considered.
*Nine times this year Clemens has thrown his last pitch for the game and walked off the mound on the loss. In all but two of these starts, his bullpen and offense have rescued him and he has avoided the loss. In two of them the offense have actually managed to put the Yankess ahead while he was still the pitcher of record, and he has ended up with the win.
*Only once this year has Garcia left on the loss and been rescued by his team mates.
*Only once this year has Clemens come off on the win and had his team mates blow it for him giving him a no result.
*Four times this year Garcia has come off on the win only to have his bull pen blow it and give him a no result.
Quite significant, I feel. Given Garcia's considerably superior performance in ERA, whip and number of complete games (as opposed to Clemens' higher number of strikouts), Clemens could really consider himself fortunate to be in such strong contention for the Cy Young. As has been said in several previous posts, all hail run support.
|
| 27 | ChicagoTRS
Sustainer
ID: 4324316
Mon, Oct 01, 08:57
|
Hawkes...don't be stupid...Clemens willed his teammates to come back in those games...he should get the credit for that too. ;-)
|
| 29 | Mattinglyinthehall
ID: 1832399
Mon, Oct 01, 09:24
|
Yes Chicago TRS, IMO he should get some credit for that. Not kidding.
Funny that nobody acknowledges KKB's post # 18. The simple and undenyable fact that Yankee hitters have rallied around Clemens all season long is obvious. Call it luck or chance or whatever if you must, but as Gump points out in post # 9, and steve houpt in post # 12, Clemens has been the more consistantly adequate pitcher. Sure, Moose is better when they are pitching well, but quite a bit worse when their sub-par games are compared.
In my opinion, the Yankee hitters are more confident when Clemens, a pitcher they have seen give them a good chance to win almost every time he steps out there, is on the mound. That confidence causes them to stay loose and focused and keeps them from pressing too much. Anyone who denies this hasn't watched much baseball, or hasn't paid much attention, or just seeks reasons not to give Clemens his due. Staying loose and confident is as much an aspect of the game as anything else. The manager of any hitter or pitcher on a hot streak or of any team on a winning streak will tell you they've got to remain confident and loose, and manager of a slumping team or player will talk about telling players not to press and building confidence to get back on track.
JKaye wrote in post # 7 that, "Mussina is the best pitcher on the team hands down." I'd agree that Mussina has had the most and the most number of dominant performances on the team (among the starters anyway). But Clemens gets the nod as the best - or most valuable - pitcher on the team this year IMO. He's simply more consistantly adequate, even if Moose shows more when he's on. That Yankee hitters rally around him with bloated run support should only add to his value as a pitcher, not take away from it.
|
| 30 | stinkypuff
ID: 2883927
Mon, Oct 01, 10:24
|
with all due respect (to anyone who will take offense), let's get to the real issue: People don't like Clemens. He's a head case; Mussina, by comparison, is a class act. If it weren't for the personalities involved, we wouldn't be having this debate (which is to say, Clemens gets the award).
I mean, would anyone get this worked up over the Randy/Schilling Cy Young debate? I bet not.
|
| 31 | biliruben
Sustainer
ID: 231045110
Mon, Oct 01, 10:33
|
Actually, Stinkypuff, I like Clemens better than Mussina.
"W-L record is not the ONLY criteria for evaluating pitching excellence. But it is one well-recognized and well-deserved item in the list of whether a pitcher might be pitching well or poorly in any given time interval. So when judging season-long awards, it certainly MUST be given a good bit of consideration IMO." -F Gump
But, even if you think W-L should be considered (which I guess I could agree with, if you don't put too much weight on it), when picking Clemens you are putting it AHEAD of stats which directly show pitcher performance, when you compare Clemens to Freddy or Moose.
Q: Why would you weigh a stat which is an amalgum of starting pitcher, defense, relief corps and offensive performance (W-L), HIGHER than a stat (ERA) that is a direct measure of how well a starting pitcher (and only the pitcher) actually performed during the year?
|
| 32 | ChicagoTRS
Sustainer
ID: 4324316
Mon, Oct 01, 11:13
|
For me this has nothing to do with Clemens vs Mussina...you can substitute any pitchers names. It is about the importance of win/loss record when evaluating a pitcher.
