Now that the Reds have begun to re-build, -tool, -load, etc. I’ve been wondering if there is a specific target, beyond being good, that they should be aiming for. That is, if you had a blank slate, is there a specific way to be good that is better than other ways?

To start, I had to define what better meant, and in the baseball world, I still think that it all comes down to the World Series. Yes, the regular season is probably a better measure of overall talent, but if the Reds won the Series again, and someone told me that it didn’t matter because they weren’t the most talented team overall, I would laugh all the way to the Fountain Square parade.

So, to see if there were better ways to be better, I decided to look at how each of the last 30 World Series participants have stacked up against league average in four categories: Offense, Defense, Starting Pitching, and Relief Pitching. The measures that I used (which of course we can quibble over, but for this exercise let’s not) were wRC+, fangraphs’ defense metric, and xFIP+ for starters and relievers.* You can follow those links for explanations of each.

*I scaled the Def measure so that league average is 100 rather than 0, so that it would be consistent with the other measures. For each of the measures I used, 1 point over 100 means the team was 1 percent better than the league average, and vice versa. Fangraphs uses xFIP-, where lower is better, but I inverted it for consistency.

I think just looking at the full results is interesting, so I’m going to include a long table here. See you on the other side.

World Series Teams

First impressions

Only 3 of the 30 teams were above average in all four areas of the game (’04 Cardinals, ’07 Red Sox, and 2011 Rangers), and while I’m not really differentiating between World Series winners and losers for this review , it’s worth noting that only one of those went on to win. So being good at everything clearly isn’t necessary to get to or win the Series.

Conversely, by combining the four measures I found that only four teams were below average overall (’03 Yankees, ’04 Red Sox, 2010 Rangers, and 2010 Tigers). So, while teams don’t have to be better than average at everything, 87 percent of recent World Series teams were good enough at some phases of the game to compensate for a weakness, and make them better than average overall. This is sort of a weird thought to follow, but the average World Series team was 9 percent above average overall. Make sense?

My belief going into this was that good offense and good starting pitching were going to be the best and most common formula, but only 11 of the 30 teams were above average in both, and about half of those went on to lose. The ranges for both starting pitching and offense went from about 10 percent below average to 17 percent above.

How to be good

The most common ways to be good were defense and relief pitching (24 of 30 teams), while only 18 teams had above average starting pitching, and only 17 teams had above average offense. That last one was especially surprising, given that it’s only 57 percent of the sample. That relief pitching seems so important does not bode well for the current Reds, since they have a terrible bullpen right now, but I think bullpens can be turned around more quickly than other parts of a team.

The weird thing is that while defense was tied for most frequently above average, it also varied the most. The range was from 83 percent below average, to 108 percent above average. There’s a lot for front offices to think about there. Clearly defense is important because World Series teams often are good at it, but a team can still be great and have a terrible defense. Some of that variance is due to the still-evolving state of defensive metrics, but still…

The big takeaway for me (from this admittedly crude analysis) is that even in the data-intensive modern era, there are still many different ways to make a great baseball team. The next time you here a play-by-play guy or TV analyst say “you absolutely cannot have a team that doesn’t do X and think you can win the World Series,” you’ll know that they’re probably wrong. For every World Series team with three aces at the top of their rotation, there are the 2015 Royals. For every team with MVPs up and down the lineup, there are the ’05 Astros.

As a GM for a rebuilding team, I think the takeaway should be to first, make sure your bullpen is at least decent (the worst World Series team was only 4 percent below average in relief pitching), and then to look at where your team excels and where it falls short to decide the best way to get the team to that 10 percent better-than-average-overall level. Maybe, based on trades and free agents, the best course is to shore up the weakness and be a generally-good-at-everything team, or maybe it’s to hit the throttle on your strength and blow other teams away with one super power.

13 Responses

  1. Fish

    I might’ve taken this analysis a bit farther although I like the line of reasoning. I’d say the a team should attempt to be within one standard deviation of the mean in each area. Also, it would be interesting to look at records and the correlation with win totals and each category, overweighting the category with the highest win correlation (by win total I mean regular season and not just playoffs. It would give you a larger pool which, IMO, would give you a better data set).

    • Jeremy Conley

      You’re always welcome to take it further, let us know what you find.

      The avenues you suggest aren’t that interesting to me, but I see where you’re going. There are a lot of tests you could try to run to test the correlations of different statistics with win totals, and many people have already done those. At the end of the day, we know that runs differential is highly correlated with win totals, so you can also just start testing which statistics are the best predictors for runs scored and runs given up if you want.

