Batter's Box Interactive Magazine Batter's Box Interactive Magazine Batter's Box Interactive Magazine
It is one of my Principles of Baseball Thinking that you must make up your mind slowly. Make judgements slowly. Come to conclusions slowly.

No, more slowly than that.


One of my own longterm gripes about Toronto's current management crew is that they often seem to make decisions in a somewhat random, haphazard manner. They often seem to be acting on the spur of the moment, without bothering with second thoughts or reconsideration. Flying by the seat of their pants. Reacting to events, rather than charting a course and preparing for contingencies. Maybe it's an AL East thing.

By contrast, one might consider the example of Bobby Cox and John Schuerholz in Atlanta. It's barely May, and other teams are already making on-the-fly adjustments, already tossing overboard the carefully laid plans of the off-season - and all based on the results of a handful of games, over the first few weeks of the season. Not Cox and Schuerholz. You want proof? Last night the Braves ran Mark Redman out to the mound. Again. Redman had gone 0-3 with a 10.13 ERA in his first four starts, and he didn't disappoint last night - the Phillies knocked him out of the game in the second inning. I think Redman's chances of making it through the end of May as part of the Atlanta rotation are somewhat less than zero - but imagine if he worked for one of the AL East outfits. Two weeks ago he'd have been in the bullpen, or in Triple A, or designated for assignment, or (if he played for the Yankees) entering a witness protection program somewhere. You have to admire Cox and Schuerholz' commitment to their own judgement, and their willingness to stick by it until forced to act otherwise. Until persuaded by the actual evidence. These are not qualities I associate with the Toronto management team.

Well, what the hell do Cox and Schuerholz know about winning in the major leagues?

Mark Redman is doing his best to force the Braves to make a judgement, and I expect he will soon force their hand - but no one will be able to say the team acted hastily, or impulsively, or irrationally. And sticking with him for an extra start or two or three may very well have cost them a game or two. A close pennant race, like the 2007 season in the NL East promises to be, may very well come down to a game or two. But the principle of judging and deciding and assessing in this manner, as a way of conducting your business, is ultimately is worth much, much more than two particular games in April or May. The investment pays off elsewhere, and it does pay off. Abundantly. And Bobby Cox has an awfully good record in close pennant races anyway.

At any rate, nothing's happened to make me change my mind since I wrote:

I very much hope the Jays stick [McGowan] in the Syracuse rotation and forget about him. Just leave him alone until September. No bouncing back and forth between AAA and the majors, between the pen and the rotation. Let the kid pitch.

Just to recap:

2004
Reconstructive elbow surgery on May 13.

2005
While rehabbing in Florida, is diagnosed with Type II diabetes
Returns to active duty on June 1 at Dunedin. Goes 0-3, 3.70 in 11 starts at A and AA.
Promoted to Toronto on July 30. Went 1-3, 6.35 in 13 games, 7 starts.

2006
Reports to spring training 15 pounds underweight, adjusting to his new nutritional program.
Optioned to Syracuse on March 18. Will be converted to a relief pitcher.
Recalled by Toronto on April 27. Makes 5 relief appearances (1-0, 7.94)
Optioned to Syracuse on May 11. Oh, forget this bullpen crap. Back to the rotation.
Recalled by Toronto on July 29. Makes one start, loses, and goes to the bullpen. Makes 6 relief appearances (10.91 ERA) and...
Optioned to Syracuse on August 23.
Recalled by Toronto on September 2.

So yeah, I still think it's a good idea to just leave the kid alone for a while, and let him pitch.

Granted, somebody has to start on Thursday. Granted, Ricciardi and Gibbons would probably rather have red hot pokers inserted somewhere tender than give the ball back to Towers. Granted that the day after Victor Zambrano (he of the 5 BB per 9 innings over his career, he who has worked half a dozen innings in the past month, he who is less than a year removed from TJ surgery himself) faces one of the most patient lineups in all of baseball, there's just a chance that the bullpen may be a little depleted and a real live starting pitcher will be required on Thursday...

I'd still have left McGowan alone. If absolutely pressed, I'd rather call on Taubenheim or Banks or Brad Arnsberg himself for an emergency start to get the team through the road trip, and then consider my options.

Now maybe this will work. It could work. Maybe after all of the turmoil, maybe after all the serious and frightening health issues (surgery on his pitching arm! an incurable disease!), maybe after being bounced around like a goddam yo-yo between the majors and the minors, between starting and relieving... maybe, after all of this nonsense, maybe all McGowan needed to settle his life and career was just one uneventful month in upstate New York. Sure, this could work.

This had better work.

Finally, a little cut-and-paste, presented without comment.

Toronto Blue Jays Team Pitching Statistics (as starter)

NAME G GS W L IP H ER R HR BB SO K/9 P/GS WHIP ERA
Roy Halladay 6 6 4 0 47.1 35 12 13 3 7 33 6.27 107.0 0.89 2.28
Josh Towers 4 4 1 3 23.0 27 12 17 4 4 21 8.22 88.0 1.35 4.70
Tomo Ohka 5 5 2 2 28.2 29 16 18 6 11 12 3.77 91.0 1.40 5.02
A.J. Burnett 6 6 2 2 33.0 30 20 20 5 21 26 7.09 99.8 1.55 5.45
Gustavo Chacin 5 5 2 1 27.1 29 17 17 6 7 11 3.62 84.4 1.32 5.60

Blue Jays 26 26 11 8 159.1 150 77 85 24 50 103 5.82 95.0 1.26 4.35

Well, whatever. I'm sure they know best.

But I'm beginning to wish Dave Berg was still on the team. And that scares the hell out of me, too...
2 May 2007: Here's Your Hat, What's the Hurry? | 107 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
Chuck - Wednesday, May 02 2007 @ 07:15 AM EDT (#167050) #

Finally, a little cut-and-paste, presented without comment.

When Burnett's numbers suggest he's indistinguishable from 3-4-5, it's hard to be thrilled that A.J. is a Jay.

John Northey - Wednesday, May 02 2007 @ 08:07 AM EDT (#167051) #
The more I look at the numbers the more I wonder about this team.  Josh has, by any measure outside of run support, been the second best starter on the team.  If you had to pull him then put in Janssen instead as he has shown some ability so far and has actually thrown a few innings.

I have a feeling we'll see someone knocked out early, Gibbons giving in and giving Josh the ball and Josh pitching his butt off then the other guy gets a second start while Josh sits gathering splinters in his butt. 
Jacko - Wednesday, May 02 2007 @ 08:22 AM EDT (#167052) #
Speaking of Janssen, what the heck happened to his strikeout rate?  

After mowing guys down effciently in spring training, he's managed on 2 K's in 13.2 IP.  This goes along with a .95 WHIP and a 1.32 ERA, so I should not complain.  But is this a conscious choice, or a ticking time bob? Lord knows the rest of the relief staff (save Accardo) have blown up pretty badly this year.


AWeb - Wednesday, May 02 2007 @ 09:23 AM EDT (#167053) #
It's a small sample for Janssen so far. Expecting a strikeout rate around 6-7/9IP in the long haul wouldn't be unreasonable, but sometimes that'll include stretches of 13 innings with only 2K. At least his K/BB is at 2, that's a good peripheral, right? For instance, Halladay only had 2 strikeouts in consecutive starts already this year, including his 10 inning game.

The more I look at Towers' numbers, the more baffling it is. Somehow, Burnett throws 115 pitches last night, gives up 7 runs and 2 HR. In his last start, Towers gives up 3 HR in the 3rd, gets pulled in the middle of the 5th inning at 70 pitches, after getting 2 outs and someone reaching on an error, striking out 7 and walking none. How can that start possibly be "the last straw"?
Squiggy - Wednesday, May 02 2007 @ 09:29 AM EDT (#167054) #
I am no Towers fan, but looking at that chart it is very difficult to make the case for removing him, of all people. Just another knee-jerk decision by JP and the gang, not too surprising in that light.

Also, Burnett's numbers are a huge disappointment - given all the talk about new attitude, hanging out with Halladay, feeling healthy and strong etc. He has had good early-inning run support and mediocre oppostion so far and despite this is walking more people than ever. Small sample size, granted, but not a good start for the #2 on a supposed contender.
Squiggy - Wednesday, May 02 2007 @ 09:37 AM EDT (#167057) #
One more point - the Redman analogy reminded me of a very similar case closer to home, when our own Josh Towers was allowed to muddle along in similar fashion for an even longer period. In hindsight, this patience with his abysmal performance was a true difference-maker in the standings. As many others have said, this is likely colouring judgements for 2007.

Which is all to say, it's not that JP and the gang are not ever patient. Perhaps they just are not very good at picking their spots.
actionjackson - Wednesday, May 02 2007 @ 11:14 AM EDT (#167059) #
I recall reading somewhere that during Halladay's wanderings in the desert of re-building everything from the ground up in 2000-2001, Mel Queen would constantly repeat the phrase (paraphrasing): "That's a lot of talent going to waste." I think the same can now be said of Burnett, although his start to this season is nowhere near the disaster that Doc's 2000 was. It's a shame that they can't re-build Burnett from the ground up, seeing as 1) They'd lose him on waivers  2) They desperately need him where he is and  3) He's too pig-headed and proud to learn and re-learn the art of pitching. I really hope Roy gave it to him after the latest debacle last night, but I'm not sure anything will help at this point. A.J. may be able to keep up with Roy physically, but he's a baby when it comes to the mental/psychological side of pitching. The problem is if/when the light goes on for him, he will dominate and it may not be here in Toronto. Then, the Griffins and Elliots of the world will have another example to hold up as an exhibit of JP's ineptitude, despite the fact that they've excoriated Burnett since he got here. It's fun to be a Monday morning quarterback isn't it boys?

Speaking of JP's ineptitude, it is in the way he handles young talent primarily. Not only McGowan has suffered his jerking around, but Marcum has too, and perhaps Adams. Fortunately, they haven't messed too much with Janssen's head. It makes me worry about Purcey and Romero. Are they the next ones to suffer his impatience and his "Strat card" GM mentality? I use the Strat card analogy with reference to the story in "Moneyball" about Beane burning cards that were underperforming for him. I find it incredibly ironic that a so-called "old school" GM like Schuerholz seems to understand the concept of small sample size better than a "new breed" GM like Ricciardi. JP's roster construction also leaves a lot to be desired, but that puppy has been beaten to death so I won't touch it here. Ironically, considering how inept he can be at dealing with people and PR in general, he seems to be able to acquire team oriented players (sometimes at the expense of ability, but nonetheless), who from this distant vantage point seem to get along with and pull for one another. Then again the Yankees won a lot of games while openly fighting with each other, sometimes right in the dugout, so who knows if that's important or not? As mentioned above, the leash that players, particularly pitchers are given seems to depend on the potential or ceiling of the player rather than actual performance.That, plus being overly patient with Towers last year is the only reason I can find for his abrupt banishment to the bullpen. As for the Gus injury: "There's a lot of moving parts in his delivery" doesn't cut it for this fan. Obviously his wacky delivery is placing strain on his forearm/elbow/shoulder, so try to cut down on the number of moving parts, particularly that twisting thing he does with his arm behind his back. Ouch!

