Baseball Articles | Index
Mid-Season Review 2000
By Tom Tippett
I've been a little too busy writing software to do a lot of writing in
the past few months, but as the All-Stars take the field for this year's
mid-season classic, I decided to take a few moments to comment on what
we've seen to this point in the season.
Scoring is up . . . scoring is down again
The media has wasted few opportunities to talk about the surge in homeruns, and there's no denying the fact that we're on pace to shatter a whole bunch of records this season. But a closer look at run production reveals a couple of interesting things.
Last year, teams scored an average of 5.17 runs per game in the AL and 5.00 runs per game in the NL. Through May 21, scoring was up significantly, to 5.43 and 5.39 runs per game in the AL and NL respectively. That's an increase of 5% in the junior circuit and 8% in the senior circuit.
To put this in perspective, the modern (post-1900) record for runs per game is 5.55 for the two leagues combined in 1930. The scoring average was between 5.10 and 5.20 in three other seasons in the '20s and '30s, without the aid of the DH. And there were plenty of seasons in the 1800s when scoring was much higher than it is today. So, in one sense, while we're at the high end of the range for scoring over the past hundred years, we're not yet breaking new ground.
There is one big difference, though. A lot of the runs in the 1800s and the early 1900s came as a result of errors, and error rates have been declining steadily for the entire history of the game. So earned runs are indeed at record levels.
Last year, the teams produced about five runs a game in an environment where the league batting average was .271 and the slugging average was .434. By comparison, in 1891, the league batted only .252 and slugged only .342, but still managed to produce over 5.5 runs per game. The difference? The 1891 teams made an average of 423 errors each (compared with 117 today), and that was enough to push scoring above today's levels even though the league ERA was only 3.34. Today, fewer than 8% of all baserunners reach via error, but that figure was over 20% in the 1800s.
In the weeks since May 21, run production was only a hair above 1999 levels, dropping the overall scoring averages to 5.33 (AL) and 5.21 (NL) through the break. Still up from a year ago, but not nearly as much as before.
I'm not sure what to make of this slowdown. Probably nothing. On the other hand, the Commissioner's office recently announced that scientific research has shown that this year's baseballs are legal but near the upper end of the permissible range for liveliness. Is it possible that some action has been taken to produce balls that are closer to the middle of that range?
If Pythagoras was Commissioner
The Pythagorean Theory, developed by Bill James, states that a team's winning percentage is almost always very close to the square of the team's runs scored divided by the sum of the squares of runs scored and runs allowed.
In other words, there is a strong and quantifiable relationship between runs and wins. If you score 800 runs and allow 700 in a season, you can usually expect to finish with a 92-70 record. Outscore your opponents by 200 runs (900-700) and you can expect to win 101 games. Most division titles are won with 90-something wins, so a team normally needs to outscore its opponents by 120 runs or more to have a legitimate chance to take home a division flag.
But it doesn't always work out that way. Sometimes you win a laugher and then lose a pair of one-run games, dropping a series two games to one despite outscoring your opponent by a large margin. These things tend to even out over a long season, but there are always a few teams each year that defy the odds. Last year, for example, the Royals won ten fewer games than their run margin predicted because their bullpen blew a ton of leads. This season, thanks to a series of amazing come-from-behind victories, those same Royals have won three more games than their run differential would indicate.
So here are the standings as they would be if wins and losses were totally in synch with scoring:
W L Pct GBL
Despite their recent six-week slide, the Red Sox continue to lead the major leagues in ERA (facing the DH and playing in Fenway, no less) and have outscored their opponents by 49 runs. But they haven't been able to translate that performance into wins because they haven't been hitting in the clutch (7-16 in one-run games, 1-6 in extra innings) and they've squandered four good performances by Pedro Martinez.
On the other hand, the Blue Jays sit atop the division despite allowing 21 more runs than they've plated themselves. A Toronto rooter might argue that some of those enemy runs were meaningless, that they simply chose to conserve their good pitchers in games they lost by scores of 17-6, 19-7, 16-3, 18-6, and 15-7. Red Sox and Yankees fans can look at the same information and make a reasonable claim that the Jays don't have enough pitching to hold the lead in the second half. Time will tell.
W L Pct GBL
The AL Central race would be a lot closer were it not for the fact that
the White Sox have won three more games than projected and the Indians
two fewer. But let's not downplay what Chicago has accomplished to this
point. They are second in the majors in run differential (can you guess
who's first?), and if they can maintain this pace, they'll finish the
season with 994 runs scored and very impressive run differential of +171.
