Baseball for the Thinking Fan

Login | Register | Feedback

You are here > Home > Primate Studies > Discussion
Primate Studies
— Where BTF's Members Investigate the Grand Old Game

Friday, March 08, 2002

Rating the Defenses: AL West

Robert Dudek analyzes the Defenses a division at a time.


Fielding Run Average (FRA) is something that I developed last year, to answer a specific question. Separating out all the fielding-independent elements of defense (strikeouts, walks allowed, hit batsmen and home runs allowed) is there a way to rate the performance of pitchers/fielders in everything else?

That’s what FRA aims to do. For a detailed explanation of how it works, please read my article Measuring Team Defense.

To evaluate team defense, I used each teams’ raw FRA, adjusted for park and averaged together 2000 and 2001. Keep in mind that there is evidence that particular pitchers can influence FRA, so it can’t be considered a pure measure of fielding prowess. However, I am confident that a large majority of a team’s FRA can be safely attributed to the fielders.

A typical FRA is in the neighborhood of a run better than ERA. In 2001 the AL average was 3.596; in 2000 it was 3.754.

Anaheim Angels


2000/2001 FRA data (AL rank)

20003.025 (1).9663.142 (2)
20013.100 (3)1.0243.027 (4)

2000/2001 ranking: 1st in the AL 

The Angels defense has been excellent the last few years. Adam Kennedy has great range at second base (last year he had an incredible Zone Rating of .888 versus a major league average of .824 for 2B). The outfield covers a lot of ground, though perhaps the optimum arrangement is for Anderson to play center and Erstad to play left as their team’s defensive performance relative to league was a bit better in 2000 than in 2001 (I don’t think that was Mo Vaughn’s doing).

Kevin Appier should have a fine season with a good defense supporting him - the Mets’ 2001 defense was below average and the Angels of 2002 figure to be one of the top defenses in the league once again (even if they make Brad Fullmer their regular 1B).

Seattle Mariners


2000/2001 FRA data (AL rank)

20003.372 (2).8933.778 (6)
20012.624 (1).9352.806 (1)

2000/2001 ranking: 3rd in the AL 

Safeco, as seen from its FRA Park Factors, is a great place to pitch and the Mariners have put up great defensive numbers the last 2 years. John Olerud, Mike Cameron and David Bell are all very good defensive players. Last year, the Mariners defense soared to new heights with the addition of Ichiro and Bret Boone. Even adjusting for park, they ranked #1 in the AL. 

With Bell’s departure I don’t expect them to play quite so well, although Cirillo is a good defender in his own right. Ruben Sierra, the other major newcomer on D, is not known to be a quality glove. Overall, I expect the Mariners’ defense to stay in the top 3, which will give James Baldwin’s numbers a boost.

Oakland Athletics


2000/2001 FRA data (AL rank)

20003.724 (7).9423.954 (10)
20013.083 (2).9273.326 (6)

2000/2001 ranking: 7th in the AL 

The addition of Damon and Dye improved the outfield defense by leaps and bounds. I’m sure that the improving numbers of the Oakland “young gun” starting rotation was aided by the defensive improvement.

T Long is apparently moving back to center, which I think will cost the A’s some. The infield is not bad, but it remains to be seen if Carlos Pena et al will be better, or worse than Big G defensively. I just hope that David Justice doesn’t get a lot of playing time in left field.

Like Safeco, the Coliseum helps the defense get outs (mostly because of the large foul territory area).

Texas Rangers


2000/2001 FRA data (AL rank)

20004.281 (14).9894.328 (13)
20014.206 (13).9484.438 (13)

2000/2001 ranking: 14th in the AL 

The Rangers pitchers of 2000 and 2001 suffered enough and to show what a kind-hearted soul John Hart is, he’s brought in a whole new starting rotation so that the previous bunch can hook on with teams that actually play defense.

When you think about how good I-Rod is at cutting off the running game, you start to really appreciate how badly things go when opposition hitters put the ball in play. Carl Everett should help a bit, but Izzy Valdes is probably in for the shock of his life (expect his ERA to go way up this year). Chan Ho and Dave Burba shouldn’t be as drastically affected, since Cleveland and LA were as bad defensively as Texas was in 2001 (though Chan is moving from an excellent pitchers’ park to one which seems to favor hitters).

The defense could actually get worse if Catalanotto becomes the regular 2B in Mike Young’s place.

Throwing out base-runners

As we all know, both pitchers and catchers differ in their ability to cut off the running game. What I like to do is compare a catcher’s throwing performance to the other catchers on the team. This is not a perfect approach, since in some cases particular catchers are in the lineup when particular pitchers are on the mound. As well, the backups themselves differ in their ability to throw out base-runners so we can’t really make firm conclusions looking at this data. All the same, it is interesting to look at.

As it happens, all 4 teams in the AL West had the same regular catcher in both years, so that is helpful. I couldn’t find pitchers only CS for 2001 so I lumped all CS together instead of segregating them into “normal” and the 1-3-6 kind.


Ben Molina2121986865.7

The other starts were made by Jorge Fabregas (42), Matt Walbeck (38) Shawn Wooten (20) and Jose Molina (12). None of these guys had trouble throwing out base-runners and Wooten and J.Molina were good in limited trials.


Dan Wilson1821153867.0

Not only did Wilson extinguish runners at a greater rate than the subs, the opposition didn’t try to run on him as often. Ben Davis threw well for San Diego the last few years, so he should represent a defensive upgrade over Tom Lampkin who allowed runners to steal at a .705 clip.


Ramon Hernandez2452377170.0

Ramon was the busiest catcher in the division. His throwing did not differ radically from his teammates (mostly Fasano and Myers). For Oakland’s sake, I hope Hatteberg isn’t behind the plate too often in 2002 (with the non-slide-stepping Red Sox he produced an .872 over the past 2 years, versus a rest of team .744).


Now we come to the King of kings, or should I say the Gun of guns.

Ivan Rodriguez186975444.3

Without I-Rod in there, opposition runners saw the green light much more often. They were below break even, so the Rangers catching subs were not bad. The main back-up, Bill Haselman, allowed a .708 PCT. This was the largest difference I found between the main catcher and the subs in the American League (I haven’t checked the NL yet).


