DMB News October 2006
Diamond Mind Email Newsletter
October 31 , 2006
Welcome to the fifth edition of the Diamond Mind email newsletter for the year 2006. 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.
Topics for this issue:
2006 Season Disk
Upgraded Products for 2006
Simnasium Offers for Diamond Mind Owners
2007 Bill James Handbook
DMB in the Media by Tom Tippett
Tech Tips from Luke Kraemer
To Bunt or Not to Bunt by Tom Tippett
Although a majority of our customers now order their Diamond Mind products through our web store, a good number prefer to order by mail.
So we've begun sending our annual October mailing to registered owners of Diamond Mind Baseball. That mailing includes an updated order form that includes the 2006 Season Disk and the 2007 Bill James Handbook.To order by mail without waiting for your letter, you can print an order form via the "How to Order" page of our web site.
Work is underway on the 2006 Season Disk, which will begin shipping around December 14th, and we are now taking advance orders.
As usual, you'll receive a ton of information with this season disk, including everything you need to start playing games immediately upon installation:
- full rosters with every player who appeared in the big leagues
- official batting, pitching and fielding statistics, including left/right splits for all batters and pitchers and modern statistics such as inherited runners, holds, blown saves, pickoffs, stolen bases versus pitchers and catchers, and in-play batting averages
- games started by position versus left- and right-handed pitchers
- updated park factors
- a full set of real-life transactions and game-by-game lineups for season replays
- two schedules, the original (as-scheduled) schedule and another (as-played) reflecting rainouts and other rescheduled games.
- real-life salaries for all players
- complete manager profiles for all teamsYou can place a credit card order now through our web store (follow the link from www.diamond-mind.com) or by calling us at 800-400-4803 during business hours (9-5 Pacific time, Mon-Fri). The 2006 Season Disk is priced at $29.95.
We know that, at this time of year, many customers are putting together their Holiday Wish Lists. To help with that, here are the products we upgraded this year:
In August, we released a major upgraded, version of one of our most popular products – the All-Time Greatest Players disk. We've added more than 630 new players, bringing the total to 1760. In the original 2003 edition, we didn't include anyone whose career fell mostly before 1894, but the 2006 edition includes stars from the 1870s, 1880s, and 1890s.
We've also updated the stats and ratings for players who were active from 2003 to 2005 and relaxed the thresholds for earning a rating at a defensive position, so you'll see some existing players with an extra position or two.
The larger player pool allowed us to expand the number of teams from 32 to 48. We were able to create standalone teams for Toronto, Montreal, Atlanta, Milwaukee, Minnesota, Los Angeles (A), Baltimore, Oakland, and San Francisco.
Because of the new players and teams, we developed new manager profiles for every team, added ratings for 16 historical parks (all of which have images available for free download from our web site), updated the ratings for modern parks to reflect the 2003-2005 seasons, organized the teams into two leagues with four divisions, and created two new league schedules.
Even though this version is bigger and better than the 2003 edition, we're holding the price for new customers at $29.95. Registered owners of the 2003 edition can upgrade to the 2006 edition for $17.95.Upgraded Classic Seasons – In 2006, we added real-life transactions and game-by-game lineups to three more Classic Seasons: 1954, 1961 and 1973. This brings the total number of upgraded Classic Seasons to twelve – 1934, 1946, 1954, 1955, 1961, 1965, 1966, 1973, 1974, 1975, 1976 and 1977. These seasons are priced at $19.95 each.
Many of our registered Diamond Mind owners have already taken advantage of the offer of a free team credit to try Simnasium’s Total Baseball. Total Baseball uses Diamond Mind technology and allows you to scout and choose from among thousands of players from baseball's glorious past, then test your ability to draft and manage a team of historical players through 162 games, competing against 11 other team owners in the race to the pennant.
Now is a great time to add an online baseball experience to your baseball simulation play (obsession may be closer to the truth for some of you). It’s easy to try Simnasium – if you have not played Total Baseball before, get a free team credit by sending an email to firstname.lastname@example.org with the subject line “DMB free team offer” and your name in the body of the message. Once you have registered on the Simnasium site (www.simnasium.com) and created a user-id, you will receive the free team credit.