W L Pct GBL
The Mariners lead the majors in run margin -- 501 to 382, or +119 --
and could easily be sitting on a big lead if not for a 6-13 record in
one-run games. By the way, the biggest surprise that came out of the computer
simulations we ran back in March was that the Mariners were projected
to lead the league in fewest runs allowed. The reasons: a full season
in a pitcher's park, a much-improved defense, and more depth in the pitching
staff. At the break, they do indeed lead the AL and the majors in this
category, having allowed only 382 total runs and an astonishingly low
15 unearned runs.
W L Pct GBL
The Florida Marlins have provided the second-biggest upside surprise of the season, having finished the first half in third place with a 45-43 record thanks to a 21-12 record in one-run games, a pitching staff that ranks third in the league, and Antonio Alfonseca's 28 saves. But this could be the high-water mark for a team that has been outscored by 27 runs and has a pretty tough schedule over the next month.
W L Pct GBL
Houston has not had a good season by any measure. You've heard the reasons why -- injuries, a 3-19 record in one-run games, plus an ace (Jose Lima) and a closer (Billy Wagner) who were expected to be among the best at their positions but have been hammered night after night. But I think the mood in Houston might be a little different had they come anywhere near the record that their -51 run differential would normally produce. They would still be a very long shot in the wildcard race, but fourth place would feel better than the cellar, and management might be a little less inclined to throw in the towel.
W L Pct GBL
The NL west race looks like it will be a good one, but look how close it would be if every team was matching their Pythagorean projection. Even the last-place Padres would still be within hailing distance of the front-runners.
But the bottom line is that more than half the season is in the books already, and nothing is going to change the fact that Arizona leads the west, Toronto is in first place with 48 wins despite being outscored, and the Astros have the worst record in baseball and no hope of making a serious run at anything this year.
I'm not suggesting that we abandon wins and losses and start handing out playoff berths based on run margins. The whole point of the game is to win, not to amass impressive statistics. But I do think a quick look at runs scored and runs allowed can shed some light on where things are now and where they might end up come September.
The new ballparks
Much of the media's pre-season speculation was devoted to the three new parks (PacBell, Comerica, and Enron) and the first full season in Safeco Field. I was curious to see how closely the reality was matching up with those educated guesses, so we computed park factors for the period through June 30 using play-by-play data from STATS, Inc.
Comerica Field in Detroit was expected to be a tough park for right-handed hitters, and it has turned out to be just that. Righties have homered there at only half the rate in the other stadiums around the league. But it's been quite average for homers by lefties to this point, and the spacious outfield has also produced a significant increase in singles and a small increase in doubles, so the park has not had a drastic effect on overall offense. It appears that most of the blame for the Tigers' last-place standing in team scoring must be placed on the hitters, not the ballpark.
An early offensive explosion earned Enron Field several nicknames, including Ten-Run Field. But things have settled down a bit in recent weeks. Through June 30th, Enron has produced 33% more homers than the average NL park. That's a lot, but still well below the levels we've seen from many of the other homer-friendly parks of days gone by (such as Fulton County Stadium, the Baker Bowl and the Polo Grounds). It does, of course, stand in stark contrast with the Astrodome, which was always one of the worst longball parks in the game. And, like the Astrodome, Enron has also yielded a lot of doubles and triples, especially for right-handed hitters, making it a very good park for hitters.
So far, it appears as if the Giants have an extreme pitcher's park on their hands, despite the short fence in right field. Their team batting average is 22 points lower and they've scored 84 fewer runs at home than on the road. The staff ERA is 3.42 at PacBell and 6.41 everywhere else. And our park factors report shows that it has depressed singles, doubles, triples and homers for both lefties and righties by 8-23%.
Note: This is a good place to point out that these park factors are quite preliminary, mainly because just about every team has played more games in some parks than in others. San Francisco has already visited Coors Field seven times, so the park factors for the full year will almost certainly be less extreme than these figures suggest. By the way, the Giants and Rockies recently completed a home-and-home series in which 85 runs were scored in the four games at Coors and only 25 runs in four games at PacBell.
Finally, the Mariners played half a season in Safeco Field last year, and the early returns indicated that it too was a haven for pitchers. Another half season of data shows that it was no fluke. The park continues to take a big bite out of batting averages for both lefties and righties. Like last year, however, it has given up its fair share of homers. It's not the anti-Coors, but don't be surprised to see a gaggle of free agent pitchers choosing to settle in the Northwest for a few years.