Robert Dudek Posted: March 08, 2002 at 05:00 AM | 36 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Christopher Posted: March 08, 2002 at 12:25 AM (#604916)
I think this is one of the big question marks for the A's this year. I imagine the 2002 A's outfield to look much more like the 2000 version than last year's. Going from Long - Damon - Dye to Justice/Giambi - Long - Dye (coming off a major injury) is a downgrade at each position. Dye has yet to play this spring, which isn't good news. Long I think is underated but still doesn't have the range of Damon. And the only reason no one says Giambi is the worst outfielder in the game is because he's played with Ben Grieve. Fortunately for the A's, only Zito shows real flyball tendancies among the starters.
   2. Charles Saeger Posted: March 08, 2002 at 12:25 AM (#604918)
Robert, something's wrong with your methods. Seattle, not Anaheim, had the best defense in the AL last year, even after adjusting for park. IIRC (the spreadsheets are at home), Anaheim was about average. Statistically, Seattle had the best defense I had ever seen.
   3. Charles Saeger Posted: March 08, 2002 at 12:25 AM (#604919)
Robert: oops. I read the article wrong. Mea culpa.
   4. Mike Emeigh Posted: March 08, 2002 at 12:25 AM (#604926)

I'm not sure that applying general park factors is the right thing to do here. The one largest influences on park factor is the HR factor, which is defense independent. Case in point: Comerica Park ranks as a pitchers' park largely because it's extremely difficult to go deep there - but there tend to be more non-HR hits there than in Tiger road games, and virtually the same number of doubles and triples in Comerica as there are when Detroit plays elsewhere. PNC Park in Pittsburgh seems to have the same characteristics; looking at the non-HR hits, it would be a decent hitting environment, but because it appears to suppress HR it comes in as being slightly favorable to the pitcher. Houston is the flip side of this coin; it's virtually neutral on non-HR hits but looks like an extreme hitters' park because of the HRs.

-- MWE
   5. Robert Dudek Posted: March 09, 2002 at 12:25 AM (#604930)

The park factors I used were not general park factors. I calculated park factors for the FRA stat itself.

I broke down FRs and then calculated FRAs at home and road separating the homeside from the visitors for each. I then calculated park factors in the same way I approach general park factors: I adjust for park based in the number of innings each team played at each road park.

For example, Minnesota is not a particularly good homerun park, and so, even though the park overall favors hitters, it fvaors high FRAs even more.
   6. Robert Dudek Posted: March 09, 2002 at 12:25 AM (#604931)

The answer to your question is "a lot", though I don't think anyone really knows. The precise answer to that question lies in Voros territory.

The FRA stat is influenced by particular pitchers, but I would estimate that less than 10% of all pitchers influence the stat significantly over the long term. It's possible that a particular team will have an inordinate number of these pitchers, but that doesn't happen too often.

Using two-year ratings, I'm confident that the stat effectively separates out the good defenses from the bad. It was necessary, however, to park adjust team FRAs to come up with an adequate rating.
   7. TeddyA Posted: March 09, 2002 at 12:25 AM (#604934)
I've always been amazed by Ivan Rod's caught stealing percentage because, presumably, managers think about who the catcher is before trying a steal. If steal attempts were determined rationally and managers had adequate information about the catchers ability, then all catchers with larger enough sample sizes should have very similar caught stealing percentages. That percentage should be just above the break-even point of 66% (or likely higher given current scoring). It's amazing that year after year managers let anyone other than the handful of great basestealers run against I-Rod. Are there data on how many of these steal attempts were botched hit and run attempts or other special cases that might make an "attempted steal" more rational?
   8. Chris Dial Posted: March 10, 2002 at 12:25 AM (#604938)
My problem with this methodology is that it doesn't attempt to differentiate good pitching from bad fielding. The Texas individuals aren't below average fielders. *Most* hits are a function of the pitcher (or hitter) and not the defense.

The Texas team is not loaded down with Lead Glovers. ARod, Palmeiro, Kapler are *good* fielders. Cattalanotto was average in LF last season, and Young was about as good as Bret Boone. I don't think it's accurate to say that the Texas defense is bad.
   9. Mike Emeigh Posted: March 10, 2002 at 12:25 AM (#604940)
*Most* hits are a function of the pitcher (or hitter) and not the defense.

Then how does one explain the results of Voros McCracken's research? How does one explain Dick Cramer's findings that the results from conventional measurements of fielding skill (errors, DPs/opportunity, and passed balls) correlate more closely with a team's ability to prevent hits than conventional measurements of pitching skill? How does one explain the research that Pete Palmer has done, demonstrating that year-to-year variations in BA on balls in play against pitchers are, while not entirely random, much closer to random than year-to-year variations in batting average among hitters? Almost anyone who has looked at the data in any significant detail has concluded that hits are much *less* a function of the pitcher than of either the defense or the hitter.

-- MWE
   10. Chris Dial Posted: March 10, 2002 at 12:25 AM (#604941)
Actually, Mike, it's pretty simple. The zones as assigned by STATS, while not perfect, draw an accurate picture of which balls in play (BIP) that a fielder can be expected to make a play on at least 50% of the time. The data can be taken and analyzed for hit location. If the hit location is dominated by GBs through the holes (56 or 34) and line drives betweent he IF and OF, then those hits are *the pitcher's* fault - or the hitter's credit. They cannot and should not be assigned as defensive shortcomings. The pitcher threw a ball the hitter could put in play in a fashion that would result in a base hit, and no fielder could make a play on the ball with ordinary effort.

Seriously, you have all the hit location data for 1999 and 2000. Where did the hits land - were they GBs in the hole and line drives betweent he IF and OF - or long hits down the line and over the OF heads? Sometimes there plays are made, but if the play on a ball is made (MLB-wide) by less than 50% of the time, it doesn't seem reasonable to me to assign the "accountability" to the defense rather than the pitcher (or give credit to the hitter).

*Most* hits are not reasonably defensible. You can take that to the bank.
   11. Robert Dudek Posted: March 11, 2002 at 12:25 AM (#604942)

It's true that most hits are not reasonably defensible, but the difference between good and bad defensive teams are the minority that are.

I can also add that defense isn't only preventing hits from happening.

It's possible that the Texas Rangers pitchers have given up a lot of line drives over the past two years, and, as I've said, FRA is not a pure measure of fielding. It could well be that the Rangers have had a collection of pitchers that are predisposed to having high FRAs. Then again the ZRs of individual players may not have any relevance to the overall defensive quality of a given team.

You've yet to come up with a satisfactory explanation for why Voros' DIPS works. To me, it's clear that batting average on balls in play is more a function of the fielders than what the pitcher throws to the plate.