Cox and Schuerholz understand that in order to win, you need contributions from everyone on the roster. But there's a limit to everyone's patience. I suspect Redman will be given 8-10 starts to work things out and then will be in the bullpen or on the waiver wire. Ricciardi despite it being his sixth year on the job is still learning when to "hold 'em and when to fold 'em". Schuerholz has been in baseball for 40 years and learned from some of the greats in Baltimore and KC. Cox has been managing or GMing for 30 years. Maybe it's time to force JP to take on a veteran baseball advisor type in order to help with decisions requiring patience and development. There's a lot of youth on the management staff, but maybe a greybeard or two is required. You know, the baseball equivalent of Wayne Embree. Of course, that might take us right back to pig-headedness, which is where we came in.

Mike Green - Wednesday, May 02 2007 @ 12:09 PM EDT (#167062) #
I suppose I owe a comment on the starting pitcher numbers.  When Burnett was signed, we ran a poll and I gave the signing a "1", the lowest of all ratings.  As it has turned out, Burnett has now thrown 168.7 innings in the AL, with an ERA of 4.27 and a 144/60 K/W.  That is about what I expected that he would do.  That said, in hindsight, I was wrong.  Burnett's performance doesn't seem like a decent value for $10 million a year, but it is.  He's not a great pitcher, but he is a good one and the Jays could use a couple more of those. Now, if he can get back to throwing a few more strikes...

Magpie's dissertation on McGowan is entirely accurate.  As others have noted, this is not the first time that the Ricciardi/Gibbons/Arnsberg team has had difficulty sorting through pitchers with skill.  It is probably the most difficult aspect of the job, and one that takes time to learn.  At some point, ownership will have to make a judgment whether the "pitching management" team is learning, and whether they have the capacity to handle the job. 

ayjackson - Wednesday, May 02 2007 @ 01:13 PM EDT (#167063) #

Right now, Lilly or Meche at $12 or $13m looks like good value, and I wouldn't have thought that.

McGowan's first start figures to be a tough one - in Cleveland.  I'm not sure I can think of a worse place for him to make his return.  Given that, I hope they give him a few starts while Chacin is on the DL - and not rush Thomson up if he has a good outing at AAA.

As for the curious treatment of Towers, I think it was predictable from JP's perspective, if not from the "bench-boss perspective".  Towers is in the last year of his contract and does not figure to be here next year.  If he were to continue to pitch badly this year, the Jays won't offer him a contract next year.  If he were to pitch well this year, he would go to the highest bidder in free agency - and I'm not sure the Jays would feel comfortable being that team, given his track record.  Zambrano, on the other hand, has a cheap team option for next year - $4m, I believe - and JP must find out if he can be a back of the rotation starter for this team.

 

ChicagoJaysFan - Wednesday, May 02 2007 @ 01:26 PM EDT (#167065) #

I think there are a lot of biases going on in some of these posts: selection bias, confirmation bias, recency bias, attribution error, etc.. 

To reference actionjackson's post specifically, the only examples given are McGowan and Adams and I don't think they're reasonable at all.  I think that McGowan has suffered much more from Tommy John surgery, diabetes, and being a pitcher, which is an inherently unpredictable development path than he has from being "jerked around".

Russ Adams was a starting shortstop throughout his minor league career, 2 seasons in Toronto ('04 and '05) and then until he all of a sudden stopped being able to make the throw in '06.  I don't see the jerking around at all with him.
robertdudek - Wednesday, May 02 2007 @ 02:41 PM EDT (#167067) #
There has to be something else going on with the Towers decision, other than the merits of his pitching.

As for acquiring "team players" - Shea Hillenbrand - he sure turned out to be a team player, eh?

I disagree somewhat with the "jerking around" comment. Every team sends out some of its young players to AAA and recalls them at various points. It's better that Adams plays in AAA than sitting on the bench in the majors.

Rios and Hill have come up, and despite struggles at various time, have remained in the lineup, so I don't think it's quite fair to paint JP with the "serial yanker" brush.


Mylegacy - Wednesday, May 02 2007 @ 02:46 PM EDT (#167068) #

I agree with more of CJF's post than AJ's.

Burnett has morphed into Ted the Tease. Actually, he was closer to Ted (results wise) all along and we just sort of ignored it. Interesting to see if he takes his post-08 "walk," interesting to see if we care. Remember when El Artista came and we all thought we'd be the ones that get him to the next level...hope springs eternal.

Last night was one of the few games this year were we actually got beat.

McGowan has literally been to hell and back, physically, mentally and professionally. It would be better for HIM to be in AAA all year, maybe a September call-up. HOWEVER, this year and next are win now years. With Sparky, Ryan, League and Zaun down we need some guys to come forward. Lind is welcome, McGowan... we need you to be who we hope you can be...and we need you to be there now. In my heart I feel he is ready, I hope he does in his. By mid-June we'll know.

Tonight's game...I wonder if Zambrano will be wearing a helmet on the mound and if Gibby is starting to dig his foxhole yet. Strangely, I feel the ZBomb will be a pleasant surprise tonight. But then, I thought Dewey would beat Truman too.

Ron - Wednesday, May 02 2007 @ 03:20 PM EDT (#167070) #
I would be very interested to see the Jays W-L record since 2003 if you remove all of Doc's starts. I'm guessing the Jays are below .500 without him.

Anybody else having trouble listening to Jays talk? The fan 590 website doesn't broadcast it. Even the 980 Kruz website hasn't streamed Jays talk this season (at least when I tried to tune in). Sometimes it's tough being a Jays fan that lives outside of Ontario.





actionjackson - Wednesday, May 02 2007 @ 03:59 PM EDT (#167071) #
Ron, Jays' talk has been unavailable on Raptor playoff game nights because the Raptors take over the flagship airways and bump the Jays up to 610 CKTB in St. Catherines, which I'm finding to be tough to get in Toronto due to the high pitch whirring interference. I don't know if that means Wilner has to travel to St. Catherines, but it does mean he can't do the Jays' talk portion of the wrap up show.

Probably some of what I said this morning was out of frustration with inconsistency, injuries etc., but I stand by the assertion that a senior advisor or two to the GM wouldn't hurt. It's great to see all the youth in the front office, but I think 1 or 2 people that have been in baseball for 25+ years might help balance it out. Theo Epstein has Bill Lajoie, Brian Cashman has Gene Michael, and Billy Beane learned from Sandy Alderson, but who does JP have, even if it's only to bounce ideas off?

actionjackson - Wednesday, May 02 2007 @ 04:16 PM EDT (#167072) #
Also Ron, ask and you shall receive:

Blue Jays overall record since 2003:   333 W  340 L  .495 PCT

Blue Jays record since 2003 when Halladay pitches:   77 W  37 L  .675 PCT

Blue Jays record since 2003 when Halladay does not pitch:   256 W  303 L  .458 PCT


There you are Ron, as we suspected all along, Roy Halladay is a god. 180 points above the overall winning percentage and 217 points above the winning percentage in games when he does not pitch. In a word: Wow!

Ron - Wednesday, May 02 2007 @ 04:33 PM EDT (#167073) #
I forgot about the Raptors being on the same station as the Jays.

Okay next time somebody tells me Vernon Wells is the most valuable player on the Jays because he plays everyday, I will show them those stats.  The W-L records are like day and night.


VBF - Wednesday, May 02 2007 @ 04:45 PM EDT (#167074) #
I would be very interested to see the Jays W-L record since 2003 if you remove all of Doc's starts. I'm guessing the Jays are below .500 without him.

I'd also guess that most teams that haven't made the playoffs since 2003 are below .500 when their ace doesn't pitch.

On a different matter, let's not believe Ricciardi's propensity to flip flop players here and forth is anything old. This is a man who gave Josh Towers v. 2006 12 starts last year.

Again, I think Josh Towers was perhaps the wrong guy for the quick pull, but I like the idea of it anyways. Before mid-May rolls along, we need to know who our best starting pitchers are and we're not going to know unless we give them all a shot to see if they sparkle. I for one, am looking forward to getting Zambrano, McGowan, and hopefully Janssen, Marcum, and Thomson into games. Because I guarantee you that at least one of those guys is going to sparkle for us.
Mike Green - Wednesday, May 02 2007 @ 04:51 PM EDT (#167075) #
Here is a classic illustration of the difficulty in developing accurate defensive metrics.  BP's Rate and the Jays shortstop/second base combination will illustrate.

Aaron Hill had rates of 120 and 116 in 2005 and 2006. So far, this year, it is 100.  The reason is that he has participated in fewer double plays than in previous years.  We know subjectively that his pivot is very, very good, so what's going on?  Royce Clayton has been poor on the DP for years, and has been much worse than that this year.  He has a Rate of 84, and that is likely an overstatement of his current defensive ability. John McDonald is good enough, especially when he is not overworked, and he and Hill make an excellent DP team.  He has a Rate of 117 as a shortstop so far this year, and that is also an overstatement. Clayton has played much more than McDonald, and so Hill's overall DP rate is off.  Rate apportions a significant part of the decline in 4-6-3 and 6-4-3 DPs for the Clayton/Hill combination to Hill. 

Rob - Wednesday, May 02 2007 @ 05:02 PM EDT (#167076) #
Roy Halladay is a god. 180 points above the overall winning percentage and 217 points above the winning percentage in games when he does not pitch.

If Halladay is .217 above the others, Santana is eerily identical at .216:

Twins overall WPct since 2004, when Santana became a full-time starter: .559 (286-226)
When Santana starts: .729 (78-29)
Other games: .513

Peavy, to take but another option from the current poll, is .071 above since 2003.

I'd also guess that most teams that haven't made the playoffs since 2003 are below .500 when their ace doesn't pitch.

Oh yes, it's an amazing concept, isn't it? Sportsnet showed a graphic on Monday night detailing how the Jays have performed worse in the games they lost, leading those I was watching the game with to wonder if they might have performed worse on the exams they failed.
Mike Green - Wednesday, May 02 2007 @ 05:23 PM EDT (#167078) #
Halladay was named MLB AL pitcher of the month for April, and will receive a HD-TV.  At this stage, do players really need to receive loot with their awards?  It seems superfluous to me.
Magpie - Wednesday, May 02 2007 @ 06:02 PM EDT (#167079) #
Are they the next ones to suffer his impatience and his "Strat card" GM mentality?

I absolutely do not believe Ricciardi has a Strat card mentality. I believe Ricciardi is an old-school GM. I believe he has way more in common with Pat Gillick than Billy Beane. Like Gillick, he was a scout for a long time, and he has all of a scout's prejudices and tendencies. He may have learned a few things from Beane, but... you can take the girl out of the trailer park but you can't take the trailer park out of the girl.

A little Jerry Springer wisdom, there.