      The point of this exercise was less about being as precise as possible in predicting a team’s win totals, but to see if (in an admittedly small sample) there was a particular mix of strengths that was more common than others in recent World Series teams. There’s a lot more you could look at in that direction though, and I’m glad the article got you thinking of some things you’d be interested in checking out.

      • Jack

        So… I think the test now is to use your analysis to predict (at the end of the season) which of the playoff teams should make it to the WS.

      • Jeremy Conley

        Not a bad idea. Maybe I’ll do an update when we’re at the half-way point and have some stats to go on.

    • Jeremy Conley

      Some interesting points TCT. I’m not sure I totally buy the first part, about managers using their bullpens differently in the playoffs, I’d have to look into that. I think it’s quite possible that they stick to the generally the same patterns of use, especially if those patterns were successful all year.

      The second part, on the starting rotation, I do think is a good point and makes sense with what the numbers show. I would suspect that for the 12 World Series teams in the sample that had below average starting pitching in the regular season had at least two really good starters. Then, in the playoffs, when 5th and sometimes 4th starters aren’t used, that weakness becomes less of a liability and those teams’ strengths can come through more.

      Conversely, if you have bad defense and bad relievers, there’s really no way to hide that, unless all of your starters can pitch 15+ strikeout complete games. A tall task in the playoffs!

  2. Jeremy Conley

    The defensive side is very interesting. The average team in the sample saved 25% more runs than the league, while the average team in the sample was only about 3% better on offense and in pitching.

    But you can also see that defense is the most variable part of the game for teams in the sample. None of these World Series teams had an offense, starters, or bullpen that was even 15% below league average, let alone 76 like the 2001 Yankees or 66 like the 2004 Red Sox.

    The first part of that makes it seem like the thing that defines these teams is their defense (as you were generally suggesting), while the second part makes it seem like defense is totally inconsequential.

  3. Chuck Schick

    I found the Giants to be interesting as they’ve done well, consistently, without spending Yankee money. They seem to be great at 1 thing, but they aren’t bad at anything.

    The takeaway, to me, is to be as great as possible at something and just don’t suck at anything and you have a shot.

  4. Scott Carter

    Interesting article. I never really thought of it this way, but looking at it I would say you would want to be above average in at least two of these areas and it may not make much difference which and not much below average in the other twi. With that said, and I don’t always understand the statistical analysis data, but doesn’t a great defense make your pitching better. From what I see Billy Hamilton and Zach Cozart and BP has certainly made some of our pitchers better than they might be on other teams.

  5. Streamer88

    Provocative read! I know your time is limited but your data is comparative only to the means of the entire league (setting all at 100]. Might be fascinating to compare these data to those teams in those years who were in the playoffs but did NOT make the world series. Those findings could then answer an entirely new question: what separates playoff caliber from WS caliber teams, I.e. what increases chances for success in the post season? As I said – provocative stuff. Nice work!

  6. RedFuture

    Since there was no commonality as to how to be good to get to the WS, it makes sense to me to lower the goal to be good enough to win the best of 5 game divisional series. We know the Reds have not accomplished that since 1995. Anything can happen to any team able to win any playoff series, as a higher level of confidence ensues. How do you measure confidence? We know that confidence is a roller-coaster ride. A lot rides on how hot the team is entering the playoffs. Maybe your numbers do have more significance when you tally them up for the last 10, 20, or 30 games of the regular season as opposed to the whole season.

  7. cfd3000

    I agree that three of these four measures are appropriate, but that the starting pitching component needs tweaking to properly evaluate WS teams and especially to identify WS winners. Two studs and three scrubs is way more valuable in a 7 game series with off days than four or five solid but not spectacular starters. Think MadBaum a couple years ago, or even Jose Rijo (and some Nasty Boys) in 1990. I’d be curious to see the xFIP+ numbers on the top two starters for each WS team, particularly the winners. I suspect they were mostly ridiculously good. You have to get to the playoffs of course, but after that starters 4 and 5 actually become relievers. Either way, cool to see confirmation that there is more than one way to win.

  8. brunsfam

    The critical take-away seems to be “How do small market teams find a way to compete?” So the question of which metrics are more cost-effective must be considered for us Reds rooters. Relief pitchers and solid gloves must be part of the small-market formula. The elite starting pitchers and power hitters will command big money by comparison. When a steady, competitive defensive-focused team makes a run, then it can be supplemented with a free-agent ace or bopper – or maybe a couple develop in the farm system and create less need for a trade. Pitching & defense win championships – although they don’t necessarily see tickets unless accompanied by winning.