Just because the hitter makes a good swing doesn't mean you can absolve the defense of a debit for what happens after that. If a pitcher makes an excellent pitch and the hitter happens to hit it out of the park, the pitcher gets the "blame" all the same. Whatever credit is given on the offensive side has to be mirrored on the defensive side.

I don't think zone rating is a very good stat to use when discussing team defense.
   12. Mike Emeigh Posted: March 11, 2002 at 12:25 AM (#604945)
The distribution of balls in play - where the balls are hit - is almost entirely a function of the hitters. All hitters - LH hitters and RH hitters - tend to pull ground balls and hit fly balls the other way. They do this regardless of the type of pitcher on the mound, whether it's an LHP or RHP, ground ball pitcher or fly ball pitcher, high-K pitcher or low-K pitcher.

The pitcher tends to control two things: (a) the frequency with which the batter swings and misses, and (b) the type of ball (ground ball or fly ball) that he hits when he does put the ball into play. The hitter tends to control one thing - the direction. And because the direction of the ball in play is usually the largest determining factor that decides whether the ball is a hit instead of an out, the hitter has far more control over the results of the play than does the pitcher.

Assigning large roles to hitting and defense and small roles to pitchers on the results from balls in play explains a lot of things that we see. We see hitter performance varying much less from year-to-year than does pitcher performance; if the hitters have more control over what happens when they put a ball into play we'd expect to see that. We'd expect to see power pitchers be more successful over a longer period of time than finesse pitchers - and we do. We'd expect to see the existence of a class of pitchers like those dubbed the "Tommy John" class, who need to be on good teams in order to pitch well - and we do. And when we see some pitchers who can succeed with "trick pitches" like the knuckleball or the spitball (there's some evidence of a "spitballer's advantage" in early 20th century data), we can suggest that one reason they can do so is that they neutralize the batter's ability to control the direction of a ball put into play because of unnatural movement on their pitches.

Hitter/fielder control fits the available data much better than does pitcher control.

-- MWE
   13. Chris Dial Posted: March 11, 2002 at 12:25 AM (#604949)
Robert -
there's no such thing as team defense in baseball. No more so than team pitching or team hitting. It is a collection of individual plays for the most part. That's why individual ZRs define how the team plays defense. If each player turned in a perfect ZR, there would still be hits, and there would still be a FRA, even though the defense fielded ever ball they could (within general expectations).

Okay, ZR isn't perfect, and individual ZRs don't purely define the defense, but the data set *could* be used in that manner.

As for why DIPS works, I don't know what you mean by "works". What's the spread of $H? And how many hits is that? Can those hits be explained by the quality of defense? Personally, I don't know what DIPS tells us. All I get from it is that we can't predict a pitcher's hits in year 2 based on year 1.

Excellent pitches don't get hit out of the park. Pretty much by definition. The occasional hit? Sure - HRs? No.

Mike -
what pitchers control (a) leads to balls batters swing and almost miss. These result in pop-ups or weak GBs. That's a pitcher function, as much as a hitter missing is.

I am aware of these things you talk about wrt where balls go when hit, and I don't have alot of qualms about a hitter being mostly responsible for hits. You aren't arguing against my position. I'm not saying hits are the pitcher's fault - I'm saying they are NOT the defenses. As such, a "team defense" stat assigns blame (and hope) where it doesn't belong.
   14. Robert Dudek Posted: March 12, 2002 at 12:25 AM (#604953)

To my knowledge, ZR doesn't measure a team's ability to turn a double play, a team's ability to prevent doubles and triples, nor outfielders' or catchers' abilities to restrict runners from advancing. Not only is ZR not close to perfect in terms of measuring range (e.g. I've never seen ZRs properly park adjusted), but it leaves significant aspects of preventing runs outside it's scope.

To me, DIPS shows that pitchers do not have a huge influence on BIP AVG, and yet teams will sometimes differ significantly on this score. The logical explanation is that the fielders have more effect on this stat than pitchers do.

Team FRA doesn't appear to be a random stat: teams that do well one year tend to do well the next and teams that don't do well one year tend not to do well the next. There are real differences between teams on defense and this shows up in the FRA stat.

It's true that various pitchers have a tendency to be above or below average in this stat. There are some things that the stat measures that are mostly due to the pitchers (e.g. pitching out of a jam). As such FRA is not a pure measure of fielding ability and I never purported it to be.

FRA is far from a perfect stat: it's purpose is to estimate how good or bad a team is on defense (pitching and fielding) taking the independent elements (W,K,HBP,HR) out of the equation.

Texas was below average the last two years in team ZR, and since this stat imperfectly measures only a portion of what constitutes team defense, I don't see what evidence you have for saying the Texas Rangers' defense was not "bad" in 2000 and 2001.

Using ZR to rank players on defense is sort of like using unadjusted OBP to rank hitters - you'll be missing some huge pieces of the puzzle. Actually, it's not quite as good because OBP precisely measures the frequency of reaching base, whereas ZR imperfectly measures defensive range.
   15. Chris Dial Posted: March 12, 2002 at 12:25 AM (#604954)
No, ZR doesn't measure the ability to turn a DP. That's a separate stat. It isn't too difficult to account for. ZR doesn't define range - it defines how often a fielder turns balls into outs. It misses some aspects of defense, but it includes teh vast majority of defense - like OPS covers offense (okay, not to quite the same degree), and it can be modified to include some problems. I don't believe in "park-adjusted" infields. Every major league field is manicured. The OF issues are isolated - Coors, Fenway LF, Minnesota ceiling. I'll have to see some evidence that there is something going on in other parks before I buy into it.

As for DIPS Avg - what's the status of DIPS ISO? Until I see some evidence that there's no skill there, I'm not buying that hits are 85% batter. A pitcher like Greg Maddux prevents hard hit balls - that's what I'm putting out there. Okay - one season he gives up an extra gork per start than the previous season, but what is the predictability wrt how often he surrenders extra basehits? I re-read DIPS the other night and this stands out as missing. I've asked Voros, and he doesn't have the complete study done, AFAIK, and this is a critical aspect to it.

As for Texas, if the individual fielders on the team are above average, then their "team defense" cannot be the worst in the league. I don't buy it.

