The Colorado Rockies are putting their title hopes on the line Sunday, when they host the Milwaukee Brewers in Game 3 of the National League Divisional Series. The Rockies trail the Brewers 2-0 in the best-of-five series, meaning another loss will equal elimination.

With so much riding on Sunday's outcome, the Rockies' decision to start German Marquez over Kyle Freeland is certain to raise questions. It was Freeland who, during the regular season, posted a better ERA than Marquez (2.85 versus 3.77) in more innings (202 versus 196 in 33 starts each). Both will enter Sunday on a full rest, too. So what are the Rockies thinking?

Let's try to figure it out.

Colorado wants three wins, not one

If the Rockies are going to come back and advance to the NL Championship Series, they'll need good starts from Marquez and Freeland -- not just one. Viewed from that perspective, it doesn't really matter who starts Game 3 and Game 4, so long as there is a Game 4.

Of course, this logic doesn't hold up to scrutiny. Yeah, the Rockies need three wins ... but they have to win Game 3 (then Game 4) to have the opportunity for those three wins. That means they need to empty their bag of tricks (within reason, anyway) in Game 3 to maximize their chances of winning. That means starting Freeland.

Verdict: Nope.

It's a rest thing

While both Marquez and Freeland enter Sunday with at least four days of rest, perhaps the Rockies made their decision with health ramifications in mind.

Remember, Freeland started Tuesday's Wild Card Game on short rest. Marquez, meanwhile, started Monday's NL West tiebreaker game and is going on long rest. With both needed, the Rockies may have decided to give Marquez a nod before he went too long between starts -- and to give Freeland an additional day off to make up for cutting his recovery process short last time.

The argument against here is that the Rockies were willing to start Freeland on short rest to begin with, so they're not above prioritizing their needs above his well-being.

Verdict: Nah.

Marquez is a better matchup

What if the answer is that the Rockies believe Marquez is a better matchup against the Brewers than Freeland? That wouldn't bode well for Colorado in Game 4, but it would explain why they'd prefer to start Marquez in Game 3, right?

Figuring out why teams prefer one matchup to another has become more difficult as teams have loaded up on proprietary data. Still, let's give it a shot.

Let's start with the most obvious difference between the two: Marquez is a righty, Freeland is a lefty. Do the Brewers have a large platoon split? Not really. They were seven points of OPS better against right-handers during the regular season. They were also seven points of OPS better against right-handed starters, suggesting this isn't a situation where they feasted on weak middle relievers. The Brewers have a lot of interchangeable parts based on handedness, among other factors, so it makes sense that their play is about even against both lefties and righties.

Marquez and Freeland have similar grounder-to-flyball rates, so we're going to skip that exercise. What about strikeouts? Marquez had a higher strikeout rate by eight percentage points. The Brewers were worse against power pitchers than finesse pitchers, as defined by Baseball-Reference, so that's something. But, uh, the Rockies have already shown their hand so far as that split is concerned. They started Antonio Senzatela in Game 1, and he had one of the lowest strikeout rates on staff. Beyond that, they didn't carry Jon Gray on the roster, and he had the second-highest K rate among Rockies starters -- and a lower walk rate than Senzatela.

How about individual arsenals? Obviously this is harder to nail down -- not every slider is built the same -- but could Marquez's higher velocity and breaking balls make him a better fit than Freeland's reduced heat and changeup? Nah. The Brewers had the sixth-highest wOBA against fastballs clocking in above 95 mph (18th-highest on heaters below 95 mph) and were third in wOBA on breaking balls, as opposed to 10th on changeups. 

Verdict: As the kids say, this ain't it.

Marquez is better

This feels a little too simple -- and perhaps a little too galaxy brain-like -- but this might be it. At minimum, this is the answer that feels closest to the truth. That's because there is a case to be made that Marquez has performed better than Freeland this year.

Yes, really.

While Freeland has the better ERA, Marquez has a better FIP (3.40-3.67) and a better DRA (3.23-3.89) by healthy margins. He's also held the opposition to a lower OPS since the All-Star Break by more than 30 points. None of this is necessarily predictive, and none of it is to suggest Freeland is a bad pitcher -- each is a top-32 pitcher, per Baseball Prospectus's WARP measure -- just that Marquez had him beat in the metrics that are supposed to measure pitcher performance better than ERA, which is prone to being influenced by defense and luck.

In other words, you probably go with Marquez if you're analytically inclined and you're tasked with choosing one to start in a make-or-break game -- just as the Rockies are on Sunday. Whether it works out -- and whether the Rockies made their decision because they believe Marquez is the better pitcher -- is anyone's guess.

Whatever led the Rockies to Marquez had better prove to be a good predictor, however. If it's not, the Rockies will have all winter to wonder what would've been had they went with Freeland.