For those of you who have received your free team and want to build a baseball empire, you can claim a 2 nd team credit with the purchase of one team. Send an email to email@example.com or firstname.lastname@example.org with the subject line “DMB 2-for-1 offer” after you have purchased a team from Simnasium. One team is only $19.95.Interested in setting up a custom league and testing your managing abilities against your friends’ abilities? As a registered owner of Simnasium, you can earn free team credits for yourself and your friends with our referral discount. When you refer a friend to Simnasium, he or she will receive a free team credit to try the game. When your friend purchases a team, you will receive a free team credit.
Since 1990, the annual Bill James Handbooks have formed the backbone of our baseball library. For a complete, well-organized reference that includes every active player, you won't find a better book.
You can order the paperback edition from Diamond Mind for only $19.95, two dollars off the cover price. The spiral-bound edition, which lies flat on your desk, is $24.95. We are taking advance orders now and both editions will begin shipping the week of November 6th.
Among the many great features of the Bill James Handbook are:
- career registers for every active player, including minor-league stats for players with little big-league experience
- complete fielding statistics for every player
- expanded pitcher stats include hitting, fielding, and holding runners
- park factors and rankings
- left/right splits for all batters and pitchers
- conventional and sabermetric leader boards
- team standings, augmented by many team performance splits
- team rankings for batting, pitching and fieldingNOTE: Because of the added weight, first-class and air mail shipping rates are not available for orders including this book. Priority Mail and Overnight shipping rates are available.
In recent weeks, we were asked by two major media entities to simulate portions of the 2006 postseason before they happened.
The first request came from Jonah Keri, a contributor to ESPN.com and the YES Network web site, who was writing a story about the Yankees postseason chances for the YES Network.
At the time we were contacted, Detroit had the lead and the tie-breaker going into the final weekend, so we expected the pairings to be Minnesota-New York and Oakland-Detroit. Instead, that final weekend saw Minnesota rally to win the division when Detroit was swept by the lowly Royals.
So it was that Minnesota hosted Oakland and Detroit went on the road to New York in the division series. Our simulation results made the two home teams the clear favorites, but as we know now, the underdogs came out victorious in both encounters. In fact, the favorites won only one game between them.
This is one of the reasons I always have mixed feelings when we're asked to predict a playoff series. On the one hand, it's very satisfying when a major television network, web site or newspaper comes to us. That means we've achieved a certain level of competence and trust over the years.
On the other hand, anything can happen in a short series.
Let's suppose that we were given an opportunity to talk to the baseball gods, just for a moment, and the gods told us that team A had a 70% chance to beat team B in a seven-game series. Further, let's suppose that those gods oversee a world where the outcomes are not predetermined. In other words, the gods possess the perfect knowledge that the series has a 70% chance to go in favor of team A, but they don't possess the power to make it come out that way.
There's still a 30% chance that the weaker team will win the series. Three times out of ten, it will be team B that is coating their locker room with champagne spray a week later.
Of course, we're not gods, so the best we can do is rate the players accurately, simulate a series many times, and develop an informed estimate of the true winning probability for each team. And while we believe we know enough about baseball to come up with an estimate that is very close to the real thing, we'll never know for sure.
When the real-life series is played, it is played only once, so we never get to find out whether that one outcome is representative of what would happen if these teams contested the series a thousand times. And because this is the only real outcome that people can see, some will conclude that it was the "right" outcome and anyone who predicted anything else is an idiot.
Despite these mixed feelings, we quickly said yes when Allen St. John of the Wall Street Journal called three weeks ago to ask us to simulate the World Series. In addition to simulating the four potential matchups, he and his editors asked us to run several what-if scenarios involving talented players who in real life are unavailable due to injury. During this project, I spoke with Allen many times and found him to be very knowledgable about sabermetric matters.
Allen's story appeared in the By the Numbers column in the Weekend section on Friday, October 20, and it said lots of positive things about Diamond Mind Baseball and the methods we use to simulate games.