Research like this can be had, it just resides in the hands of the Evil Empire.
   16. Mike Emeigh Posted: March 12, 2002 at 12:25 AM (#604955)
ZR doesn't define range - it defines how often a fielder turns balls into outs.

when those balls are hit in the vicinity of the fielder as defined by STATS. If a ball isn't hit in the pre-defined vicinity of the fielder, it isn't counted against him if he doesn't make the play even if under the game circumstances he should have been positioned to field it; if a ball is hit in the pre-defined vicinity of the fielder, it is counted against him if he doesn't make the play even if under the game circumstances he shouldn't have been positioned to field it.

The plays a fielder should be able to make are dictated by game circumstances - the identity of the hitter and pitcher, the runners on base, the number of outs, the game score. There is *no* static map that can adequate define the expected coverage areas for every game circumstance - and zone-based methods will always be suspect for that reason.

what pitchers control (a) leads to balls batters swing and almost miss. These result in pop-ups or weak GBs. That's a pitcher function, as much as a hitter missing is.

Some pitchers *might* be able to do this (Maddux, knuckleballers) but it doesn't appear to be a repeatable skill for most pitchers. This is another area where some study is needed, where $H rates are evaluated for particular ball types.

I'm not saying hits are the pitcher's fault - I'm saying they are NOT the defenses.

And what I'm saying is that the evidence suggests that the defense does prevent some percentage of hits once the ball is put into play. Not all hits, certainly, but some - enough so that you can evaluate defenses by using the differences in the rate at which balls in play are converted into outs. Furthermore, the evidence also suggests that the pitcher's role in hit prevention is limited primarily to keeping the hitter from putting the ball into play in the first place, so that you can ascribe most, if not all, of the differences in the rate at which balls in play are converted into outs between teams to the defense. That still leaves a fairly substantial role for the pitcher; the non-defense affected aspects of pitching account for anywhere from 45-55% of the impact on run scoring.

-- MWE
   17. Chris Dial Posted: March 12, 2002 at 12:25 AM (#604958)
"even if under the game circumstances he should have been positioned to field it" Huh? Who defines *that*?

The plays where a fielder is not in position to cover his zone are the extreme minimum. In your post regarding plays hit to DA's Zone 6 you state 95% of these plays are made. If SS's were playing out of position with any frequency, that couldn't happen. It should be testable. If playing at DP depth were really out of position, teams that had more DP situations would see a decrease in ZR. I don't see it. As I've stated, ZR isn't perfect, and I've had the discussion re: putting GPS systems on players in asb-sdp. Does it describe the value of performance *mostly* accurately? I say yes (well, the data - the present use is even less than ideal).

As for the hits the defense could prevent, I'd suggest that the number is already defined - missed ZR chances. You have the DA raw data - how many hits (percentage-wise) fall in-between the infield and outfield? How many GBs in the holes and up-the-middle are there? How many basehits go *right through* Zone 6? You already have this data. If less than 5% of balls hit to ZOne 6 are hits, and 95% of balls hit to 67(?) are hits, that goes a long way to saying that those balls are unplayable and *not the defense's fault*. What are those percentages? I was talking to Voros about it last night. The hit differential between a good typical DIPS season (BIP avg .260) and a bad one (DIPS Avg .320) is about 30-40 hits. If DIPS avg differences is due to luck/defenses, and there's no predictive value, that indicates that the level of defensive impact is in the range of 30-40 hits (per 200 IP pitcher). That's about 25% of hits can be assigned to the defense.

You seem to say something similar with this: "you can ascribe most, if not all, of the differences in the rate at which balls in play are converted into outs between teams to the defense".

I think I agree. What are those differences, and how many runs does that make? Let's see - 1500 Hits allowed and 160 HRs - so 335 *hits* attributed to defense (at 25%). 335 hits is about 285 runs (at average). Or 285 runs in a season - or 1.75 rpg. For an average team. Maybe that coincides with Robert's work. I'd do it this way though.

Please note, I'm doing this on my office desk calendar. I would also take as many players as I could get on a given team (well, every team) and see what the total number of ZR plays not made and see what % of hits allowed (Less HRs) is. I'll work on this this weekend - unless someone has more time.

   18. Robert Dudek Posted: March 12, 2002 at 12:26 AM (#604964)

Not park adjusting ZR is just plain wrong. Parks differ in any number of ways which will affect this stat.

There are still a few turf parks out there and surely the plastic has a different effect than grass on infielder's ZRs. Visibility is an issue as well. The outfield dimensions of parks vary quite a bit, especially now with the new retro parks.

Because of wind currents, there are some parks where the ball carries better than others - which will likely change homeruns (no ZR chance) to flyball outs (successful ZR chance).

The park adjustments I've figured for FRA suggest that there are parks like KC, Minnesota and Toronto that have a tendency to inflate FRA and others like Seattle, Oakland and Baltimore (the usual suspects) that tend to decrease it. It's inevitable that this will impact ZR as well.

You cannot expect anyone to take ZR seriously as a rough guide for defensive ability, when, in addition to its system of measurement (which is contestable to say the least) it leaves very important skills like preventing extra base hits, turning the DP and throwing out baserunners out of the equation.

If you then leave it unadjusted for park , it's usefulness as a TEAM defense metric becomes highly suspect.

BTW, I checked, and Texas had a below average team ZR in both 2000 and 2001. I also found that Texas was 12th in 2000 and 14th in 2001 in the AL in terms of park adjusted non-homerun extra base hits allowed per flyball (meaning they gave up many more doubles and triples than would be expected given the park and GB/FB tendencies of the pitchers).

Those two facts suggest that Texas is likely a below average defensive team. I venture to say that my FRA studies add more weight to that side of the argument.

As far as Maddux is concerned, I've found that his FRAs have been consistly better than his team as a whole, but not by a large margin (I believe that the data is posted on for Maddux and 15 other pitchers who have pitched the most innings since 1990).

As I said, it's possible that a team could acquire a large number of pitchers that have a large effect (positive or negative) on the FRA stat. There aren't many pitchers like that, and it would be very unusual for there to be such a huge imbalance of them on one team. I've used data from 2000 and 2001 and so that should reduce the chances of those odd cases ending up on the same teams because there is normally significant pitching staff turnover from year to year.

I haven't park adjusted FRA for 1999 and earlier yet, but here are Texas' unadjusted FRA rankings since 1996:

1996: 3.575 (5th); 1997: 3.846 (9th); 1998: 4.468 (last);1999: 4.102 (13th).