I do agree that win/loss should be a factor in determining the Cy Young or the performance of a pitcher but I just think it should fall down the list of factors. If two pitchers like Clemens and Mussina have nearly equal stats then the win/loss record should be used for a tie breaker.
For example lets say Randy Johnson is in the AL this year and we will take 2 wins away from him because lets say his bullpen blew a few extra saves...that would leave him with a record of 18-6 clearly inferior to Clemens 20-2 but I would without a second thought give the Cy Young to Randy Johnson because his ERA is a full run lower, has many more strikouts, more IP, opponents BA .40 lower, more complete games/shutouts. RJ was clearly the superior pitcher.
My opinion is there are far more valuable stats for evaluating a pitchers performance than win/loss. It is a team stat that has a lot to do with run support, defense, bullpen, etc...
|
| 33 | ChicagoTRS
Sustainer
ID: 4324316
Mon, Oct 01, 11:17
|
BTW I have nothing against Roger Clemens...I am mostly an NL fan so I have no hatred or anything for Clemens. I think Clemens is clearly one of the best pitchers in the history of the game and definitely top 4 in the past 20 years with RJ, Maddux, and Pedro.
|
| 34 | stinkypuff
ID: 2883927
Mon, Oct 01, 11:49
|
Yeah, I hate Clemens -- but even i would have a hard time voting for mussina over him. Just thought i'd throw the personality factor out there.
|
| 35 | F Gump
ID: 578382615
Mon, Oct 01, 12:15
|
A few answers to what has been addressed to me:
1) to HAWKES (16)
You said "Nine times this year Clemens has thrown his last pitch for the game and walked off the mound on the loss. In all but two of these starts, his bullpen and offense have rescued him and he has avoided the loss. In two of them the offense have actually managed to put the Yankess ahead while he was still the pitcher of record, and he has ended up with the win."
Interesting, but a mangling of stats and manager style to make your point. There were 8 times rather than 9, that I could find. In 3 of those, Clemens was "bailed out" of a loss, so he was indeed "rescued". In 2 of those, he lost. So he was actually "rescued" only THREE times, in the course of a year, which is not a big deal either way.
The other 3? They were times when he was a subject of Torre's managerial style. Late in games, with the bullpen they have, when the Yankees score to tie or take the lead, Torre pulls his starter. So the offense scores, and Clemens gets pulled without pitching another pitch.
If a pitcher is EVER behind in a game, you could try to make the point that he was somehow "rescued" by his offense, I guess, but you would not because that is obviously absurd. So the fact that his team scored late, rather than early, is not relevant - and actually, you could alternately make a case that Clemens hung in there to give his team a chance, and did so long enough to where they won.
Again, it is this sort of analysis and examination that is TOO MUCH DETAIL. Sometimes you get bailed out, sometimes you don't. Chance is a part of the game, just accept it. Clemens has a ton of wins and few losses, and he deserves credit.
(2) to biliruben
To some degree, ALL stats are team stats. Some teams are better fielding teams than others, in ways that never show up in fielding percentage, which has a HUGE effect on ERA.
The official scorer can make a difference. There are hit-error decisions that are often borderline or questionable, yet the decision can hit a pitcher with 4-5 ERs, or alternately erase them as unearned.
Managerial style and even bullpen can greatly effect a pitcher's ERA. (For example, I saw El Duque get absolutely hammered in Texas in a start this year. His ERA got screwed cause he was not sharp yet was "taking one for the team" to rest the bullpen. If that was only one L, no biggie, a loss is a loss. But if ERA is the ONLY stat, then he is penalized all year for one bad outing).
Home park makes a big big difference. Early in the year, Hampton had a rather mediocre ERA, but he was the best pitcher in the league or close to it, cause he was pitching well at home. Raw ERA comparisons don't give him credit for that; W-L record does. And conversely, if ERA is the prime factor, teams that play in "pitcher's parks" will be the only ones that can have a Cy Young winner. The W-L gives some balance to that skewing from a park. So ERA could NEVER tell us how a Rockies pitcher compares to a Dodgers pitcher.
