DMB News December 1999
December 16, 1999
Written by Tom Tippett
Welcome to the fifth edition of the Diamond Mind email newsletter. Through these newsletters, we will try to keep you up to date on the latest product and technical information about the Diamond Mind Baseball game, related player disks, and our ongoing baseball research efforts. Back issues are available on our web site (www.diamond-mind.com).
If you don't wish to receive these messages in the future, please send an email response with the subject line "unsubscribe". We'll immediately remove your email address from the list. And if you know someone who would like to subscribe to this newsletter, we'll be happy to add them to the mailing list if they send us an email message with the subject line "subscribe" and their name and street address in the body of the message.
1999 Team Reviews
We've begun publishing our team reviews for the 1999 season. Each review consists of a detailed comparison of projected to actual stats for all key players on the team, team and player comments, and a brief look ahead to the 2000 season. They'll be published more-or-less simultaneously on our web site (www.diamond-mind.com) and on ESPN.com. Two new teams will appear each week until early March.
1999 Season Disk Update
I'm happy to report that we'll begin shipping today, as previously announced. If you've been waiting to place your order for the 1999 Season Disk, please keep in mind that any new orders will have to wait until we've shipped all of the advance orders. It'll take us a few days to get everything out the door, so we cannot guarantee delivery of new orders by Christmas. We will, of course, do everything we can to ship everything by Tuesday, December 21, so most items should be there for the holidays.
For 1999, we've added real-life salaries to the season disk. A few years ago, we made space in our player file to store the salary and contract expiration year for each player. It was never our intent to fill in these slots with information on real-life contracts. Rather, we added them so Diamond Mind Baseball leagues that use salary cap systems would be able to enter their salaries, see those salaries on screen and in reports, and have those salaries carried forward from year to year by our season disk migration feature.
But we've been asked by quite a few of our customers to add the real-life salary information anyway. And that's what we've done this year. We're grateful to Doug Pappas, a longtime Diamond Mind customer and member of the Society for American Baseball Research (SABR), who agreed to let us use a salary database that he has meticulously compiled for the past few years.
NOTE: Because the salary slot was originally designed to hold the salary assigned in your league draft, not the real-life salary, the migration feature carries forward the salary from last year's league disk. This means that if you run migrate, the real-life salaries on your 1999 disk will be wiped out and replaced by the salaries from the league disk you are migrating from. We are aware that this may not be the behavior you wish to see, so we plan to enhance the migration feature in version 8 of the game so you have more control over how salaries are handled in the migration process.
Player disk notes
Please remember to read the player disk notes when your 1999 Season Disk arrives. They contain a lot of useful information about how we develop our ratings, and they talk about specific players whose ratings might differ from the past or from the perception created by the baseball media. Many of the questions that people ask us this time of year have already been answered by those notes.
In version 7, you can find them by choosing Info from the main menu, choosing Source and picking the player directory, then using the View command to display the notes. From there, you can print the notes if you choose.
Reminder for league commissioners and managers
The 1999 Season Disk is a copyrighted product of Diamond Mind and contains real-life statistics that are copyrighted by STATS, Inc. If you give a copy of this season disk to someone else, you're violating the law and harming both Diamond Mind and STATS.
Having said that, it's also true that our recommended procedure for running a Diamond Mind league is for the commissioner to set up the league and make all transactions on a master copy of the season disk, and then to send copies of that disk to league members, PROVIDED THOSE LEAGUE MEMBERS ARE REGISTERED OWNERS of Diamond Mind Baseball and the season disk.
We don't want to put league commissioners in the position of having to police the copyright laws on our behalf. All we ask is that the commissioner send us a list of league members so we can check our database of registered customers and indicate which individuals are entitled to receive the league disk.
I believe the vast majority of our customers are honest, and that's one of the reasons why we have never been forced to copy protect our disks. But we do get calls from people who believe they don't need to buy the season disk because they'll be getting it from their league. In most cases, the individual is simply unaware that this is a violation of our copyright, and when we explain the situation, they're happy to buy a legal copy.
Frequently asked season disk questions
This seems like a good time to answer a few of questions we get each year about the season disk.
Q: How do you decide whether to rate a player at a defensive position?
A: Generally speaking, a player gets a rating at any position where he's started at least one game or played more than a few innings. If someone has played only a few innings at a position, it depends on the player and the position. If he has established his ability to play that position in recent years, we'll generally give him a rating. If he got only a cup of coffee in the big leagues and doesn't have any other defensive ratings, we'll give him a rating if it's his primary position. But if it's an especially difficult position (such as CF, SS, or C), we generally won't give the rating to someone who hasn't established the ability to play there, especially if he's a very good hitter. For that reason, Chipper Jones isn't rated at SS this year, even though he played an inning there.
