Sports Reference Blog
Why Can WAR Change From One Year to Next?
Posted by sean on October 4, 2012
One of the unsettling things for fans who are trying to get a grasp on Wins Above Replacement is that the numbers can and have changed over time. Setting aside the 1000's of historical errors in the baseball record book that are being searched out one-by-one, Miguel Cabrera's batting average is and will always be .330. His WAR (for baseball-reference.com) of 6.9 Wins however may change slightly over time.
Why is this? While we want to think of WAR as a statistic like batting average, it is an estimate rather than a precise measurement. Lots of factors are put used in estimating this value and sometimes they change as better estimates become available either due to more advanced research or new data. Now obviously we think it is a pretty good estimate (at least as good as any other measure of player value). Each step has been rigorously researched and justified and is available for you to review and poke holes in (see the link below).
I think an interesting parallel is stock valuation. The techniques used to value stocks in 1980 or even 2005 are different from the techniques used to estimate the value of stocks in 2012. More data is available now, new computing techniques and even newly discovered mathematics. If we were to go back and apply 2012 techniques on 1980 stocks we would have different valuations using 2012 techniques than what we had in 1980 using 1980 techniques, and we'd probably be a lot more accurate.
Another example would be estimates for the size of the earth. This number has been refined and improved over 1000's of years as new techniques for making this estimate have become available. But even now this size is not a direct measurement (there is no measuring tape or scale big enough), but an estimate.
Now the difference is that probably few people care about that difference when it comes to stocks or the size of the earth, but we continue to be fascinated by past baseball seasons. So when we go to a great deal of study to estimate the effect of not having batter strikeout data on the value of outs in early 1900's baseball or the value of an infield single or an IBB in 2005, that affects our view of how valuable that player was in that season. At the time in 2005, we didn't consider IBB's as different from non-IBB's and we didn't differentiate between infield and outfield singles. Now we know the value of those differences and we apply it to our understanding of 2005, 1955 and 1905 baseball.
Previous to this season, we made several large changes to how we calculate player value. They are all listed below in the link, but the major one is the use of Baseball Info Solutions Defensive Runs Saved. Switching from Total Zone to DRS for 2000-present caused some very large swings in defensive value.
Defense is hard to measure as there are dozens of factors that go into its measurement, but we feel confident that their system is the best. They also are continually trying to improve the system. For instance, this past offseason they added batted ball timer data to refine their estimates of player defense. That means every ball in play for a substantial number of years was reviewed and timed (by hand). This then changed the defensive estimates for nearly every player. The stat got better as newer techniques and more data was applied to the question.
And even then if you think all defensive measure are bunk, use oWAR. It is every part of WAR, but assumes everyone is an average defender.
And if you think replacement level is bunk use WAA or wins above average. For single season measures like MVP races it works just as well as WAR. For careers, you'll probably undervalue average players with long careers.
Now could Cabrera and Trout's numbers for 2012 change next year? Yes, park factors are one factor in how batting is considered and we use 3-year park factors, so ideally the 2012 park factor includes 2011, 2012, and 2013, so if Comerica or Anaheim play much differently next year that could cause a change (albeit small--like half a win at most extreme) in their WAR totals.
Let me say one other thing, because of the fuzziness, I would never look at a WAR of 7.6 and a WAR of 6.5 and say the first player is "clearly better than the second". I would say that the first player is "probably or likely better than the second". However in the AL MVP race we have a 4 win difference which as far as WAR goes is huge, so in my opinion (and yours may be different) Trout was clearly a more valuable player than Cabrera this year. And, of course, if playoff appearances to leaderboard troikas are super important for you and overrides whatever else happened in the regular season then WAR isn't really applicable.
[...] [...]
what if changes to ballparks are made such as Safeco is planning in 2013?
Cabrerra had less than 2% of his plate appearances at Safeco in 2012. Trout had over 7% of his plat appearances at Safeco.
