Batter's Box Interactive Magazine

Truth be told, I'm not all that big on these projection systems. For instance, I buy Baseball Prospectus' book every year and read it cover-to-cover, but I rarely even look at more than a couple dozen of their player projections. For some reason it just doesn't do it for me.

I did, however, read BPro's recent article on their PECOTA projection system and how it compared to, among other systems, Baseball Primer's ZiPS. After reading the BPro article, I looked at some of the ZiPS projections for this year and then thought "hey, this could be my first ever entry at the Batter's Box!"

So here it is...

Your 2004 Toronto Blue Jays:



                          AVG      OBP      SLG
 C   Greg Myers          .278     .347     .460
1B   Carlos Delgado      .288     .421     .561
2B   Orlando Hudson      .291     .353     .430
SS   Chris Woodward      .270     .331     .435
3B   Eric Hinske         .276     .366     .485
LF   Frank Catalanotto   .303     .364     .473
CF   Vernon Wells        .316     .361     .523
RF   Reed Johnson        .301     .359     .438
DH   Josh Phelps         .291     .375     .545

To be blunt, that is one scary-ass lineup, and none of those projections look particularly unlikely to me. I'm not sure where everyone would bat, but the Cat/Hinske/Wells/Delgado/Phelps portion of the lineup is downright nasty. Of course, you all knew the offense was good.

(BTW, I just traded Greg Myers away in my Diamond-Mind keeper league, so I'm going to be very upset if he goes .278/.347/.460 this year)



                          IP      ERA
SP   Roy Halladay        234     3.31
SP   Miguel Batista      176     4.14
SP   Ted Lilly           153     3.94
SP   Pat Hentgen         127     5.39
SP   Josh Towers         169     4.69
RP   Kerry Ligtenberg     59     4.12
RP   Justin Speier        72     4.13
RP   Aquilino Lopez       84     4.07
RP   Terry Adams         133     3.52
RP   V. de los Santos     51     4.94
RP   Jason Kershner      109     5.12
RP   Bruce Chen          123     5.41

Setting aside the fact that Kershner isn't going to throw 109 innings, those 12 guys total up to 1,490 innings with a 4.31 ERA. The 40 or so innings that Kershner won't throw will probably be made up by Hentgen (I would think), who has a very similar projected ERA. Other than that, most of the IP totals look reasonable, with Lilly and Batista likely throwing 20-40 more than their projections, and Terry Adams and Bruce Chen throwing about that many less than their's.

So, here's my big question...

The Fighting Jays had a 4.69 team ERA last season, which ranked ninth in the American League. Let's say their 2004 ERA is exactly what the projected totals suggest, 4.31. Is that 0.38 improvement enough to push them from 86 wins into the low-90s?

In addition to that, does everyone here expect the offense to remain the same (or improve) in 2004? They scored 894 runs last year and if those projections hold up, I would think a similar total is likely. But let's say the offense goes down to like 850 runs (which would have been good for 4th in the AL last year), is the pitching-staff going to improve enough to make up for that and still get them to 90+ wins?

I think it's likely the Red Sox and Yankees will both score 925+ runs in 2004, so I think if the Jays are going to do what many people here (including myself) think, which is surprise some people and seriously compete for a playoff spot, they are going to have to do one of two things:

1) The offense will have to actually be significantly (25+ runs) better than last year
or
2) The pitching-staff is going to have to be better than even the above projections suggest.

I say the chances of #1 happening are fairly slim, simply because of how good the offense was last year. The chances of #2 happening? I'm not so sure...

Posted by Lucas on Sunday, January 18 2004 @ 03:59 AM EST.

Fun with ZiPS | 44 comments | Create New Account

The following comments are owned by whomever posted them. This site is not responsible for what they say.

_Geoff - Sunday, January 18 2004 @ 04:15 PM EST (#15198) #

Trever Miller's 83 innings will be pitched by Terry Adams

Lucas - Sunday, January 18 2004 @ 04:22 PM EST (#15199) #

http://www.aarongleeman.com
Thanks Geoff. I made the switch and updated the team totals (and ditched Durocher altogether).

_Geoff - Sunday, January 18 2004 @ 04:23 PM EST (#15200) #

Other than that, excellent work Aaron!