Q: If a player is used at a position where he has no range rating, how well will he perform defensively?
A: It depends on the position and the player's other defensive ratings. The game looks to see whether the player is rated at a similar position and then assigns an adjusted rating accordingly. That means that players can move to similar and easier positions without much of a penalty. For example, you can move a good defensive CF to another outfield position and he will continue to play well at his new position. And you can move a good shortstop to second or third and he'll continue to play well. But if you move someone to catcher, or put a first baseman in the outfield, or move your right fielder to third, you can expect to pay a price defensively.
Q: Our draft league uses players from both leagues. Does that mean we should use the Neutral Era for league play?
A: The Neutral Era represents the 20th century average in several ways -- overall level of offense, composition of offense (rates of doubles, triples, homers), pitcher durability, and error rates at each position.
If you use the Neutral Era instead of the "1999 A" or "1999 N" era, you'll get less offense, fewer homers, more complete games, and more errors. In short, you'll get "20th century average" baseball instead of 1999 baseball. There's nothing wrong with using the Neutral Era for your league, and some people do this just because they don't like the offensive explosion we've seen in the past few years.
But there's no need to use the Neutral Era just because you're using players from both the AL and NL. After removing pitcher hitting stats from the NL, the AL and NL stats have been virtually identical almost every season in recent memory. That means your league can use either the NL era or AL era from the current season and get essentially the same results, because Diamond Mind automatically adjusts for the effects of the DH.
Q: How well does Diamond Mind adjust for the player's home ballpark in real life? For example, last year Pedro Astacio had a 5.04 ERA overall but it was 7.16 at Coors Field and 3.60 away from home. If Pedro plays in Dodger Stadium in a Diamond Mind Baseball league, will he be closer to the away numbers or close to the overall numbers?
A: Our system for rating players takes full account of real-life ballpark effects. Last year, Coors Field increased scoring by 63% and Dodger Stadium was close to neutral. Astacio's personal home-road splits were wider than for the park as a whole, so he won't be projected to go as low as his road ERA of 3.60, but the change in parks will help him a lot. In fact, I've been thinking I might try to draft him or trade for him in my league.
On a similar note, our system also adjusts for the DH rule. The ERA in the AL generally runs about a half a run higher than the NL, so a pitcher with a 4.00 ERA in the AL performed about as well as someone with a 3.50 ERA in the NL, assuming their home parks are similar.
When you're evaluating players for your league draft, keep the park and DH factors in mind. Colorado hitters won't be nearly as good in other parks, Rockies pitchers will be much better in other parks, AL pitchers will see their ERAs drop if they move to a non-DH league, and so on.
Q: STATS, Inc. publishes a player's batting average depending on where in the order he hits, how he hits against certain pitchers, how he hits with certain counts, etc. Some our managers are convinced that where a player hits in the order on his Diamond Mind Baseball team will be affected by his ML average in the same spot. Also some managers think that if player A is 0-20 vs a certain pitcher that that will be reflected in the game. Are they right?
A: The statistics that determine batting and pitching performance in Diamond Mind Baseball are park-adjusted totals and left/right splits. If you start adding in other splits -- such as batting order position, day vs night, grass vs turf, and month-by-month -- you quickly reach a point where the data is statistically meaningless because the samples are too small.
Chances are, a given player had only a few appearances in a season against a left-handed pitcher, on artificial turf, in a day game, and while he was batting in a certain position. And anything can happen in a few plate appearances. Heck, anything can happen in a week's worth of games (just look at the weekly leaders in USA Today) or even in a month's worth of games (check out my article on the meaning of monthly stats at www.diamond-mind.com/april.htm). And it's bad game design to build a factor into a game when that factor is based on data containing huge amounts of random variation.
It's even worse for individual batter/pitcher matchups. In the 1999 season, there were 189,692 plate appearances, and there were 72,438 different batter/pitcher matchups. That means each matchup involved an average of 2.6 plate appearances. There's no way to do anything meaningful will small amounts of data like this. And even if there was, it would still only make sense to build this feature into the game if it was true that past performance in a matchup was an indicator of future performance, and I'm not at all sure that it is.
So the factors we use are the ones that actually mean something: overall performance, park effects, and left/right splits. Even so, some of the sample sizes are on the small side for part-time players. That's one reason why I like our projection disks, where player performance is based on three years of major-league and minor-league data. In a single real-life season, there are always a few guys who hit .385 against left-handed pitching solely due to chance, and these guys become too valuable in your Diamond Mind Baseball games. That doesn't happen with the projection disk.
