Park factors, sample size, and more..

This is another area I wanted to touch on, sort of as a follow up to my “Prospect v Non-Prospect” post a few days back. As I mentioned in that discussion, when looking at a minor league player’s stats, you have to really look at things through a number of different lenses. A .260 batting average tells you relatively nothing if you don’t know how old the player is, what league he plays in, what position he plays, and what his home park is like. Not all .260 averages are made equal, not all 15 HR seasons are equal, and not all 3.00 ERA’s are equal. Understanding what goes into figuring this out will tell you a lot about the minor leagues in general.

The first thing to consider is the park factor for the team the player plays for, and then look at the park factors for the other teams in the league.  Some leagues are generally thought of as hitter friendly (the CAL League), while others are generally viewed as being pitcher friendly (the Midwest League), and some are fairly neutral, to an extent (the SAL). To keep this relative, we’ll look at the Phillies minor league affiliates and how their home parks and their league affects their numbers. The only thing you need to know to understand what these numbers mean is that “1.00” is neutral, anything less than 1.00 favors pitchers, and anything above 1.00 favors hitters.

Lakewood: 

Lakewood plays in one of the better pitcher’s parks in the SAL, and in terms of cutting down on home runs, one of the best minor league parks in all of baseball. Here are the park factors, weighted over three seasons from 2004 to 2006 for Lakewood

Runs = 0.86
Hits = 0.95
Doubles = 1.08
Home Runs = 0.51

As you can see, it’s nearly twice as hard to hit a home run at Lakewood than your average minor league park. Run scoring is down, and the park is above average in terms of doubles allowed, but that’s partly due to the extreme difficulty to hit home runs. These numbers would tend to indicate that pitchers numbers probably look a bit better than they are, and hitters numbers might look a bit worse than they really are. The league average for the SAL are

Runs = 1.00
Hits = 1.00
Doubles = 1.02
Home Runs = 0.97

If you remove Lakewood and get an average for all the other teams in the SAL, the factors look like this

Runs = 1.00
Hits = 1.00
Doubles = 1.01
Home Runs = 1.00

It may not seem like a major difference, but it is easier to hit away from Lakewood than it is to hit there for half the season, and that will negatively impact hitters and positively impact pitchers.

Clearwater: 

Clearwater, on the other hand, is more of a hitter friendly park. The park factors are

Runs = 1.03
Hits = 1.02
Doubles = 1.10
Home Runs = 1.01

The League average park factor is

Runs = 1.00
Hits = 1.00
Doubles = 1.01
Home Runs = 1.02

If you remove Clearwater, the averages are

Runs = 1.00
Hits = 1.00
Doubles = 1.00
Home Runs = 1.02

So, Threshers hitters have an advantage when it comes to driving the ball into gaps for doubles, and they experience roughly the same home run advantages as other teams.

Reading:

Reading is a somewhat neutral overall park, but it’s the opposite of Clearwater, in that it’s much easier to hit home runs than it is to hit doubles. Reading’s park factor

Runs = 1.03
Hits = 0.98
Doubles = 0.85
Home Runs = 1.33

Eastern League averages

Runs = 1.00
Hits = 1.00
Doubles = 1.00
Home Runs = 1.02

If you remove Reading, their away park factor is

Runs = 1.00
Hits = 1.00
Doubles = 1.02
Home Runs = 1.00

Based on that, you could say Reading is a decent hitters park. Home runs are much easier to come by, and that factor outweighs the suppression of doubles.

Ottawa: 

Ottawa plays in a fairly neutral park as well

Runs = 1.03
Hits = 1.01
Doubles = 1.00
Home Runs = 0.90

The International League average

Runs = 1.00
Hits = 1.00
Doubles = 1.00
Home Runs = 1.01

Remove Ottawa, and their away league factor is

Runs = 0.99
Hits = 1.00
Doubles = 1.00
Home Runs = 1.01

Fairly average numbers, 10% tougher to hit home runs, but no other area is really suppressed moreso than league average.

