clock menu more-arrow no yes

Filed under:

The 2015 Mistake Index

Is there a correlation between the number of mistakes teams make and their success?

Whoops!
Whoops!
Ed Szczepanski-USA TODAY Sports

Ever since I started seriously writing about baseball around two years ago, there are certain areas I like, such as the business side of the game and anything to do with base running and the very strong correlation between the number of base runners and success. By viewing play-by-play data, I've also attempted to measure the number of mistakes teams make to see if making fewer mistakes correlates well with success.

My conjecture is simple--the fewer mistakes teams make, the more success they'll have. The turnover battle is a staple of football pre-game shows, turnovers in basketball are a major subject as well, and a mistake in baseball is a turnover of another base, and as such something to be reduced as much as possible.

My selection of mistakes can be considered arbitrary, but in general I look for those controllable items that can be reduced with effective coaching and dedicated practice. I must be clear that I'm not suggesting that individual events can be eliminated, but I do believe the total number of mistakes can be reduced.

There are three general types of mistakes, the first of which focuses on pitching. These include hit-by-pitch, wild pitches, passed balls, and balks, as well as walks that eventually score and any pitching event which directly allows a run to score--for example, if a pitcher balks and a run scores right then, not later in the inning. In addition, I'm pretty sure I count walking the opposing pitcher as a mistake, and if I don't, I should.

Hitting mistakes include pickoffs and bad bunts, which in general are bad sacrifice bunt attempts. Attempting to bunt with the bases empty is not considered a mistake, since this often works well--teams typically bat around .350 or so when bunting (they sure aren't this year, but it's still early). Baserunning mistakes are the TOOTBLANs (Thrown Out On The Bases Like A Ninny), and my standard is harsh and unyielding--any time a player is thrown out trying to advance, it's down as a baserunning mistake, no matter if he's out by 20 feet or if it took a Yoenis Cespedes miracle throw to get him. Forceouts are not included, since it's not the runner's fault he was forced out.

Fielding mistakes include errors and unearned runs allowed. This double-counts an error, but I can live with that because it's my index, and also because unearned runs are the true gifts to the opposing team. If I've added up correctly (and I've made a couple changes), that should be eleven different categories. One I would absolutely love to add but can't figure out how to do so through play-by-play data would be throws to the wrong base or missing the cutoff man, but it's better to be happy with what can be measured than cry over what can't be.

This data viz shows team's total mistakes:

<script type='text/javascript' src='https://public.tableau.com/javascripts/api/viz_v1.js'></script><div class='tableauPlaceholder' style='width: 700px; height: 745px;'><noscript><a href='#'><img alt=' ' src='https:&#47;&#47;public.tableau.com&#47;static&#47;images&#47;20&#47;2015MistakeIndex&#47;TeamMistakes&#47;1_rss.png' style='border: none' /></a></noscript><object class='tableauViz' width='700' height='745' style='display:none;'><param name='host_url' value='https%3A%2F%2Fpublic.tableau.com%2F' /> <param name='site_root' value='' /><param name='name' value='2015MistakeIndex&#47;TeamMistakes' /><param name='tabs' value='yes' /><param name='toolbar' value='yes' /><param name='static_image' value='https:&#47;&#47;public.tableau.com&#47;static&#47;images&#47;20&#47;2015MistakeIndex&#47;TeamMistakes&#47;1.png' /> <param name='animate_transition' value='yes' /><param name='display_static_image' value='yes' /><param name='display_spinner' value='yes' /><param name='display_overlay' value='yes' /><param name='display_count' value='yes' /><param name='showVizHome' value='no' /><param name='showTabs' value='y' /><param name='bootstrapWhenNotified' value='true' /></object></div>

Teams want to make fewer mistakes than their opponents, so if they do, this number will be positive.

The third tab, Win Pct Corrrelation, gets to the heart of the issue, because it's one thing to quantify a team's mistakes, but doing so makes sense only when compared to how many mistakes their opponents make. If the Mistake Index is to have any value, teams that make fewer mistakes than their opponents should have more success. Like any other measure, there will be outliers, but in general, it's a correlation that holds up well.

An argument can be made that I've left some things out. The first is walks--there's a pretty decent correlation between walks and team success (this data viz is through only half of 2014 and is among the first I ever did, so view it with a grain of salt), but in general, all teams give up walks, and in the end adding them in doesn't really change the index all that much. For example, in 2014, the Nationals gave up the fewest walks with 352 and the White Sox gave up the most with 557. That's a difference of over 200, but that's at the extremes. In general, if all teams do something, it's difficult to pin down exactly how much one team was affected more than another. In addition, the sheer number of walks would overpower the index and I've have to introduce some type of weighting, and that gets into areas I don't want to enter.

Stolen bases are another area I've included in the past but no longer do. Not all teams take the same approach on the base paths, and there's absolutely no correlation between stolen base success and team success (here's a different tab from the data viz referenced in the last paragraph), the 2014 Royals notwithstanding. In 2014, the Royals and Cardinals had a similar number of caught-stealing (36 and 32, respectively), but the Royals had almost 100 more stolen bases (153 to 57). Without being able to figure out a baseline of what is "right," I'm more comfortable just ignoring them.

Another thing I'd like to do but is beyond my ability is to see if these mistakes changed the Win Probability Added (WPA) dramatically. For example, an error in the 8th inning of a 12-0 game has far less impact than one in a 5-5 game in the bottom of the ninth that leads to an opponent's victory. Generally speaking, it's not that big of a deal, since most of these mistakes would be constants, except for those extreme cases where a mistake happens late in a close game. This isn't as much of an improvement of the index as a refinement, since I strongly suspect the outcomes would be similar.

For this index to have value, it has to measure something that can be controlled. If these mistakes are outside team control, then it makes no sense to measure them since there's nothing that can be done to change the outcome. It doesn't appear teams take this approach, as they increasingly hire coaches to work with individual players on specific competencies like baserunning, fielding, and other specialties.

Individual mistakes will always occur, but what hopefully can be accomplished is that they can be reduced in the aggregate. Practice and better positioning can cut down on errors, paying more attention to the situation when running the base paths can reduce getting thrown out, and not trying to nibble the corners with the bases loaded can reduce runs walked in.

Correlations between mistakes and winning from 2009-2014 can be seen in this Tableau data viz. There are some differences in the measures I use, but for the most part it matches up well with what I've shown for this year.

Anything teams can do to reduce the number of free bases opponents get is a positive step, and teams are trying to reduce these mistakes. The Mistake Index data viz will be updated daily, so bookmark the link and refer back to it often to see how teams change over the year as well as see the general trends in mistakes.

Some of these measures are not easily found and can be illuminating, and in particular, finding opponent mistakes can generally only be done by analyzing play-by-play data. For example, it's easy to find team's runs and runs allowed, but while seeing the number of errors a team makes is fairly easy, finding the number their opponents made isn't as simple. Every team makes mistakes, and the ones that make fewer than their opponents seem to have more success.

All data from Baseball-Reference

Scott Lindholm is a contributor to BP Wrigleyville and lives in Davenport, IA. Follow him on Twitter @ScottLindholm.