Unfortunately, Allen, or his editors, chose to lead with this paragraph:
"With the World Series set to begin in Detroit tomorrow evening, Tom Tippett knows something important. He knows who's going to win."I was absolutely stunned when I read that. Anyone who knows me knows that I would never make such a claim. And anyone who listened to my phone conversations with Allen or read the emails we exchanged would know that it's the polar opposite of what I had been telling him. Over and over, I pointed out that the most interesting aspect of playoff simulations is the frequency with which the weaker team wins, and I sent him links to two articles I'd written on the subject.
At one point, Allen called me to get a quote for the article. It was clear from the way he posed the question that he was setting me up to make a very strong claim in favor of the Tigers. When I told him that I felt like he was trying to put words in my mouth, he backed off a little, and I again talked about the chances for the underdog to pull off an upset.
In the end, Allen couldn't resist the temptation to sacrifice truth for a strong lead paragraph, so he made me appear to be a know-it-all who doesn't have a grasp of the underlying probabilities. That's so far from the truth that it would be laughable if not for the fact that lots of people who don't know me will read this article and think I actually said that.As I write this, the Tigers are trailing the Cardinals two games to one. Maybe they'll rally, maybe they won't. But if the Cardinals win it all, I won't be stunned. Even though we believe the Tigers are the superior team, St. Louis was able to take them in a seven-game series in 30% of our simulations. That's the real story here.
If you are an Internet league commissioner and just finishing your season using the 2005 player ratings, consider using the Migrate command (File menu) when you're ready to set up your league using the 2006 ratings. First you'll need to install the 2006 season disk using a meaningful league folder name. For example, if last year's league folder was named ABC2005, you might name the new one ABC2006. After the 2006 season is installed, go to the File menu and choose Migrate. Set the Source database to ABC2005 and the Target database to ABC2006. The ABC2006 database will then have the same league(s), teams, rosters, etc. as the 2005 one. Players not used in your 2005 database and new ones added to 2006 will be in the free agent pool.
Flash Drives: Gamers regularly ask what's the best way to play a season on two systems, typically a desktop at home and a laptop when on the road. In the past I'd recommend emailing backup files between the two systems or burning CDs. Both techniques work but they have their drawbacks. I have a new, preferred recommendation using a Flash Drive, also known as a Thumb Drive or Memory Stick. These drives are the size of a Bic lighter and can hold an incredible amount of data. The smallest I've seen holds 256MB of data, room for lots of DMB seasons. The lowest price I've seen for this capacity is only $10. I just saw a 2GB one in the paper for only $25 and a 4 GB one for $45! You can get Flash Drives with lots more space but they can run in the hundreds of dollars. A Flash Drive plugs into a system's USB port(s) which is typically on the front panel of your PC under the standard drive bays. Laptops could have their USB port(s) on the sides or the back. I believe all systems sold today come with at least one USB port. This is the same port you'd use for uploading pictures from a digital camera.
Typically after you plug in one of these drives, Windows will pop up a message and announce that a new drive has been detected. After a few seconds, ‘My Computer’ pops up showing the contents of the drive. You'll also be able to see what drive letter Windows has assigned to it. My system already has A:, C:, D:, and E: drives so my Flash Drive is assigned as an F:. You don't actually need ‘My Computer’ at this point so go ahead and close it.
Start up DMB. Go to the File menu and choose "Install season disk". Select a season to install such as AGT2. After you accept the license agreement, DMB will ask for a location for the database and a name. The location will default to "C:\dmb9".
To install the season on the Flash Drive, change the location to its drive designation, in my case, "F:" (without the quotes). You can stick with the default database name, in my case AGT2, or you can give it a different one. Click on OK and the season will load onto the Flash Drive. You can now play that season just like you would one on your hard drive.
To play this season on another system, either shut down DMB or switch to a different season database on your hard drive. It's now safe to remove the Flash Drive. Plug the Flash Drive into the USB port on the other system. Windows will announce there is a new drive, just like the first system, and ‘My Computer’ will pop up. You should see the season folder you just installed on the drive from the first system. While still in ‘My Computer’, note the drive letter designation assigned to the Flash Drive. Close ‘My Computer’ and start DMB. Since the season wasn't loaded on this system, you'll need to let DMB know where it is. Go to the File menu and choose "Add reference to existing database". Set "Drives" to the letter designation assigned to the Flash Drive. The contents of the drive will be displayed in the "Folders:" window. Double-click on the season folder to open it and then click on OK. After a few seconds, DMB will report that the database is added and is now the active one.