Also, I wasn't thinking of Towers being pulled from the rotation after four starts as being an instance of the management group's tendency to make sudden decisions. If you believe the official explanation, which I certainly don't  - they can't possibly be that stupid  - it's much more like the Dave Bush demotion in May 2005. Which as I recall made roughly the same amount of sense on the surface, i.e. none at all.

But I don't believe them anyway. I think they had their minds made up about Towers before spring training. So the real mystery is why this group had him in the rotation to open the season in the first place. I suppose it's possible that the fact that Towers outpitched every starter in the spring not named Halladay may have had a slight influence on events, but I doubt it.  And believing in that one day and not believing in it the next also makes them look dumb and flighty, if you ask me.

So the answer to that little question can only be because John Thomson got injured and they were sure Zambrano wasn't ready to go into the rotation as early as Opening Day. It's actually got nothing to do with Towers, and has very little to do with how he performed. He's finished here, he's now auditioning for the other twenty-nine teams, whether he knows it this time or not. And that's a judgement I can certainly understand, and it's not one I would call rash or sudden, even if I may not share it myself.

I was thinking in particular of the handling of McGowan, the Aaron Hill shuffle last summer, and the Miguel Batista business in spring 2005.
ChicagoJaysFan - Wednesday, May 02 2007 @ 06:47 PM EDT (#167082) #
I know I said I wouldn't comment on Towers anymore, so I apologize - but couldn't yanking him have to do with the fact that he had 3 bad starts in 4 outings.  He has been noticeably worse than the others in the rotation, but it just doesn't show up in those selected statistics (I know that someone else probably chose which stats to use, so I'm not trying to accuse Magpie of anything there).  Furthermore, the fact of the matter is the back end of the rotation has not been good enough so it's time to start seeing if the Jays can fix it.  They've got a lot of potential fixes on hand (Zambrano, Jansen, McGowan, et al), so why not start to try them out.  Then the question becomes:  who do we replace?  Well, I think it comes down to who do you think is least likely to improve over the rest of the season.  I think the guy that gave up 5 runs in his last 7&2/3 innings in 2006 to lower his ERA to 8.42 and hasn't done anything to indicate he's not the same pitcher is the ideal candidate.  Here are his 4 starts so far:

Against the Royals - 5&2/3 innings, 10 hits, including a homer, double, and two triples, to cause 6 runs (only 3 earned though).  The 3 unearned are not indicative of bad luck in my book however.  Sweeney reached base on error to lead off the 6th.  Towers then gave up a homer, followed by two outs, followed by a double and triple (Downs gave up a single for the last unearned run).  Remove the error as if the at-bat never happened and he still gives up two of those runs (I have issues with the unearned runs are disregarded entirely).  That's a bad start.

Against the Tigers - great start.  7&2/3 innings, only one unearned run with only 3 hits and a walk.  The only average or better start this year.

Against the Orioles - 5 innings with 5 runs, only 4 earned (but considering it's Josh's error, I think that's something that should count against him).  9 hits (but interestingly all singles) and 3 walks.  I'd call that another bad start.

His last start, against the Rangers - 4&2/3 innings with 5 runs (all earned and none occuring after he was pulled).  Yes, he had a great K:BB ratio (7:0), but he also gave up  3 homers, a double, and two singles in less than 5 innings of work.  I'd call this a bad start.

Finally, if you attribute the ER's on an as-responsible basis (i.e. give Towers the run caused by his error in the Oriole game and two from the Royals), then Towers ERA is all of a sudden 6.26.  To be fair to him and charge everyone else for all of their unearned runs, he still ends up with an ERA that's half a run higher than everyone else.  His K/9 and BB/9 may be great, but look at how he gets beat like a rented mule in those games.

Why aren't those enough reasons to be yanked?  Why can't the simplest and obvious explanation work? 


As to why the Jays have said they had other reasons for moving him to the bullpen - I don't see how they gain at all if they tell the media they put Josh in the bullpen because he just isn't a good starting pitcher.
robertdudek - Wednesday, May 02 2007 @ 07:50 PM EDT (#167083) #
It's actually got nothing to do with Towers, and has very little to do with how he performed. He's finished here, he's now auditioning for the other twenty-nine teams, whether he knows it this time or not.

Well, I CAN'T understand it. If management is not willing to allow a player to change their minds with his performance on the field, then they are absolutely LOST AT SEA.

For anyone with open eyes, Josh Towers has pitched better this season than 2 of the other 4 members of the rotation. Performance on the field should trump any and all preconceived notions.
robertdudek - Wednesday, May 02 2007 @ 07:57 PM EDT (#167084) #
Chicago Jays Fan,

The stats quoted by Magpie ARE the relevant ones. Particularly strikeout and walk data - the best indicator of future effectiveness bar none.

There is no question but that Towers outperformed Okha and Chacin - and by a comfortable margin. No question.

GregJP - Wednesday, May 02 2007 @ 08:23 PM EDT (#167085) #
Josh Towers walking out of the bullpen just now looked like a 15 year old boy walking into his first high school dance.  :)

Oh, yeah, and.................Zambrano sucked like we ALL new he would and will in the future.

Magpie - Wednesday, May 02 2007 @ 08:28 PM EDT (#167086) #
For those curious, ESPN selected the stats. Just basic Blue Jays pitching stats, choose the split "as starter." I deleted any columns that applied to relief pitching, like Saves or Holds.
Mike Green - Wednesday, May 02 2007 @ 09:04 PM EDT (#167087) #
Ahh.  5th inning, bases loaded, down by a run.  Stairs pinch-hits for Clayton and walks.  McDonald pinch-runs.  Just like we hoped would happen.  Finally.
GregJP - Wednesday, May 02 2007 @ 09:04 PM EDT (#167088) #
Pat "genius"Tabler........"I wonder if there is something wrong with Clayton"  (as Stairs pinch hits for him)

Ah, Pat, ya there is something wrong with him.  He can't hit, and based on his performance so far this year he can't field any more.  So if you have a career OPS in Manny Lee's general area and routine ground balls are eating you up.................................

GregJP - Wednesday, May 02 2007 @ 09:24 PM EDT (#167089) #
Alex must not have prayed enough today, as God just made him drop a fly ball.
GregJP - Wednesday, May 02 2007 @ 10:03 PM EDT (#167090) #
I really like Janssen, but I'm confused.

You pitch Accardo and Frasor yesterday in a blowout, and now you pitch Janssen in the late innings in a tie game. 

Rob - Wednesday, May 02 2007 @ 10:27 PM EDT (#167091) #
If my roommates were subjected to this long, boring game, I'd imagine they would update Mrs. Named For Hank's classic line: "Come on, can't we turn this off and play Wii Tennis?"
VBF - Wednesday, May 02 2007 @ 10:34 PM EDT (#167092) #
Ahh.  5th inning, bases loaded, down by a run.  Stairs pinch-hits for Clayton and walks.  McDonald pinch-runs.  Just like we hoped would happen.  Finally.

I can look past all the managerial choices I took issue with in the past. This move was just perfect. He gets it.
Magpie - Wednesday, May 02 2007 @ 11:18 PM EDT (#167093) #
Except I think there was something wrong with Clayton (besides him being, well, Clayton). Something's up with that. Maybe he had Vernon's flu or something.

May I be the first to predict that after this game Towers is bound for Syracuse or DFA'd and Taubenheim or someone else is coming up? At this moment, Marcum is entering the game. Tallet and Frasor are all that's left in the pen. A kid from AAA is starting tomorrow.

A move will be made after the game. Mark my words.
John Northey - Wednesday, May 02 2007 @ 11:23 PM EDT (#167094) #
Dang, went for a walk and was going to post that Towers is doomed after today and Magpie beat me to it.

A good question is what does Cleveland do now?  7 relievers used with just 2 of them under 20 pitches (Mastney so far and Fultz at 16 pitches).  Crazy stuff.

Magpie - Wednesday, May 02 2007 @ 11:23 PM EDT (#167095) #
On the other hand, Marcum should still be available tomorrow.
AWeb - Wednesday, May 02 2007 @ 11:27 PM EDT (#167096) #
I ws going to comment on the possibility of releasing Towers (I say do it, if yo're not going to use him as a starter, he's taking up a roster spot).

And then Hafner dribbles one down the line for a game winning RBI. How can any defense, no matter how much it's "playing the percentages", leave room for someone to essentially hit a hard bunt to win the game with a runner on first? I assume Lind was in "no doubles" defense, which meant he had no chance from there on that play, and the Cleveland third base coach knew it (he was waving him home as the ball went past him). Fracking frack frack. Grumble grumble.

GregJP - Wednesday, May 02 2007 @ 11:29 PM EDT (#167097) #
Baseball is sometimes strange/cruel.

Janssen has 3 or 4 balls scalded against him but escapes unscathed, while Marcum gives up a flare and a dribbler down the 3rd base line and is the loser.

I agree, that was Josh's last appearance as a Jay.  I hope he resurfaces in the National league in a pitchers park as a 5th starter.  I think he can be decent in that role.

Thomas - Wednesday, May 02 2007 @ 11:30 PM EDT (#167098) #
You finally give Gibbons his precious second lefty for two left-handed batters who hit 150 and 225 points of OPS worst against lefthanded relievers and what does he do? Go with Marcum.

And the third batter due up in the inning is a switchy who is 60 points worst against lefties.

Thomas - Wednesday, May 02 2007 @ 11:37 PM EDT (#167099) #
I'm not blaming Marcum for that loss because Hafner hit that ball perfectly, but I still think that Tallet is the better option for at least Dellucci and Hafner there. I also don't understand why Gibbons pinch hit Smith for Fasano and used up the last two bench players with two outs in the ninth inning. Fasano's just as much of a power threat as Smith and you burn your last two bench players for what is, at best (and I don't believe it is) a marginal hitting improving and also defensively downgrade your catcher, which almost came back to haunt us as the umpire blew the call.

I agree with the consensus that Towers is done and it's not his fault. The Jays management of the roster this year has been more frustrating than I can remember in any of the past few years.

Ron - Wednesday, May 02 2007 @ 11:43 PM EDT (#167100) #
Marcum throws a pitch out of the strike zone and Hafner still hits it to left field. Baseball can be a cruel game sometimes.
I'll be a happy man if I never see Victor Zambrano pitch in a Jays uni ever again. He needs glasses to find the strike zone.

Chuck - Wednesday, May 02 2007 @ 11:47 PM EDT (#167101) #

And the third batter due up in the inning is a switchy who is 60 points worst against lefties.

And the 4th batter is Nixon, who has never hit LHP. If a RHB comes in for him, you flip Tallet out for a RHP.

I'd suggest not only that Towers is done, but that Marcum is AAA-bound to regain his presumably shaken confidence. I'd suggest both, but are there enough replacement arms ready to go?

Magpie - Wednesday, May 02 2007 @ 11:52 PM EDT (#167102) #
I'll be a happy man if I never see Victor Zambrano pitch in a Jays uni ever again. He needs glasses to find the strike zone.

It's his fatal flaw, and at his age he's unlikely to get over it. But he does some other things well enough to have been a perfectly average starting pitcher, and in the AL East.

