Thursday, May 6, 2010

RePost - Remember the Era!

Eras are by far the most important concept to understand in order to put forth a successful ATB campaign. Eras (not Earned Run Average, but DMB's name for the statistical average of a given League and Year) put a player's story in context, and most importantly from a gaming standpoint, are required to make a realistic baseball simulation. From the DMB help sections:

It is not possible to play realistic games among teams of different eras without adjusting for these changing playing conditions. For example, without these adjustments, a 1912 team would have almost no chance of beating a 1984 team because it would make two to three times as many errors. Are the 1984 fielders really that much better? Of course not. They just have the advantage of using modern gloves and playing on artificial turf.


DMB uses eras to adjust for these factors. In DMB:

  • A .280 hitter in 1968 (when the league batting average was under .250) is a better hitter than someone who hit .280 in 1930 (when the average player batted .300).

  • A shortstop making 40 errors in 1912 is a better fielder than a shortstop making 30 errors in 1993.

  • A starting pitcher completing 30% of his games in 1984 is more durable (relative to his peers) than someone who completed 50% of his starts in 1920.

  • A DH-league pitcher with a 3.30 earned-run average is a better pitcher than someone with a 3.00 ERA in a non-DH league.

From our standpoint, it bears repeating - eras have an impact on every aspect of the game: Hitting, Pitching, and Fielding. Each of these need to be taken into consideration when drafting.

DBM uses several required statistics create "normalized" player records. DMB uses the league and year, along with the core player stats, to create a "hidden" normalized stat line. This normalized stat line is what is used to determine the results. It is "hidden" from users, including commissioners, as it is proprietary to DMB and perhaps more importantly, displaying it would ruin a lot of the fun in playing the games out.

I am not going to go into era adjustments for fielding since the About DMB section tells you how to precisely calculate Fielding Error rates. Instead I'll focus on batting and pitching. First thing's first, the following statistics are required whenever I add a player to the database:


AB, H, 2B, 3B, HR, HBP, IBB, SH, GDP, SO, SB, CS



This is critical. Hopefully you noticed that R, RBI, W, L, and even ERA are not part of the required stats for DMB to generate the normalized player record. Instead, it uses more sophisticated stats that can be gleamed from the ones above, such as WHIP, K/BB, K/9, H/9, SLG, AVG, OBP, etc. This is just to say ignore certain stats when determining which players to draft.

Please note that in the end these stats correlate exactly to what makes a player a good and don't be concerned, for instnace, that ERA isn't a core statistic. Said another way, if a pitcher strikes out a lot of batters, minimizes his walks and hits against, and makes sure not to give up the long ball - his ERA will be excellent.

Also note, in cases were some of these statistics are not known, DMB uses a built in equation to estimate them for us.

Now that we understand how DMB works, it's time to take a look a some real life era data

This chart depicts the runs scoring environment for the two main leagues in baseball history, the American and National Leagues of course. There were 5 other leagues throughout history, none lasted more than 10 years (most much less), and the last of these was the Federal league in 1915.

Right now, as we know, we are in the midst of a elevated run scoring environment, although it not as prolific as what was experienced in both leagues during the early 1930's and in the American League in the last 1930's; and of course, nowhere near the levels of the late 1890s. The lowest runs scoring eras were around 1907-1909, 1917, and 1968.

And now the key and the only official warning you'll get from me - several great players made a name for themselves in these extreme eras. 1968 is home to Bob Gibson (1.68 ERA) and Denny McLain (31 Wins); 1930 is home Bill Terry (.401 AVG), Hack Wilson (191 RBI); and the 1907-1909 period sports Three Finger Brown (1.31 ERA) and Addie Joss (1.16 ERA). This is not to say all or even some of these players will not perform, rather it is to say don't expect those insanely low ERA's or high batting averages to translate into our game, and do a little research comparing them to their peers before drafting them. The goal, really, is to find good batters in low offensive years, and good pitchers in high offensive years. Two excellent, but crude tools to use are the ERA+ and OPS

leader boards found I personally like the math aspect, and download the historical "Lahman Database" found at the Baseball Archive, and perform some calculations on my own.

Finally, I will leave you with two more points and two more charts.

  • I literally have no idea how DMB values each season in terms of run scoring. I would imagine it is much more robust than the chart I used above that simply depicts runs per game, and they in all likelihood the DMB engine gets down to the individual stat level - League Average SLG, OBP, AVG or perhaps even HR Rates. But for all intents and purposes, unless you have a lot of time on your hands AND are a complete freak, you don't need to get down to this level of detail.

  • Don't forget parks. The era is only part of the normalization process, and DMB also uses proprietary park factors in the process

No comments:

Post a Comment