Using the A's as a Model for the Jays, Rays and O's

by Athletic Supporter on Apr 15, 2008 9:47 AM PDT

I've written a piece called "Davids and Goliaths" about how the Blue Jays, Devil Rays and Orioles can compete with the Yankees and Red Sox. It uses the 2000 - 2006 A's as a model for success. I hope you'll give it a look and leave some feedback either here or on my blog.

Here's the link:

Roto Journal: Davids and Goliaths

It's Part one of a Four-Part Series, with Parts 2-4 (coming soon) detailing each of the 3 teams.

Mods: If this is posted in the wrong section, please move it to wherever it's appropriate.

Thanks!

Scott

Tweet Comment 5 comments | 0 recs |

Do you like this story?

Read More: a's, blue jays, devil rays, orioles

Comments

Some critiques

why are you limiting the time period of your research from 2004 onwards?

And then using 200-2006 when you use Oakland as your model. That seems very arbitrary. Cherry picked, in fact.

“The correlation between pitching and winning games is stronger than that of hitting, base running or fielding. The most important single statistic turns out to be ERA.”

Also, ERA is a very crude measure. Firstly, no adjustment for park. Which is true of all your the numbers that you used. Secondly,

Defense is another area that does not show a strong tie to winning games. (Guess who had the majors’ best fielding percentage in 2007? Baltimore.)

Why are you using fielding percentage?

Therefore, I would rank the priorities of our roster assembly as follows:
1- Starting pitching
2- Hitting (emphasis: On Base Percentage over Slugging Percentage)
3- Bullpen
4- Defense
5- Speed (all-around)

You should replace defense with Fielding Percentage, since you are evaluating fielding percentage, NOT defense.

This agrees with the method the A’s used to attain their success. Those teams combined low ERA and high OBP to win games, and were average or lower in other areas. As an example, the A’s 2006 squad placed 4th in ERA and 7th in OBP. In the other major categories mentioned earlier, they were 9th or lower.

Have you considered that ERA is not entirely dependent on the pitchers who throw the ball, but also the fielders who catch the ball? Replace a defender like Tulo, or Reyes, with a defender like Michael Young, and see if you still believe that defense does not matter.

It’s pretty damn easy to say that the way to build a good team is to have a team with a high OBP and low ERA. The issue is HOW you get that high OBP and low ERA.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Apr 15, 2008 11:10 AM PDT reply actions

Critiques ...

Of course ERA is going to correlate better than the various offensive stats you mentioned - for two reasons.
1. It more or less directly and necessarily correlates with runs allowed. Anything based on runs, obviously, are going to correlate more closely to wins and losses than component statistics.
2. ERA is a comprehensive stat - incorporating all aspects of pitching. Slg, OBP, Avg, etc individually only deal with a small portion of offense as a whole.

For this type of analysis, you need to look at more than just the value of various components (eg SP ERA is more valuable than RP ERA because of difference in innings, which, btw, is not necessarily true, due to higher leverage innings pitched by some relievers)—you need to look at the cost as well. You’re not looking for the best players, you’re looking for the best bargains.

As RFLOH mentions, using Fielding% for judging defense is basically worthless. Using ERA alone for the bullpen is hardly any better.

"It's for your own good. Big strong Devo knows whats best for Poppy" -- Mossback

by devo on Apr 15, 2008 11:32 AM PDT reply actions

OK

rfloh & devo, you make some good points.

1) I recognize that using only the rankings from the last 16 AL playoff teams might seem like a small smaple size, but I thought it was adequate to illustrate the correlation to winning.

2) I think the A’s of that period are a great model for any team trying to become better. That’s why we see other GMs buying in at least partially to the A’s philosophies (see Moneyball).

3) I agree that ERA is a crude measure. That’s why I was surprised to find that it had more correlation to winning than any other stat that I reviewed. You can say that ERA is derived from many factors, such as the influence of the catcher, pitching coach, manager, stadium, etc. However, I’m using it as an easy gauge of past performance and predicted future performance.

4) Defense is toughest to value because of the lack of pertinent stats other than fielding percentage. Range is definitely a big factor, but I feel that attempts are possibly as much a function of pitching than as of fielding ability. A more relevant stat with rankings that can be traced back to prior years would be welcome.

5) Of course, you are right in saying that HOW you get to the good OBP and ERA is important. This article isn’t as much about player development as it is about taking the existing teams and looking at the known performances of existing players to see if those teams can turn themselves around. I’ll address that more in the next chapters.

Thanks for the input. I am always striving to make my writing more thorough and well-thought-out.
:)

by Athletic Supporter on Apr 15, 2008 11:36 AM PDT reply actions

ERA

It is much more work, but if you want to assess the quality of a team’s pitchers, you need to break things down further, Ks, BBs allowed, or OBP, SLG, OPS against. Or consider using one of the pitching stats that attempt to isolate pitching from defense: Fielding Independent Pitching from THT / Fangraphs. or team Pitching Runs above Average from Baseball Prospectus, which is easy to obtain for past years. Team FIP unfortunately does not appear to be easy to obtain. Of course you can calculate it on your own.

Sure ERA correlates well with winning. That is similar to saying runs allowed correlates well with winning.

Secondly

However, I’m using it as an easy gauge of past performance and predicted future performance.

The problem is that a pitcher’s one year ERA is not a good predictor of future performance.

Defense is toughest to value because of the lack of pertinent stats other than fielding percentage. Range is definitely a big factor, but I feel that attempts are possibly as much a function of pitching than as of fielding ability. A more relevant stat with rankings that can be traced back to prior years would be welcome

Defensive Efficiency Rating is still fairly crude, it does not account for park and pitcher quality, but it is a better measure of team D than fielding percentage. It basically measures the percentage of balls in play converted into outs.

Another stat you might want to look at is Baseball Prospectus’ team Fielding Runs above Average. It has its issues, but it is easily available for any time period.

If you want to get more sophisticated, look at team zone rating.

As for the value of range, a player with great range can get to balls that a player with poor range cannot. I do not see how that is a function of pitching ability.

ZIPS: Milledge: 466 HR, 485 2B, 2282 hits, 278-379-524

by rfloh on Apr 15, 2008 12:17 PM PDT up reply actions

The biggest problem with this analysis ...

is that you’re comparing fundamentally different statistics.

You could do the reverse and find that team runs scored correlates much more closely to winning than do, say, team strikeouts (pitching).

Do a real analysis and you’ll find that runs scored and runs allowed, individually, correlate very poorly with winning percentage:
For Runs Scored:
y = 0.1042x
R2 = 0.3403
For Runs Allowed:
y = -0.1044x + 162
R2 = 0.4652

Basically, you see that the correlation between preventing runs and scoring runs with winning is exactly the same (preventing a run is worth 2 thousandths of a win more than scoring a run).

Scored or prevented—a run is a run. The question you should be asking is it cheaper to score a run or prevent a run.

"It's for your own good. Big strong Devo knows whats best for Poppy" -- Mossback

by devo on Apr 15, 2008 12:30 PM PDT up reply actions