What I don't get is this. If you're thinking of Zambrano as a starter, at some point - which the team obviously was - why on earth wasn't he in Syracuse actually being used as a starter? Instead of pitching 5.1 innings of relief in five weeks. He's going to have his second spring training and stretch out his arm in the major leagues, in May. Why?
Thomas - Wednesday, May 02 2007 @ 11:55 PM EDT (#167103) #
Magpie, I think this issue came up the day the move was announced and nobody had a good answer or even speculation as to why the Jays would do this. They took a guy who threw six innings in his last start of spring training and had him throw 5.1 innings in April. Which is also a perfect illustration of the uselessness of having a permanent seventh reliever. It might be useful in particular series for example, such as this one, but again, there's no need to have one permanently entrenched in the bullpen.
Ron - Thursday, May 03 2007 @ 12:09 AM EDT (#167105) #
It's his fatal flaw, and at his age he's unlikely to get over it. But he does some other things well enough to have been a perfectly average starting pitcher, and in the AL East.

I know Zambrano had some success pitching in the AL East in 2003 and 2004. Too bad we are in the year 2007. Zambrano has no business starting for a team that thinks it's a playoff contender. The sooner the Jays get rid of him, the better off the Jays will be.

What I don't get is this. If you're thinking of Zambrano as a starter, at some point - which the team obviously was - why on earth wasn't he in Syracuse actually being used as a starter? Instead of pitching 5.1 innings of relief in five weeks. He's going to have his second spring training and stretch out his arm in the major leagues, in May. Why?

JP said the best 12 pitchers were going to make the Jays out of Spring Training. I'm guessing JP thought the Jays were better off with Zambrano pitching for the Jays instead of the Sky Chiefs.
China fan - Thursday, May 03 2007 @ 12:48 AM EDT (#167107) #

   I think it's time for Janssen to be converted to a starter.  He's the only pitcher -- aside from Halladay -- who has shown consistent success throughout 2007, both as a starter in the Grapefruit league and as a reliever in the regular season.  I know he is valuable in the bullpen, but he'd be much more valuable in the starting rotation if he can make the switch.

    The other issue is Burnett.   For the second year in a row, he is the guy who can make or break the Jays.  His injuries last year -- and his inconsistency this year -- have been the biggest reason for the team's failure.  If he was healthy and consistent, the Jays would have two aces and they'd be contenders.  The team can patch together a 3, 4 and 5 starter if they were getting strong performances from the number 2 slot. 

Avail - Thursday, May 03 2007 @ 08:42 AM EDT (#167109) #
China Fan, couldn't agree more.

I can only assume the sentiment is that we don't have a long list of reliable relievers so keep Jansen in the pen. I certainly hope its not actually some residual punishment being dished out for his failure to inform the team of his injury last year.

The part I cannot understand is this: Jansen (when healthy) has shown he can pitch deep into games consistently. This is the guy who had two 1-hit 7+ inning performances last year as a rookie. His deportment, variety and quality of pitches, and his effortless delivery scream starting pitcher. Wouldn't he be MORE valuable in a starting role where he can give the bullpen rest from having to pitch 4 innings 3 of 5 days.

As for Burnett, he was really inconsistent at the beginning of last year but then had a good stretch in the middle/end of the season. We can only hope that happens here, but for a supposed 1B starter who he HOPED gave us one of the strongest 1-2 punches in the league he's not living up to billing.

Arnsberg and Doc both need to drill into AJ that throwing 100mph is only effective if/when you can hit the zone consistently. Dial it down a notch, gain the accuracy, and get ahead in counts. When you're up 0-2 in a count then you can let it rip on a pitch without as much pressure to get it in the zone, and on an 0-2 fastball there is a lot more chance you get the batters to chase if the pitch is a borderline strike.

AWeb - Thursday, May 03 2007 @ 09:24 AM EDT (#167110) #
See, I thought the problem with Burnett in the last start was that he had dialed it down too much. He seemed to be in love with his tailing 2-seamer that goes about 90. The problem being, of course, that if a tailing 90 mph fastball doesn't tail enough, or isn't kept down in the strike zone, it gets pulverized. But that was just my impression based on seeing a couple of innings...I couldn't watch it anymore after a while.

I would rather see Burnett rear back and try to blow people away, personally. Let's be realistic about this..he's 30, has never been able to locate his pitches consistently, and his fastball is only a good one because it goes very fast. For every Randy Johnson type who figures out how to stop walking so many people at the age of 30 (most similar at age 28: Burnett, oddly), there's a hundred who don't. Burnett has been an above average pitcher pitching the way he always has, I'm content for him to keep doing that. Well, I'll still complain sometimes...but Burnett is not Halladay. The only thing they have in common (as players) is being tall right handed white guys who can throw hard. As it has been noted before, location is a natural talent. If pitchers could be taught to reliably locate their pitchers, don't you think they'd do it?

Frank Markotich - Thursday, May 03 2007 @ 09:37 AM EDT (#167112) #

Towers now sports a 5.33 ERA with 5 HR allowed in 25 innings; this on the heels of an 8.42 ERA last year. Does it matter what his strikeout to walk ratio is?

In 2005 his strikeout to walk ratio was an excellent 112/29. How's his future effectiveness been since then?

Mike Green - Thursday, May 03 2007 @ 10:32 AM EDT (#167118) #
Frank, I think that Robert was referring to the K rate.  Towers' K rate in 2005 was 4.9.  That is a little close to the line, and that is why many people anticipated a decline from him in 2006.  His current K rate of 7.8 (admittedly in only 25 innings) is definitely sustainable. 

Towers did pitch poorly in 2006, but this year his ERA does not reflect how well he has pitched.  It does seem that he is going to have change his pitching pattern when runners are on. 

A - Thursday, May 03 2007 @ 11:02 AM EDT (#167121) #

At this stage, do players really need to receive loot with their awards?  It seems superfluous to me.

Mike, it's not about rewarding the player as much as it is good marketing for the maker of the television. Who wouldn't want everyone to know Roy Halladay uses their product?

Mike Green - Thursday, May 03 2007 @ 11:12 AM EDT (#167122) #
Yeah, I know.  I guess people are so numb that they will buy ABC's products because ABC sponsors the MVP trophy in baseball, basketball or ringette rather than because of any intrinsic merit to the products.  Oh for the days when a cigar company sponsored the out-of-town scoreboard on the radio broadcast!
Rob - Thursday, May 03 2007 @ 12:10 PM EDT (#167125) #
Oh for the days when a cigar company sponsored the out-of-town scoreboard on the radio broadcast!

"We're back in the General Motors broadcast booth here on the FAN Radio Network, brought to you by Rogers, a proud supporter of the Toronto Blue Jays and GM: Drive To Win. While the new pitcher gets ready, let's take a look at the Home Hardware Out-of-Town Scoreboard. At AT&T Park in San Francisco, it appears Barry Bonds has homered in the fourth inning to put the Giants up on the Rockies 2-0 and we remind you, by entering at bluejays.com, you can win either a 2005 Pontiac Pursuit or a 2005 Chevrolet Cobalt whenever a Blue Jay hits a homerun that strikes the car display case in the 100-Level seats in right-center field."
Chuck - Thursday, May 03 2007 @ 12:31 PM EDT (#167126) #
On the topic of advertising, let me heartily endorse Terry O'Reilly (no, not the puglist). His radio show, the Age of Persuasion, runs on CBC-1. Here is his blog.
ChicagoJaysFan - Thursday, May 03 2007 @ 12:39 PM EDT (#167127) #

The stats quoted by Magpie ARE the relevant ones. Particularly strikeout and walk data - the best indicator of future effectiveness bar none.

There is no question but that Towers outperformed Okha and Chacin - and by a comfortable margin. No question.

Obviously I disagree with this, so I'll explain why.  First, a starting pitcher has two purposes: not allow runs and pitch a lot of innings - those two data points are a pitcher's performance.  K/9 and BB/9 are useful to predict future performance, but that's it - they're not an end in and of themselves.  So, using IP and R (minus the Sweeney error run), Towers has been significantly outperformed this year by Ohka and Chacin - so yes, it definitely can be questioned that Ohka and Chacin have outperformed Towers.  The reason one would use K/9 and BB/9 is to determine if that outperformance is likely to continue in the future.

So, as far as predicting Towers future performance, K/9 and BB/9 are good predictors of future performance (as are HR/9 and GB/FB).  However, you now have to ask yourself if you think that 2 starts in which you ignore his HR/9 (a gaudy 3.5 in those 2 starts, which you have to ignore to say that he is effective with his K/9 and BB/9 ratios) are a better indication of his future performance than the rest of his career.  I tend to look at the strikeout performances of those 2 games as statistically insignificant and look at the pitcher's ratios over a longer period of time.
actionjackson - Thursday, May 03 2007 @ 12:41 PM EDT (#167128) #
Ah yes, the El Producto out of town scoreboard, back in the days when there was no stinkin' Alice Fazooli's and Tom Cheek would rave about Trader Vic's for reasons that were baffling to a naive 10 year old growing up in Willowdale. Oh yes, the advertising was still there, it just didn't take every second word of the broadcast or maybe I just didn't hear it in between Dave Stieb's brilliance or Otto Velez' 4 HR double header against the Indians. Just trying to stir up some mojo for tonight's game. Of course these Indians are not the Indians of yore, but you can't blame a guy for trying.

Off topic: Speaking of Mr. Velez, in a program I have from 1977 (Vol.1 No.18), there's a picture of him in the stands pre-game, "chatting with some fans" (page 72) on the road. Unless they have doppelgangers, those two fans are a very young Roberto and Sandy Jr. (as in Alomar). Who knew that 15 years later, one of those "fans" would be an integral part of Back to Back Championships for Otto's team. Cool.

Manhattan Mike - Thursday, May 03 2007 @ 12:49 PM EDT (#167129) #

Curious: what do you guys think that the Jays can get for Towers? Certainly, the Jays would end up eating a large portion of his salary.  I cannot expect that the return will be anything more than a minor-league roster filler so what's really the point of paying for him to play on another team for now, until it's clear that there's a better alternative available from Syracuse?

I've thought this year that Towers biggest problem was his consistency throughout a game. Even in his last appearance as a starter, he looked OK save for that single inning. And let's not forget that it was on these pages that there seemed to be a widely held view that the move of Towers to the 'pen was rather hasty, given his K/BB ratio and given that his ERA (prior to the game yesterday) was relatively attractive.

At what point do you conclude that perhaps the best option is to let him try to stick it in the bullpen for a bit and see whether he can be useful in mop up roles, when the game isn't really on the line any longer (read: when you aren't worried that JT will have that bad inning) and Gibby doesn't want to waste an arm that can be used in a more pressured situation the next night or so. Seems that he can be an effective inning eater when the game is out of reach, which isn't something worth discounting entirely.

Chuck - Thursday, May 03 2007 @ 01:02 PM EDT (#167131) #

Curious: what do you guys think that the Jays can get for Towers? Certainly, the Jays would end up eating a large portion of his salary.

