Now that the 1930 Replay is finished it is time to review what happened, what went right, what went wrong, and what changes do I need to make to my process before starting a new replay. There are several different areas to cover here, so let's see if I can put these in some semblance of order.
Replaying the 1930 season is something I had considered for a long, long time, but with other family and work obligations, something had to give. I decided to energize around doing this in late 2017 and started putting together a detailed plan of attack. The first decision - cards and dice, or computer (BBW)? I do prefer dice and cards as I really feel like I am much more involved in the game at hand. The BBW version is much more sanitary in that regard, but by taking care of all the post-game paperwork it became obvious to me that BBW was the way to go.
The other benefit here was using the actual lineups resolves about 90% of your issues with regards to player usage, and especially player usage by position. I expected this chore to be a little more arduous, but the rosters were not large to begin with, so all I really had to monitor was pinch-hit and reliever usage. It is a fault of mine - in my draft leagues and in other replays I have been involved in, I am pretty miserly with relief appearances early in the season because I am afraid to run out before the end, and I did it here too. In the last few games of the replay, I ran a lot of relievers in and out to make up for this, but this is something I definitely need to pay more attention to.
I went through every player to review their player usage, and identified any ten-game absences and included that in my ATMgr roster manipulation. The actual ATMgr tool roster manipulation part went well - it just did what I told it when I set it up before play started. However, there were several occasions during the season when a team only had one active catcher. It was accurate to say that one active catcher caught every one of those games during that time period, but it is also accurate to say that there was likely a second catcher available to play if needed. There were a couple of cases where a miscellaneous infielder needed to be available, and there was a week where a team was down to four active pitchers on the roster. Again, while accurate, and it was something that I was able to work around, but I was probably a little too strict. A lesson learned for next time is to review this a little more in-depth and massage the roster manipulation file a little more to avoid these type of extreme cases.
As part of my pre-replay prep I did go through the (non-revised) 1930 players disk and added a card for all 92 non-carded players, plus I went through the player disk and added missing stats: player age, games started for pitchers, defensive numbers for multiple position players, steals/caught stealing numbers for catchers and pitchers. Because I was going to manage player appearances and minimize injuries all players were set to injuries ratings of J1 or J0, and because I was going to monitor steals all players got an A steal rating.
Of course when using the BBW version of the game it really isn't as customizable and a Cards and Dice replay, so you kind of have to accept that going into this. I was able to limit injuries to "rest-of-game" injuries only, which worked well for my purposes. I did turn off rain-outs as well. There were a surprising number of actual rain-shortened games that I wasn't able to account for, but that didn't seem like a large sacrifice.
I did regrade pitchers using the Weighted Averages guidelines. I thought this went well also. APBA has a habit of downgrading pitchers for lack of appearances, and this corrects that as well. Most pitching grades bumped up or down a level or two, so other than the corrections for low usage pitchers there wasn't much of a change, although pitchers that appeared on different teams were graded separately, and there was some variation there.
So, it was a replay of the 1930 season, a simulated re-creation of something that already happened. How did it turn out? How accurate was it? To summarize, I think the defenses were just too strong. Earlier seasons featured players with two error numbers on their card, but not for 1930. The distribution of good, average, and poor fielding grades was pretty normal, but both leagues came up around 600 errors short of their real total. And it's not just errors - that ability to get that third out and get out of a potential big inning was impacted as well. If you were a good pitcher with a great defense, you were a great pitcher. If you were a good pitcher with a lousy defense, you weren't a great pitcher. This also provides some level explanation as to why I had so many shutouts and no-hitters.
The net result of this was a significant number of missed at-bats, and then the corresponding number of hits, run, and extra-base hits that would come from those extra at-bats. Both leagues had a slightly lower batting average than expected, and both had a slightly better ERA than expected, but I think this is all part and parcel of the same thing.
So, in my eyes, that was the problem. Do I have a solution? If I were to turn around and replay 1930 again, I think the first thing I would do is downgrade the SS10's, 3B6's, 2B9's, 1B5's to the next lower fielding one rating. I realize APBA makes teams, not players, but no one in 1930 really deserved those type of ratings. I am not personally in favor of going through and re-grading every player, but when an SS8 that really made 40+ errors suddenly makes less than 20, it is something consider for some cases.