Wells' slugging percentage jumps out at me as quite low, but that is offset by a number of optimistic offensive projections

I think the numbers beside Lilly and Hentgen's stats could easily switch, the important point being is that one will match optimistic expectations and one will disappoint

I really can't imagine Durocher and Chen pithcing a comibined 175 innings in 2004, but I think you acknowledged that already

To answer the question posed, I don't think the offense will be as good as in 2003. I think that will be offset by what can be expected from our upgraded staff, but to compete for a playoff spot either both Lilly and Hentgen will have to match optimistic projections or, if they fail, we'll need a McGowan/Arnold type to step up in the rotation (much less likely) - If Terry Adams represents a 4.20 ERA as opposed to T-Mil's 5.20, then I think the bullpen will be fine

_Geoff - Sunday, January 18 2004 @ 04:26 PM EST (#15201) #

Obviously my above comments were made prior to the alterations...

_Jabonoso - Sunday, January 18 2004 @ 04:46 PM EST (#15202) #

Chen as spot starter-long man and Towers solid full season seems less likely to me than big improvements in Cat's-Hinke's-woody's batting

_Tassle - Sunday, January 18 2004 @ 04:51 PM EST (#15203) #

I'm just curious, does anyone know how accurate ZiPS projections usually end up being? Has anyone ever done a study on this after the end of a season?

Lucas - Sunday, January 18 2004 @ 05:00 PM EST (#15204) #

http://www.aarongleeman.com
I'm just curious, does anyone know how accurate ZiPS projections usually end up being? Has anyone ever done a study on this after the end of a season?

Check out the link to BPro that I gave at the very beginning of this entry.

_sweat - Sunday, January 18 2004 @ 08:53 PM EST (#15205) #

For the most part, those projections look to be about where I would expect guys, althought I expect a better year from Halliday (last years 3.25era included a pretty awful april); Hentgen, who had a 4.09 era last year; and Kershner, who had a 3.17 era last year. Just my thoughts. Nice to see you postin here Aaron, as I'm sure a lot of Boxites are regular visiters to your site.

Mike Green - Sunday, January 18 2004 @ 09:58 PM EST (#15206) #

Welcome, Aaron. The ZiPS projections would have the Fighting Jays significantly improving team OBP over last year, despite the loss of Shannon Stewart, and even allowing that Kevin Cash will have a significantly lower OBP than Tom Wilson. The projections, if they prove accurate, would result in the Jays scoring a tremendous number of runs.

But, both Jonny and I feel that the offensive projections are on the whole optimistic. Each of the individual players could perform at the level projected, but for all of them to match the projection would be unlikely.

The wild card is the performance of the Syracuse crew- Gabe Gross, Rios, Quiroz, Bush, McGowan and Arnold. While none may start the season in Toronto, it is quite possible (some would say likely) that by mid-season one or more will be contributing significantly to the Fighting Jays as Hudson and Phelps did 2 years ago.

Pepper Moffatt - Sunday, January 18 2004 @ 10:38 PM EST (#15207) #

http://economics.about.com
The ZiPS projections would have the Fighting Jays significantly improving team OBP over last year

It's not that big a leap for the ten players on Gleeman's list:

		03OBP	04OBP	600PA
C	Crash	0.374	0.347	-16.20
1B	Delgado	0.426	0.421	-3.00
2B	Hudson	0.328	0.353	15.00
SS	Woody	0.316	0.331	9.00
3B	Hinske	0.329	0.366	22.20
LF	Cat	0.351	0.364	7.80
CF	Wells	0.359	0.361	1.20
RF	Reed	0.353	0.359	3.60
DH	Phelps	0.358	0.375	10.20
	AVG	0.355	0.364	5.53

The "600PA" column represents how many more times a player would have to get on base in 600 plate appearances to go from his 2003 OBP to his 2004 projected OBP.

Hinske and Hudson are really the only ones that jump out at me. The projection from Hinske makes sense if you assume he'll be healthy this year. I'm not sure why ZiPS is so optimistic about Hudson.

The phelps one isn't unreasonable, given that it's only 10 extra times on base in 600 plate appearances and you'd expect Phelps to improve as he's a young player.

I agree the projections for minor league players do look wildly optimistic, but given that they probably won't get much playing time on the big club in '04, it shouldn't make much of a difference.

Cheers,

Mike

_John Neary - Sunday, January 18 2004 @ 10:58 PM EST (#15208) #

I agree with Mike -- pulling a number out of my a... er, out of my hat, those look like 60th percentile projections to me. None of those projections is implausible by itself, but I wouldn't put even money on those nine guys hitting as well or better than their ZiPS projections in aggregate.