I think the Yankees are so hard up for healthy arms that they'd take Towers and his salary. In these hyper-inflationary times, I'm not sure if Ricciardi wouldn't rather burn the $2M than see the Yankees potentially improve.

robertdudek - Thursday, May 03 2007 @ 03:11 PM EDT (#167138) #
Josh Towers' lifetime K/W ratio is in excess of 3 to 1, among the best in major league history for a starting pitcher.

His increased K rate this year is an indication that he is improving.

robertdudek - Thursday, May 03 2007 @ 03:25 PM EDT (#167139) #
Short term ERA is heavily influenced by fielders and the random bunching of hits (i.e. dumb luck).

The point I'm trying to make is that there are enough statistical differences in the relevant categories between 2007 and 2006 to conclude that whatever was ailing Towers in 2006 is no longer applicable.

There has never been a starting pitcher with Towers' K/W data (I'm talking career stats) who has not been at least reasonably effective. Considering the lack of other viable options at this time, it makes NO SENSE, ON THE MERITS OF IT, for Towers to be bumped from the rotation.

I don't buy Chicago Fan's explanation and I don't buy Magpie's either (Zambrano would have started in AAA as a starter if it were inevitable that he would return to the rotation).

There MUST be something else going on behind the scenes that has to do with personality issues. That is what I have maintained all along.

AWeb - Thursday, May 03 2007 @ 03:43 PM EDT (#167140) #

I agree with the "there must be something else going on" view, but do note that Towers has given up 21 runs in 25.3 innings this year, after 62 in 62 innings last year. Towers ERA is excessively kind to his actual run prevention performance so far this year. And in the end, as someone noted up thread, pitchers get paid to prevent runs from scoring. Lots of Ks and few BBs is almost without fail a great method to accomplish this, but guess what, it has failed Towers so far.

To pull out an obscure name from recent blue jays past, Towers has pitched like Nakamura did in 2004. (Actually, their numbers are almost identical. I swear I thought of him before I checked the numbers.)  Some guys are just way too prone to the longball to stick in the majors...Towers may be one of them.

Mike Green - Thursday, May 03 2007 @ 03:51 PM EDT (#167141) #
Ricciardi did say at the start of the season that Towers would be on a short leash.  If Plan B at the start of the season was Zambrano, it really did not make sense to have him throwing 5 innings a month to prepare.  If personality issues arose between Gibbons/Arnsberg and Towers, that doesn't reflect particularly well  on Gibbons/Arnsberg either.  If it happens once, that's normal, but once a year (Bush/Lilly/Towers) is not a good record.



robertdudek - Thursday, May 03 2007 @ 03:55 PM EDT (#167142) #
So you are comparing a guy who pitched less than 40 innings in the majors with another who has pitched well over 600?

What Towers has done in past years is only marginally relevant. How he has performed this year (which, by the way, was pretty well in his 4 starts, despite the attempt to manipulate less relevant stats by some posters) is slightly more relevant.

The main issue is what is Towers likely to do the rest of the year (and future years). Nothing speaks to this more relevantly than K/W data. Nothing.


China fan - Thursday, May 03 2007 @ 04:00 PM EDT (#167143) #
     I fully agree that Towers has an incredibly impressive K/W ratio.  But surely there are intangibles at work here, which maybe cannot be quantified by statistics?    I don't think it's necessarily a personality issue.   Ricciardi and Gibbons are looking at Towers and Zambrano every day -- not just in the games, but also in their bullpen sessions and side sessions and everything else -- and they could be looking at a whole variety of factors that can't be reduced to numbers.   I know that this is a tough case to make, because it's so difficult to quantify, but isn't it possible -- for example -- that the Jays might believe that Zambrano is cooler under pressure, has a better focus on the mound, has more life on his pitches, is less likely to give up the big inning, etc?   I'm sorry for putting this in such a vague way, but the Jays must have internal assessments that go beyond the raw data.  Even if you look at their performances last night, for example:  Towers had more strikeouts than Zambrano, fewer walks, and fewer pitches in a similar number of innings -- and yet, once again, he gave up a home run in a key situation and surrendered more runs than Zambrano.  The Jays might have a feeling that this is a recurring pattern that Towers just cannot shake.  Maybe they feel that Towers loses his focus at key moments in a game and will be ever thus.  Maybe they feel that Towers has no potential to improve beyond his current performance, whereas Zambrano has a greater likelihood of reverting to his peak years of the past.   I guess I'm willing to give the benefit of the doubt to Ricciardi on this one -- but I'll admit I'm wrong if Zambrano is a bust and Towers does well on another team.
Barry Bonnell - Thursday, May 03 2007 @ 04:12 PM EDT (#167144) #

"We're back in the General Motors broadcast booth here on the FAN Radio Network, brought to you by Rogers, a proud supporter of the Toronto Blue Jays and GM: Drive To Win. While the new pitcher gets ready, let's take a look at the Home Hardware Out-of-Town Scoreboard. At AT&T Park in San Francisco, it appears Barry Bonds has homered in the fourth inning to put the Giants up on the Rockies 2-0 and we remind you, by entering at bluejays.com, you can win either a 2005 Pontiac Pursuit or a 2005 Chevrolet Cobalt whenever a Blue Jay hits a homerun that strikes the car display case in the 100-Level seats in right-center field."

You forgot to tell them to drop by Alice Fazooli's after the game. They have DJs four nights a week dontcha know?

laketrout - Thursday, May 03 2007 @ 04:19 PM EDT (#167145) #
With all the love affair on this site with K/BB rates there's one important thing that "control" pitchers like Towers have to have and that is the stuff to get batters out.  It's all well that you walk very few batters but what good is it if your opponents are slugging .500 off you?  This is a problem all the "control" pitchers similar to Towers (Lieber, Millwood, etc.) have is if their pitches are moving in the strike zone they can be brilliant but as soon as their pitches start to flatten out they get crushed.

This is what sets a pitcher like Halladay apart from that bunch - his pitches have great movement and he also intuitively knows how to mix his pitches to keep hitter off balance.  It seems like batters know what Towers is going to through and what he's throwing is totally hittable. That's a recipe for disaster.
robertdudek - Thursday, May 03 2007 @ 04:26 PM EDT (#167146) #
I guess I'm willing to give the benefit of the doubt to Ricciardi on this one -- but I'll admit I'm wrong if Zambrano is a bust and Towers does well on another team.

It will be too late if that happens. The idea is to KEEP the good pitchers.

I'm not saying Towers is going to be an all-star, but he has value as a starting pitcher. It is very important for a team not swimming in pitching talent to nurture guys like Towers. That means supporting them with a good defense and not yanking them out when they have a couple of bad games.

Clearly Ricciardi and company don't like Towers. That's obvious. But it isn't based on his performance on the mound. Towers has only had one full season in the majors and he pitched very well. This season, he seems to be attacking the strikezone better, as evidenced by his increased K rate. He has shown improvement, and his gem start this season shows the potential for excellence is there (Okha has not pitched well in ANY start, despite what numbers may say).
It's Towers, rather than Okha or Zambrano, that has the breakthrough potential.

Thus I conclude it is a "personality" issue.

I'm not willing to give any GM the benefit of the doubt - I've seem too many cases of perfectly good pitchers cast aside based on insufficient sample size.

Your argument would make sense if it applied to someone like McGowan, someone who has potential to be better than he's shown. But nothing in Zambrano's history suggests this. He's been around the majors a long time and he keeps getting chances because he has a good arm. But it takes a lot more than a good arm to make a good major league starter (just ask Dewon Brazelton).

Zambrano is the new Tanyan Sturtze. A guy with a decent arm, but without a decent K rate and with poor command.


robertdudek - Thursday, May 03 2007 @ 04:29 PM EDT (#167147) #
Millwood and Lieber. Boy, you sure picked two washout pitchers there, didn't you.




Magpie - Thursday, May 03 2007 @ 04:33 PM EDT (#167148) #
Towers had more strikeouts than Zambrano, fewer walks, and fewer pitches in a similar number of innings -- and yet, once again, he gave up a home run in a key situation and surrendered more runs than Zambrano.

Now, now. That's mainly because when three of Zambrano's baserunners were filling the bases, Towers retired Peralta on a come-backer. And when two of Towers' baserunners were aboard (after one clean single, an infield hit off the bag and a dropped fly ball? What's up with that?), Towers gave Peralta some room service. Reverse those two at bats...

But I'm moving on! I still don't get the Zambrano move. The team didn't and doesn't believe in Towers - obviously true, otherwise they wouldn't have pulled the plug after four starts. Towers only broke camp in the rotation because Thomson got hurt and Zambrano wasn't ready. It's not an unrerasonable judgement for the team to make.

But if you don't believe in Towers, why wasn't there a Plan B?
robertdudek - Thursday, May 03 2007 @ 04:47 PM EDT (#167149) #
Towers only broke camp in the rotation because Thomson got hurt and Zambrano wasn't ready. It's not an unrerasonable judgement for the team to make.

If Zambrano wasn't ready, then why was he languishing in the bullpen instead of stretching out at AAA? That is the fly in the ointment in your theory, Magpie.

dalimon5 - Thursday, May 03 2007 @ 04:52 PM EDT (#167150) #

Gibby has to go. I get more and more frustrated just thinking about his approach in this "must-win" year. For Pete's sake, the players all finally seem to be intent on winning and giving it their all, maxing out so to speak. Even Burnett and others who haven't had their A game. But Gibbons, boy has he ever picked a time to NOT be the manager he is capable of being. HE is the one who seems lost at times. The Towers situation is what it is, but Gibbons certainly could have squeezed the towel to see if there was any water left in it. Instead, he just seems to be making moves sometimes to "set the tone" and "the message"...to manage so to speak, but...the moves are not ones any good manager would make! It breaks my heart, again, seeing managers like Gaston and others out of work.

* I still think McGowan comes through with 12 wins this year, and sets the last three spots with Zambrano and someone other than Ohka at the end of the all star break. Zambrano is a better pitcher than Towers. In 2005, Towers had 13 wins in 33 GS. But I recall a good portion of these wins coming from non AL-East teams. Not that that matters in the end. On the other hand, Zambrano, working with a MUCH MUCH weaker TB team, in comparison, won 12 games in 28 starts in 2003, presumably with less run support. To be sure, can anyone check to see if any of Zambrano's wins came outside of his starts (I think he had 5 bullpen games).

Also, Towers is costing the Jays 2 million dollars more this year, which could explain his removal, if and when the Jays get rid of him.

But, in my opinion, the Josh Towers saga was born in and has ended in mismanagement. Last year with JP's inability to find another spot starter to fill in for Josh, and this year with the mismanagement by Gibbons to more or less set him up for failure. "Our team isn't winning when you pitch," is the apparent reasoning. Come on, why not put blame on the players not making the routine plays behind Josh Gibby? Gibbons more and more is managing like someone tring to manage rather than a big league thinker getting things in motion in order to maximize productivity and team growth.
Kieran - Thursday, May 03 2007 @ 04:59 PM EDT (#167151) #
While Towers does have an impressive 3:1 K:BB ratio, it's his H:IP ratio that is problematic. Not only does he give up a lof of hits, but many of them for extra bases (including the dreaded HR).