You can now play games with the season on the Flash Drive. Like the first system, when you want to remove the Flash Drive, first change DMB's active database or shut it down before you remove the drive. Plug the drive back into the first system. Make the season database on the Flash Drive the active one and the results of all the games played on the second system are there. After you initially install the season on one system and issue the Add Reference on the other one, all you'll have to do from that point on is plug in the drive and issue a "Change active database" command to play that season. If you want to autoplay a large block of games while automatically saving boxscores and game-by-game statistics, the games will run rather slowly compared to a season on a hard drive but should be acceptable.
If you'd prefer to play all the games for a season on a hard drive yet still be able to move it from system to system, install the season on the hard drives of both systems like you would normally. When you're ready to move the season to your other system, shut down DMB. Plug in the Flash Drive. After ‘My Computer’ pops up, navigate to your hard drive. Open the DMB9 folder. Copy the season folder and Paste it to the Flash Drive. After the copy is complete, close ‘My Computer’ and unplug the drive. Plug the drive in the other system. When ‘My Computer’ pops up, Copy the season folder and Paste it to the DMB9 folder on that system's hard drive. You'll be warned that the folder already exists and asked if you want to replace it. Choose "Yes to all" and after the copy is complete, the season will be identical to the one on the other system.
This technique has the advantage of faster autoplayed seasons plus, the Flash Drive and the other system will serve as backups. If you use the technique of playing all your games on the Flash Drive, you won't have to copy season folders back and forth to your hard drives. You should, none the less, make periodic backups of the season database folder on the Flash Drive in case you lose it or it goes bad. Even if you don't want to use the Flash Drive for use on multiple systems, it's great to use for season backups.
Last week, the Mets were two runs down entering the ninth inning of game seven of the NLCS, and the Cardinals put the game and the series in the hands of their young closer, Adam Wainwright.
Wainwright, a rookie, was a starting pitcher in the minors but had spent the entire season in the Cardinals bullpen, pitching mostly in the middle innings before taking over as the closer when Jason Isringhausen was lost to injury.
Looking a little shaky, Wainwright started the biggest inning of his life by giving up a pair of singles to the Mets #7 and #8 hitters, Jose Valentin and Endy Chavez, putting runners at first and second with nobody out.
Generally speaking, first-and-second with nobody out is the best bunt situation in the game. On average, it's worth the risk of a failed bunt to try to get two runners in scoring position with one out. When the pitcher is due to bat, as was the case in this game, bunting the runners over is a no-brainer. Or is it?
Faced with this decision, New York manager Willie Randolph went in a different direction, summoning Cliff Floyd to pinch hit for Aaron Heilman. Floyd struck out, the rally fizzled, the Cardinals went to the World Series, the Mets headed home, and Randolph was the target of a great deal of criticism for eschewing the bunt.
Few things in baseball produce more discussion than the bunt. Traditional baseball people, including legions of fans, are strong proponents of bunting in these "obvious" bunt situations. Meanwhile, much of the sabermetric community decries the bunt as a complete waste of a plate appearance in all but a few limited circumstances.
But this case provides an excellent opportunity to take a closer look.
The traditional method of evaluating this bunt versus swing away decision is to consult the run probability tables. By studying tens of thousands of innings from real-life games, one can determine the average number of runs scored and the probability of scoring at least N runs in any game situation.
That's how we concluded that the most attractive bunt situation is first-and-second with nobody out. The average number of runs and the probabilities of scoring one or two runs are improved by a successful sacrifice.
But those run probability tables are based on the average of a large number of innings, so they represent the most likely outcome when you have an average series of hitters facing an average pitcher in an average ballpark during an average era. Randolph wasn't dealing with an average situation. He was confronted with one very specific situation. Can we really use the long-term averages to help us make this decision?
No, we can't. We have to look at the skills of the bunter and the pinch hitter, the skills of the pitcher, and the effect of the ballpark and the era in which the game was played.
We're playing in a relatively high-offense era, which reduces the value of the bunt and raises the value of swinging away. But Shea Stadium is a good park for pitchers, so that tends to offset the era effect.