Now if Safeco is more hitter friendly in 2013, as I understand it Trout's 2012 WAR will go down slightly. But should changes in the ballpark in 2013 have anything to do with 2012 WAR?
If a parks changes significantly we will probably treat it as a new park.
http://seattle.mariners.mlb.com/news/article.jsp?ymd=20121002&content_id=39364000&vkey=news_sea&c_id=sea
are these considered significant changes?
I'll have to consider it more carefully.
This, and your follow-up piece about WAA, are excellent. I think there are two major problems people have with WAR in general:
1) They don't understand it (and neither do I), and people fear what they don't understand. However I don't NEED to know the calculations. WAR tables almost always pass the "smell" test; for exampe the NL WAR table is Posey, McCutchen, Braun, Wright, and Molina right now. I don't think it's a stretch to say those are the five most reasonable NL MVP candidates.
2) A lot of times on the SR/B-Ref site, WAR gets cited as a be-all, end-all stat, and I know that's not the intention but I think people misinterpret the point when it may be the only "stat" in a particular post or comment.
Anyway thanks for the great site and content this season, again.
Here is a ? then on WAR. From what I read the WAR crowd says Trout deserves the MVP over Cabrera. But since WAR can change over time would it be fair to vote for Trout based on his WAR when 20 years from now a new WAR formula may say Cabrera had a better WAR with the new calculations? BS HR RBI with minor exceptions do not change but can you base an MVP vote on a stats that can have a big swing as new calculations are determined?
Please note in the line above
BS HR RBI with minor exceptions do not change
that BS should have been BA :-) LOL
Well, AVG and RBI are horrible ways to judge a player's overall value. WAR, whether it changes or not, at least attempts to encompass all facets of a player's game.
You admit that WAR and all the other new stats are fuzzy. And you say that a difference of one run may not make a difference. However, you then say that Mike Trout is "clearly a more valuable player than Cabrera this year."
Will you then take Trout's numbers, compare them to Mantle's in 1956/1957 or Mays' in 1964/1965 (all between 10.7 and 11.1) and say that Mantle's 1956 or Mays' 1965 was NOT clearly better than Trout's this year? Because if that is what you are saying, this does not pass the smell test.
Mantle, Robinson, and Yaz led their teams to pennants, and with the exception of one voter in 1967, won unanimous MVP awards. MVP means "most valuable." It does not mean someone has a higher set of spurious statistics. It means that someone took his team where it would otherwise not have gone. Cabrera has the record to be a unanimous MVP choice.
not sure I understand that last rant. Cabrera is going to win MVP. There's no doubt of that. I mean... I've seen a guy win MVP with a 3.7 WAR on a last place team.
But I doubt Cabrera's win will be unanimous.
Alanruns, the point of statistics is that you don't need the smell test if the stats can tell you what's actually going on. Besides, I don't even agree with your smell test (another reason they're iffy) and can't for the life of me figure out what in the stat line is causing you to say that. I don't think it's outlandish to say that Trout's 2012 featured the best fielding and base running, and the hitting seems at least close enough that such edges should put Trout's season in the conversation with the seasons you mentioned.
Sure, Trout's career shouldn't be compared to Mick's or Willie's yet, but individual seasons are fair game... heck, it wouldn't be fun to keep the records if we treated them as sacrosanct and refused to honestly compare them to modern numbers.
The problem with relying on any of these statistics is that they are smoothed out over an entire season. Mike Trout batted 392 in July but only 257 in September. Cabrera's worst month was April (298) and his best was August (357). The problem with Trout is, day in and day out, you don't get a guy who batted 325 with 30 homers, because you might get the 392 hitter and you might get the 257 hitter. But with Cabrera, you pretty much get the same guy every game. The bottom line is that Cabrera doesn't disappear and his season statistics aren't hiding ups and downs. And that's why he's more valuable than Trout. Incidentally, there's a linear relationship between Cabrera's OPS and the Tigers' winning percentage. As Cabrera goes, so goes the Tigers.. But the Angels were only 14-11 when Trout had his monster July and went 18-9 when he batted 257 in September. Still think Trout's more valuable that Cabrera?
