Alivox on Baseball

One of the joys of being a baseball fan is second-guessing the managers: he should have gone to the bullpen one batter earlier (always), he should have pinch-hit for that bum, he should have called for the steal/bunt/hit-and-run.

This page is a slightly more abstract variant: second-guessing the commissioner and the MLB front office. The good part is you're much more likely to be right; the bad part is you're much less likely to see your wishes granted. But that doesn't detract from the pleasure!

Automating the Strike Zone
Revisiting Foul Balls
Starting Relievers
The Designated Hitter
The Season
The Net Statistic
Text Scoring
Sample Game

Automating the Strike Zone top

Illustration by Scott Mattern

Baseball and football are unique among sports in that almost everything has to be called. In other sports, play continues without interference from the officials until a rule is violated, but baseball umpires rule on every play, even on every pitch, whether or not it's obvious.

In the course of this, nobody works harder than the home plate umpire, who must call every ball, called strike, swung strike, foul tip, hit by pitch, or dropped third strike (he has help for foul balls). A pitched ball moves at around 100 mph, and the strike zone changes for every batter. It's an almost impossible task to perform consistently, and he has to do it from a crouch, wearing a mask and other protective equipment which interferes with vision, mobility, and comfort. And yet it's a crucial task, and his mistakes make a big difference in a game.

There is a new technology called QuesTec which uses multiple cameras to reconstruct the path of a pitched ball. This technology was developed for television, but is also reportedly being used by MLB management to evaluate the performance of umpires. However, it's easy to imagine using it to make the primary call, with the ump reserving the right to overrule - apparently, it has its technical imperfections.

But it wouldn't be hard to install an automated strike zone under home plate itself. Of course, this wouldn't tell us what QuesTec now does about the trajectory and speed of the pitch, but that information isn't relevant to the umpire's job, anyway. All we need to know is the horizontal and vertical position of the ball as it crosses the plane of the front of the plate.

The technical challenges are not trivial, but they are eminently solvable given the importance of the task (and the money available to solve it). The key to the success of such a system would be a few slight changes in the rules. For instance, we should rule that a pitch which crosses the plate behind the front, e.g. a backdoor slider, be out of the strike zone. We should also standardize the vertical dimensions, perhaps from 17 to 34 inches, so that the front-to- back depth of the plate helps the hitter visualize the strike zone.

We could even extend the use of automation to call swung strikes. One proposal would register a strike if the bat penetrated the strike zone, regardless of whether the batter's wrists had broken. This also prevents a batter from claiming that it was the bat, not the ball, that broke the plane. Since this rule change results in more strikes, we could compensate by allowing batters to swing all they wanted at balls out of the strike zone, as long as the bat doesn't break the strike zone.

But it would also be acceptable to install a system that required of the home-plate umpire the same kind of judgement that is required of the other umpires. While I'm sure they would protest vigorously against the introduction of an automated strike zone, I think they would come to like it - it certainly makes their job easier. And pitchers and hitters would definitely like it: nobody likes the inconsistency of the current system.

Revisiting Foul Balls top

As it now stands, a foul ball counts as a strike for the first two strikes, reflecting the batter's failure to hit a good pitch back onto the field. But then, foul balls become the batter's friend, as they permit him to protect himself against a third strike while waiting for a pitch he likes or a fourth ball. The change in batter's styles with two strikes is very evident, and somehow inelegant.

An alternative would be to recognize that foul balls - failures by the batter to direct a hit onto the field - are comparable to balls - failures by the pitcher to direct a pitch into the strike zone. Here's one idea : we count foul balls separately, and award a batter a walk when he has three more balls than fouls. Meanwhile, he would foul out if he had three more fouls than balls. This treats all foul balls consistently, and it brings the balls per walk in line with strikes per strikeout: three outs, three strikes, but four balls?

I would also count each throw to first (or another base) as a ball, in which case we might not need the balk rule.

Starting Relievers top

I'm sure there's some good reason why this is a bad idea - maybe one of you can tell me.

I've noticed the following facts:

So why doesn't some team try an "all-reliever" pitching staff? Each pitcher would be expected to go an inning or two, but not to face the same batter twice. You'd use four to seven guys a game, out of eleven-twelve pitchers on the roster, but they'd all be available again tomorrow. And fewer relievers would warm up without getting into the game.