Secondly, I think I might add a second error number to some of the cards. Which cards? Not catchers or first baseman - they already have an extra Rare Play number, but maybe to all players who hit below .300. I would have to convert an out number to an error number somewhere, and my estimate this would mean that about half of the at-bats would now have the extra chance for an error.
But what about double plays? A double play requires someone being on-base to begin with, and this should increase that opportunity. The other thing is … the way double play resultants are counted. Basically, a GDP and a DP are not the same things. If a player grounds into a standard type of double play, it is counted as a GDP. However, if I player pops up on a sacrifice attempt and a double play occurs, it is not counted as a GDP. Similarly, if an outfielder guns down a runner at home on a failed sacrifice fly attempt, it is not counted as a GDP either. When I look at the final totals I can see that the counted number of GDP's are down as compared as to what is provided in Baseball-Reference.com, but I am not sure what I am really counting. My perception is that I had plenty of double plays throughout the replay when I add in these other uncounted instances.
After 35+ years of APBA I have often thought it would be nice if each player had two defensive ratings, one for regular plays, and another for errors. And since I am dreaming anyway, it would be nice if an "E" could be added to a players card as an error number and just have the game randomize the error number resultant.
One more comment on defense - I am planning to replay 1949 next. When I look at the 1949 defensive numbers I see a similar distribution of defensive ratings, but I also see that the actual number of errors decreases by about a third (~1500 to ~1000 per league). Given that each of my 1930 replay leagues had ~900 total errors, I would expect the final replay resultant to be much closer as is from a defensive standpoint.
Confession: I am not an advocate of the sacrifice, so I acknowledged to myself at the beginning that this was something I would have to be purposefully cognizant of for the entire reason. The result was that I was pretty close in both leagues, coming ~50 attempts short in each league. The other thing with sacrifice hits is that they are not automatic. It is my perception that ~20% of SH attempts failed to produce a sacrifice hit resultant - double play's, fielders choices, stolen bases or caught stealing, and errors are all possible outcomes, I just never realized to what extent.
What I didn't do - intentional walks. Again, I am not an advocate, but at some point about halfway through the season, I realized I had not used an IW all season. Mea culpa. It's not like there were a lot of these anyway, so for next time, it is something I will have to specifically monitor and account for.
As soon as I wrap up the 1930 World Series and the write-ups for 1930 it will be time to move on to my next project - 1949. Why 1949? I did a C&D 1949 replay in the mid-to-late 1980's and got it published in the APBA Journal. However, I didn't use XB's much less XC's, so at only twenty players per team I did get some odd results. I want to do it right this time.
- I have purchased the updated season disk already. All of the players are included, and the missing stats have been added, so I am saved some effort there
- I will update the pitcher grades using the Weighted Average guidelines
- I have started reviewing player usage in preparation for getting my ATMgr files set-up
Once this is all set-up I will be ready to go. However, I am going to approach this leisurely, take my time, take some time off as well, and then be ready to hit it, probably around January 2019.
Once play gets started, as far as the day-to-day process went, I was pretty happy with it. Besides the game, I have a window open with the actual box score for the day, and my Microsoft OneNote in another window. I use the actual box score for "suggestions" on who to pinch-hit with or what relievers to use on a given day. As soon as the game was completed I drafted a quick 2-3 sentence write-up of the game, and then I could immediately go on to the next game because of the time I spent with ATMgr before the replay started. I am really pleased with how well this all worked.
The once-a-week snapshot that I post in the blog posts seems about right. Every day would be too much, once a month would not be enough, and doing it weekly allows me to capture the events of the week and the pennant races pretty well. Google Blogger as a blog tool has its annoyances, but I can’t beat the price, plus I have learned how to work around its weakness. The only thing I am really considering changing is quitting Facebook, which means I wouldn't really be able to share this with the APBA community. Both Google and Facebook have grown increasing creepy, so we'll see I guess.
Comments
Post a Comment