How about the skill of the bunter? Suppose Heilman had gone to the plate to lay down the sacrifice. In six professional seasons, Heilman has batted a grand total of 107 times, roughly half in the minors and half in the majors. He has been credited with 12 sacrifice bunts, 5 at the big-league level.
Those 5 successful sacrifices came in 10 tries. Four times, he fouled off one or more bunt attempts before the plate appearance was resolved on something other than a bunt. Six times, he got the ball in play, and five of those six moved the runner(s). His rates for getting the ball in play and advancing the runners are slightly below the league average, but it's fair to say that he's been an average bunter. Ten attempts are not enough to prove otherwise when his rates are fairly close to the norm.
More interesting is the fact that Heilman had not batted once in the 2006 season, and I can't imagine that Randolph would have asked Heilman to do something so crucial for the first time in more than year.
But let's not give up on the bunt option without considering the possibility of using another player to lay down the bunt. One obvious candidate would be Tom Glavine. He's been a very good to excellent bunter throughout his career, he's had plenty of practice, and he could be used without burning a position player.
Oddly enough, Glavine had failed (bunting into a force out) in his only previous bunt attempt in this series. Still, had Glavine been called upon, I believe he would have had a very good chance to succeed this time.
The chosen pinch hitter was Cliff Floyd. On paper, he was a very good choice, mainly because of how he matched up with Wainwright. In his brief big-league career, Wainwright has dominated right-handed batters (.185 average, .523 OPS) but has been hit hard by lefties (.301 average, .845 OPS). Floyd is a lefty who has hit righties very well. Overall, I'd say this matchup is very favorable to the swing-away case.
So far, we've identified the major factors in the decision, indicated that the bunt would be called for if all of those factors were deemed to be near the long-term averages, and identified which factors favor the bunt and which favor swinging away. But we haven't tried to quantify the impact of each of these factors.
Fortunately, we have a tool that enables us to do exactly that. I call it our lineup-dependent expected runs calculator. With this tool, we can create a specific game situation -- a certain batting lineup against a certain pitcher in a certain park in a certain era with a certain set of baserunners and a certain number of outs -- and calculate the average number of runs that can be expected to score in that inning along with the probabilities of scoring at least one or two runs.
This tool allows us to find out when a specific situation differs enough from the long-term averages to point to a different conclusion. In this case, it supports Randolph's decision to use Floyd as a pinch hitter.
If we assume the bunt was attempted and was successful, Jose Reyes would have stepped to the plate with runners on second and third and one out. From that point, the Mets would have a 43% chance to score at least two runs in the inning, which is what they needed to prolong the game.
If the bunt was attempted but was not successful, Reyes would have seen runners at first and second with one out. Getting no runner advancement in return for that out would have reduced the two-run probability to 31%.
With Floyd swinging away against Wainwright, the two-run probability rises to 49%, which is even better than the successful bunt scenario. In other words, even if the bunt was guaranteed to succeed, it would still be better to swing away. Furthermore, swinging away rather than giving up an out makes it more likely that you'll get the third run that wins the game right here, right now.
What we don't know, of course, is whether Floyd was capable of performing at his normal level. Because of a foot injury, he had been in and out of the lineup, mostly out, and may have been a shell of his former self.
At the same time, we don't know whether Wainwright was capable of performing at his normal level. He was a rookie in a huge situation who had just let two of the enemy's weaker hitters to reach base. Maybe nerves would get the better of him. If I'm Randolph, I don't want to give the kid an easy out.
And we don't know whether nerves or the wet weather would have caused the defense to botch the bunt and leave the bases full with nobody out. That would have pushed the two-run probability to 66%.
While those uncertainties mean that we cannot know for sure, I believe Randolph made the right choice, and I'm glad to have this opportunity to give him a little support at a time when many others are being critical.
My other goal with this little bit of analysis was to point out how complicated these decisions can be. Randolph had very little time to make the call and a lot of factors to consider.Just because it didn't work out this time, it doesn't mean he made a bad decision. I would criticize his move only if it turned out that he had good reason to believe that Floyd's injury had sapped much of his hitting ability, and since I wasn't there and didn't have access to the people involved, I can't say anything about that.