Sunday, December 14, 2003

Twins vs. Royals = BAP Revelations

Quick Message To those BAP Believers - You're gonna love this.

Bases Advanced Percentage was originally designed to measure all of the things that happen in a ball game that don't end up in either On-Base Percentage (OBP) or Slugging Percentage (SLG). It seemed to me that things like going from first to third on a single, or sacrificing successfully with runners on first and second, were game events that were very relevant, and to ignore their importance because they were not convenient was a mistake made by nearly the entire stat community. I devised a way to measure the number of extra bases gained (EBs) and dove deep into game logs to see if BAP could do a better job of predicting game results than OBP, SLG, and OPS, on the theory that any stat that was worth anything must prove itself to be very important in determining the winners and losers of ball games. The tests were run three times during the season, the most recent being the entire 2003 post-season, and each time BAP proved to be the most accurate.

It was while I was discussing the Playoff results that I came across the following stats for the Minnesota Twins and the Kansas City Royals:

OBP - KC .336, Min .341
SLG - KC .427, Min .431
Runs - KC 836, Min 801

How can a team that has a higher OBP and a higher SLG than another team get outscored by 35 runs? Where is all of the missing offense for the Royals? Is it EBs?

I set about to find out. The goal became this, to find the total number of EBs for both the Royals and Twins, and to calculate the BAP scores for each, to see if BAP would explain the difference between the run totals. In theory, if BAP is the difference, then the Royals would need much higher EB totals than the Twins to make up the differences in OBP and SLG.

Now of course, I'm not done yet. This is going to take some time. I've only finished 18 games for each team, but the results are so interesting that I needed to share them, so here we go:

After 18 games, the Royals have scored 94 runs, and have gained 103 EBs.

After 18 games, the Twins have scored 72 runs, and gained 79 EBs.

Well, so far so good. The Royals are scoring runs and accumulating EBs at higher rates, which is what we wanted to see. As a matter of fact the rates by themselves are interesting, too. When I first saw these numbers, I couldn't help but notice that both teams were gaining EBs at a little more than one per run. So I decided to calculate it. What I found was that the Royals, with 103 EBs and 94 runs, are scoring 1 run for every 1.10 EBs. The Twins, with 79 EBs and 72 runs are scoring 1 run for every . . . no . . . that can't be right can it?

1.10 EBs?!

Here are the stats extended to four places beyond the decimal point:

Royals - 1.0957 EBs/Run
Twins - 1.0972 EBs/Run

What the hell does this mean? Well first of all, there seems to be a direct relationship between the number of runs a team scores and how many EBs it accumulates. But that leaves us with the ol' chicken-and-egg theory: Does the scoring of runs cause more EBs, or does more EBs cause more runs?

My first feeling is that this couldn't hold true all of the time, that it is too small a sample. I mean, I've scored some Red Sox games, and they always seemed to score lots runs without any EBs. And if the scoring of runs was causing more EBs, then we would never see something like the Twins-Royals Scenario (from now on to be called TRS, I can see where I'm going to get tired of typing the whole thing out), where the team that has a higher OBP and SLG has scored fewer runs. But it's still possible that the gaining of a high number of EBs is ONE WAY of scoring a lot of runs, which could explain the TRS.

But none of this explains those two rates. What are the odds that both teams are accumulating EBs at the exact same rate as runs scored? I know it's only 18 games, but if a player comes into the majors with no prior stats to judge him on (think Roy Hobbs) and gets hits in his first 18 games, it's safe to say he's a pretty good hitter.

Lets take this one step further. Suppose the rates stay the same for the entire season. Will that explain the differences in run totals in the TRS?

The Royals scored 836 runs in 2003. If they stay at the rate of 1.1 bases per runs scored, they will accumulate 920 EBs for the season.

The Twins scored 801 runs in 2003. If they stay at the rate of 1.1 bases per runs scored, they will accumulate 881 EBs for the season.

Projected BAP Scores for 2003:

Royals .617
Twins .616

It appears that BAP will in fact be able to explain the differences in run totals between the Twins and Royals.

I'm going to leave this post up until next weekend. After that, I have 16 days off, so I'll have more updates then. In the meantime, I'd love to hear from you. What do you make of all this?

BAP For Beginners

Dave's Email


Post a Comment

Links to this post:

Create a Link

<< Home