Sunday, April 18, 2010

The Account-Form Scoresheet

Cross-posted at Weekly Scoresheet, for which I still want your scoresheets.

Most scoring systems endeavor to keep a chronological account of the game (at least in theory--only the Project Scoresheet system is purely linear, but one can reconstruct the game in order from a traditional scoresheet). However, what if one was only interested in recording the basic box score statistics of individual players, and not about preserving the game play-by-play?

For starters, you could certainly save a lot of paper. You could eliminate a lot of codes--there would be no need to record wild pitches, passed balls, balks, and the like (remember, I said basic box score statistics). It would be possible to describe each batter's plate appearance in just a few characters.

As is the case with so many other baseball topics, one can turn to Bill James for a creative idea about how to do this. In the 1983 Abstract, he introduced what he called the account-form box score. It was a box score that retained much of the data from a scoresheet, yet was compact enough to be printed in a space no larger than that devoted to traditional box scores.

I recently threw together a new scoresheet that records only the information necessary to create an account-form box score of a game. I made a few tweaks to James' coding system, but for the most part I just used it as is.

First, the ways to reach base:

S = single
D = double
T = triple
H = home run
W = walk
I = intentional walk
B = hit batter
E = error
f = fielder's choice (at end of code)

Three baserunning events are noted, by use of a lowercase letter after the on-base code:

s = steal
c = caught stealing
x = other out on base (That is charged to the runner, like a pickoff or out advancing when not forced. If the runner is retired on a force, no "x" is added).

Outs are recorded in typical scorekeeping notation--a ground out to short is "63", a fly to right is "9", a strikeout is "K". Double plays get a D prefix--"D643". A sacrifice hit gets a b suffix (for bunt); a sac fly needs no modifier because it will involve a RBI. If a fielder's choice results in a batter-runner not reaching base (i.e. it is the third out of the inning, so the batter-runner never actually gains first), then it is recorded with a "f" suffix (for example, 46f).

When a batter gets a RBI for driving in someone other than himself, James uses a ', so D' is a RBI double. D" would be a two-run double. When three other runners score, an exclamation point is used, so a bases-clearing double is D!. Thus, a solo homer is just H, a three-run homer is H", and a grand slam is H! A sac fly to left is 7'.

In James' system, batter-runners that reached base and do not end up scoring are given lower case symbols...s, d, t, etc. If you are keeping a paper scoresheet in this style, this is problematic, as you'll have to do a lot of erasing. So what I did instead was embolden the codes of batter-runners who scored. When paper scoring, this is easily accomplished by tracing a pencil marking with black ink. When keeping a paper scoresheet, it is also helpful to mark off the innings, which I did with a number in the far left portion of the box.

The biggest drawback to the account-form box as James envisioned it was that pitcher records are hard to read. James simply noted the pitchers by the number of batters they faced each time through the order. Suppose for instance that CC Sabathia faced 32 hitters, Phil Hughes faced one, and Mariano Rivera faced five. James notation would be:
Sabathia (9995), Hughes(xxx1), Rivera (xxx32)

Sabathia faced nine hitters the first three times through the order, than five the fourth. Hughes faced no batters the first three times through, and one on the fourth. Rivera faced no batters the first three times, the final three on the fourth, and two on the fifth.

This makes it a chore to read through the account and tally up a pitcher's statistics (the same issue exists for batters of course, but it is a lot easier to read across a row to take in the result of three to six plate appearances than it is to read up and down column to take in the result of potentially dozens of plate appearances). There's no easy solution--the best option might be to just give each pitcher a basic line, perhaps in the format of (IP H R ER W K Dec), like:

Sabathia (7.2 8 3 2 2 6 W) or Sabathia (7.2, 8, 3, 2, 2, 6, W)

Also, I tend to think that a better notation to indicate batters faced for the above example would be something like:

Sabathia (1-32), Hughes (33), Rivera (34-38)

This does not solve the problem with reading a pitcher's performance I described above, and it may force you to count off by nine (9, 18, 27, 36, etc.) in order to figure out where to start reading the scoresheet to find a particular pitcher, but I think it's more intuitive and it saves a little space that is being given back by listing out pitcher lines. Of course, counting off by nine shouldn't be a problem for those versed in systems (such as Project Scoresheet) that use numbered boxes.