This would also encourage situational pitching, since your pitcher isn't going to be in there for very long anyway. You could bring in your lefty to face their big lefty, then leave him in for a couple of innings. Or you could pinch- hit for your pitcher in the second inning (in the National League).

What's wrong with this idea? It's true that starters are better pitchers, often much better, and it makes sense to use better pitching as much as possible, but I think this is the way to do it: I think it would result in more pitches thrown by the better pitchers, in better circumstances (the first time around), and when needed more.

The Designated Hitter top

Illustration by Stephen Schildbach

I dislike the DH rule, but I am not outraged by it. It is inelegant, but it eliminates the rally-killing automatic out that most pitchers represent, and it has allowed great hitters like Edgar Martinez to stay in baseball a few years longer. What bothers me most is the difference between the two leagues, but I also wouldn't mind seeing a little more bullpen strategy in the American League. The ideal would be a rule that offered the best of both worlds, and I think I have one.

Picture the right to designate a hitter for your pitcher while keeping him in the game as a physical object like a flag or scepter: let's call it the Baton. In my proposal, the first time either pitcher comes up to bat in a game, his team gets the Baton. If they wish, they may use the Baton to allow a designated hitter to bat for the pitcher, but they then have to give the Baton to the other team. But if they don't use it, they keep it, and the other team either has to let their pitcher hit, or pull him for a pinch-hitter.

This rule would provide for numerous strategic decisions:

In this plan, the designated hitter is not a regular position like, say, the second baseman. Instead, he is a kind of pinch hitter with the special power to share a slot with the pitcher. So, for example, he need not be named in the starting lineup - the manager can wait until he's needed to name him. Once someone has been brought into the game as the DH, he can only bat in that slot, but the pitcher can also bat in the same slot without the DH leaving the game. The DH is also allowed to bat without using the Baton as long as the pitcher is pulled while he's batting; in other words, he can also pinch-hit normally for the pitcher. The Baton is only needed to keep the pitcher in the game while the DH bats for him.

It's also possible to move a DH into the field, but it requires the use of the Baton. To do so, the manager must pull his pitcher and put the new pitcher in another slot in the batting order, as if he were doing a double substitution. The batter (new or old) in the new pitcher's slot becomes the new DH. This can be done anytime you have the Baton, which then passes to the other team.

For example, imagine that Hernandez is pitching and Martinez is his DH. Late in the game, their slot comes up with one out and you (the manager) decide to send Hernandez to the showers to keep the Baton. You don't need to name your new pitcher until the beginning of the next inning. Martinez singles and represents the winning run, so you send Gonzalez in as a pinch-runner. Unfortunately, the next batter, Rodriguez, grounds into a double play and the inning is over. Rather than bring Enriquez, the new pitcher, into the slot vacated by Hernandez, you could use the Baton to bring him in in Rodriguez' slot, leaving the speedy Gonzalez [sorry] in the field and postponing your next need for the Baton a little: Rodriguez becomes the new DH. Or you could keep the Baton, leave Rodriguez in the field, and let Gonzalez and Enriquez share the original pitcher's slot. But once you're into the opposing bullpen, the Baton isn't so valuable.

As the example illustrates, this new DH rule offers lots of strategic thinking, but also avoids having pitchers bat most of the time: the best of both worlds.

But maybe a much simpler idea would do the same job : just don't count foul balls as strikes when the pitcher is hitting. In other words, give him a small advantage at the plate, and then let him hit.

The Season top

The problem with the current season is simple to state: too many teams are out of the running too early in the year. To fix this, we need to allow some teams with lousy first halves to make it to the playoffs. If you can't accept this heretical idea, you won't like the details. Of course, you probably hate the postseason, too, and just want to give the World Championship Trophy to the team with the best regular-season record.

But the postseason is great baseball, even though the eighth-best team isn't much worse off than the best, going in. Playoff games are worth much more than regular-season games towards a championship, and we value a team's ability to perform under pressure. After all, a grand slam is worth more than a solo homer, too. So I've stolen an idea from the game shows, where the final round of questions is always worth much more than the first round.