Understanding our Player Ratings
When interpreting our ratings, please keep the following things in mind:
An Average rating is a compliment. We rate players relative to other major leaguers at the same position. An average rating means that the player has performed at a level attained by only a handful of other professional baseball players.
Don't read too much into a range rating. When we say that a player has Very Good range at a position, it means he gets to more balls than the average player. It's not an overall evaluation of his defensive ability. We have separate ratings for errors, throwing, and passed balls, and it's not unusual for someone to have an Average range rating and much better ratings in the other categories.
Our ratings reflect the ability to make plays, not raw athletic skills. A very fast outfielder might still be rated Average or below if he doesn't get a good jump on the ball. An infielder needs many skills -- positioning, quick feet, good hands, a strong and accurate arm, and good judgment -- to make plays in the big leagues, and a deficiency in any of these areas might be enough to turn a very flashy player into an average playmaker. Someone with average speed might be an excellent baserunner because he has great instincts about when batted balls will fall in for a hit. An outfielder who doesn't have a strong arm may still be successful in slowing down the running game if he gets to the ball and gets rid of it quickly. That's why we study real-life play-by-play data to see which players are actually getting the job done.
Some positions are harder to play than others. If we give player A an Average rating in center field and player B a Very Good rating in left field, it doesn't mean we think player B is better than player A. The standard for center field play is much higher. Conversely, a player who is not regarded as a great outfielder may still get a decent rating in left field, because that's where managers often put the guys who can't play a more demanding position.
If someone did not make any errors at a position in real life, it doesn't guarantee that they won't make any in Diamond Mind Baseball. Beyond a certain amount of error-free playing time, we feel they've earned an error rating of zero. Below that level, the rating is based on a weighted average of zero (for the time they played) and the league average (for enough playing time to bring them up to the level where a player would earn the zero rating).
The error ratings that appear on the game screen (and the roster report) represent the projected number of errors this player would make at this position in 100 full games (900 innings) in the era the league is playing in. If you move this player to another era (or run the roster report from the Exhibition League, which is linked to the Neutral Era), the e rating changes accordingly.
The Prone injury rating doesn't necessarily mean that the player will miss a lot of playing time. In fact, our injury system is pretty mild compared with real life, and you won't see Diamond Mind Baseball players going down with season-ending injuries in April. We give the Prone rating to anyone who (a) was on the disabled list at any time during the season or (b) missed 15 or more games due to injury without going on the disabled list.
The Clutch and Jam ratings are given to players who (a) performed at a very high level in the late innings of close games during the season and (b) performed at a level higher in these situations than in other situations. But the most important thing to remember about these ratings is that they DO NOT play a large role in the game. Personally, I would never use a weaker player over a better one just because the weaker player has a better Clutch or Jam rating.
Manager profiles for draft leagues
One very important thing to remember is that if you draft new rosters (either by hand or using the draft utility from our web site), you need to create a manager profile for each team. The computer manager is designed to work hand-in-hand with the information in the profile, and it will get confused if you do not at least set up a starting rotation, bullpen roles, and saved lineups. And remember that the computer manager will always choose lineup #1 against a lefty starter and #2 against a righty starter, no matter what you happened to name those lineups.
On the pitcher profile, there are three mode available to you -- Strict, Skip, and Time. The Time setting is used to give each player the number of starts they had in real life. But this usually doesn't work well in draft league settings because most teams don't have a group of pitchers who started exactly 162 games. Instead, set up a rotation and use either the Strict or Skip mode, and the computer manager will follow your rotation, making adjustments only when necessary to handle injuries.
On the batter side, you can choose GameByGame or TrackStarts for your depth charts. TrackStarts mode was designed for replays using real-life rosters, and it makes sure that every player gets the right number of starts at each position against left- and right-handed pitchers. The GameByGame setting was designed for draft league situations. In that mode, the computer manager will faithfully use your saved lineups, making changes to those lineups only if a starter is injured or if you've established non-zero spot start percentages for some players.
When I set up my draft-league team, I usually start by using the manager profile generator to create a profile. (I would never use that profile as-is, because the generator is designed to produce profiles for real-life season replays. But it's faster to start with that profile and make changes than to create a profile from scratch.) Then I set up a Strict rotation, put my depth charts into GameByGame mode, and change all of the spot start percentages to zero except for a few positions where I'm trying to spread the playing time among a few players to avoid going over our leagues playing time limits.
In parallel with our work on the 1999 Season Disk, we've been working on version 8 of the game. We'll discuss more of the new features in the next newsletter. Until then, I hope you enjoy playing with the 1999 Season Disk as much as I enjoy working on the ratings.
- Jim Wheeler