And finally, the game is about Ws, in a major way. ERA is an average of how you did in general; W-L tells how you performed each time there was a contest. Teams don't add up their stats at the end of the year and then have the pennant awarded on compiled stats - each day stands on its own. As the MOST IMPORTANT player each day, the pitcher needs to get credit or blame in a unique way for whether each day's game goes into that W or L column.
(3) to CHICAGO TRS
I don't say W-L is the only stat. Nor is ERA. Both are significant. A pitcher who has a sparkling ERA but has only won 7 games would never and should never be considered. A pitcher who wins 20 games with a 7.50 ERA should be disregarded also. But the nature of the game (since it is about daily wins and losses) says from past awards that the top pitchers as to W-L are then evaluated as to ERA and Ks and sometimes a few other minor factors, and then a CY winner is selected therefrom. And I think that is proper.
In your example, there would be nothing wrong with selecting RJ over Clemens because of the big difference in Ks and ERA. 20-2 and 18-6 are both great records. BUT - if you move it the other direction and say RJ has only 13-14 Ws, then you don't consider RJ, cause Ws ARE important.
I don't think you just give it to the top winner. Or the top winning percentage. Or the top ERA. Or the K leader. All are MAJOR factors, though, and all are relevant.
In the Clemens vs Mussina comparison that started this, the whole thesis was that W-L should be IRRELEVANT somehow. Or explained away. Or disregarded. But in that case, with similar ERAs and similar Ks, the big difference in W-L makes an easy case for Clemens over Mussina - and rightly so, Neyer and all the so-called "experts" notwithstanding.
|
| 36 | ChicagoTRS
Sustainer
ID: 4324316
Mon, Oct 01, 12:58
|
Neyer did pick Clemens over Mussina also...
"If I were voting for the Cy Young and had to pick between Clemens and Mussina, I would pick Clemens because Mussina has pitched poorly in more games than Clemens has."
Fgump...home ballpark is a good point and is definitely something that should be taken into consideration. A 3.5 ERA at Colorado is probably close to a 2.0 ERA in LA. After reading your last post I do agree with most everything you said. I guess we are about on the same page on this. It comes down to all stats should be considered and people should not just look at a gaudy win/loss record and automatically award the Cy Young to that pitcher.
|
| 37 | Mr. Budweiser
ID: 9911112
Mon, Oct 01, 13:32
|
Can we face reality please? ;) In Cy Young voting, it truly rarely matters what a pitcher's ERA is; the bottom line is the big W for a winning team.
Clemens could have an ERA of 10.78 and would STILL win the Cy Young solely because of his and his team's record.
|
| 38 | biliruben
Sustainer
ID: 3502218
Mon, Oct 01, 13:49
|
Sounds like Jamie Moyer could be your man, then.
20 wins in sight.
Now this guy knows how to win, baby. ;)
|
| 39 | Blizz
ID: 54721318
Mon, Oct 01, 18:48
|
Everyone knows that the high run support helps Clemens' record but all I've read about the high run support effecting his ERA is Neyer saying that it has little or no effect on the ERA. Neyer doesnt know what hes talking about when he says that pitchers dont pitch to the score. If a team is winning 16-1, no matter how bad a pitcher doesnt like giving up runs, if there is a runner on third and one out he will gladly trade a run for an out. With low run support the pitcher could be winning 2-1 with a runner on third and one out and he will try to bear down to get an out without letting the run in. Those meaningless runs add up and I'm sure with the big leads Clemens has had, he has given up more meaningless runs this season than anyone else competing for the Cy Young award. He could be leading the league in ERA if he worried about the meaningless runs or if he was in a pitchers park.
Also, someone mentioned complete games. Clemens has gone 8 innings 4 or 5 times this year and he could probably have 3 or 4 complete games but he is a power pitcher and throws a lot of pitches plus he has Mariano Rivera as the closer. If I was Torre and after 8 innings I could choose between Rivera or a Clemens who has thrown 115 pitches I wouldnt think twice about putting in Rivera. I realize that Mussina has Rivera as the closer too but in his 3 nine-inning complete games he has had a low enough pitch count going into the ninth to not be tired yet.
|
| 40 | jumpball, Co-Guru Jr
ID: 258453019
Mon, Oct 01, 19:09
|
Great discussion!!