_Kyle S - Monday, January 19 2004 @ 01:11 AM EST (#15209) #

I guess it's impossible to know without seeing the data the various systems use, but are there any guesses as to how ZiPS factors in previous minor league seasons for players with 3 or fewer years of ML experience? Perhaps that could be the source of optimism. It would also be interesting to know how ZiPS and PECOTA differ in that regard.

The most interesting part of the Silver article to me was his comment on PECOTA using weak DIPS assumptions as well as defense adjustments, while he claimed that ZiPS used strong DIPS. This may be true, but I could swear that Pettite's numbers on the Szymborski disk got a lot better once he was traded to the Astro's. Maybe I'm misremembering, and ZiPS uses only park factors, whereas PECOTA employs some sort of DER adjustment for computing a pitchers' BABIP. Is the ZiPS formula public, or is it too complex and time-consuming to make available to us? I haven't been able to find it on the Oracle.

Mike Green - Monday, January 19 2004 @ 08:59 AM EST (#15210) #

Mike M, I'm betting that ZiPs is optimistic on Hudson because it takes into account the MLEs from his performance in Syracuse and Tennessee in 02 and 01, as well as his youth.

As Robert has pointed out elsewhere, the reported MLEs are somewhat high historically, at least for Jay prospects, as the comparison between minor league stats and year later major league stats fail to take into account expected age improvements.

I might be Hudson's biggest fan, but I would be very happy if he gave up switch-hitting and ended up with a .340 OBP and a .410 slug this year.

Finally, nine points of OBP for a team is a very, very big deal. It's the difference between a mediocre offense and a good one, and between a good offense and a great one. Using the ZiPs numbers, the Fighting Jays figure to score about 950 runs, instead of 894. I'm not counting on it.

Pepper Moffatt - Monday, January 19 2004 @ 09:26 AM EST (#15211) #

http://economics.about.com
Finally, nine points of OBP for a team is a very, very big deal.

Statistically, it really isn't. As far as run creation, it doesn't seem to be.

Over the last 3 years, I found 32 teams with identical SLG, and I compared their OBPs and runs scored. One team had identical OBPs as well, so I threw them out and that left us with 31 teams.

Of those 31 teams, I compared their OBP to their run scored, to see how many extra runs a point of OBP leads to.

On average, it looks like one additional point of OBP is worth 1.14 runs. (R/OBP) is actually (R/OBP*1000) Here's the teams:


TEAM	R	OBP	SLG	R/OBP	
NYY	897	0.354	0.455	0.6002002	
TEX	843	0.338	0.455	3.3752003	
TEX	826	0.33	0.454	2.5002001	
CHW	798	0.334	0.451	3.7692001	
SEA	927	0.36	0.445	4.6252001	
ARI	818	0.341	0.442	-11.6672001	
BOS	772	0.334	0.439	10.1822001	
MIN	771	0.337	0.433	20.0002003	
MIN	801	0.341	0.431	-0.8002001	
CHC	777	0.336	0.43	-4.0002001	
TOR	767	0.325	0.43	0.9092001	
LAD	758	0.323	0.425	1.9332002	
COL	778	0.337	0.423	4.5562001	
FLA	742	0.326	0.423	3.2732003	
MIL	714	0.329	0.419	5.5002002	
SEA	814	0.35	0.419	4.7622001	
CIN	735	0.324	0.419	3.0382002	
HOU	749	0.338	0.417	4.0002003	
OAK	768	0.327	0.417	-1.7272003	
ANA	736	0.33	0.413	3.3332001	
ATL	729	0.324	0.412	-3.3332002	
LAD	713	0.32	0.409	-0.4552001	
KCR	729	0.318	0.409	-8.0002001	
DET	724	0.32	0.409	-2.5002001	
ANA	691	0.327	0.405	-13.0002002	
BAL	667	0.309	0.403	1.1432003	
CLE	699	0.316	0.401	1.2002002	
NYM	690	0.322	0.395	-1.0002002	
TBD	673	0.314	0.39	-7.6672001	
TBD	672	0.32	0.388	0.4622002	
PIT	641	0.319	0.381	10.500	
AVERAGE				1.146

So the extra 9 points of OBP should be worth 10 or 11 extra runs, not 56. Mind you, that assumes that team SLG will stay exactly the same.

I'm sure if I ran threw a longer sample, I'd find a number different than 1.146, but I promised myself I'd spend less than 10 minutes on this. As it took me 6, I succeeded. :)

Cheers,

Mike

Mike Green - Monday, January 19 2004 @ 10:12 AM EST (#15212) #

Mike M, I'm trying to follow your chart, but not succeeding. I catch the first set of matched teams, but then the second set doesn't seem to match. Maybe I'm missing something. Must have another coffee...