As noted in the original text, The Ballpark has actually tended to decrease FRA in 2000 and 2001, so I feel safe in saying that thee 1996-1999 would not be significantly changed by park adjusting them.

Only 5 Rangers pitchers had 20+ inning for the Rangers in 1999 and 2001 (Helling, Zimmerman, Venafro, Glynn, Crabtree). They acounted for 495 innings in 1999 and 416.3 innings in 2001. That's quite a bit of turnover and since most pitchers do not have a significant effect on FRA, we'd expect that the Rangers poor showing in this stat to be mostly the result of substandard defense.

I think the addition of ARod and MYoung probably improved the infield a bit (but Lamb was another newcomer). I think the outfield was bad the last two years and this is the main reason for the Rangers poor FRA.

   19. Chris Dial Posted: March 13, 2002 at 12:26 AM (#604965)
Robert -
unless you have evidence that turf makes a difference, I don't believe it. I don't believe there are park effects for the infield. Unless you can demonstrate it, or offer a solid hypothesis, you probably shouldn't assert it. Can you prove visibility is an issue anywhere but Minnesota? Is there even any anecdotal evidence? I can state categorically that OF dimensions are not an issue with ZR (outside of Coors and Fenway LF). *NONE* at all. And wind currents? Please. Conjecture has no value here.

As for your FRA PF - are they different from STATS PF for XBH? No? Doesn't sound like the defense. And unless you actually hve some evidence it impacts ZR, I don't buy it. The zones aren't there. If anything the corner OF where there is extra foul territory could get a ZR boost - but that doesn't appear to happen - or at least not in a significant manner (significant meaning one I can see).

As for taking ZR seriously as a measurement of defense, you can like what you want. Consensus is that ZR is easily the best metric available - FRA notwithstanding. What do you think the difference is in preventing XBH on the infield is? How about the difference in DP turns? Do you know what the difference in DPs is from top to bottom? It's not alot in terms of runs. Throwing out baserunners (from the OF?)? That's a virtual non-stat. There are exceptional OFs, but for the most part it's not significant considering how many plays the OF makes overall.

There's *no such thing* as TEAM defense. It doesn't exist.

The Rangers '00 and '01 ZR was .846 and .847 while the league averages were .848 and .851. They were just below average both seasons. And *really* far from the worst. Your metric doesn't indicate that Texas is just "below average" - you rate them 14th - last. I don't buy it.

If a park is a good hitters park, then it should follow that it is a poor pitching/fielding park. However, we know that Texas is an excellent hitter's park, and yet, you have it as being a good fielders' park. I don't understand that at all.

I looked at the correlation of team FRA in your original article. It was 0.25. I looked at team ZR (which is a cumulative number- not a "team" ZR) for '00 and '01 it was 0.5. I think ZR better describes how a team performs on defense.

"we'd expect that the Rangers poor showing in this stat to be mostly the result of substandard defense." If by "we" you mean "you", then I'll agree. And substandard (assuming below average) isn't 14th. Your "expectation" makes a huge assumption that FRA means anything.

I'm most interested in how a hitter's park can also be a fielder's park.
   20. Robert Dudek Posted: March 13, 2002 at 12:26 AM (#604967)

Frankly, other than the creators of the stat, you are the only person I've ever heard of being a fan of ZR. In any case, popularity does not make a stat accurate or useful.

The Ballpark has been fairly neutral the last two years. And it is possible for a park to be homer friendly, but restrict singles, doubles, and triples. The Ballpark is close enough to neutral as an overall hitters park that it's not suprising for it be a shade under neutral as a fielder's park. I speak only of the last 2 years, as I have not calculated FRA PFs pre-2000.

I'm very happy for you that you don't "buy" park effects influencing ZR. Ideally EVERY stat in baseball should be park-adjusted. There's no such thing as a stat that takes place outside the context of park influence. If you favor the use of the stat, it's up to you to come up with park factors for it - or else PROVE that they are unnecessary. To assert that the stat doesn't need to be park adjusted is a little bit comical. Take a look at event PFs for singles and doubles: the differences between parks are real and hold up quite well over time. How can that NOT impact ZR?

As I said, the Rangers defense may not be as bad as FRA says, but what makes you think it's as "good" as ZR says it is. I've looked at DP rates in the stats 2001 handbook (p.246). The best in the AL was KC at .641 - they were 14.1 DPs turned above average. Boston was the worst at .495 - 17.6 DPs below average. If we use a factor of -.37 for each DP that's a range of 11.7 runs.

Regarding extra base hits, in the same league the difference between best park adjusted figure (Minnesota) and the worst (Oakland) was about 92 extra base hits. I don't know how much weight you'd want to give to each extra base hit but let's assume it's about 0.35 runs(subtracting the value of a single from a double and adding a bit for the triples). That's 32.2 runs.

In 2001, Texas finished last in preventing extra base hits adjusted for park and were 99 XBHs below average. I think that could go a long way in explaining Texas' poor FRA that year. The difference between them and Minnesota (1st) was about 163 extra base hits. That's about 57 runs.

Preventing stolen bases (in which Texas is very good of course) in 2000 had Boston as the worst (159-47) and the Chisox as the best (79-58). Using +.18 and -.32 as the weights we get a difference of about 18 runs. In 2001, the difference between Texas (best in the league) and Baltimore (2nd worst in the league - Boston was a total outlier) was about 19 runs.

Outfield assists we can only guess at, but that could be around 10 runs.

All these skills are not covered in ZR and cumulatively they are important.

The range between the best overall team pitching and fielding in 2000 was about 200 runs (Boston and Texas). A lot of that difference is going to come from the fielding-independant stats.

The drawback to FRA is that it does include things that the pitcher is at least in part responsible for. Its strength is that it collects all these "secondary" defensive skills in one basket.

The strength of ZR is that it "attempts" to measure the ability of fielders to make outs - a very important skill. The drawback is that it doesn't address several important defensive skills.

I think that if you adjusted ZR for park, you'd have a fairly useful stat.

   21. Chris Dial Posted: March 13, 2002 at 12:26 AM (#604969)
Robert -
I scored for STATS. I know what ZR is capable of and what its limitations are, and most importantly, what the data it is based on means. Remember Defensive Average (DA)? That is so close to ZR it is goofy. In fact, ZR could be converted to DA with minimal effort. I believe that's what Mike Emeigh uses. The biggest problem with ZR is it's poor publicity and lack of access to raw data. And how about Mitchel Lichtman's Super LWts? That's based on ZR scoring.