Another issue is how to list substitutions. When printing a box score, it is easy to insert an extra line for substitutes and indent it. But with a pre-printed scoresheet, it's trickier. What I've done is use letters (a, b, c, d, etc.) to show substitutions. The timing of the substitutions is noted by using the first time through the batting order that the player appeared. For example:

[a Giambi 4] indicates that Giambi pinch hit, with his first PA coming the fourth time through the order in the batting order slot where a is listed.
[d PR Guzman 5] indicates that Guzman pinch-ran the fifth time through the order.
[c 6 Everett 3] indicates that Everett was a defensive replacement at shortstop, and that his first PA (if in fact he got one) would come the third-time through the order. The exact time he entered the game defensively is not recorded, as it isn't on a typical basic box score.

This allows the scoresheet to be more compact, but it does make it a little tougher to quickly read batting lines for lineup slots in which multiple players appeared.

Honestly, I don't have a lot of use for an account-form scoresheet myself; it's not detailed enough for my tastes as a scorekeeper. It's the box score that I want--I want to be able to get on the internet in the morning and read a mini-scoresheet rather than a box score. Which line tells you more, and more importantly IMO tells you a better story:

63, K, H", 9, W

or

4 1 1 3 1 1
with a note below saying HR: Player X

or

4 1 1 3 1 1 HR
as the more helpful Baseball-Reference box score does?

It's true that the account-form style requires you to do a little counting in your head if you want to know H-AB, or runs scored, or the other box score categories. However, I'd argue that it's very easy to do that sort of addition, and that it also is a cosmetic improvement--there are an awful lot of zeros in box scores.

Incidentally, this is the same reason that all of the scoresheets I design myself omit columns in which to record batting lines. I'd rather devote that space to larger scoreboxes and do some math in my head than waste a bunch of space on columns I'll have to fill in to create a box score that I could just as easily read on the internet. Pitchers are a different story, as it is much harder to get a quick read on their performance as the number of scoreboxes under consideration is much larger (and more spread out) than for a batter.

Any site that implemented an account-form box score would instantly get all of my morning box score hits. I'm a little surprised that someone hasn't implemented it or something similar on online just to stand out, as there are so many places that you can go to get box scores. As it stands now, only Baseball-Reference offers something truly unique.

Below is an account-form scoresheet for Cincinnati @ Milwaukee, 4/10/2008, copied from a scoresheet I kept in a different format. I simply typed the account into Excel, although ideally I would have paper-scored it and scanned. It will be easier to read this way though. A drawback is that since I am using the cells in the spreadsheet, if a leadoff hitter scored, the inning number is in bold as well. I did not include pitcher lines, just the batters they faced:




Let me go through batter-by-batter and write what they did in English:

Patterson: grounded to second, grounded to second, grounded to second, grounded to second, flied to center
Keppinger: grounded to short, popped to short, grounded to pitcher, intentional walk, grounded to pitcher
Griffey: struck out, flied to center, walk, hit into double play
Phillips: flied to right, flied to right, struck out, grounded to second
Dunn: grounded to first (a pop to first would have been marked 3^; although James used 3 for a pop to first and 3- for a grounder to first), grounded to first (pitcher covering), single (scored), grounded to short
Encarnacion: grounded to third, walk (scored), two-run home run, single
Hatteberg: walk, single (put out on base), double (scored), grounded to second
Bako: struck out, RBI double, RBI single, walk
Harang: sac bunt to first, struck out, single
Votto: flied to center

Weeks: popped to third, flied to center, hit into fielder's choice at second, popped to second
Gross: double, flied to center, grounded to second
Counsell: hit into fielder's choice at short
Fielder: grounded to short, flied to left, flied to right, popped to third
Braun: grounded to short, popped to third, grounded to third, struck out
Hall: double (scored), grounded to short, grounded to short, grounded to second (via pitcher)
Hart: sac bunt to first, grounded to short, struck out
Villanueva: sac bunt to pitcher, struck out
Kapler: grounded to first (pitcher covering)
Kendall: grounded to short, single, single

Here is a link to download the account-form scoresheet in Excel format (scroll down to Account-Form scoresheet).

No comments:

Post a Comment

Comments are moderated, so there will be a lag between your post and it actually appearing. I reserve the right to reject any comment for any reason.