A season is 162 games, which I propose to divide into 54 three-game series. Assuming the leagues are balanced, I propose that each team play two series against each of the 14 other teams in its league, one home and one away. They should also play one series against each of the 15 teams in the other league, alternating home and away every year. This is more interleague play and less division play than we now have, but that's a good thing, making the division into divisions less important. Why should the poor Tampa Bay Rays have to beat the Yankees and the Red Sox every year, while the Cardinals coast against the Brewers, Pirates, and Cubs?

With an odd number of teams in each league, there will always be some interleague play going on. In fact, there will usually be five interleague series at any one time, representing a third of the games (five times a year, there will be seven interleague series going on at the same time). One way to organize interleague play is to stage nine five-game tournaments between divisions, e.g. all five AL West teams would play all five NL Central teams over a five-series period. These nine tourneys would take 45 series; since we only have 43, the last two series of one tourney have to be spread out over the rest of the season.

These 43 League and Interleague series would take us to about Labor Day each year, and leave 11 series left to play. Up to this point, results haven't mattered to the schedule, but from now on they would. We now divide the teams in each league into five Classes of three teams each: Gold, Silver, Copper, Nickel, and Tin, based on their standings within their league.

In Class play, each team plays the two other teams in its class, the three teams in the class above, the three teams in the class below, and the three teams in the same class in the other league, for a total of 11 series. The Gold teams don't have a class above, so instead they play the Tin teams from their own league, who don't have a class below - this mismatch rewards Gold and penalizes Tin, as it should be.

Postseason invitations would be awarded as follows, based solely on Class play (i.e. ignoring the League season before September):

The eight teams left will thus represent various routes to the postseason. Some will have done well all season, but some teams who do well early will falter when the competition gets better. Some might have gotten off to a bad start, but done well in Class play. At any point in the season, winning is better than losing - nobody wants to drop into a lower class just to face easier opponents, since the ratio of winners gets so much worse.

There is also room to adjust the postseason invitations. For example, we might take pre-September records into account but double the point award for victories in class play. The idea is to keep everybody in the running as long as possible, even if their chances fade with each loss.

In the postseason, the first round should be two four-team tournaments, with each team playing a series against all three other teams from their league. This offers more games (18 instead of a max of 17 now) in a shorter time (nine game days instead of a max of twelve) than the current division and pennant series format. The World Series can stay as it is.

Ownership top

Baseball teams are called "clubs", but the membership is very exclusive: the owners. As a group, baseball owners contribute less to baseball than players, staff, officials, or even the fans, and yet they reap more and exercise near- total control.

So why don't we "take each team public" by selling its shares to its fans, one share per seat (with the shares varying in value)? Each shareholding fan would receive the right of first refusal to buy "his" seat for a game at the normal price, and would get a vote and a dividend, if the team were profitable. Shares could be traded on the open market, both privately and publicly, and a new stadium or seating plan might call for new shares or a buyback.

This is kind of like how professional rugby works in Australia, where the clubs are not so different from American golf or yacht clubs. A clubhouse would be a good place to mingle with players, to stage events, or even to watch away games with other fans. And it would breed a more sophisticated loyalty among fans than can be expressed with hats and bumper stickers, a loyalty bred of investment and involvement.

The Net Statistic top

Batting average, slugging average, and on-base percentage are all measures of a players individual contribution to the success of his team, beyond merely counting home runs and RBIs. My version of this measure is called the Net.

In one phrase, the net is the ratio of bases earned (by hitting, advancing runners, walking, stealing, etc) to outs caused. There is no direct measure of at-bats. Pitching statistics are the same as hitting, except of course that pitchers want lower nets, while hitters want higher ones.

On the good side, a single counts as 1 base, a double as 2, a triple as 3 and a homer as 4. You also get credit for any runners you advance, so a single with a man on first counts as 2 bases (assuming he stops at second). A walk is as good as a single, and so is getting hit by a pitch. You get credit for any bases you steal, even on a wild pitch or a passed ball, and for any bases you reach on an error. However, any bases you give up by committing an error are subtracted from your total!

On the bad side, you get blamed for every out you cause hitting: strikeouts, forceouts, fly balls, fielder's choices, etc. If you hit into a double play, you get marked for two outs. You also get docked for getting caught stealing or picked off. You even get docked for sacrifice outs, although you also get credit for the bases your teammates advance. However, you get no credit for outs you help make in the field - that's a team effort.