TBIA has been neutral? Hardly. I looked it up last night before commenting. Sure, it has a high HR PF, but the XBH data is also high (105-112). It's run scoring is high. It's not close to being a neutral park. It's the biggest hitter's park in the AL. And do you have TBIA as a "shade under neutral" for your PF? Where do you have Texas wrt the rest of the league?

I said I don't buy PF for *infield* ZR. Ground balls?

The difference between the best team at DPs and the worst is *12* runs (I'm accepting your calc because I don't have the data in front of me - although I think your estimate is way too low). Doesn't that tell you that DPs are a small part of defense? A SS prevents some 400 runs just on GBs (ZR). He turns another 75 DPs and catches a bunch of pop-ups. These extra plays are insignificant relative to what he prevents doing his routine job.

As for XBH prevention, this is where we disconnect - *fielders - as a group - cannot significantly prevent XBH*. This is easily testable with raw ZR data (which of curse I don't have). Doubles are in the gaps (not covered by ZR) and down the lines and off the walls. They are plays not made by *all* defenses. The TBIA has a high PF for XBH. That includes *both* teams performances at TBIA and on the road - that says the XBH in TBIA are a function of the park, not the guys playing defense.

Stolen bases are not a function of the team - they are a function of the pitcher and catcher. And we know Texas is excellent at this. Of course, I don't know what your park factor for SBs is (since all aspects need to be park-adjusted).

Personally, I adjust ZR based on the pitching staff - not the park. And that's for OF only. My entire methodology and *thorough* explanation of ZR and it's usability is on this site.

   22. tangotiger Posted: March 15, 2002 at 12:26 AM (#604975)
You guys should read the great work MGL has done looking at types of balls the pitchers give up

ZR is a good stat, but UZR is better (this is what MGL uses).

The missing part in this discussion is "luck". Just because a pitcher has a low responsibility with $H doesn't mean that the fielder is responsible for the rest. It could be luck as well, as far the the defense is concerned.
   23. tangotiger Posted: March 15, 2002 at 12:26 AM (#604976)

I talk about balls in the dead zone here. Balls that pitchers give up that fielders do NOT have a reasonable chance to get to. Texas Rangers lead the league here.

If I had the data, I would do the following: determine the "success rates" for every STATS zone. Determine the "frequency rates" for those same zones. By using league average rates for one, and the player's/team's rate for the other, we can "isolate" the fielders from the pitchers, insofar as turning balls in play into outs.

This is what UZR does for fielders. Now, you jsut have to do the opposite for the pitchers.

You can also add more variables (such as GB,FB, LD, and LH/RH, LP/RP) to get even more accurate results.

   24. Mike Emeigh Posted: March 17, 2002 at 12:26 AM (#604983)
I talk about balls in the dead zone here. Balls that pitchers give up that fielders do NOT have a reasonable chance to get to.

In the breakdown of data that I have from the 1999 and 2000 seasons, fielders made plays on at least 10% of the balls hit into every zone on the field. That includes the gaps. That includes the "down-the-line" shots. That includes the "no-man's land" between the infielders and the outfielders. Even when you take the bar up to 33% (where fielders make a play one in three times), you still have most of the field covered; the only areas of the field where fielders don't convert at least 1/3 of the balls hit into those areas into outs are the pop-up zones between the infield and the outfield. Maybe you don't think a 10% chance is a reasonable opportunity, but certainly I would think a one-in-three chance would be more than reasonable.

My opinion is that we have no business making assumptions about whether or not it is *reasonable* to expect a fielder to make a play on a ball. A ball in play is either converted into an out or it isn't. When it is, the defense should be credited for making that play; when it is not, the defense should be penalized for not making that play. We don't excuse a batter when he hits a ball where it wasn't *reasonable* to expect him to get a hit, and we shouldn't be that lenient with the defense, either. The defense should be expected (from an analytical standpoint) to convert every ball put into play against it into an out.

-- MWE

   25. Robert Dudek Posted: March 18, 2002 at 12:26 AM (#604985)
I agree with Mike on the above point.

I also agree with "tango" when he says that luck influneces these results, but luck influences hitters and pitchers too. One reason I looked at two years worth of data is to increase the sample size.

Going back, the Rangers have had below average FRAs since 1997 (5 years) and it's unlikely that this is because of bad luck.

Everything that happens on offense has to be debited or credited to the defense (pitching plus fielding).

DIPS shows that pitchers do not exert a strong influence on BIP%, but I think it is nevertheless true that they do influence it somewhat. I've also noticed a remarkable consistency among pitchers in terms of the percentage of hits in play that go for extra bases. I feel very safe in asserting that preventing extra-base hits is a task that the fielders are mostly responsible for.

In sum, every hit against has to be assigned to either the pitcher or the fielders.

   26. tangotiger Posted: March 18, 2002 at 12:26 AM (#604986)
Robert: I agree that pitchers exert little influence on balls in play being converted into outs (as the $H in DIPS shows). However, we have NOT yet shown if the fielders exert any more or less influence on balls in play being converted into out. We cannot make the assumption that since the pitchers exert little influence, then the fielders must exert the remainder. In fact, the fielders may exert even less than pitchers! And the unaccounted for is luck. There's no need to distribute all of the luck portion to the fielders (if the intent is to have everything add up).

DIPS makes bold assertions backed up by lots of research. And still we have many questions. To make assertions about the fielder's responsibilities with little research is to be on thin ice.

Mike: As I said, I much much prefer UZR. Every zone has an average conversion rate. Say, Zone SS-left is 30%, Zone SS-middle is 95%, and Zone SS-right is 40%. Therefore, we should not expect the fielders to convert every ball in play into an out. We should expect fielders to convert the balls in play based on where the balls are hit. If Jeter has the bad luck of getting more than his fair share of balls hit to Zone SS-left AND HE CONVERTS THEM at a 50% clip, then this is a plus and not a minus. With ZR, it is very possible that a player can be above average in his conversion rates in every zone, and still be below average overall, because of the frequency in which balls are hit to each zone.

Therefore, why not simply establish what the league average conversion rates are for each zone, and compare that to what each player actually converted. This is what UZR is all about. We can further refine this to look at different conversion rates based on the type of hitter/pitcher matchup we have.

All I am saying is to do a similar thing for pitchers, to isolate their skills from the fielders.