Steady Eddie or Streaky Sam for players could mean nothing in the final tally if they finish with the same numbers. The same thing for teams.
I could have this amazingly consistent team... they hit a double every inning and never strike out or walk. They have .500 slugging. Unfortunatly they finish with 0 runs scored on the season.
I being rediculous but what I'm saying it it's the final result matters. The last post convinced me of nothing.
Trout was phenomenal while his team did average and stunk while his team did well. So he had good stats at the end of the season, once averaged out. The better Cabrera was, the better the Tigers were. We're not talking hypothetical players hitting doubles here - this is what actually happened this season. And the MVP is about who brought more value to his team.
I hope Doug B is right and that Cabrera wins the MVP. But that's not what the guys on TV are saying.
Regarding Seamus: I know that no one is comparing Trout to Mays or Mantle over a full career. But the numbers compare him to Mantle in 1956 or Mays in 1965. It doesn't matter how many bases Trout advanced runners on outs (one of the crazy stats I saw in the definitions of WAR); Mantle and Mays advanced plenty of runners with homers and shots up the alley.
Don't forget that they were absolutely sweet center fielders.
And finally, they inspired their teams in a way that no stats could ever show. Just look at a picture of Mantle's bloody abscessed hip in the 1961 World Series, or the layers of tape he put on every day.
Inspired their teams in a way stats could never show? What does that mean? Was he giving speeches after every home run? I'm a layman so please help me understand?
That's why I prefer cumulative stats to averages. Averages just say you have the potential to do something if you're ratio of doing it in the past stays the same over a certain number of at bats or whatever, whereas cumulative stats (hits, runs, RBI, wins, etc) say that you've actually done something.
Lately I've heard a lot that RBI's shouldn't carry that much weight in WAR because they are "plays of opportunity" and not every player see a similar amount of ABs with runners on.
But then wouldn't that mean the same for the "runs grounded into double plays" part of WAR? A cleanup hitter (like Cabrera) is much more likely to see players with runners on than a lead off guy (like Trout). In fact, Cabrera hit .340 with runners on in 291 at bats, while Trout hit .306 with runners on in 180 at bats... but Cabrera has a Rdp of -5 and Trout has a Rdp of +1!
The entire park factors thing defies logic. Angels stadium was the 17th easiest park to hit homers in for 2008, magically became the 2nd easiest in 2009, and went back to 23rd in 2010. How is that possible - the park didn't change. And supposedly it's more of a pitcher's park than Comerica, which is remarkable because Comerica is 10% bigger and almost 20 degrees colder in April and May. That's because it relies on factors that have nothing to do with the park itself. It's a ridiculous calculation and underscores that crafting a formula does not automatically mean that it represents what you want it to. The sabermatricians need to lift their head up and start introducing some real world into these stats.
An interesting stat that you can compute from info already provided on Baseball Reference is RBI percentage, for lack of a better name. That is, you add up the plate appearances and the number of baserunners during those plate appearances, and then just divide RBI by that number. In other words, the actual RBI divided by the number he would have had if he hit a homer in every single plate appearance. It's a cool stat. Addresses the whole complaint that some guys have more RBI opportunities. Although some guys could have more runners on third base than others, but nothing's perfect.
They have something similar to an rbi % already calculated in situational hitting. They list baserunners when at the plate and the number who scored, although not necessarily by RBI. The "BRS%" is the percentage of all baserunners who scored on the batter's play.
Unless one is Lawrence Olivier reciting the St. Crispiens Day speech in a way that it never really happened, speeches do not inspire; we lead and inspire by our actions. That's what Mantle did.
More importantly, I am amazed by the growth of the new statistics; RBI percentage sounds like one of them. You bat a guy like Cabrera third because you want him up with more runners on base; he drives the ball, so you hit him where he can do that. And yes, the pitchers pitch differently to him when he has two on and none out than you do to a leadoff hitter.