Every base and every out you make hitting or running gets charged to the opposing pitcher, too, except fielding errors. If he throws a walk, a homer, and three strikeouts, he will show 8 bases for 3 outs. Most nets are around 1: the ratio of bases to outs is about even. But it takes a few at-bats for the stat to settle down. Within a game, it varies wildly from 0 to infinity (for a guy who makes no outs), so it's only shown to two decimal places when fewer than ten at-bats are included.

Sabermetrics (The Society for American Baseball Research) reveals that on base percentage plus slugging percentage is the best quick estimate of runs scored, and the Net could be adjusted to reflect this by counting first base as worth two bases; in other words, bases + non-outs.

Text Scoring top

"Scoring" a baseball game means keeping score, not just of the runs, but of all the events in a game. Traditional baseball scoring involves marking symbols on a special form, with a grid of at-bats showing blank diamonds. This makes it easy to see how far each runner got, but difficult to reconstruct events in the order they occurred. It also makes it difficult to recap a game in a newspaper, for example, which is why line scores were developed.

Text scoring solves both problems. It also uses two grids whose lines are the nine slots in the batting orders, but the cells contain conventional alphanumeric text: numerals and upper- and lower-case letters, and the occasional punctuation.

The digits 1-4 represent bases earned by the hitter and runners: 1 means a runner reaching first base, 2 a runner reaching second, 3 a runner reaching third, and 4 a runner scoring. The digit 4 is usually boldface to make it more visible. If you single with a man on first, your score would show 12 - the lower numbers are always written first.

The uppercase letters represent players handling the ball and some other elements, as follows:

Digits and uppercase letters are combined to describe play using a few simple principles. If the batter gets a hit, the first symbol is the digit representing the base he reached, e.g. 1 for a single. This is followed by the digits for bases reached by other runners, e.g. 24 for an RBI double. These digits are followed by the letters for the defensive players that played the ball, e.g. 24R for an RBI double to right field. If there had been, for example, a throw from right field in to second base, it is typically omitted if the play were already over, e.g. a stand-up double, so its inclusion (24RX) indicates that there was a play at second. Home runs and ground-rule doubles just show the nearest defensive player for reference.

If the batter is out, the first symbols are the letters for the defense. A grounder to third would be TF, and a fly to center would be a simple C. If the first baseman plays the ball and steps on first for the forceout, this would be indicated FF. Any runners that advance on the play are indicated by following digits, e.g. TF2 for a grounder that advanced the runner from first to second.

Uppercase E is used to indicate an error, followed by the letter of the player at fault. So for example, 1EPb would indicate an error by the pitcher while fielding a bunt. I use lowercase e as a comment to indicate an error that the scorer misjudged - he didn't agree with me! For example, if the scorer had ruled the above a hit, I might still have scored it 1ePb.

Lowercase letters are used to add color to the action. Here are some examples:

And here are some less common occurences:

If you are fanatic enough to record pitch results, they are shown with small letters up front:

Sometimes, you need to show a base itself, for example to indicate who a pitch runner is replacing or when a fielder is covering an unusual base. In those cases, the base is a superscript, with ° indicating home plate. For example, when the second baseman covers first during a wheel play, it would be scored FS¹b.

If there is a play during an at-bat, for example a steal, it is noted in the same cell, followed by a slash. For example, a batter who walks, then steals, might show 2KX/ before the next batter's result. If he's caught stealing, it would be KX/. If he's picked off at first, it would be PF/. The end-of-inning line might run through the slash.

Parentheses are used for substitutions, enclosing the position and the new player's number. Bringing in #77 to pitch would be (P77), and pinch-hitting #11 would be (H11). When multiple substitutions take place or a substitution also moves another player, both numbers are enclosed in the same parentheses, e.g. (L43R42) means #43 is now playing left and #42 is now playing right. When a defensive substitution is made, the number alone is used in the offensive side to remind you that it's a new batter. For example, if #43 comes in to play left field (L43), put (43) in the line for his slot in the batting order, so you'll know later who batted. In contrast, when a pinch hitter stays in the field, use the letter alone to indicate the position.

