03 June 2015

What the Orioles' Record Means

The Orioles are 23-26 as of the end of May (and now 23-28) and fans are beginning to get restless. Alex Conway from Camden Chat wrote an article noting that while it is still early in the season, the team is in a particularly beneficial position to sell if necessary. Chris Stoner from Baltimore Sports and Life stated that while the final epitaph of this season hasn’t been written, that so far the Orioles haven’t played well. This begs the question of when teams in general, and the Orioles in particular, should decide that they’re unlikely to make it to the playoffs and therefore prepare to sell at the deadline.

An article in the Wall Street Journal in 2011 claimed much of the drama of the season is pretty much over after 50 games – by June 1. By then, about one-third of the teams are out of it and another half dozen will join them if they don’t get hot quickly. Their research noted that just 9% of teams with a losing record on June 1 wound up with 90 wins from 1996-2010, suggesting that teams below .500 after 50 games should sell.

There are some issues with this. The author wrote this article in 2011 before the creation of the second wild card and grouped all teams under .500 into a single category without considering actual record. Only 4% of teams with a losing record on June 1st wound up winning 90 games from 1996-2010 despite the article claiming it was 9%. Updating this research will help determine which teams should decide to be sellers at the end of May.

For all seasons from 1998-2014, I determined how many wins each team was above .500 by June 1st, their total wins for the season, whether they won 90 games, whether they would have made it to the playoffs or a tiebreaker game if there were two wild cards, and whether they at least tied for the division lead or a wild card spot presuming that there were two wild card spots. While there weren’t two wild cards in 2000, there are two wild cards now and therefore knowing which teams would make it to the playoffs using today's rules is more important than determining teams that actually made it to the playoffs. It would be possible to look at all teams from 1901-2014 and thereby increase the sample size. However, using data from even 1990 to determine playoff odds in 2015 is difficult due to all of the changes in the playoff format because of expansion and the creation of wild cards. At minimum, it would be difficult to normalize data from 1990 to make sense today if not outright impossible.

This link leads to the complete results and below is the data in a more summarized table.

The 2005 Astros are the only club that was 10 or more games below .500 on June 1st and still made it to the playoffs despite winning fewer than 90 games in the studied sample. Only 4 out of 82 teams finished the season with a winning record. Miami, Philadelphia, Milwaukee, and Oakland should expect to be sellers at the deadline and can start planning to do so now.   

Teams that are 1 to 9 games below .500 entering June will probably not make the playoffs. 34 out of 143 (23.8%) teams finished the season with a winning record, 18 (12.6%) would have at least made it to a tiebreaker, and 10 (7%) won over 90 games. Clubs in this category should make a decision to play to sell based on whether they are expecting a significant infusion of talent. Playing in a weak position can also help as 26 of these teams played in a division where the division leader won fewer than 90 games and 5 (19.2%) ended up making it to the playoffs. It’s too early for these teams to actively plan to sell but they need to realize that their chances of making the playoffs have decreased since the beginning of the season. 

Future playoff teams are likely to be at least three games above .500 going into June. Each category of teams that were three games above .500 or more going into June had at least a 50% chance of making it to a tiebreaker game. Of the 20 teams that were exactly three games above .500, 13 or 65% ended up with a winning record, won an average of 85 games, have a 35% chance of winning at least 90 games, made it to the playoffs 35% of the time and would likely have made it to the playoffs 50% of the time if there was a second wild card. These clubs have higher chances of making the playoffs than they did at the beginning of the season (when they had a 1/3 chance of making the playoffs) and should plan to be buyers at the deadline. 

Teams that are 3 to 9 games above .500 at the end of May would at least tie for the second wild card spot 55.5% of the time. Out of a sample of 137, 111 ended the season with a winning record (81%), 63 (46%) won 90 or more games and 47 (34.3%) had at minimum tied for the division lead. Their playoff chances are slightly better than a coin-flip. Teams that are 10 to 14 games above .500 at the end of May are beginning to look like locks. Only 5 out of these 49 teams have losing records, including the 2005 Baltimore Orioles. As the data show, these teams have roughly a 50% chance of winning or tying their division and a 71% chance of at minimum playing in a tiebreaker game for the wild card. 

Teams that are 15 or more games above .500 at the end of May can be reasonably certain that they’ll make it to the playoffs. None of these 23 clubs have ever had a losing record for the entire season and only 5 such clubs, including the 2014 World Series Champion San Francisco Giants, failed to win at least 90 games. Only 2 of the 23 would have failed to make it to the playoffs using today's rules. 

What does this mean for the 2015 Baltimore Orioles? For starters, they should probably be getting slightly nervous about being under .500 this late in the season. However, the 2014 Baltimore Orioles entered June at .500, July at only 4 games above .500, and still ended up winning 96 games total on route to a division title. There may be lower chances of a team around .500 going on a run but it is likely that at least one team in the AL East will do just that. The Orioles are probably hoping that they’ll be that team.


Anonymous said...

Shouldn't we think about games back as an issue as well?

Like how many GB a team is behind the leader on June 1st.

Do Fangraphs and ZIPS do this already?

Matt Perez said...

I had originally thought that also. I looked at divisions where the leaders had poor records (not quite your criteria but reasonably similar) and surprisingly didn't find conclusive results. Maybe a larger sample would make this possible.

I don't believe that ZIPs tries to predict playoff performance. Now that you mention it, I do think Fangraphs 'Season-to-Date mode' does something similar although their methodology isn't really explained.

Anonymous said...

If it does come to selling, with the way Ubaldo is pitching, do you think the Orioles could get a respectable return?

Good article.

Matt p said...

No because he's hit or miss. Can't trust him to perform. At twelve million dollars per year he is risky.