With free agency arriving, it's everyone's favorite time of the year again: those precious few months when we can argue nonstop about whether this player or that player is worth their money.

Disagreement is part of the evaluation game, of course, and this particular free-agent class has a number of players who are likely to inspire spirited conversation. As such, let's take a look at five free agents likely to be the most polarizing. (Note: Players are listed in alphabetical order.)

The disconnect between Andrew Cashner's shiny ERA (3.40/138 ERA+) and his less-shiny peripherals (1.34 strikeout-to-walk ratio/4.61 FIP) is the driving force behind his inclusion.

Cashner's strikeout and swinging strike rates each plummeted, with the latter moving from 18.8 percent to 14.9 percent. He didn't make up for the reduction in swings-and-misses with improved control or upped groundball tendencies, either -- he walked more and generated a lower percentage of groundballs than his career norms.

It's worth noting Cashner did seem to tweak his arsenal, junking his slider for a harder cutter. Batters hit nearly 70 percent grounders on the cutters they put into play, so perhaps Cashner can keep the ball on the infield more often heading forward. For now, there's no reason to think he's anything more than a back-end starter whose ERA is going to cause some fights.

This is Welington Castillo's second consecutive winter on the open market. Last year he was non-tendered by the Arizona Diamondbacks and ended up with the Baltimore Orioles for $6 million. He went on to author the best offensive season of his career, setting new highs in home runs (20) and OPS+ (115). Based solely on his bat, you could argue he was a top-10 backstop in 2017.

The catch is that ... well, Castillo generally isn't good at catching. He has a strong arm, and he's typically above-average at blocking balls in the dirt. But when it comes to framing pitches, Castillo has historically been a well-below-average performer. In fact, Castillo had lost more than 50 runs on framing from 2013-16, per Baseball Prospectus. That's more than 12 runs annually, or more than enough to make him an overall negative behind the dish.

Typically, this would end with a comment on how Castillo will be relegated to the bargain bin again because teams are valuing framing more and more these days. Alas, this story is more complicated than that -- to the extent that Castillo actually graded as an above-average pitch framer in 2017. How legitimate was Castillo's offensive growth? How sustainable is his defensive improvement? Was everything just a fluke? Now you see why Castillo is listed here.

Everyone likes to dig into advanced metrics to figure out if the surface-level numbers are maintainable. In Eric Hosmer's case, doing so just creates more headache.

Hosmer posted new career-highs across the board in 2017, finishing with the best batting average, on-base percentage, and slugging percentage of his career. His 132 OPS+ tied with Anthony Rizzo as the eighth-best among qualifying first basemen. Solid stuff.

Any attempt to suss out how much of Hosmer's improved play is retainable is ill-advised, however. To wit, FanGraphs data has him hitting nearly as many groundballs as usual, meaning he wasn't part of the so-called "flyball revolution." Hosmer pulled the ball less often than normal, too, and -- for the freaks out there who are into murky, subjective measures -- made less hard contact and more soft contact than in recent years. Oh, and by the way -- Hosmer's average exit velocity went from 91.7 mph in 2016 to 89.6 mph in 2017, per Statcast.

In other words, Hosmer's long-awaited breakout looks nothing of the sort under examination. Some team is going to pay big bucks for an up-close look-- and arguments are going to ensue from that point forward.

A Rorschach test for how people evaluate offense. To a certain set of eyes, Eduardo Nunez is a quality contributor who masks his shortcomings with a high average and good speed. To others, he's a limited hitter whose reliance upon batting average makes him a potential liability.

Nunez has experience at numerous positions -- last season alone, he played more than 10 games at third base, shortstop, and left field -- though he's not a highly skilled fielder. He is, however, coming off the best offensive effort of his career.

In a different day and age, Nunez would probably get a sizable free-agent deal (remember Chone Figgins?). In this day and age, he's likely to be scrutinized to the point where he settles for a one- or two-year pact.

Fernando Rodney's 39th and final save in 2017 doubled as the 300th in his career, the 26th-most all-time. One more double-digit save season and he'll get into heaven -- or, at least, into a club of 27 pitchers who have amassed 10 or more such seasons. Pretty neat either way. So, why is a bonafide proven closer like Rodney on this list? Because of his perceived unreliability.

Between Rodney's 4.23 ERA and his inconsistent command, he's close to being what some might call a top-stepper -- a pitcher who forces his manager to be perched and ready to call for a new arm. Yet Rodney's ERA translated to a 114 ERA+, and his strikeout-to-walk ratio was the best it had been since 2014. Besides, his blow-up tendencies might be overstated.

FanGraphs tracks "shutdowns" and "meltdowns" for relievers, metrics based upon how much win probability a pitcher adds or subtracts from his team. Rodney ranked fourth in the league in shutdowns, and recorded fewer meltdowns than well-regarded, younger closers like Alex Colome, Edwin Diaz, and Roberto Osuna. That isn't to suggest Rodney is Kenley Jansen or Craig Kimbrel or anywhere close to that level. But armed with his high-grade changeup and crooked hat, he probably deserves to enter 2018 installed as someone's closer.