So, to summarize. When looking at a player’s slugging %, you want to understand how much easier it is for that player to put up numbers in his home park. If you look at a guy like Lastings Milledge of the Mets, you can get a really good idea of why park factors are so important. In 2006, he played at AAA Norfolk and had a slugging % of .440. Norfolk’s home park has a Doubles rating of 0.95 and a HR rating of 0.66, meaning it was generally 39% more difficult for Milledge to rack up extra base hits in his home games than it was for a player playing in a neutral park. His translated slugging %, which accounts for him playing half his games at home and half at the rest of the parks in the International League was .537. His OB% was solid at .388, and couple that with a .537 slugging % at age 21 in AAA, and people are raving about him being one of the best prospects in the minors (even moreso than they already did), but instead his 2006 was labeled a “disappointment” by some.

One more example, this time highlighting just the opposite affect. Carlos Gonzalez, a 20 year old OF prospect in the Diamondbacks organization, posted a .300/.356/.563 season at Lancaster, their high A affiliate in the California League. Here are Lancaster’s park ratings compared to the league average, which is the number in parenthesis.

Runs = 1.22 (1.00)
Hits = 1.15 (1.00)
Doubles = 1.02 (1.01)
Home Runs = 1.60 (1.04)

So, just at first glance, you can see he was basically playing in a hitter’s paradise. It was almost 60% easier to rack up extra base hits in Lancaster, including a whopping 56% advantage on home runs, than it was the rest of the Cal League. When you plug in Gonzalez’s numbers, his translated Slugging % goes from .563 to .490….a HUGE difference. When Gonzalez was promoted to AA Tennessee in the Southern League, he dropped off to a .213/.294/.410 line. Tennessee is still a hitter’s park, with about a +.18 factor for doubles and home runs, but nowhere near the hitting environment found in the CAL League.

If you’d like to see park factors for every team in the minors at short season ball or higher, check here.

The other area I briefly wanted to touch on was sample size. Trying to predict the performance of a player in a season is really difficult to do with any kind of real accuracy. There are so many factors at play, so many wildcards and variables, that getting it really close isn’t very easy. It makes it even more difficult when you are only looking at a very small sample of at bats. The average minor league prospect will log anywhere from 600-1500 at bats in the minors before making his way to the major leagues. The cream of the crop might only spend 1 full season in the minors, while other guys may take 4-5 years before they are ready. When a player has a longer track record, two or three seasons, it is much easier to evaluate his ability and guess where he’s going from here. When a player only has 150 AB’s to go on, you’re certainty level is going to be a lot lower.

Similarly, it’s tough to really draw any accurate conclusions based on one month’s worth of action. You can see trends developing, but one month’s worth of data is worth half as valuable as two months of data, one third as valuable as three months of data, etc etc. We can look at Mike Costanzo’s walk rate and cringe after his start in Reading, but at the end of last season, his rates looked better after a slow start. At the same time, we can see Greg Golson’s .308 average after 1 month and start to get excited, but we have to remember the 1,000 AB’s he’s logged prior to this season and put a bit more weight on that performance than the last 80 AB’s or so he’s had. Generally speaking, what a player did 3 years ago is less relevant than what he did 3 days ago, but it is still important, and finding the balance there is the key to figuring out where a guy will go from here.

It’s fun to get excited over hot starts, it’s depressing to get down over cold starts, but it doesn’t tell us an awful lot, especially when a guy doesn’t have much of a track record. Justin Upton, who signed for one of the largest bonuses in MLB Draft history, struggled last year as a 19 year old in the Midwest League. People were already starting to ask if he was overrated, or if he wasn’t as good as people thought. Well, he’s tearing the cover off the ball this season, and now people are back on board. My advice on him, like my advice on most minor league guys, is to look at the big picture, then look at the smaller sample, and try to weigh them equally, temper expectations, and see how things pan out.

1 thought on “Park factors, sample size, and more..

Comments are closed.