I don't think one can look at one or two metrics in isolation. Clearly, Josh has great control but is (relatively) easy to hit for major league batters. Zambrano is likely no better, and the Jays' brass does seem to make a lot of quick, flip-flop decisions based on recent performance. That said, I'm not really going to argue with Towers being pulled from the rotation. I don't have a lot of confidence in him.

This staff has a lot of problems right now. Janssen is starting to look a lot more hittable, Marcum seems to have hit a rough patch, Chacin has been inconsistent (and hurt), Burnett has been both wild AND hittable, and of course Ryan is hurt and Frasor has faltered of late. It's time for someone to step up and perform. I really hope it's McGowan.
laketrout - Thursday, May 03 2007 @ 05:03 PM EDT (#167152) #

It's not a personality issue.   It's a matter of Josh not getting batters out when it counts. It doesn't matter what your K/BB ratio is when this is how you perform in the clutch.

Here's what he's done so far this year:

AB / AVG / OBP / SLG / OPS
None On
61 / .279 / .323 / .393 / .717

Runners On
43 / .326 / .327 / .651 / .978

AWeb - Thursday, May 03 2007 @ 05:04 PM EDT (#167153) #
So you are comparing a guy who pitched less than 40 innings in the majors with another who has pitched well over 600?

Yes I did, but just for a convenient comparison to a guy, from the Jays, who could strike people out, didn't walk many, but still got killed. That's no more (well, OK, a little more for me) a convenient cherry pick than saying: Towers has only had one full season in the majors and he pitched very well. Last year would've been a full year in the majors, except for the historical level of awfulness, the same reason Nakamura didn't get any more innings. I fully acknowledge that Towers is far more likely to pitch well (in the preventing runs sense) in the future than his replacement in the rotation Zambrano. I also am forced to note that despite great peropherals this year, he has failed to do that so far. If you aren't walking batters, and you strike them out a lot, and still fail, they must be hitting the crap out of you when they do make contact. Which is what has happened to Towers (lifetime .490 slugging against, .500 this year) so far.

I'd say removing Towers from the rotation in favour of a Zambrano is lunacy though. Zambrano, when he was healthy years ago, was a Burnett type pitcher (not that good, but similar), with lots of walks, strikeouts, and not many hits. It all basically washed out and he was thoroughly average in performance in the end. The Jays could defienitely use that guy right now. But....Zambrano is not that guy right now. Without his top fastball, it looks so far this year like he's still the bad parts (walks, HBP), without the compensating good parts(Ks, hit prevention). And if Zambrano is ever going to learn to pitch with his new reduced "stuff", which does still have a lot of movement to my eye, he can't do it 8 innings/month. It's obvious to basically everyone here, and basically separate from the Towers issue: why on earth wasn't Zambrano getting more innings in the minors if they saw him as a starter?
laketrout - Thursday, May 03 2007 @ 05:13 PM EDT (#167154) #

Millwood and Lieber. Boy, you sure picked two washout pitchers there, didn't you.

Tower's is almost a mirror pitcher of those two examples in both style and past performance, if you're looking for a metric to predict Josh's future - there it is.

robertdudek - Thursday, May 03 2007 @ 05:15 PM EDT (#167155) #
Here's what he's done so far this year:

AB / AVG / OBP / SLG / OPS
None On
61 / .279 / .323 / .393 / .717

Runners On
43 / .326 / .327 / .651 / .978



SAMPLE SIZE


robertdudek - Thursday, May 03 2007 @ 05:16 PM EDT (#167156) #
I would LOVE to have either of those two guys in the Jays rotation.
Manhattan Mike - Thursday, May 03 2007 @ 05:20 PM EDT (#167157) #

Chuck,

I don't think that there's any chance that the Yankees and the Jays get a deal done for Towers. I think that there's even less of a chance that Cashman calls JP and asked to trade Towers, offering to pay his salary as well.

I would call that wishful thinking, though I'm not sure that it's wishful on my part since I'd rather have Towers in the Jays bullpen than in the Yankee rotation.

If the Yankees are looking to add pitching, its only going to come if they fail to land Clemens. I believe that Cashman's view is that a healthy rotation consisting of Clemens-Mussina-Pettitte-Wang-Igawa is better than one that has Towers in it. Further, I believe that Cashman's view is that the Yankees are much better off giving someone from AAA the ball or trading for someone with a little bit more cache than a guy who went 2-10 last year than paying for Towers to do the same.

Richard Griffin wrote in yesterday's Star (before the game) that it's time for Towers and the Jays to part ways. But I ask again: what is meant by this? Who is willing to pay his salary? And, if so, why is it beneficial for the Jays to pay for Towers to pitch for another team this year? As best as I can tell, right now it isn't in the team's best interest and I have a hunch that JP agrees with me.

Magpie - Thursday, May 03 2007 @ 05:54 PM EDT (#167159) #
Zambrano, working with a MUCH MUCH weaker TB team, in comparison, won 12 games in 28 starts in 2003, presumably with less run support.

Well, not exactly. 2003 was indeed Zambrano's best season. As a starter, he went 11-10, 4.24. He had 14 quality starts in his 28 outings, including five in a row in June, soon after he returned to the rotation. Tampa sent him to the bullpen after he began the year 0-2, 6.91 in five starts. The Devil Rays supported Zambrano with 144 runs (5.14 per game.) His bullpen blew two saves behind him, costing him a couple of wins - the other team's bullpens blew two saves as well, sparing him a couple of losses. He went 7-5 against the AL East (his win in relief came against Baltimore.) The Orioles were the team that cuffed him around the most. He went 3-0 against Boston, and pitched very well indeed in two of those starts. (In his other win against the Red Sox he gave up 8 runs, but won 15-9)

Towers in his best season went 13-12,  3.71 in 33 starts. He ran off 12 quality starts in a row at one point, the longest stretch by any Jays starter since Roger Clemens, and had 21 quality starts in all. He didn't have quite as much run support as Zambrano (162 runs in 33 starts, 4.91 per game) and didn't get a whole lot of help from the bullpen, which had six blown saves behind him. Towers went 5-4 against the AL East (that's where 4 of the blown saves came). Curiously, the team that gave him the most trouble (besides Texas,  his career-long nemesis) was Washington. Go figure.

For what it's worth, Towers' 2005 is significantly better than Zambrano's 2003. How relevant are either of these seasons today - what a 30 year old pitcher did 2 years ago, when he was 28, what a 31 year old pitcher did 4 years ago, when he was 27?

I dunno.

ChicagoJaysFan - Thursday, May 03 2007 @ 05:58 PM EDT (#167160) #
Everything this year is sample size - it's only 4 starts.

Even with that said, Towers problems with runners on are something he has displayed throughout his career:

.308/.341/.523 (.864 OPS) with runners on
.296/.325/.469 (.794 OPS) with the bases empty

Interestingly, he benefits from a lower BABIP when there are runners on versus the bases empty - I would have suspected that people would to have higher BABIP when there are runners on.
laketrout - Thursday, May 03 2007 @ 06:00 PM EDT (#167161) #
OK THEN...

How about his three year average then? Not as dramatic as this year but still not pretty.

AB / AVG / OBP / SLG / OPS
None On 936 / .296 / .326 / .443 / .770

Runners On 643 / .313 / .350 / .516 / .867

Thats still an increase in batter's OPS against him of almost .100
Magpie - Thursday, May 03 2007 @ 06:10 PM EDT (#167163) #
What's unusual about that, ChicagoJaysFan, is where the increase comes from. It's perfectly normal for both OnBase and Slugging to increase with runners on base - it happens league wide - but normally the OnBase portion of the increase is larger than the Slugging portion. Batting averages go up, largely because the presence base-runners force defenders to adjust their positioning.

But Towers is the other way around - the increase comes largely in the slugging component. Weird.
ChicagoJaysFan - Thursday, May 03 2007 @ 06:19 PM EDT (#167164) #
He looks like a different pitcher when there are runners on base.  He's also issued 10% more walks in 30% less innings (32 AB per walk with the bases empty, 19 AB per walk with runners on).
Magpie - Thursday, May 03 2007 @ 06:32 PM EDT (#167165) #
The subject was interesting enough to take a look at the other Jays starters. Halladay behaves most like the overall league averages - in his case, the increased OPS comes mostly from an increased (by 25 points) OnBase percentage. With Burnett, however, it's the slugging percentage that jumps by 40 points. You've covered Towers. And Thomson has the most dramatic split of all - OnBase is up 37 points, and slugging is up 80 points.

The other three guys all get tougher with men on base. Zambrano shaves 11 points from the OnBase and 28 points off the Slugging. Chacin, confirming what we've long known about him, knocks off 28 points from the OnBase and 27 points off slugging. And Ohka? The least effective of these seven with the bases empty, Ohka shaves 32 points off the OnBase and 72 points off the slugging.

ChicagoJaysFan - Thursday, May 03 2007 @ 07:07 PM EDT (#167166) #
Why no Vernon today?  Is he injured?  He had a good game yesterday, so if he is hurt, it didn't affect his performance.
Leigh - Thursday, May 03 2007 @ 07:18 PM EDT (#167167) #
Pretty frustrating thread, eh Robert Dudek?

Part of the beauty of the strike that I am on is that I don't end up in these threads screaming at ERA/Clutch/Wins/Small Sample Slaves who just don't get it and won't listen to reason.

I have not watched a Jays game  (aside from a couple of at bats while flicking through the channels) or commented in the threads here since the strike.  I gather from a cursory perusal of this thread that Accardo is still not the setup man/closer/high-leverage guy, Towers is nowhere near the rotation and the everyday lineup still includes at least two OBP sinkholes.

The feelings of helplessness that I was getting from watching this managerial train-wreck are melting away the deeper I get into my strike.  I am still helpless, of course, but by not watching the games I have relegated the feelings associated with that helplessness to a latent simmer (rather than a blatant boil).

Feel free to join me on the picket line... it's great for the blood pressure.

If only we could lock Gibby in a room with Voros McCracken for an hour or two.

Gerry - Thursday, May 03 2007 @ 07:24 PM EDT (#167168) #
Didn't see all the game last night but didn't the defense play a role in the big inning?  Rios was charged with an error but I think there was another misplay as well.  Towers did hang the curve to Peralta but the defense was the difference between a solo and a three run home run.
Magpie - Thursday, May 03 2007 @ 07:37 PM EDT (#167170) #
I think there was another misplay as well.

Not in that inning (the inning before, Clayton misplayed a ground ball into an infield hit.) The big inning did feature another infield hit, that bounced off the second base before McDonald could play it.
laketrout - Thursday, May 03 2007 @ 08:20 PM EDT (#167171) #
Pretty frustrating thread, eh Robert Dudek?