As for my other posts regarding dead zone, I would not worry about "semantics" like reasonable. I was simply trying to point out that STATS draws lines as to what is playable and what is not, and that Texas Rangers pitchers have more balls land in the "unplayable" spot. Obviously, every balls is playable to some degree. I would again prefer to see the UZR for every pitcher, but this is what they've given us to work with.
   27. tangotiger Posted: March 18, 2002 at 12:26 AM (#604987)
Mike, I see from your post in the other defense article that you think (somewhat) along the same lines as I do. Looking at the Ranger pitchers, can you give us a breakdown of how many balls they give up in each zone (the frequency rate), the number of those balls that were converted into outs (the success rates), and the league average frequency rates and success rates for those zones. (Bonus for us if you have the time/inclincation: a split along LP-LH, LP-RH, etc, as well as home/away might be benefifical as well. )
   28. Mike Emeigh Posted: March 18, 2002 at 12:26 AM (#604988)
We cannot make the assumption that since the pitchers exert little influence, then the fielders must exert the remainder.

Nor can we make the assertion that it's due to luck, either. There's always the possibility that it's due to the ability of the hitters.

Dick Cramer and Pete Palmer have done some analysis (unpublished as yet; don't know how either one will release their work) that tends to support the notion that fielders have more to do with the results from balls in play than do pitchers or hitters.

As I said, I much much prefer UZR. Every zone has an average conversion rate. Say, Zone SS-left is 30%, Zone SS-middle is 95%, and Zone SS-right is 40%. Therefore, we should not expect the fielders to convert every ball in play into an out.

The problem with this analysis is that the conversion rates are based on an average *distribution* of balls into that zone - an average mix of GB/FB/LD - and an average *position* for the fielder, based on the combination of positions he will normally take up. An average SS, for example, might play at normal depth 65% of the time, at double-play depth 30% of the time, and in 5% of the time. He might shade toward the middle about 40% of the time, and toward the hole about 60% of the time. If a shortstop plays for a team that allows fewer baserunners than the norm, he might play at normal depth closer to 70% of the time, at DP depth maybe 27% of the time, and in just 3% of the time. If a shortstop plays for a team with a preponderance of LHP, he might be shading toward the hole 70-75% of the time; if the opposite, he might find himself shading up the middle 50% of the time. All of these changes affect that so-called "average mix". If my SS is shading toward the hole, then instead of making 30% of plays in SS-left and 40% in SS-right, he should be making 40% of plays in SS-left and 30% in SS-right. If he's shading up the middle, he should be making, say, 50% of plays in SS-right and only 20% in SS-left. And if he's positioned correctly, and covers plays with those percentages, he'll make more plays than a SS who fields an "average" percentage in both areas, because there will be more balls hit into the zones he's shading toward. That's why you shade toward that area in the first place.

I prefer an inclusive approach rather than an exclusive approach. Don't try to eliminate areas of the field from coverage by a fielder; rather, assume the broadest possible range that he could possibly cover and assign him 100% responsbility for that entire area. Yes, this means that there will be overlaps between fielders (which you account for by reducing the negative impact of a fielding failure on any individual fielder, as Charlie Saeger has noted). Yes, this means that a lesser fielder could benefit from playing beside a stronger fielder. Yes, this means that every ball is considered *fieldable*, and that fielders can be penalized for not making plays on balls that they didn't have a realistic chance to convert into an out under the circumstances (although as I said before I think that's a relatively small set of plays). But I think that approach, properly implemented, is far *less* error-prone than ZR or UZR.

-- MWE
   29. tangotiger Posted: March 18, 2002 at 12:26 AM (#604989)
Nor can we make the assertion that it's due to luck, either. There's always the possibility that it's due to the ability of the hitters.
*** I am not making that assertion. I am saying that we have to consider all these things, before we make the assertion.

The problem with this analysis is that the conversion rates are based on an average *distribution* of balls into that zone - an average mix of GB/FB/LD - and an average *position* for the fielder, based on the combination of positions he will normally take up.
*** I agree. This is why I'm trying to get people to publish the data along those lines, and the reason that MGL publishes his work along those lines. But all I've been working on is what's available to me. I'd love to get success rates and frequency rates for every STATS zone that considers:
- the batter's handedness (LH/RH)
- the pitcher's handedness (LP/RP)
- the batter's gb/fb tendencies
- the pitcher's gb/fb tendencies
- the numbers of outs
- the runners on base

With UZR, at least it's the first step away from ZR. It's the first step to break up the zones into different success rates. Now, we should account for more variables. All known variables should be accounted for and established as to their impact.

And the overlapping areas should also be accounted for and studies. Does Jeter play differently because he has Scott Brosius as a great fielder next to him? (I think MGL looked at this one last year).

I'm open to hearing and seeing all of the data and studies. Until then, the best we've got is UZR.
   30. tangotiger Posted: March 18, 2002 at 12:26 AM (#604990)
I also suggest that people look at MGL's work on UZR. Here is the link with a cut/paste of his methodology

First, I go through the data, and determine who fields (turns into an out) what (fly balls, pop flies, line drives, and ground balls), and how often, in each zone (I use Project Scoresheet zones (larger, fewer) and not STATS zones, although they are basically the same thing). (I do use the hit type and hit location data from STATS Inc., although, as I said, I transcribe the zones from STATS format to PS format.) Anyway, I go through the data again, and if a ball falls into a zone (no out) then any fielder who normally fields at least a percentage of the balls hit into that zone gets "credit" for a piece of a hit allowed. If a ball is fielded (turned into an out) in any zone, then the fielder who makes the play gets credit for a piece of an out (no other fielders get "penalized). These "pieces" are determined by the overall frequency that each fielder records an out in each zone (for each type of hit). BTW, I don't have the info on how hard each ball is hit. That would be useful, of course, since then I wouldn't have the problem of some fielders getting more harder hit balls than others (by chance) and being unfairly penalized for that. Anyway that is the basic methodology. The above description is not complete, but you should get the idea. As I said, I THINK I have described the complete methodology in another post. Rest assured that my methodology is very sound. The correlation coefficient (R^2) for my 1997 and 1998 UZR's is .52 (all fielders other than pitchers and catchers, who have had at least 200 "chances" in each year). For 1999 and 2000, it is .56. I think that these are very high numbers!