Did you ever wonder why Willie and Mickey hit third and fourth, and Ricky Henderson (also with power, and the owner of the most leadoff home runs) batted leadoff? Mantle once said "if I knew that 30-30 (HRs and steals) would be that important, I'd have done that every year." Mantle and Mays were most valuable driving the ball. In the history of the game after 1920, power hitters have been your most valuable. So let's not get to new stats that ignore what managers and GMs see with their own eyes.
#19 Sean adjusts for the frequency that the DP is in order.
Cabrera grounded into a DP in 19% of the plate appearances where the DP was in order, Trout only 8%.
#20 The method for calculating park factors is not new. First published in the 1980s with only minor updates since then.
And it relies on only one factor. The ratio of runs scored and allowed in home games compared to road games (inter-league games excluded from both)
#21 If you want to adjust for opportunity, Tom Ruane has a really nice paper on the subject at:
http://www.retrosheet.org/Research/RuaneT/rbipro_art.htm
(Any of Tom's papers at retrosheet are well worth the read)
#23 It's worth noting that Henderson added the power after he'd established himself as a leadoff hitter.
What's really remarkable is that Steve Boros announced that he'd have Henderson try to hit more HR and he tacked the HR power on without missing a beat.
#25 - re park factors. Thanks for the explanation and that's actually what I thought it was. But to me it's simply falsely attempting to tie a coincidental occurrence to a cause and effect relationship. Unless the park itself changes, the only other relevant factor is weather.
The fact of the matter is that there are only 3 legitimate factors (that I can think of) that influence the actual ability to hit in a certain park: field conditions (like grass length), the distance of the walls, and weather. Over the course of 80 games, year over year, you would expect those factors to have very little fluctuation.
More over, what this doesn't account for is a couple of things like some players simply play better at home and also some teams might do a better job of accumulating players that "fit" their park better than others. So, as varying teams' rosters change, so can these results.
The bottom line is that it defies logic that a specific park can fluctuate so wildly from one season to the next. I personally think attempting to do this is more harm than good. Just because we came up with a formula and we can plug numbers into it, really doesn't mean that it represents what we want it to in the real world.
"More over, what this doesn't account for is a couple of things like some players simply play better at home and also some teams might do a better job of accumulating players that "fit" their park better than others. "
Sure. This is why it's important to understand that park factors are a value in context adjustment and not an ability adjustment (though the two are tightly related)
A good example of this is Joe DiMaggio/Yankee Stadium.
In the time DiMaggio played for the Yankees it played as basically a neutral park.
(That is there were just a few more runs scored and allowed in Yankee road games than there were in Yankee home games)
And yet for his career DiMaggio hit a fair amount better on the road (.315/.388/.546 at home and .333/.401/.610 on the road. -- the numbers are slightly different on bb-ref, these numbers are Pete Palmer's "complete". BB-Ref seemingly doesn't have all of his PAs. No biggie, the two sets of numbers are close and the overall point remains the same)
The reason is pretty obvious. Death Valley takes home runs away from right-handed pull hitters (Joe Gordon for instance had similar splits)
In terms of value we don't care that DiMaggio wasn't particularly well suited for his park (I mean you can live with a gold glove CF with a 934 OPS. That's a heck of a player) The offensive context he played in is one where a league average position player could be expected to his .276/.355/.405 and that's what he's adjusted against.
Similarly, Sandy Koufax (career era of 1.38 in Dodger Stadium) was particularly well suited for Dodger Stadium. He'd probably have been a tad less valuable if he'd pitched in a different place, but that's not the question that WAR (or OPS+ or win shares or any other park adjusted value metric) is seeking to address.
Projection systems (generally) break things down and attempt to assess how a different park will affect a player. Generally with less than total success. I know Dan Szymborski has done a ton of work on the subject with relatively small improvement in his projections.