Well one last kick at the can...   I just got pretty fed up myself of an entire thread claiming Ks to BBs ratio was the be all and end all of pitching stats.
robertdudek - Thursday, May 03 2007 @ 09:05 PM EDT (#167176) #
.308/.341/.523 (.864 OPS) with runners on
.296/.325/.469 (.794 OPS) with the bases empty


Perfectly normal split for a starting pitcher.

1) most pitchers pitch better from the full windup
2) When there are lots of runners on base it generally means the pitcher doesn't have his best stuff.
3) When there are lots of runners on base it generally means the pitcher is facing a better offensive team.

Next.

robertdudek - Thursday, May 03 2007 @ 09:17 PM EDT (#167178) #
Leigh,

I'm still having fun watching Vernon Wells, Alex Rios, Adam Lind, Aaron Hill and Roy Halladay. So no strike for me. My expectations are simply a bit lower than they were a month ago.

This organisation is lost at sea at the moment, I've come to realise that in the last week or so. For the first time since the illustrious tenure of one Gordon Ash, I've come around to the view that JP must be shown the door at the conclusion of this season. Not now - that would just be piling chaos on top of chaos.

But in the cold logic of the off-season, after yet another 85ish win season, it will be clear that we need someone who can finish the job. That is, construct the roster in such a way that there is a contingency plan for every contingency. You know, the way Herzog the Great used to do it.

There is a lot of talent on this team. And several glaring, massive holes.




laketrout - Thursday, May 03 2007 @ 10:04 PM EDT (#167184) #
Robert,

You keep making claims with nothing to back yourself up. ..and continue to shoot down my assertion's when I do back them up.

So... one last time.

Perfectly normal split for a starting pitcher.

Is it? Here is the league average for the above stats for the entire league this year:

AB / AVG / OBP / SLG / OPS
None On         
15427 /  .254  / .320  / .401  / .721
Runners On      
11994  / .260  / .343  / .402  / .745


And all of MLB in 2006:

AB / AVG / OBP / SLG / OPS
None On        
94661  / .264  / .325  / .428  / .753
Runners On     
72680  / .276  / .351  /  .437  / .788

You'll notice that there is not a big jump in slugging percentage for the league. Yes, average and on-base-percent go up as expected but when compared to Josh's stats it shows that he is prone to the "big inning". He doesn't simple give up more hits but more doubles, triples and home runs when runners are on base. 

If you thought this year's example for Josh was too small a sample size you can take a look at his last year's stats too and see that this is a trend for Josh.  Management is pulling the plug on him now, not because of personality reasons, but because he's exhibiting the same patterns as he did last year.

Why is he failing with runners on base?  Maybe he's trying too hard not to lose the next batter by attacking the strike zone and leaving all too hittable pitches in the strike zone.  Is his psyche shattered? You bet it is.
ChicagoJaysFan - Thursday, May 03 2007 @ 11:16 PM EDT (#167190) #
Ok, this is quasi-non Towers related, but I'm curious where you are coming from Robert, so I'd like to just ask you some general questions about this so that we're not going around in circles.  Below I've stated the question and my thoughts on the question. 

First, I assume that you feel that R and IP are the ultimate goals for starting pitchers.  K and BB ratios are useful because they are good indicators of future performance in the R and IP categories, correct?  (If that's the case, then we agree almost completely here - K and BB ratios need to be supplemented by HR ratios).

Second, and this seems to be an area we disagree on, why is the best measurement of past performance not R and IP?  (This is actually one of the main reasons I disagree with a lot of people in Sabrmetrics.  It cannot be said that an actual event that occurred (i.e. a start) would result in a different outcome if repeated because a given model does not explain how that outcome occurred.  That situation means that the model is incomplete or does not work for all situations, it does not say that what occurred should not have occurred. Thus, while our model (K and BB ratios) do not explain the outcome of the start, the outcome itself must be considered in evaluating that start).

I do have a very solid background in statistics, regression, and econometrics and you seem statistically-oriented, so if you want to go in to the technical terms that's fine.

Also, on a note related to Leigh's comment - there has been at least one study (Drinen and Bradbury) that suggest that clutch pitching may exist - as there is a negative correlation between line-up protection and batting performance (i.e. having a Frank Thomas hit behind a Vernon Wells will cause Wells' numbers to get worse - the theory is the pitcher pitches better in response to the upcoming threat). 
brent - Thursday, May 03 2007 @ 11:42 PM EDT (#167191) #

Since Joe Lawrence was occupying second base and Orlando Hudson was in the minors, I have not liked the way J.P. handles prospects. I also hate the way Gibbons will burn through his bullpen pitchers. However, let me say that until I see better replacements for J.P. (who I think is around a league average G.M.), I am still firmly on the bandwagon and support him.

AWeb - Friday, May 04 2007 @ 01:18 AM EDT (#167192) #
Second, and this seems to be an area we disagree on, why is the best measurement of past performance not R and IP?  (This is actually one of the main reasons I disagree with a lot of people in Sabrmetrics.  It cannot be said that an actual event that occurred (i.e. a start) would result in a different outcome if repeated because a given model does not explain how that outcome occurred.  That situation means that the model is incomplete or does not work for all situations, it does not say that what occurred should not have occurred. Thus, while our model (K and BB ratios) do not explain the outcome of the start, the outcome itself must be considered in evaluating that start).

The thing leaking into some of the arguments for Towers is that some seem to be saying that he has pitched well, because his K rates and walk rates are good, and because good rates like those usually indicate success, he must be pitching well. This is a common mistake (I've got some stats background too), and I'll try to break it down here:
1. K rates and BB rates correlate very strongly with runs allowed. Put in Hr rate, and I believe they predict future performance in terms of runs allowed better than the actual runs allowed in the past
2. Towers has good K/BB rates, therefore Towers has pitched well.

This is not true. The correct conclusion is that it appears Towers is a good bet to perform better in the future than he has. K/BB rates are useful at predicting the desired outcome (runs allowed), but you already have the runs allowed for past starts. When you have something that predicts an outcome well, and the actual outcome those things are trying to predict, the actual outcome (runs allowed) is the right thing to use. To use an extreme example, if someone starts a game, strikes out 10, walks 2, and gives up 8 runs in 5 IP, this is not a good start. But it does give an indication that they might perform better next time.

I think what robertdudek is arguing (and if I'm off on this, correct me) is not that Towers has pitched well (he hasn't prevented runs well, and has therefore not pitched well...I don't see how that can be disputed), it's that looking at his stats there is good reason to think he can and will pitch well in the future. Using things like K rates, BB rates, and HR rates to predict future performance is slightly better than using runs allowed in the past (It can only be slightly better since the rate stats tend to correlate very strongly with the runs allowed rate , which of course is the whole point of using them in the first place). The past starts are gone though, the important thing is the future starts. I do think leaving out HR rates for Towers is a major oversight, since that's always been his problem, when he's had one. But even with a poor HR rate, there's good reason to think he'd improve. Especially when compared to Zambrano. I'm of two minds on this. One part of me sees his last two years and cringes. The other part sees 8K/9IP with few walks and thinks he gotta' do better than he has, right? Well, consider other pitchers off to similar starts.

For instance, in another thread, I posted some pitchers who are off to terrible starts, such as Eric Bedard (34 IP, 13 BB, 42 K, 6 HR, 6.09 ERA), and Brett Myers (23.7, 12BB, 32K, 6.46 ERA). Both of these pitchers have performed badly this year. But given their underlying numbers, there is good reason to think they will perform better in the future. I'd say Philadelphia is insane to put Myers in the pen as they have if he's still capable of starting, although part of that is because his last two years have been very good as well. I've come around on this again, actually. I say give Towers more chances. He's not killing the season like he did last year; he hasn't been that terrible so far.
robertdudek - Friday, May 04 2007 @ 02:23 AM EDT (#167193) #
AWeb,

It is not quite as simple as "a pitchers job is to prevent runs".

Suppose a pitcher makes a really good pitch, but the hitter manages to hit the ball out of the park. Has the pitcher performed poorly? No - in this instance, the hitter's performance was much better. It is the RELATIVE performance of batter/pitcher/fielder that determines the outcome of the individual pitch.

In the long run, the pitchers that make the best pitches will have the best results and will be classed as "the best".

But in the short run, park, fielders and opposing batters all shape the raw stats that people use to make judgments.

Yes, of course Towers has had problems with the gopher ball. And yes Towers probably has problems making some of the appropriate adjustments with men on base (this should take care of itself as he gains experience). If he didn't, with his K/W numbers he'd be a solidly above average starter and we wouldn't be having this discussion.

But there are positives there too. And the Jays don't exactly have quality starting pitchers falling from the trees, so why would you abandon a pitcher that has shown flashes of brilliance in his (still) brief major league career.

I play a lot of poker, and the first thing you learn on your way to being a good player is not to be results oriented. You have to train yourself to make the correct decision at every decision point, even if the result is sometimes catastrophic.

In this sense, the pitcher's job is not to prevent runs - that is the macro view. The pitcher's job is TO MAKE GOOD PITCHES, and let the results take care of themselves.


robertdudek - Friday, May 04 2007 @ 02:50 AM EDT (#167196) #
why is the best measurement of past performance not R and IP?

The short answer is that the pitcher is not 100% responsible for R and IP. There are no stats that the pitcher accrues in a vacuum. Therefore, all performance has to be evaluated in context, which means looking at the park, the fielders, opposing batters, weather conditions and (over the short term) dumb luck involved in producing R and IP.

This is the longer answer ...

Over a small sample size, direct observation is the most important evidence. We look at the pitcher, we watch and we observe  the pitches he makes to a given set of hitters. We compare this with our experiential portfolio of pitches we've seen major league pitchers throw. We can grade the pitches as GREAT, GOOD, OKAY, BAD, AWFUL. If we are objective and skilled "pitcher watchers" we ought to come very close to the truth about how a pitcher is performing.

The next step down is to look at walks and strikeouts, because we are removing one layer of outside influence (namely, the fielders). The park also has less effect on K and W than it has on other batting events.

The next step down is homeruns allowed, which are affected greatly by the park. But since HR are rarer events than W or K you are more likely to be deceived by a fluke clustering of HR allowed. This is the kind of stat you have to look at over a period of 3-4 years before it starts to mean anything.

And the step down from that is all the other batting events, influenced as they are by the fielders.

After that, fluke clustering of particular batting events (e.g. the sequence "K,K, single, single, HR, K" leads to 3 runs, but  "HR, single, single, K,K,K" leads to 1 run)  introduces a huge randomizing factor over number of runs allowed in a small number of starts.


 
Mike D - Friday, May 04 2007 @ 07:19 AM EDT (#167197) #
I'm glad this debate has gotten back on track.  I generally agree with ChicagoJaysFan, and the one quibble I have with Robert right now is that observationally, Towers seems to struggle with his mechanics (and, accordingly, his location) from the set position. 