Anyways, I'm quite sure that over a larger sample, 1 point of team OBP is worth significantly more than 1.1 runs. The average team has over 6000 PAs in a year so 1 point of OBP is over 6 baserunners, which is worth about 1.5 runs. Still, you're right that the OBP difference between the 03 actual performance and the 04 ZiPs projection doesn't account for anything like 56 runs.

ZiPs does project a 20 point increase in slug for the Fighting Jays, which would put them overall within spitting distance of Boston's OBP and Slug for last year. That said, I won't be laying 50 bucks on it.

Pepper Moffatt - Monday, January 19 2004 @ 10:20 AM EST (#15213) #

http://economics.about.com
Mike M, I'm trying to follow your chart, but not succeeding. I catch the first set of matched teams, but then the second set doesn't seem to match. Maybe I'm missing something. Must have another coffee...

The chart seemed to cut off the matches and is only displaying the first team. I'll have to try it again later. The final column on the LHS is really the only important one. How much one extra point of OBP is worth. There's a lot of variance in there, so quite often the figure is negative.

Anyways, I'm quite sure that over a larger sample, 1 point of team OBP is worth significantly more than 1.1 runs. The average team has over 6000 PAs in a year so 1 point of OBP is over 6 baserunners, which is worth about 1.5 runs.

You're violating ceteris paribus.

The whole point is that we want to know what happens when slugging doesn't change. But in your case, slugging *would* change, because you're adding an extra 6 base runners, so the numerator would get smaller.

In your case you have increased *both* OBP and SLG by 1 point.

So to offset that, you'd have to turn some homers into doubles and doubles into singles to get SLG not to change. Once you do that, it's going to be less than 1.5.

Cheers,

Mike

Mike Green - Monday, January 19 2004 @ 11:19 AM EST (#15214) #

Mike M, I'm not following this either. Maybe I need a double espresso. Slugging is total bases/AB. If a team's OBP increases by 1 point because it has 6 more walks in 6000 ABs over the previous year, with no change in the number of singles, doubles, triples and homers, slugging should be unaffected.

Pepper Moffatt - Monday, January 19 2004 @ 11:32 AM EST (#15215) #

http://economics.about.com
Mike M, I'm not following this either. Maybe I need a double espresso. Slugging is total bases/AB. If a team's OBP increases by 1 point because it has 6 more walks in 6000 ABs over the previous year, with no change in the number of singles, doubles, triples and homers, slugging should be unaffected.

The number of AB has to go down by 6, as walks are not included in AB.

Remember, it's TB/AB, not TB/PA.

Cheers,

Mike

Pepper Moffatt - Monday, January 19 2004 @ 11:35 AM EST (#15216) #

http://economics.about.com
Ahh.. I see your confusion. In message 16 I wrote:

"so the numerator would get smaller"

That should clearly be "so the denominator would get smaller".

I think I'm the one that needs the coffee. :)

Cheers,

Mike

Coach - Monday, January 19 2004 @ 12:03 PM EST (#15217) #

Thanks, Aaron. Please don't limit yourself to Jays-related ideas; drop by any time.

The offense will have to actually be significantly (25+ runs) better than last year

I can't see it. Kevin Cash isn't as horrible at the plate as he looked last fall, but he won't come close to matching Tom Wilson's 289/356/462 first half. I don't think Myers will revert to his pull-happy ways, so I don't anticipate a huge decline, just a slight step back from his career year. Likewise, those are very high standards for Wells and Delgado just to maintain, so it's hard to project improvements. My most optimistic scenario is, with good health all around and Hinske making up most of the difference, they match last year's output. As Mike Green says, the "wild card" could be Rios and Quiroz replacing Sparky and Cash.

The pitching-staff is going to have to be better than even the above projections suggest.

I'm interested in ZiPS, DMB, PECOTA and other measures, but I ignore them when they contradict my guesstimates. Just 127 IP for Pat Hentgen with a 5.39 ERA? It doesn't compute, if you saw him pitch in August and September. I'm expecting 200 innings, of much better quality than that. He didn't miss a turn in a 6-3, 3.10 ERA second half while facing mostly playoff-bound teams. His control was great, his stuff was surprising. Given how so many others have come back from that surgery over similar time frames, it's possible that Hentgen is the soundest he's been since 1998, when he may have begun to pay the physical toll for those back-to-back 265 IP campaigns. It's always possible that he pulls a groin and doesn't tell the training staff, but I think we suffered through enough of that last year.