While I was reading through this thread, I got the urge to look at some history. Let me take you back to 1968 . . .
In that year there were 3 dominant pitchers:
Bob Gibson was 22-9 with a 1.12 ERA (an amazing number; how did he lose 9 games?),
Juan Marichal was 26-9 with a 2.43 ERA, and
Denny McLain was 31-6 with a 1.96 ERA.
Unfortunately, I don't have run support numbers for that year, but something tells me that Gibson's was pretty low. ;-)
Now let's look at the Cy Young voting for that year. Upon first glance, it would seem that McLain and Gibson might have split the vote (there was only one award covering both leagues back then). But the fact remains that McLain was a unanimous winner.
Seems like there's a precendent here for the award being based on a large number of wins. And maybe that's the way it should be, considering the career record that is held by the namesake of the award.
|
| 41 | Wignall
ID: 238552323
Mon, Oct 01, 20:50
|
Re 40 - "...Seems like there's a precendent here for the award being based on a large number of wins. And maybe that's the way it should be, considering the career record that is held by the namesake of the award"
-Yes, but the namesake of the award also holds the career record for losses.
|
| 42 | Wignall
ID: 238552323
Mon, Oct 01, 21:44
|
Post 39 - “Everyone knows that the high run support helps Clemens' record but all I've read about the high run support effecting his ERA is Neyer saying that it has little or no effect on the ERA. Neyer doesnt know what hes talking about when he says that pitchers dont pitch to the score. If a team is winning 16-1, no matter how bad a pitcher doesnt like giving up runs, if there is a runner on third and one out he will gladly trade a run for an out. With low run support the pitcher could be winning 2-1 with a runner on third and one out and he will try to bear down to get an out without letting the run in. Those meaningless runs add up and I'm sure with the big leads Clemens has had, he has given up more meaningless runs this season than anyone else competing for the Cy Young award “
Lets have a look at the games in which the Yankees gave 7 or more runs in support – that is total support in full game, not just when he was on the mound.
7-3 – gave up 2 runs to trail 1-2
16-5 – gave up 5 runs when up 12-0
(Yes, I suppose he thought he could give up some here)
12-5 – gave up 1 run to trail 0-1, 2 runs to trail 1-3, 1 run when up 5-3, 1 run while up 9-4.
(Hardly think he wanted to give up the first 4)
9-4 – gave up 1 run to trail 0-1, 3 runs when up 5-1
(Ditto)
9-3 – gave up 1 run to trail 0-1, 3 runs to trail 1-3
10-1 – gave up only run bottom of 1st inning to trail 0-1
7-5 – gave up 4 when 7-0 up
(I don’t think he would consider these 4 meaningless)
8-5 – gave up 5 to trail 3-5
12-1 – gave up 0 runs
7-2 – gave up 1 run while up 4-0, 1 run while up 4-1
(Takes a game to within 2 runs, after being up by 4)
10-3 – gave up 2 runs when 5-0 up
(Again brings it close, but I guess he knew that the extra five runs of support were coming right? ;-))
9-5 – gave up 2 runs to trail 1-2, 3 runs to trail 2-5
7-5 - gave up 1 run to trail 0-1
Really, from the above there are only two instances in which he has a considerable lead at the time of giving up runs, 12-0 and 7-0. However, I hardly think that someone as intense as Clemens would be happy with allowing 4 or 5 runs when up any score. To him there would be NO meaningless runs. Yes, some pitchers allow one run in order to get an out, from the stats above we see that there is more to that in Clemens’ games.
|
| 43 | Blizz
ID: 54721318
Tue, Oct 02, 01:07
|
I didnt really look into those insignificant runs from Roger, I was just thinking that in theory a pitcher with a big lead would give up some runs that he normally wouldnt have. I agree with those runs you say are significant so I guess his ERA is probably around what it should be, although I do believe if he was in a pitchers park his ERA would drop(but then again he wouldnt get the high run support). Also I'd think he and the rest of the team probably is content with leaving with any lead big or small and having reliable guys like stanton and rivera close it out for him.
|
| 44 | jumpball, Co-Guru Jr
ID: 0922620
Sat, Oct 06, 20:39
|
So now that Clemens has lost his last 2 starts to finish at 20-3, does that hurt his Cy Young chances?
