For the past year or so, one of my favorite toys on all of the internet has been FiveThirtyEight’s Complete History of the NBA, which is a compendium of ELO rankings for every franchise in NBA history across every game they’ve played since the NBA began. ELO is a relatively straightforward concept that FiveThirtyEight ported in from world class chess rankings; at the beginning of the NBA, every team opens up with an average ELO ranking (FiveThirtyEight uses ~1500 for their average baseline), and then their rankings change based on the outcome of games. As time moves on, expansion teams are dropped in at a below average baseline. Wins raise your ELO, losses drop it. As time goes on, the difference in ELO between the two teams dictates the shift in ELO from the result. To keep using the NBA as an example, when the 55-5 Warriors lost to the 12-51 Lakers, the two teams’ ELOs shifted nearly 40 points in opposite directions. Just a few days before, the Warriors beat the Thunder in San Francisco; their ELO didn’t increase at all.
I love ELO as a tool because it’s both less sensitive to the whims of the schedule, and because it carries a memory with it that we as fans often lose over the course of a long season (or even over a really good or really bad week). Idly navigating through the NBA’s ELO history in little pieces here and there over the last year, though, bore out a truth for me about the NBA that’s usually apparent if you only pay a little bit of attention from year-to-year (which is how I watch professional basketball); the teams that win championships tend to also be the best teams. I wondered how the Complete History of Major League Baseball would look; it’s often not true of baseball that the best teams win the World Series. I assumed at that point, though, that the MLB project would be a gargantuan project an order of magnitude larger than the NBA project (more teams, more history, more games, more day-to-day variability due to differences between starting pitchers). Yesterday, FiveThirtyEight released the Complete History of MLB, along with their current ELO and projection systems, and I’ve been enthralled every since.
I have, of course, spent most of my time on the Pirates’ page. Like most baseball teams that have existed for forever, the peak of Pittsburgh Pirate baseball came a long time ago. On September 27th, 1902, the Pirates beat the Reds by a score of 13-6 to run their record to 101-34 and to raise their ELO to 1624. That is as good as its ever gotten for the Pirates. It makes sense that their peak came so early; if you take the long view of MLB history, you can see a flattening of the graph’s extremes starting in the 1960s as baseball expanded, and as the draft and free agency began their rise to prominence. Most franchises that existed before that era had their best teams somewhere between the beginning of baseball and 1960; that’s true of teams that have had sporadic success like the Pirates and A’s as much as its true of teams that have had more consistent success like the Cardinals and Yankees. There are exceptions (the Reds have the Big Red Machine, the Phillies spent over a century as the Phillies before their sustained run of success from 2008-2011).
As much as I’ve had fun picking through the high and low points of Pirate history prior to my own understanding of the team (for example: though some Pirate fans might consider much of the 1970s to be lost opportunities in terms of having only won two championships despite a number of playoff appearances, the Pirates really only spent 1972 on top of the ELO rankings without winning a championship; they didn’t return to the top of the heap until shortly before the playofffs began in 1979), I’ve mostly focused on the years that I suspect will fascinate me for the remainder of my life as a Pirate fan: 1993-2012.
The Pirates ended 1991’s regular season as baseball’s best team, by ELO. Their loss to the Braves in the NLCS resulted in the Twins replacing them atop the heap. They bounced around in 1992, but came close enough to the top of the list before the season ended that a decent start to 1993 put them back atop the rankings. ELO doesn’t care about free agency or salaries or anything of that nature; it simply self-corrects over time; the Pirates lost the top spot in late April of 1993 and haven’t been back since.
By the end of 1993, the Pirates had dropped below the average ELO water mark of 1500. Throughout the remainder of Cam Bonifay’s run as GM, they only flirted with that average line from 1997-1999, but never consistently crossed it. Shortly after Bonifay’s firing in 2001, the Pirates finally bottomed out. Dave Littlefield’s Pirates were even more frustratingly sub-mediocre than Bonifay’s; they were rarely the worst team in the league, but they were only ever even above average in 2004 (remember; ELO is not based solely on record — the Pirates had an awful start in 2006, but the NL Central was consistently sending at least one team to the NLCS or World Series in the early-mid 2000s and in both 2004 and 2005 the Astros and Cardinals played in the NLCS). This was right around the time I started blogging, and smack in the middle of the era that myself and a few other Pirate internet pioneers nicknamed “The Drive for 75.” The Pirates were never good, but they were never bad enough to matter. Following the pattern set by Bonifay, Littlefield was fired right as the Pirates started to bottom out in 2007. Neal Huntington took over, and they spent the beginning of 2008 and the end of 2009 as baseball’s worst ELO club. In 2010, the bottom of the ELO chart is a trace of the Pittsburgh Pirate line. They were terrible. By mid-season 2011, though, they’d picked themselves off of the bottom of the line and started to approach average. In 2012, they spent more time above that line than in the first 19 years of the losing streak. In 2013, they made the playoffs. Last year, their ELO peaked towards the end of the season at 1559; just about where it was in 1991 and 1992.
The Pirates are only one team, of course, but this serves as a reminder that there’s no honor or glory in being blandly sub-mediocre. It wasn’t until Huntington took over and the Pirates gave up on the pretense of 81 wins as a goal worth striving for that they were able to find their way out of the wilderness. Of course, the ELO lines can also teach us a lesson about what happens on the other half of the line that demarcates above-average and below-average, too. In 2014, the Pirates flirted with the 1500 average ELO marker in the same way that they flirted with a .500 record for most of the season. They pulled themselves together and won a Wild Card with a hot finish and a good-but-not-great ELO of 1537. They met a Giants team that had done the exact same thing; flirted with mediocrity all season, but pulled themselves into the Wild Card Game at 1515. The Giants won the game, and went on to win the World Series. Despite their impressive playoff run, they didn’t come close to threatening the Orioles as baseball’s top ELO team.
ELO isn’t necessarily perfect or even predictive, especially this early in the season, but it’s a really interesting lens to view the history of baseball through. I think there are a lot of insights that can be gleaned from it, given the right context, and it’s an easy page to lose a lot of time on if you’re not careful. Enjoy it, though, and bookmark it, because I’m sure it’s something I’ll be coming back to at various points during the season.
Photo by Rick Stewart/Getty Images
