The Rockies are Historically Road-Averse by Ben Clemens August 25, 2021 On Monday night, the Rockies lost to the woeful Cubs, 6–4. It brought their road record to 14–46, good for a .233 winning percentage. That’s the worst mark in baseball, but the Rockies aren’t the worst team in baseball — merely the worst road team. At Coors Field, they’ve gone a spectacular 43–22, the third-best home record in the game. It’s hard to imagine that this huge discrepancy comes down to a roll of the dice. Are the Rockies this bad on the road? Probably not. Are they this good at home? Also probably not. But this gap calls out for an investigation, so I set out to answer: what in the heck is up with their home field advantage — if that’s even what’s going on here? Obviously, I’m not the first person to try to answer this; earlier this year, Neil Paine tackled the subject when Colorado was a woeful 6–32 on the road. The Coors hangover effect is real, and it gives us a good reason to think that the source of the Rockies’ problems might be the road side of things rather than the home side of things. I won’t try to solve the issue of what ails the Rockies today, but still, we can gawk at their incompetence and speculate about what it means for their true talent, which sounds fun enough to me. The Rockies’ road woes are nothing new. Since the franchise’s first season in 1993, they’ve run up a .396 winning percentage on the road and a .545 mark at home. That 150-point gap is the biggest in the majors, and it’s not particularly close. Here are all the teams taken together: Home/Road Splits, 1993-2021 Team Home Away Differential Rockies .545 .395 .150 Pirates .501 .400 .101 Rangers .550 .450 .100 Marlins .506 .417 .089 Rays .524 .438 .086 Athletics .560 .475 .085 Nationals .533 .448 .085 Cardinals .582 .499 .083 Yankees .627 .544 .083 Astros .556 .474 .082 Padres .513 .434 .079 Tigers .498 .419 .079 Blue Jays .534 .456 .078 Giants .560 .483 .078 White Sox .540 .465 .075 Phillies .533 .462 .071 Dodgers .582 .512 .070 Diamondbacks .523 .453 .069 Cleveland .565 .497 .068 Mariners .530 .464 .066 Cubs .527 .461 .066 Angels .547 .482 .065 Brewers .514 .449 .065 Twins .516 .453 .063 Orioles .494 .433 .061 Reds .513 .457 .057 Mets .523 .467 .056 Red Sox .574 .519 .055 Braves .585 .531 .055 Royals .472 .423 .049 The Pirates aren’t that close to the Rockies, and they’re the closest. In fact, Colorado has only had a differential less than league average in seven of its 29 seasons. That brings us to an obvious question: is it home strength or road futility? There’s plenty of research suggesting some effect, but finding the exact magnitude has always proved problematic. In fact, some studies don’t find a hangover effect at all. If you ask me, it’s almost certainly a road disadvantage. The Rockies have the worst road record in baseball since their inception. Could it be that they’re actually that bad? I guess so, but it seems far more likely that Colorado is a middling team (as their home record would indicate) with a huge penalty when it descends from altitude. In any case, that’s not the question I’ll try to answer today. After a few hours of shuffling numbers around, I ran into the same problem that all these studies run into: it’s difficult to separate an advantage at home from a disadvantage on the road. Instead, we’ll just take the split as a given and focus on this year in particular. Are the Rockies actually a .661 team at home? That’s not a good or accurate read of their splits. But let’s say you’re curious, and perhaps credulous. If their true home/road split is 150 points of winning percentage, what are the odds that they’d put up this season’s 428-point gap? Rather than attempt to use mathematical identities to solve that, I took the easy way out by running a Monte Carlo simulation of Colorado’s season. I took the Rockies’ overall winning percentage (.456), then applied the life-to-date home/road splits to it. From there, I simulated the season, a lot — 100,000 times, to be precise. Then I asked my simulation how many times they finished with a home/road split at least this extreme. Friends: it never happened. Never ever ever, more or less. In 100,000 simulations, it happened 65 times. So I ran it a million times — and it happened 690 times. I ran it a million more times — 706 seasons at least this extreme. The re-running was mostly for dramatic effect; something weird is going on in Colorado. Something weird was going on in Colorado last year, too. The Rockies were meaningfully better on the road in 2020; they were a .400 team at home and .467 away from it. It was a shorter season, but even so, when I ran that a million