Among all those things that have to go the Jays' way to contend, the health of the rotation may be most significant. The above numbers suggest that Batista and Lilly will each take a turn on the DL, so Josh Towers gets more work than I anticipate and Bruce Chen is the only available option. It's also possible that Batista will be a workhorse and Lilly avoids injury, so that's another way these projections could be exceeded. I admit that lefty relief could turn into an issue, but it's hardly as bleak as predicted. In other words, I'm pretty sure that ZiPS has underestimated the improvement to the Toronto staff, so there's hope.

_S.K. - Monday, January 19 2004 @ 12:37 PM EST (#15218) #

Good to see you posting here, Aaron.
I think that seeing the Jays offense match last year's 894 is not unreasonable, but asking them to better it is a little farfetched. That being said, I can see the pitching staff bettering that projected 4.31 mark - I strongly doubt that Lilly, Batista, and Hentgen will total only 456 IP (which, at 6 innings a start, suggests only 76 starts total for the three of them). I don't really see any of them as a huge injury risk so unless someone totally breaks down, 85 starts seems more realistic (and 90 doesn't seem completely out of leftfield). Once you add in that extra 50-80 IP from the avg-or-better SPs, a domino effect down the pitching staff removes some poor innings from the ledger and the possibility for a 4.10 ERA is plain.

Wow, that is nearly unreadable... alright, I hereby give everyone permission to skip this entry if you see fit. I need some lunch.

_John Neary - Monday, January 19 2004 @ 12:48 PM EST (#15219) #

Mike Moffatt,

I'm not sure why the six walks would come at the expense of at-bats. The team still makes the same number of outs, so I think you're looking at six extra plate appearances, each of which ends in a walk. Your OBP is then (old number of times on base + 6)/6006, which is more or less equal to (old OBP + .006). BA and SLG are unchanged. I would assume that this is what Mike Green meant.

But I haven't had my coffee yet either ;)

John

Pepper Moffatt - Monday, January 19 2004 @ 01:01 PM EST (#15220) #

http://economics.about.com
I'm not sure why the six walks would come at the expense of at-bats. The team still makes the same number of outs, so I think you're looking at six extra plate appearances, each of which ends in a walk. Your OBP is then (old number of times on base + 6)/6006, which is more or less equal to (old OBP + .006). BA and SLG are unchanged.

That's fair, but we were assuming that each player would get 600PA (see message 10). If we add those 6 PAs, we're dropping that assumption. I'm fine with that (the assumption is unnecessarily rigid), but if you're starting to drop assumptions, it'd be wise to announce it.

I agree using outs makes more sense then using PAs. I also think we'd need to weight players at the top of the order higher those lower. This is just back of the envelope type stuff because I already had Excel open anyway. :)

Cheers,

Mike

Mike Green - Monday, January 19 2004 @ 01:58 PM EST (#15221) #

Coach, I completely agree about Hentgen. I do think that among Chen, Bush, Arnold and McGowan, there is a pitcher who will provide the necessary support when the seemingly inevitable injury occurs among the starters in mid-season. The Jays are in much better shape in this respect than last year.

_Spicol - Monday, January 19 2004 @ 02:23 PM EST (#15222) #

Moffatt, I ran your study for 2000-2002 (except I used Runs/Game instead of Runs), found 50 matches and ended up concluding that each extra point of team OBP over a season is worth 2.93 runs on average. An improvement of 9 points of OBP is about 26 runs.

There's a LOT of noise in this. Obviously, there are other influences on runs scored besides OBP, even if the SLG of the teams in question are comparable. So, I wouldn't use this as the impetus to make any bets. But if the question is whether or not the Jays can score 25+ more runs next season, then ZiPS is saying that they probably can (somewhere between or in and around Moffatt's 10 runs and my 26 anyway).

On the pitching side of the ledger, ZiPS is forecasting a 0.38 improvement in Team ERA, or about 60 runs. Adding in the offense, that's an overall run differential improvement of somwehere in the neighbourhood of 70 to 90 runs, or 7 to 9 wins.

That would put the Jays at about 94-96 wins next year. My gut says that's a little high. There will be injuries and depth is a concern on this team. My official January forecast is 91 wins, subject to change as we approach the spring, of course.

Pepper Moffatt - Monday, January 19 2004 @ 02:29 PM EST (#15223) #

http://economics.about.com
Moffatt, I ran your study for 2000-2002 (except I used Runs/Game instead of Runs), found 50 matches and ended up concluding that each extra point of team OBP over a season is worth 2.93 runs on average. An improvement of 9 points of OBP is about 26 runs.