Not quite sure where you're going but my point is that you can't assess the difficulty of a hitting in a park by using a ratio of runs scored at home vs on the road. Using "park factors" to say one player has a better "adjusted OPS" or even WAR is using a variable that is dependent upon things that have nothing to do with the park itself. Just because that ratio fluctuates doesn't mean it's suddenly easier or harder for hitters to hit there. This model fails the reality test because it tries to make factors of things that are ancillary to the park itself. Simply put - Comerica park is one of the biggest and coldest parks in the AL, yet it is considered a "hitters' park" based on "park factors". That simply doesn't mesh with reality. The fact that ratio and park factors themselves fluctuate as much as they do should tell you that the model does not work - there simply isn't that much actual variability in the real world from year to year in these stadiums.
I think it would be a more legitimate model to assign values to field dimensions, wall heights, and weather - however arbitrary - and use them across the board. This would be much more meaningful and would translate across eras as well. You might need some tweaking (moving from linear to parabolic) to account for particular anomalies like the Polo Grounds or the Green Monster, but at least you'd have something that actually translated the physical world into a factor that could be applied.
#30 "The fact that ratio and park factors themselves fluctuate as much as they do should tell you that the model does not work - there simply isn't that much actual variability in the real world from year to year in these stadiums."
It tells us that changes in park factors from year to year are mostly random noise (Eric Walker in one of the BBBAs did a good study on this. Harold Brooks did some very good work on this at rec.sport.baseball. Good luck finding it though with the new google groups) which is what convinced most people (Nelson Lu being a notable exception) to use multi-year park factors.
We do know that certain that Wrigley field plays as 3 very different parks depending on the prevailing wind. One or two more days with the wind blowing out will make a big difference. Other parks haven't been as well studied but it's totally plausible that other parks have the same thing going on.
And if you want to argue that a Cubs pitcher with bad luck with the weather can end up with a park factor that doesn't reflect his reality, you'd be correct. Tom Ruane has done some initial work on individual park factors, but it's not something that anybody's very interested in.
Ron Johnson -
You are clearly quite knowledgeable on the history and depth of this and I really appreciate the education!
I like the fact that people are recognizing that the numbers shouldn't be changing that much. I'm just not convinced that the existing model, while it may be internally consistent, is effectively modeling reality. Some of the "rankings" (is there a list of them, by the way??) don't seem to pass the "smell" test, if you will.
#32 One of the things that you can do to get a general notion of the uncertainty inherent in park adjustments is to recalculate the offensive stat (say OPS+, but this applies to any park adjusted stat) with single year park factors. The "truth" lies somewhere between the two results (likely to be closest to the number based on the 3 year factor)
Most people have a very tough time accepting uncertainty, but no offensive metric is any more precise than +/- 5 runs (for a full time player).
Incidentally it's not quite accurate to say that park numbers shouldn't be changing that much. Harold's work showed that ~70% of the changes were probably just noise. Sometimes something very real happened and we've got a very imperfect understanding of the factors that influence this.
I mean we know that if you bring the fences in or raise/lower the fences offense will rise. But we can only make a WAG in advance as to how much any given change will affect things.
But I'll predict in advance that the Angels will have a slightly lower park factor next year -- because of the changes in Seattle.
Amazing: I leave the field of play for a few days, and I see more statistics being thrown around. Back to Mark Twain: "there are lies, damn lies, and there are statistics." But that's for another day.
Park adjustments: so why in Maris' 1961 did he hit 30 homers on the road and 31 at home (or vice versa, I forget!), when he had the short porch in right field? And if it was so important, why didn't they trade DiMaggio for Williams in about 1949 or so, when the two owners had a bit too much to drink and had a hand shake agreement? And finally, why was Wally Moon the only guy to really solve the short porch in the LA Coliseum in 1958?
And in terms of changing the WAR stats: did anyone notice that with the change Mantle lost many wins or runs, and is now ranked below Rickey Henderson? How bogus is that? Change the way some meaningless stat is computed and one of the top ten players of all time is rated below a pretty good player who couldn't carry his jock. Amazing.