times, it only happened about 3.4% of the time. Clearly, that form didn’t do much to predict this year’s form. That brings up an old nemesis of mine: the Bonferroni correction. Basically, there are a lot of chances for teams to have weird seasons. Saying that the Rockies were extremely unlikely to have such a crazy home/road is true, but incomplete. Every team was unlikely to have such a crazy home/road split, and most teams didn’t! So rather than speculate about the likelihood of this happening — it happened already, the ex-post likelihood is 100% — I thought I’d ask a different question. Teams have had remarkable home/road splits before; how many of those have persisted to the next year? If you were hoping that a huge home/road split in one year was predictive of a similar one in the future, you’re in for a disappointment. I went back to 1993 (I had the Rockies data handy, after all) and looked at every team’s home/road splits for each consecutive pair of years. In other words, I looked at how each team performed in 1993 and compared those to ’94, then at how each team performed in ’94 and compared it to ’95, on and on up through 2019. The 60 teams with the best home win percentage differential were 218 points of winning percentage better at home. The next year, they were 89 points better at home. The next 60 teams were 160 points better at home than on the road. The next year, they were 83 points better at home. After taking into account that the overall mean level is 80 points, the regression to the mean was near-absolute. How about on the downside? The 60 worst teams were 72 points of winning percentage worse at home than on the road. The next year, they were 74 points better at home. The next slice up? They were 20 points worse at home in year one, 64 points better in year two. On average, your best bet for predicting a team’s next-year home/road record was to use 8% this year’s record and 92% league average. What does that mean for this year’s Rockies? We know that their long-term average home/road split is roughly 150 points of winning percentage. Let’s take that as a naive “fair value” for a Coors team’s home/road numbers. Using 92% of that and 8% of their record this year (we’re approximating here, so these are loose numbers that you should absolutely not take as perfect), we’d expect them to post a 172-point home/road split next year. That’s a boring conclusion, I know. The Rockies of next year will likely just be another Colorado team: bad on the road and decent at home. I truly wish I had something better for you. Honestly, this effect might be overselling it. Those numbers I quoted included the Rockies, and we know that they have a persistently high home/road split. The actual mean reversion number might be even higher. There’s still one interesting outstanding question. Well, there are actually a ton of outstanding questions, but there’s a particular one I have in mind. When we model our playoff odds, we treat homefield advantage as a static 4% edge; if two evenly matched teams play, the home team will win 54% of the time. That’s not the case with the Rockies, though. We know that they exhibit larger splits. It doesn’t matter for their playoff odds; those are at 0% either way. But for other teams in the race, where they play the Rockies absolutely matters. Who are the biggest beneficiaries? The Dodgers, indirectly. They have six games left with the Rockies, with three at each site. That has essentially no net effect. The Giants, on the other hand, have six games in Denver remaining. We have the Rockies as a .471 team at home. In practice, though, they’re likely something closer to a .506 team at home. That’s a sizable gap; over six games, that works out to an extra Rockies win something like 20% of the time. In a race we project to be incredibly close, that’s a big difference. On the other coast, the Phillies get a similar boost. They have three games remaining with the Rockies, all in Philadelphia, which is worth a small boost. The Braves, who lead the NL East, still have a series in Denver and a series in Atlanta, for little net effect. We don’t think that race will be as tight as the NL West, but if a game matters, the Rockies’ two-faced nature might help swing the race. In conclusion, I’m sorry. I really hoped to find some reason why the Rockies were truly and sustainably a disaster on the road. I wanted to find something convincing, or at least something plausible. I didn’t find anything like that. I hope you’ll settle for a meandering look at how bad Colorado has been on the road over the years, with a sprinkling of the Dodgers somehow catching a break mixed in for spice.