APBA Journal Samples on APBA.com

Very soon you’ll be able to download a sample issue of the APBA Journal from the APBA Game Company’s website.  Depending on how much interest this generates, we may offer more issues in the future.  If you’ve already spent good money buying the entire collection, first, thank you very much, and second, don’t worry because new back issues won’t be available for free download any more frequently than once per month.  So, assuming we keep a steady pace it will take over 30 years for someone to complete their free set.

Keep watching the APBA site for developments.

The Evolution of an Innovation

I’m slowly but surely working my way through creating an index for the Complete APBA Journal, and a pleasant surprise was uncovering a series of contributions that show how the Journal fostered innovation in its own slow but sure way back in the day. I’m speaking of what eventually came to be known as the Coxx Pitching System. For those of you unfamiliar with it, the Coxx system was a method of introducing some granularity into the pitching grades in APBA’s basic baseball game. One of the early wishlist items of contributors to the Journal was a way to distinguish pitchers beyond the essentially four basic grades: A, B, C,and D. (The A&C and A&B grades have always been exceedingly rare) The Coxx system took advantage of the 30 pitching grades assigned in APBA’s Master Game by assigning one of a set of 30 corresponding pitcher cards to each pitcher for use in the basic game. Whenever a hit result might be influenced by pitching grade (e.g., results 8 and 9 with bases empty) the pitcher made a roll against his pitcher card to determine his pitching grade for that particular at bat. You can see a refined version of the Coxx cards on Makojo’s site.

This seemed like a brilliant innovation when it was published by “Smiley Coxx” (aka, AJ Editor Tom Heiderscheit) in the February 1978 issue of the APBA Journal, and it certainly was. But did it represent a flash of genius, or merely the last stage of evolution of an idea that had long been percolating in the pages of the Journal. I suggest it was the latter.

In the October 1971 issue of the Journal, AJ founder and then-editor Len Gaydos got the ball rolling with this idea:

How about a 36 dice number card for every pitcher. Each of the four grades would be represented in varying proportions depend1ng on each individual pitcher’s sk1ll. The present Grade B might have a card with a half dozen A’s and a couple of D’s with plenty of C’s and mostly B’s. The pitcher rolls for his grade then the batter rolls for the result.

A good suggestion! Every pitcher would have a unique pitching tablet, much like the hitter’s tablet.  But not only would this increase the number of cards in the set, I can only imagine the extra hours Richard Seitz would have to spend completing it.

In the next issue of the Journal, December 1971, future professional sports statistician Jeff Sagarin praised Gaydos’s proposal and had this to say about its implementation: “As for the actual printing, APBA could print the grades right next to the hitting numbers on each pitcher’s card.” A good proposal that solves the set size issue, but this might clutter the cards, no?

Then in February 1972, the breakthrough. What was to be dubbed the Coxx Pitching System six years later was proposed in a modest letter by Armand Erickson, Bainbridge Island, Washington:

A compromise production method might be to produce a set of different rated pitchers cards with abilities from superstar to say something less than major league ability. Each pitcher could then be assigned a certain card by having the card’s number printed besides the pitcher’s name or on the roster sheet. Thus, those who wished a more complete game need only order a set of cards, say a packet of 36, and those who wouldn’t want to bother could use the present grading system.

The idea needed only the introduction of the 30-grade master pitching ratings by the Game Company in 1976, and all the elements were in place for Smiley to complete the innovation. Of course, an idea like this might develop in a matter of days in a robust online forum, but without the Journal, who knows how long it would have taken the idea to mature in those days?

What’s to be Gained from a Loose Computer Translation?

Before computer renditions of baseball boardgame sims came along, the player had complete information about the simulation model, and therefore could learn to play the game on its own terms.  That is, for example in the case of APBA, all chart results for all base situations, plus all player-card outcome data was known to the board game player.  The only variable was the outcome of a dice roll, but the universe of possibilities for that was known as well.  A manager, therefore, could make fully informed decisions on strategy with knowledge of the likelihood of any given outcome.

Not so with any computer sim for which the simulation model deviates from the boardgame and is not disclosed to the user.  For example, according to an article written by Wesley Hall which appears in the “APBA Primer” Reference Pack, tests have determined that a pitcher’s propensity for allowing walks depends on more than just the pitcher’s APBA-assigned control rating–a rating which exclusively governs the pitcher’s control (or lack of it) in the board game. According to Hall, the pitcher’s actual statistics for the year being simulated will influence his control (up or down) regardless of his APBA-assigned control rating.  So, a given Z-rated pitcher who in reality yielded walks at a higher rate than another Z-rated pitcher is likely to yield walks at a higher rate in the computer game.  This could not happen in the board game.

Perhaps this leads to more realistic outcomes statistically, but it does so at the expense of the player’s understanding of which factors affect outcomes and how. As a computer sim’s model deviates from that of the board game, one may question why a fan of the boardgame would necessarily choose that loosely-based computer sim as his game of choice.  To the extent the computer model becomes less accessible to the player, other factors such as program features and appearance design will be necessary to maintain the player’s loyalty.