Try this stat on for size: RBIs + runs scored - Home runs. That's all you need, that's all you ever needed. If you want more, OPS (which we always thought was important even when no one computed it) is helpful.
OOPS! Wally Moon was 1959-61 in the Coliseum. I just looked it up. Those 3 years he had 37 homers at home, and 12 on the road. For his final 4 years after they moved to Dodger Stadium in 1962, he had 4 at Dodger Stadium, 4 in NY against the hapless Mets (3 off Tracy Stallard--quick, what's he famous for--and 3 in the Polo Grounds), and 7 in other road games. Also, the first Frank Thomas, with the Mets in 1962-1963, did a great job with home parks.
Alanruns - quite funny input. You have just exhibited evidence that these things don't seem to pass the "smell" test. Like I said above, trying to model something, even if it's internally consistent, doesn't mean you actually modeled it.
#35 Not sure I understand your question about Maris, 1961. Park factor (the ones we use to adjust things like batting runs) don't care about how easy it is to hit HR in a given park, but rather how run scoring is affected. Yankee Stadium actually played as a pitcher's park in 1961. It played as a slightly better hitter's park in 1961 than it did in 1960, but it stands to reason that park factors would change in an expansion year.
When it comes specifically to home runs and Yankee Stadium, it was never a great HR park for lefties, just a really tough HR park for right-handed pull hitters.
(RBIs + runs scored - Home runs).
Dumbest (frequently posted) stat ever.
Runs and RBI are half runs (yes, there are a few runs scored with no RBI. No many)
Think of it this way: Home run followed by single. Net one (RS+RBI-HR)
Triple followed by Single. Net two (RS+RBI-HR)
As to why we don't use RS and RBI in evaluating players. RBI are a function of power (most simply represented by SLG) and opportunity (most simply represented by AB with runners on base). Since a player doesn't control opportunity, why use a measure that is so profoundly dependent on it. (And I've checked. The correlation between AB with runners on base * SLG is just over 98% at the career level)
RS is profoundly affected by your lineup position and how good the guys behind you are.
The earliest work by what we'd now call sabrmetricicans was by Steve Mann (working for the Phillies) about how counter stats and batting order affect runs and rbi.
As to the specific point about WAR and Mantle/Henderson. First of all, a difference of 1.3 wins is not significant at the career level. The standard error for a career as long as Henderson's is in the range of 4 wins.
Second, Henderson got those extra 1.4 wins in 541 extra games. In other words, the extra value lies in simply showing up (specifically, he has 413 runs above replacement and Mantle has 271)
So: Mantle significantly the greater player (comparing how good they were at their respective best and how long they sustained that level) while you just might be better off having Henderson's entire career rather than Mantle + 541 games of a replacement level player (but it's close here).
Not that coming below Henderson in a career ranking is an insult. At his best a sensational player, and a very good player for a real long time.
Also, here's a useful way to understand park effects. Bill Terry and Mel Ott played for the Giants in more or less the same time frame. Both were good left-handed hitters
Ott's the better hitter (career OPS+ of 155) but has a huge home/road HR split (323 at home vs 188 on the road) leading to career splits of .297/.422/.558 at home and .311/.408/.510 on the road,
Terry on the other hand (career OPS+ of 136) hit 76 HR at home and 78 HR on the road and hit .330/.382/.487 in his career at home and .352/.403/.525 on the road.
In other word, the Polo Grounds suited Ott's skills (he and Babe Ruth were the only two hitters who really mastered hitting the fairly cheap HRs available right down the line. Most major leaguers just couldn't do this on a consistent basis -- Ruth's career SLG in the Polo Grounds was .840) and hurt Terry to a fairly unusual degree.
Doesn't matter. In years where they were teammates they get the same park factor.