If memory serves, on Wednesday Josh pitched from the stretch even with no runners on base, which to me strongly suggests that he is working out his mechanics from the set position.  I don't think it's appropriate to blow off his struggles with men on base as part of a universal runners-on split trend.
Mike D - Friday, May 04 2007 @ 07:31 AM EDT (#167198) #
(This is actually one of the main reasons I disagree with a lot of people in Sabrmetrics.  It cannot be said that an actual event that occurred (i.e. a start) would result in a different outcome if repeated because a given model does not explain how that outcome occurred.  That situation means that the model is incomplete or does not work for all situations, it does not say that what occurred should not have occurred. Thus, while our model (K and BB ratios) do not explain the outcome of the start, the outcome itself must be considered in evaluating that start).

I've had many debates with Leigh (whom I personally like and respect) over the years, and this parenthetical by ChicagoJaysFan expresses the very point I was usually trying to make -- albeit more concisely than I ever did.
AWeb - Friday, May 04 2007 @ 08:38 AM EDT (#167204) #
The pitcher's job is TO MAKE GOOD PITCHES

OK, fair enough, I can't really dispute that. I think your poker analogy is a good one, although I'd like to think a good pitch turns out well a lot more often that a good bet in poker. That preventing runs isn't entirely on the pitcher cannot be disputed, especially in the short term. But to extend the analogy, if you keep making what you believe are the right decisions/pitches, and failing anyway, perhaps it's because your opponents have a good read on you? Note I'm not claiming Towers has had enough of a chance this year to make that determination, he clearly has not yet. At this point in the season, it's likely that either his K rate is a small sample fluke, or his runs allowed is, or both. But at some point if your evaluation tool (be it Ks, BBs, good pitches) doesn't match up with macro-level results (runs), you have to find out what's wrong with the evaluation tools.

I will point out that things like Ks, HRs, and BBs aren't used statistically because they correlate well with "making good pitches", it's because they correlate with run prevention. Making "good pitches" and still giving up lots of runs would be in the category of moral victory, and we all know who moral victories are for...but anyway, I'm glad we cleared up exactly where we have a slight difference of opinion here.  Which isn't really about Towers at this point (we both want him to keep starting, especially in light of the other options available).

Frank Markotich - Friday, May 04 2007 @ 08:53 AM EDT (#167206) #

I'll chime in one more time on the Towers thing.

He isn't some young pitcher who we are trying to get a handle on. He's 30 years old and has pitched 644 innings in the major leagues. He has a magnificent career strikeout to walk ratio, and a career ERA of 4.90, which is considerably less than magnificent. The ability to avoid walks has been the only thing keeping him hanging on in the major leagues. Yes, he had a good 2005, but history is full of finesse guys who put together a good year.

With this much background, it's tempting to conclude that he is what he is. Now, he has upped his strikeout rate in the early going, which is a positive but might be a fluke. Me, I would give him just a little more rope just to see if he can get it together (without a lot of expectations, but you never know), since the other options aren't exactly compelling. I don't know if there is some personality thing going on; I suspect it's more a case of the spectre of his 2006 performance and the readiness to conclude "here we go again."

Leigh - Friday, May 04 2007 @ 09:10 AM EDT (#167209) #
Lively debate here.

Chicago JaysFan, I agree that some pitchers systematically defy their peripherals - Tom Glavine is a great example - and that since the perhipherals (K, BB, HR, GB/FB/LD) do not correlate perfectly with the results, then the model is missing something.  However, and this is where we disagree (speaking for myself here; don't want to put words in Robert's mouth), I do not believe that filling in the gaps in the model with past outcomes makes it any better.  I believe that filling in the "what we can't figure out" part of the model with generic information equal for all pitchers (i.e. essentially ignoring what we don't know, rather than harmfully guessing) yields better results.

That is:  past outcomes are worse than inert, they are filled with so much noise on the individual pitcher level that they are often deceptive.  Better to take the past processes and ignore past outcomes on an individual level.

I am glad that you (CJF) articulated your position as you did (re the excerpt in Mike D.'s recent comment), because I was beginning to think that the argument here was simply that you were trying to show us which pitcher has provided the best outcomes in the past, and that Robert was trying to show us which pitcher will provide the best outcomes in the future.
Mike Green - Friday, May 04 2007 @ 09:33 AM EDT (#167211) #
Predicting the medium-term future for Zambrano isn't exactly easy either.  Here are his career splits. He has thrown 570 innings as a starter with a 4.35 ERA but butt-ugly component numbers. He has pitched better with runners on, and best with runners in scoring position or multiple runners on.  This might be a fluke, or it might be a case of a pitcher who needs additional pressure to focus.  Zambrano might be the "anti-Towers"?

The point, to my mind, is not whether Towers or Zambrano would theoretically produce better results if given a "level playing field" at the start of the season, but whether the way Towers and Zambrano were pitching, and how much they had pitched since the start of the season, and Zambrano's surgery, it made sense to make a change on May 2.  My answer is that it was obviously a poor decision.

ChicagoJaysFan - Friday, May 04 2007 @ 07:10 PM EDT (#167241) #
Leigh, if I'm reading your post correctly, you are implying that you are biasing your projections.

You seem to be stating the epsilons from the regressions to predict runs/inning based on K and BB ratios are not normally distributed and instead depend on other variables.  In this case, there are three scenarios and your preference is for arguably the statistically least valid scenario.  The best scenario would be to re-design the model so that it creates epsilons that are normally distributed (not possible given our time frame).  Of the remaining two options (ignoring that the model is wrong or trying to determine if this is a case that isn't satisfied by the model), most statisticians will argue that ignoring the errors in the model is the wrong way to go.  In fact, ignoring the errors in the model doesn't actually remove any bias, it will actually cause you to insert them.

Leigh - Friday, May 04 2007 @ 11:37 PM EDT (#167256) #
I fully admit, CJF, that I have no idea what an epsilon is (outside of college fraternity nomenclature).

As a layperson, though, I do have some ideas.

Here is what we know with certainty:  past peripherals (in conjuction with such externalities as defensive support, park factor and bullpen support) are a better indicator of future ERA than is past ERA.

Every season, though, there are pitchers who produce more positive outcomes than their peripherals would suggest.  Over time, most of these outliers regress toward to mainstream correlation between peripherals and ERA.  Some pitchers, however, persist as outliers - the Glavines.

Since we know that the Glavines are rare, and we do not know how the Glavines are beating their peripherals (at least I don't), then isn't the higher percentage play to regress everybody to the pervading correlation between peripherals and ERA?   This way, we miss the Glavines, but we get it right the vast majority of the time.

I'm probably not articulating this well.  More coloquially:  past ERA is useless because it doesn't tell us anything that the peripherals don't (except in the case of the rare Glavines, but why would we base a broad projection system on the weirdos?). 

If Pitcher A comes up for the Jays and walks 20, strikes out 30 and gives up 12 homeruns in 100 innings, with a league-average groundball/flyball ratio.  He has a 3.25 ERA over those 100 innings.  Now, we can either project him as a Glavine or as a Jerome Williams, Russ Ortiz, Sidney Ponson, Jamey Wright, Tony Armas, Horacio Ramirez, etc.  Why wouldn't we project Pitcher A to be a bad pitcher whose ERA will be bad in the future, like 95% [a made-up number, but I suspect that it's close] of his fellow bad-peripheral-types, rather than project him as a Glavine?

Conversely, if a pitcher has excellent peripherals bad has had some poor outcomes thrust upon him over a small sample size (I'm talking about Towers here), it seems like the best percentage play would be to project that his outcomes will eventually fall in line with his peripherals.  If we were to look at, say, the 100 pitchers with the closest peripherals to Josh Towers over the past 10 years, and 95 of them ended up producing excellent ERAs and 5 of them produced poor results, doesn't it make sense to project Josh as one of the 95 rather than one of the 5?

Again, I'm no statistician (and it shows!).

Thanks for the excellent conversation, by the way.
ChicagoJaysFan - Saturday, May 05 2007 @ 12:00 AM EDT (#167260) #
Leigh,

I think we've found a spot where we agree to disagree as we're now at a spot where we both admit mistakes to our process and it all depends on how one wants to make their mistakes.  I admit, I err or the side of using more feel than numbers (however not all feel) - you rely more on the model.  Both reasonable, just different.  At the point we're at right now, the pros/cons seem to be getting too academic (I follow baseball to escape that).

As to what an epsilon is ... a brief explanation (brief to hopefully prevent extreme boredom) ...

Inherently, what most of the pitching statistics attempt to do is create a predictive ERA using some form of regression. Some of the DIPS formulas act like this.  Those DIPS formulas try and predict what the true R/9 would be with a formula something like this:

Expected R/9 = v + w*(K rate) + x*(BB rate) + y*(HR rate) + z*(BABIP)

In such a formula, v, w, x, y, and z are all constants - so theoretically, if repeated enough times, you would expect the Expected R/9 to match the actual R/9.  With such a model, there is implied an inherent randomness which is captured in epsilon (essentially greek for e, so I'll use e to represent). So in a given start you will get something like this,

Actual R/9 = v + w*(K rate) + x*(BB rate) + y*(HR rate) + z*(BABIP) + e

Here, the "e" captures the noise that you referred to above.  Among other things, these models assume that e is normally distributed (i.e. bell curve).  If e is not normally distributed, there start to be a lot of issues with the use of this model.  However, that doesn't mean that the model isn't useful, it just means to be aware that it's not technically correct and thus can't be relied upon that well.

If you're having trouble going to sleep tonight, read that enough times and I'm sure it will cure things.
ChicagoJaysFan - Saturday, May 05 2007 @ 12:37 AM EDT (#167266) #
Actually Leigh, there is one thing that I disagree with strongly in your post, but I think it's more a case of us using different words to say the same thing, so I'll explain what I mean and we'll see if we're on the same wave length.

Past ERA is not useless.  Past ERA (even in its utlra-simplistic and unadjusted format, which I hate) is actually a pretty good predictor of future ERA.  What is wrong with using past ERA is that it has been shown that there are likely other predictors that are more capable of predicting future ERA both individually and in tandem with other predictors.  However, that doesn't mean that those predictors capture everything that is in past ERA.  And the predictors that we use for ERA (BB's, K's, etc.) all inherently have the same problems that past ERA does - except for the fact that they have been shown to be better predictors of future ERA.  What I mean is there is a lot of noise in K rates, BB rates, and HR rates (look how much they fluctuate), so to say that ERA is wrong because of noise while a K rate is good is not really true.  K rates are better because they have been shown to be better predictors of future ERA than past ERA and I believe that there is also less noise in K rates (but am not certain, they also inherently deal with a greater sample size). 

Finally, there is some data that is lost in the deconstruction past ERA that is not captured in all of the peripherals.  That is one of the reasons that people keep playing with the predicted formulas and keep looking for more peripherals (for 2 examples, I believe that there are efforts/attempts to refine the GB/FB/LD statistic to be more indicative of where balls in play land and there hasn't been much investigation into correlations between peripherals for players - those are data that is captured within past ERA but lost when decomposing to current peripherals).
2 May 2007: Here's Your Hat, What's the Hurry? | 107 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.