*** Now, MGL went through every STATS zone to determine the percentage of responsibility between the players. Again, I would improve upon his methodology to account for the variables I mentioned in my previous post. I "think" Mike, MGL and I are on the same page on this one, but for some reason, it looks like Mike and I are arguing!
   31. Robert Dudek Posted: March 19, 2002 at 12:26 AM (#604992)

Tell me if you think the following has occurred by chance:

Opponent Batting Average on Balls in Play - (H-HR)/(BFP-HR-W-K-HBP):

1996 .301 .305 -.004
1997 .300 .308 -.008
1998 .298 .315 -.017
1999 .301 .313 -.012
2000 .300 .314 -.012
2001 .294 .312 -.018

That seems to me to be a pretty stable pattern, involving about 4500 balls in play every year.

Is it the pitchers? It could be.

Is it the park? I calculated Arlington's overall park effect (1994-2001) on singles, doubles and triples, to be 1.029, 0.980 and 1.385 respectively. I don't know what that balances out to, but it probably favors base hits a little. What tiny effect The Ballpark has won't amount to much after road games are factored in.

The natural explanation (at least to me) is that Texas fielders have, collectively, lacked range - to the tune of an extra 50-80 hits a year.

If it is luck then I'm amazed, because the numbers are just so darned consistent.

   32. tangotiger Posted: March 20, 2002 at 12:26 AM (#604993)
First of all, let me say that I am one of the guys on Voros' side with respect to DIPS.

That said, I calculate that the simple average of those numbers to be that the AL avg is .299 while the Tex avg is .311. Given 4500x6 balls in play, the stdev of Tex's performance is 4.4. So, it is a truly remarkable performance if it was done by chance alone. If we split the difference, and assume that half the difference is due to fielding performance of the Rangers, then the stdev drops down to 2.2. Therefore, chance is a possibility. All we've done is talk about one team. I would not be surprised to see that almost all teams do NOT follow the pattern that the Rangers have. To do a (more) proper research, we should present all the data.

Furthermore, the pitchers DO have SOME control. To do a fair comparison, you would have to control for those pitchers that have multi-years with the Rangers. In this case, we are probably safe to say that the Rangers turn over their staff quite frequently.

I'm not saying that fielders don't have any control or less control than the pitchers. But before we assert that, let's study it to the best of our abilities.
   33. tangotiger Posted: March 20, 2002 at 12:26 AM (#604994)
Here are my "quick" estimates for $H for 1996-2000. That only 1/3 of the teams fall within 1 stdev shows that we are not talking about random chance.

However, we have to remember:
1 - The pitchers who do have the skill, though over a 5-year period, and with the switching of pitcher, we may "accept" that it all works out in the wash
2 - That teams selectively choose their fielders based on other fielders on the field. If you have a below avg SS, you might be more tempted to choose an above avg 2B, so that your overall defense is "average".
3 - The $H is also dependent on whether you have a GB or FB staff.

Given that, we see that some teams do stand out like a sore thumb, and it may be (mostly) because of their defense.

From this list, we'd have to say that the Reds have had the best overall defense from 1996-2000.
<FONT size="2">
Team Lg  $H  stdev
TEX AL  0.302   3.84 
OAK AL  0.296   1.93 
MIN AL  0.295   1.52 
CLE AL  0.294   1.03 
SEA AL  0.293   0.89 
DET AL  0.291   0.25 
TBA AL  0.291   0.12 
TOR AL  0.290   (0.18)
CAL AL  0.290   (0.09)
KCA AL  0.290   (0.26)
CHA AL  0.289   (0.62)
BOS AL  0.289   (0.65)
BAL AL  0.287   (1.23)
NYA AL  0.284   (2.28)
ANA AL  0.280   (3.13)
MIL AL  0.279   (2.56)
COL NL  0.300   5.07 
PIT NL  0.294   2.76 
HOU NL  0.293   2.51 
ML4 NL  0.289   1.01 
FLO NL  0.289   1.19 
CHN NL  0.287   0.53 
PHI NL  0.287   0.51 
MON NL  0.287   0.48 
SDN NL  0.285   (0.20)
SFN NL  0.285   (0.35)
SLN NL  0.281   (1.39)
ATL NL  0.280   (1.85)
NYN NL  0.280   (1.98)
ARI NL  0.279   (1.70)
LAN NL  0.276   (3.07)
CIN NL  0.274   (3.94)
   34. tangotiger Posted: March 20, 2002 at 12:26 AM (#604995)
Finally, here are the $H (non-HR hits / BIP) from 1969-2000 (only for teams playing since 1969). There's huge swings in there.

<FONT size="2">
Team  $H  stdev
CHN  0.282   7.02 
BOS  0.281   5.09 
TEX  0.280   4.15 
CLE  0.279   4.04 
MIN  0.279   3.77 
PHI  0.277   1.83 
ATL  0.275   0.76 
PIT  0.275   0.18 
CHA  0.274   (0.29)
HOU  0.274   (0.54)
KCA  0.274   (0.76)
SDN  0.273   (0.86)
SFN  0.273   (1.10)
DET  0.273   (1.60)
MIL  0.272   (1.93)
NYN  0.272   (2.19)
SLN  0.272   (2.40)
NYA  0.271   (2.87)
MON  0.271   (3.23)
CIN  0.270   (3.61)
OAK  0.268   (5.47)
LAN  0.268   (5.80)
BAL  0.266   (7.48)
   35. Robert Dudek Posted: March 21, 2002 at 12:26 AM (#604997)

Nice work. I concur with your most recent posts. It would be nice to have the 1996-2001 list park adjusted and see how it comes out.
   36. tangotiger Posted: March 21, 2002 at 12:26 AM (#604998)
Someone also made a comment about "park effects" and how there shouldn't be one (for the infield). There are certain things that cause an effect in the park you play, namely
- the playing surface (grass/turf)
- the playing area (extra foul area, backstop area)
- the temperature (carry of the ball)
- the wind (direction of the ball)
- the elevation (Colorado)
- the sight lines (day/night)

Therefore, to say that there should be no difference in parks for the infield is to say that the sum total of these variables (and I'm sure there are more) is close to zero.

Someone more enterprising that I am can study this issue by looking at "controlled environments" like the Metrodome, or "dynamically controlled" like the Skydome (where the roof may come and go) and start doing an analysis.

You must be Registered and Logged In to post comments.



<< Back to main

BBTF Partner

Dynasty League Baseball

Support BBTF


Thanks to
for his generous support.


You must be logged in to view your Bookmarks.


Page rendered in 0.6038 seconds
40 querie(s) executed