There's a LOT of noise in this.

Agreed. The variance is spectacularly high. I'd want a sample size of atleast 1000 before I felt at all comfortable with the results. A few million would be ideal, but I admit I'm a law-of-large-numbers kind of guy. :)

Three runs per point of OBP sounds reasonable. 1.2 is probably a bit low. Six seems too high.

RE: Translating into wins. The best approach is probably Robert's where you run a bunch of league simulations. His first edition had the Jays 91-71. IIRC he had injuries "off". If injuries were set on, I'd imagine the Jays would slip a bit, because they don't have the depth (or the budget) of some other teams. That being said, anywhere from 78 to 95 wins or so wouldn't surprise me.

Cheers,

Mike

_Nigel - Monday, January 19 2004 @ 03:01 PM EST (#15224) #

I have trouble seeing the offense being as good as it was last year. Just looking at Aaron's list of predictions I cannot see that coming to pass. Look at how that table is predicting the players relative to last years numbers:

Myers - moderate decline
Delgado - holding steady
Hudson - significant improvement
Woodward - significant improvement
Hinske - significant improvement
F-Cat - holding steady
Wells - holding steady
Johnson - moderate improvement
Phelps - significant improvement

Sorry, but I do not see all of that happening. As people have said, each individually could well happen but all together this is a "best case scenario" where everything breaks your way.

I think that the offense will be lucky to produce 25 runs less than last year. I would suggest the following is more realistic:

Myers - significant decline (I just do not see another .875 OPS in his future - a total of around .700 OPS - which he posted after the all-star break - is more realistic)
Delgado - holding steady
Hudson - moderate improvement
Woodward - holding steady
Hinske - significant improvement
F-Cat - holding steady
Wells - holding steady / moderate decline
Johnson - moderate decline (I just do not see a full year of nearly .800 OPS from him)
Phelps - moderare improvement

Add to this that Cash will replace Wilson's output, Wells and Delgado had terrific seasons and will be hard pressed to match last years and Toronto's 4th outfielder is unlikely to be as good as Keilty was and you have room for a decline - maybe not dramatic but still noticeable. Until the corner outfielders are a true strength (i.e until Reed Johnson is their 4th outfielder) the Jays offense will be a notch or two below that of the Yankees and particularly Boston.

_Jonny German - Monday, January 19 2004 @ 03:07 PM EST (#15225) #

IIRC he had injuries "off". If injuries were set on, I'd imagine the Jays would slip a bit, because they don't have the depth (or the budget) of some other teams.

I'm inclined to think "injuries off" would hurt the Jays more than help them, simply because the Red Sox and especially the Yankees have more injury risks and old guys. Sure those two teams can throw money at patching holes, but that's over-rated IMHO.

Yankee patches, 2003
Aaron Boone
Armando Benitez
Ruben Sierra
Gabe White

Not exactly a collection of ringers. Boston did a much better job of this than the Yankees in 2003, mostly because they had more chips to play, the Yankees having already played this game for several consecutive years. Next year, both have fewer chips. Money only takes you so far - Boone cost the Yankees their best pitching prospect.

I prefer Toronto's type of depth for July reinforcements, the Fantastic Fivesome - Gross, Bush, Arnold, Rios, Quiroz. Q-Barg?

_Spicol - Monday, January 19 2004 @ 04:21 PM EST (#15226) #

Toronto's 4th outfielder is unlikely to be as good as Keilty was

You mean the Kielty that hit 233/342/376 for Toronto? That Kielty? I like BK but he didn't play well for the Jays. It shouldn't be hard to replace him.

Nigel, in your suggestions, you've got 3 guys improving, 2 guys declining, 1 guy (Wells, in the prime of his career) who you're not sure about and the rest holding steady. How does that add up to an overall offensive decline?

_Nigel - Monday, January 19 2004 @ 04:39 PM EST (#15227) #

Spicol, actually I'm pretty sure that that line from Keilty will not be matched by Werth or some other 4th outfielder picked up off the scrap heap at spring training time. I agree that Keilty did not play well while in TO but I think the Jays will have difficulty replacing his .720 OPS (particularly given his strong OBP) from their scrap heap of 4th outfieder candidates (sad but true).