Lastly, the symbols ? and ! enable you to indicate your opinion of a play. C! might be an amazing catch, while TX?1F indicates that you thought the shortstop should have turned the double play.

A Sample Game top

Here is the text score for an actual game: Cleveland Indians at Seattle Mariners, Sunday 4 August 2002

Cleveland Indians 1 2 3 4
5 6 7
8 9
AB B O Net
R 11 Lawton R Tp Lf S XF SF24 6 2 6 0.33
X 8 McDonald SF 1W 1R 1L 0s XF 6 3 3 1.00
D 23 Burks 1PmS 12Lg 12W 12W 1Lg 5 8 0
F 25 Thome L Lf XF23 134C 0s 5 7 5 1.40
C 24 Bradley
1P? 123W 1Wi 124R(P45) KX/PP 5 8 1 8.00
S 12 Gutierrez
12C PmXS 1234L(P17) 244LXK (T9L4P22)1W 5 11 1 11.00
T 17 Fryman
124L 1L 1234W 0s PS?1b 5 9 2 4.50
L 30 Magruder
S TF2 SmRX134 R! (H9)13Rg 4 3 4 0.75
K 2 Diaz
XSoF 0s S S(P53) (H36)134R 4 0 5 0.00
P 57 Drese
28 30 15 2.00
P 47 Elder
4 12 1 12.00
P 37 Westbrook
12 14 8 1.75
H 9 Stevens
1 3 0
HL 36 Selby
1 4 0
K 38 Perez

P 39 Wohlers
3 0 3

Seattle Mariners 1 2
3 4 5 6
7 8 9
AB B O Net
R 51 Ichiro 0s 123XS 1X 1C SF SF 6 6 3 2.00
TL 4 McLemore L 0s 12W (P47)2pb/14Cg SF SF 6 5 4 1.20
F 5 Olerud 1W 1Cg FF23 FP2 1R 5 6 2 3.00
S 29 Boone 12W XS1F 0s 44L XS 5 8 3 2.67
D 11 Martinez C 0s 244R 1C 2R?(B23) 5 8 2 4.00
L 21 Sierra
0s FF 13SF (P37)0s TF 5 2 4 0.50
X 8 Guillen
R Xbb 12W 12W XF3 5 5 3 1.67
K 13 Davis
1W 1C R 344S?df 0c 5 10 2 5.00
C 44 Cameron
12b XSoF 4L TF (L36K38P39)0c 5 6 4 1.50
P 54 Halama
24 26 13 2.00
P 17 Hasegawa
8 17 3 5.67
P 45 Franklin
4 7 3 2.33
P 53 Rhodes
5 1 5 0.20
B 23 Ugueto

T 9 Cirillo

P 22 Sasaki
6 9 3 3.00

To help you decipher this, here's a recap of the Mariners sixth. Cameron led off with a homer to left. Ichiro then singled to center (a line drive, since a grounder would have been marked with small g), sending Drese to the showers. With Elder on the mound, Ichiro stole second on a passed ball. McLemore then singled to center (a grounder), driving Ichiro home: 2 in, nobody out. Olerud grounded out to first (with the pitcher covering the bag), but advanced McLemore, and Boone's home run to left scored two. When Martinez singled to center, the Indians pulled Elder and brought in Westbrook. He struck Sierra out swinging, but walked Guillen. With men on first and second, Davis popped it up to shallow right field, but the second baseman dropped it: Davis ended up on third with two RBIs. The scorer ruled it a hit, but the question mark shows I disagree (he had it in his glove, with the right fielder and first baseman within reach). Finally, Cameron grounded out to third to end the inning he started: 6 runs, 6 hits, 1 left on base.

Totals R H E LOB SO BB SB CS DP PB IW AB Bases Outs Net
Indians 10 15 0 10 4 7 0 1 1 1 1 45 51 27 1.889
Mariners 8 14 0 12 7 6 0 0 2 0 0 47 56 27 2.074

Feedback top

Please send me your comments, suggestions, corrections, questions, testimonials and invective. If I feel they're of interest to my audience, I'll post them here (without identifying you). I reserve the right to edit your contributions for length, language, and clarity.

home top

© 2002-2012 Alivox 30aug12