Ballpark Tendencies - have I got them wrong?

[Penfold]

Thanks Ray,

Sometimes I feel like a small child with one of those plastic blocks-and-shapes toys - trying to hammer the square block into the round shape and not understanding why it won't go.

Blimey, coming to BHQ from outside is a tough learning curve - there are plenty of well-established fantasy baseball prejudices to unravel!

[RAY@HQ]

As you acknowledge in your post above, you're extrapolating a conclusion that you're not proving in your data. You seem to be asking the "why" question, as if you can't accept the data without a logical explanation. But there's a ton more that goes into park factors than simply distance of the fences: local weather (temperature, prevailing wind patterns, humidity), elevation, amount of foul territory, height of the mound, quality of the batter's eye/sitelines... I could go on. At some point, the data "is what it is", especially when studied over a number of years. The small-sample exploits of a couple of elite sluggers over a small sample size in one ballpark has a ton of noise in that data.

All that said, your theory that park effects don't apply equally to all players is one that's been floated around here before, and there may well be some truth to that (in fact, it's very likely true, even if we can't quantify it). But even that statement doesn't justify the conclusion you float in your final paragraph. Well, if you've been blanketly ignoring all SD hitters because of PETCO, that's probably not the best approach. But you shouldn't swing in the other direction and just treat all SD hitters just like hitters in COL or ARI or wherever else, either.

[Penfold]

Fair enough, I get that part - but I still don't understand the logic:

There appear to be 12 ballparks with a longer distance to the LF foul-pole, 9 with a longer Left Centre wall, only Fenway Park with a shorter centre-field wall, only 4 with a longer distance to the RF power alley and 4 with a shorter distance to the right foul-pole line.

Clearly the research has been done, clearly the results show that Petco doesn't give up as many home runs as other ballparks and as a result we tend to steer clear of Padres hitters.

Yet just by looking at the distances I don't understand why that is. I'm wondering what would happen if you put a HR king into the Padres line-up: whether he would score significantly fewer HRs or whether there would be a smaller drop off.

So I tried a back-of-the-envelope exercise by taking the three top NL HR hitters from 2011 (Albert Pujols, Matt Kemp and Prince Fielder) and added their three year HR stats in NL ballparks (not including Miami).

What I found was that, over 3 years from 2009 to 2011 inclusive, those three hitters combined hit a HR every 11.5 at bats at Great American, a HR every 11.7 ABs at Turner Field and a HR every 12.8 at Miller Park.

At the other end of the scale, they only managed a HR every 33.2 ABs at AT&T, every 29.5 ABs at Citi Field and 24.6 at Petco. However, they also only managed a HR every 23 ABS at Coors Field and 22.2 at Citizens Bank. Also interesting is that they managed a HR every 18.4 ABs at Dodger Stadium.

All of which proves......nothing much; as many stats don't. It's a selective set of data and given that Prince Fielder didn't score a single Home run in 34 ABs at At&T park in those three years, based on this I wouldn't have been jumping for joy had the Giants managed to sign him as a Free Agent.

Ballparks certainly do have trends, the stats prove it, but that doesn't necessarily mean that the players follow those trends and if three stud-hitters like these can only manage an HR roughly every 6 games at Coors Field, then what do I do with my ballpark factors?

I should mention that I'm not trying to prove a point of my own, or disprove any accepted indicators, just starting to think that I shouldn't pay too much attention to where players will be playing. It seems to me that I shouldn't be dismissing any Padres just because they play half their games at Petco, yet that is exactly what I have been doing.

[RAY@HQ]

Lots to digest here:

- you definitely can't sum the HR tendencies for L/R hitters. Taking the mean is better, although the most accurate result would probably a weighted mean that accounts for the fact that more batters are RH than LH.
- I think you're understanding more than you give yourself credit for: but the idea that somewhere from 35-77% HRs in the majors would get out of PETCO a) sounds reasonable and b) does strongly suggest to me that the park is a hitter's nightmare. I mean, saying PETCO would contain an additional 35-77% of HRs that would get out of all other parks is a pretty darn pitcher-friendly place.
- as far as the idea that the quality of hitters in a park can influence the numbers, the way the calculations are done accounts for that. Basically you take the ratio of HRs hit by the Padres and their opponents in games played at Petco, vs HRs hit by the Padres and their opponents in all other parks. Then you can split that data by L/R batters, etc... but in theory that's an apples-to-apples comparison, and eliminates any bias for quality of hitters.

[Penfold]

I think I may be grossly misunderstanding the Ballpark Tendencies page...

http://www.baseballhq.com/ballpark-tendencies

This came about from a comment by KEP on the SF Giants thread:

Pretty rough to lose at SD on SD HR. Team with fewest HR in NL, and toughest HR ballpark in NL.

It was a thought that struck me as I saw the Mighty Giants give up 3 HRs in a park that, by popular perception, gives up 3HR's a week max.

So trying to establish how this was possible I asked some questions of my friend (Google) and tried to reconcile that with the data I found on the BHQ Ballpark Tendencies page.

Looking at that page, what I believe I see is that Camden Yards (for example) has a 22% advantage for LHB HRs and a 11% advantage for RHB HRs. So surely it has a 33% advantage of all HRs - or should it be the mean: 16.5% advantage?

Anyway, on the basis of my first assumption (33% advantage total HRs) I lobbed the data into a spreadsheet and came up with a best five:

League Ballclub Ballpark LHB HRs RHB HRs TOTAL HRs AL CHW U.S. Cellular Field 22% 38% 60% AL NYY Yankee Stadium 43% 17% 60% AL TEX Rangers Ballpark 31% 23% 54% NL COL Coors Field 29% 21% 50%

and a worst five:

League Ballclub Ballpark LHB HRs RHB HRs TOTAL HRs AL OAK O.co Coliseum -27% -9% -36% NL SD PETCO Park -37% -37% NL STL Busch Stadium -14% -27% -41% AL KC Kauffman Stadium -36% -11% -47% AL MIN Target Field -32% -17% -49%

And there, 3rd worst ballpark for HRs, is the St Louis Cards' home, Busch Stadium.

I've misunderstood the data, haven't I...

However, using a combination of readily available stats and ESPN's Home Run Tracker for distance (http://www.hittrackeronline.com/index.php) I reckon that if I set the distance required to get over the left field wall of PETCO as being 377 feet (distance 367 feet plus, say, 10 foot to clear the wall), then 3,492 of the 4,552 HRs scored in the majors during the 2011 season would have gone over there; 1,607 HRs would have cleared centre field (396 ft + 10ft to clear the wall) and 2,515 would have cleared the right field wall (382 feet + 10ft). The 10ft is arbitrary and is based on the presumption that a 367 ft homer would land exactly at the point where the left-field wall meets the ground and therefore would be an easy () catch for the left-fielder.

Obviously I haven't compared the point each homer crossed the wall with the corresponding point at PETCO, but the statistical suggestion that at least 35% and at most 77% of all home runs scored last season would have cleared the wall at PETCO. So how is it that the park has the reputation of being a hitters horror-show?

The simple answer is that the stats show that fewer HRs are scored at PETCO than MOST other ballparks but could that not be a reflection of the quality of who is standing in the batter's box rather than the park itself? What I'd like to see is the Padres borrow Albert Pujols, CarGo, Adam Dunn (yes, this year!), Masher Stanton, Josh Hamilton and Jose Bautista for a month and see how they get on..

There's probably a simple answer to this that I've missed so I'm all ears as to what it is.