As for why I think the offense will be behind last year. I think cather production (as a whole) will fall of a cliff relative to last year. I also expect a decline from both Johnson and the 4th outfielder position. I think Gomez will not put up the same numbers that Bordick put up (and hence collectively they will get less from the SS position). I think it will be difficult for Delagado and Wells to duplicate last season's efforts but I think they'll come close. I think this will be offset by a season from Hinske that looks more like '02 than '03 and some improvement from Hudson and Phelps. In sum, I see an offense that will have trouble matching last year.

Also keep in mind that the Jays offense last year was pretty healthy. An injury to one of the producing regulars and there will be a significant drop off to the replacement for all positions other than maybe catcher and third outfielder.

_Johnny Mack - Monday, January 19 2004 @ 05:20 PM EST (#15228) #

Here are some raw numbers courtesy of http://www.baseball-statistics.com/ 's online baseball database:

From 1903 to 2002 there were 3,276,392 PA by 5007 batters who had at least 600 PA (an average of approximately 654 PA per batter). Their total runs were 443,593 (about 89 runs each). Their average OBP and SLG were .361 and .440. BA was .289.

------

By the way, if you multiply .361 by 1.97, and then add BA and SLG and divide by 10, you come up with .144007; that's ((OBP*1.97)+BA+SLG)/10; which looks suspiciously close to 443,593/3,276.392 = 0.135391(rounded off to 6 decimal places). You could round both of those off to 14 percent, but I won't yet.

Anyway, that second number gives us the percentage of plate appearances in which those 5007 batters scored a run. Looked at another way, 100/13.5391 tells us those batters scored a run for every 7.3860 times they came to the plate. Now 100/14.4007 tells us that the above formula predicts those batters would score a run for every 6.9441 times they came to the plate. What's interesting to me is that if you divide those 654 PA per batter you get these amounts of runs: 654/7.3860 = 88.5459 (89 runs); 654/6.9441= 94.1807 (94 runs). That's a difference of 5 runs over 654 PA.

If you do round both .144007 and .135391 off to 14 percent (a rate of 7.1429) you end up with 91.56 runs in 654 PA.

I think *1.97 is too high. *1.85 gives you 1.39685 for a rate of 7.1950, or 91.3535 runs in 654 PA. If you run 1.79 in the formula you get a prediction of approx 90 runs per player. So does 1.78. *1.73 looks like the best match for an OBP of .361 as it predicts 88.5208 runs (100/13.5353 = a 7.3881 scoring rate). ((.361*1.73)+ .289 + .440)/10 = .135353.

_Spicol - Monday, January 19 2004 @ 07:08 PM EST (#15229) #

Johnny...what, exactly, are you trying to say?

_Johnny Mack - Monday, January 19 2004 @ 10:10 PM EST (#15230) #

Ummm that is a mess isn't it, Spicol? I'm saying that the formula ((OBP*X)+BA+SLG)/10 seems to bear a fairly close resemblance to R/PA. Replacing X with 1.8 or 1.85 generally gets close for an individual's run scoring rate per 100 PA. It seems as if 1.73 gets you even closer to the rate at which runs are scored when you're looking at a large number of PA such as the 3,276,392 I mentioned in my previous post.

It could be as good a measure of individual perfomance as OPS; perhaps better if it can be refined -- which is what I'm trying to figure out how to do (team, batting order position, player speed would all be factors). If it can't be refined, I'll forget about it and stick with OPS.

_Jonny German - Monday, January 19 2004 @ 10:55 PM EST (#15231) #

http://www.baseballprimer.com/articles/authors/tangotigre.shtml
Johnny, COMN for a couple articles by Tangotiger (OPS Begone! Parts 1 and 2) in which he discusses the shortcomings of OPS. He suggests not combining OBP and SLG at all, but concludes that if that's what you want to do you it should be done something like xOBP + SLG, where x is between 1.7 and 2.0.

_Johnny Mack - Monday, January 19 2004 @ 11:12 PM EST (#15232) #

Thanks, Jonny. Those articles are actually where I started from (as a result of Gleeman's GPA). I did the xOBP + SLG and got curious about whether it bore any relation to what percentage of PA resulted in runs for players. It didn't look anywhere close, so I added in BA to see what that looked like and noticed once I did so that simply moving the decimal place (/10) made the two numbers look very similar.

_perlhack - Tuesday, January 20 2004 @ 09:22 AM EST (#15233) #

I think that the Jays offense will certainly see improvements in some areas, and decreases in others.

Cash replacing Wilson is a downgrade; but let's recall that Frankencatcher also had many DH appearances, which will now go to Phelps. That should be an upgrade. Overall, the DH/C combo should even out. That is, it's not just the OBP and SLG that are important, but also the distribution of PAs among the position and bench players.