Pardon my douchey self promotion but I recently had the chance to interview Sean, he talked about how WAR is calculated for 45 minutes. He hit on things like defensive shifts, park factors, and the differences between the WAR offered here and the one used by FanGraphs. He also discussed some of the reasons why WAR can change slightly from year-to-year.
Sean was very good/honest and I think that the people on this board might enjoy the interview.
It's on my site and on iTunes. (Replacement Level Podcast) Again apologies for the shameless self promoting.
Ross
I'm sorry folks - the park factor thing is just plain irrational to me. I've read and listened to everything I could on it - and it still makes no realistic sense to me.
Let me try one example here - and I'll use made up numbers to illustrate the point. Let's say for the sake of argument that Comerica Park has a park factor of 0.9, meaning it's a "pitcher's park" (not sure if this is the right magnitude, but these are made up anyway) and Yankee Stadium is a 1.1 (hitter's park). Kansas City decides to move the walls in to 300 feet down the lines and 330 in center and softball scores abound. KC's park factor is 3.0 now but...relatively speaking, it just became harder to hit in both Comerica and Yankee Stadium, who get moved down to 0.8 and 1.0 respectively. The first question point is that, if I understand this correctly, Yankee hitters, who will play 3 whole games in KC, just got their OPS+ and everything else bumped up by the 10% decrease in park factors, right? And the same for Detroit hitters, who will play 9 games in the KC park, right? As a matter of fact, a batter could have identical stats from the year previous and let's just say they missed the KC series in each year. But in year 2, when KC is tiny, they are considered a worse hitter than in year 1.
Do I have this right? If so, it's irrational. Park factors are not "relative" to a player. If they are not being applied on a usage basis (KC factor applied to stats gathered in KC, Yankee Stadium factor to games in NY, etc), then they shouldn't be used in the manner they are used and with such importance. Please - educate me.
#42 That's why I say that park factors aren't an ability adjustment. They're an attempt to put the value of a player's contribution (offensive or defensive) into context.
Parks most definitely affect different players differently. We simply don't care.
As for your example, even a massive change to KC is unlikely to move the needle a great deal. But if you're interested in this, take a look at Detroit in the mid to late 30s (where they made frequent, massive changes). In one year it moved from a PF of 95 to 110. But then they did bring the RF fence in 60 feet. This figures to matter and it did.
Since the parks have to add up to 800 (8 parks, 100 average by definition), the other 7 parks would see their park factors go up by 2.
These days no single change is likely to have even that much of an impact.
Yeah scoring would be up across the league (this would impact the runs per win calc), but the park factors would move down about 1 point for every 13 added to KC's (has to. again, there are 1400 park factor points to spread across 14 teams -- and before raising the wrath of Sean, it's a little more complicated than that. There is a pitching factor and a batting factor -- to adjust for the fact that pitchers don't face their own hitters and vice-versa)
I will note that a massive change of that nature would make anybody question the validity of multi-year park factors. (I don't think multi-year park factors make sense for the Tigers of 1937 as a for instance)
Okay...you say they're not an ability adjustment, but that's exactly how they are used. I've seen this time and again used to "validate" Mike Trout by saying even though Miguel Cabrera led the majors in OPS, Trout led in OPS+. That's an ability adjustment.
The thing is, when I read the above, I immediately read something to the effect of "Because Angels Stadium is much harder to hit in than Comerica Park." And there's where it all falls apart.
Here's the bottom line...Comerica is bigger and colder than Angels stadium. It's 48 on average in April in Detroit and 64 in September. It's 62 and 73 in Anaheim, respectively. And Comerica has deeper dimensions and is about 10% larger. The fact that your formula tells you that Angel's stadium is (kind of dramatically) more of a pitcher's park than Comerica defies logic.
Give me a reason why Angels Stadium is harder to hit in that trumps the two things I mentioned above. And don't use your formula because your formula is attempting to create a model out of environmental observations that don't necessarily have any correlation. To me, it's like saying that Comerica is more of a pitcher's park because more guys named Bob bought tickets there.