The top 4 candidates (in ERA order)
Freddy 18-6 with 3.05
Moyer 20-6 with 3.43
Mulder 21-8 with 3.45
Clemens 20-3 with 3.51
Just using Wins and ERA, it seems to me that Mulder should be the choice. Is there any other criteria that should be considered as strongly as those two stats?
|
| 45 | slug
Donor
ID: 50491812
Sat, Oct 06, 21:25
|
I think Innings Pitched is another relevant stat. I don't see his name mentioned often, but I think Joe Mays pitched a great season and should be considered ahead of Moyer.
My order of the best pitchers in the AL this year (ignoring W-L):
Garcia
Mulder
Mays
Clemens
But if I had to vote (and consider W-L), it would be difficult to pass up on Clemens (and I hate the guy).
|
| 46 | KrazyKoalaBears
ID: 266182910
Sat, Oct 06, 21:38
|
If you're going to count Wins, then you have to count Losses. The pitcher has about as much control over one as the other. Thus, using a Win/Loss differential seems in line: Garcia: +12, 3.05
Moyer: +14, 3.43
Mulder: +13, 3.45
Clemens: +17, 3.51 I personally think the +3 differential between Clemens and Moyer/Mulder makes up for the .10 ERA difference and I think the +5! differential between Clemens and Garcia makes up for the .50 ERA difference. Still, Clemens is the only pitcher ever to go 20-1, no matter what the final results of the season. This point is not lost on the sports writers and the Cy Young is Clemens'.
|
| 47 | jumpball, Co-Guru Jr
ID: 0922620
Sun, Oct 07, 14:00
|
Looking back to posts 40, 41 and 44, since Cy holds the record for both wins and losses, it seems like the award should go to Mulder -- he has the most wins and the most losses among the top candidates.
In all seriousness though, Clemens will get, and deserves to get, another trophy for his mantle.
|
| 48 | Wignall
ID: 238552323
Sun, Oct 07, 20:01
|
KKB - Re Clemens' record start - So you are saying that if a pitcher starts the season 21-0 next year in his first 21 starts and then goes 0-14 in his remaining 14 starts to finish 21-14, he will get the Cy Young purely because of his record start? (Yes, the scenario is unlikely, but not out of the question).
|
| 49 | KrazyKoalaBears
ID: 266182910
Sun, Oct 07, 20:24
|
Wignall, that's a bit absurd. Obviously I'm not saying that Clemens should get it only because of his record start, but the fact of the matter is that after his record start he was only 0-2 and has the fewest losses of any 20 game winner. Your scenario doesn't provide that. Why don't you just ask, "So if a batter hits 75 HRs in April, then goes 0-for-2000 for the remainder of the year, he shouldn't win the MVP? It's unlikely, but not out of the question." Geez, I guess when you look at things that way, then nothing is out of the question. I mean, what if a guy has a 0.00 ERA in 30 starts, but he's 0-30? (All unearned runs! It's not out of the question) Can we please try to limit our arguments to reality? The reality is that in the current situation, Clemens went 20-1, something that no other pitcher in the history of baseball has ever done, and then went 0-2 for a grand total of 20-3, not 20-14.
|
| Rate this thread: | If you wish, you may rate this thread on scale of 1-5. Ratings should indicate how valuable or interesting you believe this thread would be to other users of this forum. A '5' means that this thread is a 'must read'. A '1' means that this is a complete waste of time. If you have previously rated this thread, rating it again will delete your previous rating. If you do not want to rate this thread, but want to see how others have rated it, then click the button without entering a rating, or else click here. |
|
|
Post a reply to this message: (But first, how about checking out this sponsor?)
|
|