[Note, last year Myers/Wilson/Cash/Huckaby had 700ABs and 770PAs between them; for Myers/Cash, I think this will drop to 600ABs and 680PAs or so, with the bulk of the difference going to Phelps]

I expect Hinske and Hudson to improve (if only marginally), Johnson and Catalanotto to decline somewhat, and Delgado and Wells to hold steady (or perform similarly to last year). I have no idea what Woody will do, but I assume it will be better than last year. Overall, the SS position will see a decline, though.

That's the way I see it anyway.

_tangotiger - Tuesday, January 20 2004 @ 10:02 AM EST (#15234) #

An extra 9 points in OBA (with a zero change in SLG) is worth an extra 26 runs per 162 GP.

Essentially, what you are doing is freezing the number of AB, H, HR, SO, etc, and simply adding more walks. To get the OBA to go up by 9 points, you add in an extra 84 walks. A quick LWTS check would give you 84 * .33 = +28 runs.

Plugging it into a more robust BaseRuns will give you +26 runs.

Mike Green - Tuesday, January 20 2004 @ 11:09 AM EST (#15235) #

Well, there you have it, Aaron. If the ZiPs projections are accurate, the Fighting Jays' offence should be produce roughly the same number of runs as the Yankees and Red Sox.

Pepper Moffatt - Tuesday, January 20 2004 @ 11:46 AM EST (#15236) #

http://economics.about.com
Well, there you have it, Aaron. If the ZiPs projections are accurate, the Fighting Jays' offence should be produce roughly the same number of runs as the Yankees and Red Sox.

How do you figure? Robert's simulations have the Jays at 861 runs, the Yankees at 882, and the Red Sox at 911. Keeping in mind that the simulation seems to reduce the overall amount of scoring, those look pretty good to me.

Cheers,

Mike

Mike Green - Tuesday, January 20 2004 @ 01:50 PM EST (#15237) #

I used Aaron's estimate of 925 runs scored for New York and Boston. I then took the 894 runs the Fighting Jays scored last year and made the adjustments that Tango calculated for the OBP improvement projected by ZiPs.

Robert's simulations are over only 15 seasons, and for some reasons the sim depresses scoring a little. The sim interestingly has the Fighting Jays allowing fewer runs than the Yankees.

As I've said before, like John Neary and Jonny German, I'm not convinced that in aggregate the Jay offence will in fact match the ZiPs projection.

_Karen - Tuesday, January 20 2004 @ 05:47 PM EST (#15238) #

Why are the Jays being constantly referred to as the "Fighting Jays" on this site? Why? What does it mean?

Mike Green - Tuesday, January 20 2004 @ 06:09 PM EST (#15239) #

Karen, somebody gave them that name when the new logo with the apparently aggressive bird appeared. No one's pushing for Tie Domi as closer yet, though...

Coach - Tuesday, January 20 2004 @ 06:24 PM EST (#15240) #

Why are the Jays being constantly referred to as the "Fighting Jays" on this site?

Mike Green's explanation is creative, but inaccurate. The name comes from J.P. Ricciardi, according to Moneyball author Michael Lewis, who, in town on a promotional tour, was telling the TSN broadcast crew about the infamous Toronto chapter that didn't survive the final edit. Jordan transcribed the highlights of the interview in a game thread and "Fighting Jays" stuck.

_Johnny Mack - Monday, February 09 2004 @ 12:52 PM EST (#15241) #

Re: my previous posts in this thread (#'s 31 and 33) are a great illustration of Richard Lederer's latest post (COMN): "There have also been some newfangled attempts to use a multiplier for the on-base average before adding this adjusted number to the slugging percentage.

Why complicate a formula that works just fine as is? I don't see the need to create additional methodologies unless they prove to be sufficiently more accurate to make up for their additional complexity. Wouldn't it be much easier for all of us to hold more intelligent discussions if we adopted batting average, on-base average, and slugging average as the three main rate stats when evaluating or comparing players? If we could agree on those three metrics, then wouldn't it also make sense to use PRO or OPS as a "quick and dirty" solo stat?" (italics mine)

Please consider me appropriately chastened and humbled [as if my incoherent ramblings hadn't already done that ;)]. From now on, in public, I'll try to stick to what I know instead of inflicting my nearly flat learning curve on others.

Fun with ZiPS | 44 comments | Create New Account

The following comments are owned by whomever posted them. This site is not responsible for what they say.