On the pitfalls of player-type analysis

Almost a month ago, Scott wrote about a Buster Olney tweet. The tweet:

Seattle Mariners have indicated winter priorities to others: No. 2-type starting pitcher (Garza?); closer; two frontline power hitters. — Buster Olney (@Buster_ESPN) November 15, 2013

Scott's (eminently reasonable) take:

Buster Olney says the Mariners have big plans this off-season. This is a good thing. Don't damn the result until you've seen the process.

In a subsequent ESPN Insider article, Olney reiterated his point, reporting that the Mariners have "indicated to others" that they will add a #2 starting pitcher, a closer, and two power hitters. Now, I started this article by referencing Scott's for a reason: it's important to remember that this is all just hearsay, and that a team saying they want the moon doesn't mean they'll overpay to get it, and that the Mariners adding talented players is a good and desirable thing. The Mariners just added Robinson Cano. That's fucking awesome. That's not a result that gives me cause to damn the process.

But Olney's tweet is a perfect example of the kind of player-type thinking that I've seen all around the Mariners blogosphere this winter. I don't want to name any names or link any links, because this isn't an attack article, but if you've spent time on Twitter or in comments sections lately you know what I'm talking about. "The Mariners need a bat." "The Mariners have to add a #2 starting pitcher." "The Mariners must sign a right-handed outfielder."

Not a single one of those statements is true. The Mariners don't need a legitimate number three, or a power reliever, or a right-handed slugger. The Mariners need good players. It's not the same thing.

The mistake being made here is a perfectly understandable one, because a faulty conclusion being drawn is based on sound premises. The Mariners are bad, and they need to get better. It's easier to improve by filling holes than by upgrading positions of strength. Therefore, the Mariners should aim to replace the worst players on their roster, those being a starting pitcher, a DH, a couple outfielders, and several relievers. This all makes total sense.

But player-type analysis has its pitfalls. I've composed a short list of them below.

Danger #1: Tunnel Vision

Dividing players up into categories and then looking for a player that fits into one or two of those specific groups can often cause teams to miss out on talented players with unusual skillsets. The Mariners shouldn't disregard (say) Shin-Soo Choo just because he's not a "right-handed power hitter". Choo should be evaluated on his own merits as a player, not on whether he falls into any one particular subgroup.

This tunnel vision can be particularly dangerous when the group being targeted is a traditionally overvalued one (e.g. power hitters, closers). By locking in on a group of overvalued players, teams sometimes confine themselves to the expensive end of the free-agent pool and miss out on more reasonably priced options. Some team is going to throw a lot of money at Nelson Cruz because he's a "slugging outfielder", but the similarly talented Davids Murphy and DeJesus signed for pretty reasonable contracts early in the offseason. Determination to add a player with a particular skill set rather than just any old player with talent is one of the surest ways to wind up overpaying.

Because the Mariners don't need any one particular skill set. Despite all the articles flying around saying that they have to improve their offense, the team could absolutely make the playoffs with the offense that they have right now. Last year's Mariners had a .307 wOBA; the Pirates made the playoffs with a .312 wOBA. Yes, the Mariners are losing Kendrys Morales and Raul Ibanez, but they're also losing Brendan Ryan and Robert Andino, and the entire crew of departing vets combined for a .298 wOBA in 2013 - lower than the team's average. Steamer projects the 2013 Mariners for a .311 wOBA before you replace Nick Franklin at 2B with Robinson Cano. The difference between last year's Pirates position player corps and last year's Mariners position player corps was that the Mariners were the worst defenders in the league and the Pirates were the eighth-best. If the M's decided to spend the rest of the offseason doing nothing but adding glovemen like Gerardo Parra, Denard Span, and Gregor Blanco, they could unequivocally improve their team enough to make the playoffs without adding a "big bat".

Of course, they shouldn't do that, because that would just be a different kind of tunnel vision. But you get the point. The Mariners should be open to adding talent in any way they can. They shouldn't lock in on one particular class of players. To be fair, the Robinson Cano signing was a terrific example of doing this right. And I myself was guilty of not considering Robinson Cano - before the Mariners actually acquired him, at least - because he was a "middle infielder" and thus didn't fit into the mold of player I was looking for the team to add. So, good job Mariners on not getting tunnel vision. Keep it up. Don't sign Nelson Cruz.

Danger #2: Losing The Big Picture

When you define players by their skillset, it's easy to lose track of whether or not they're actually good. Take Michael Morse, for example. Is he a "bat"? For sure. Is he a good player? Absolutely not. See, the important bit of player analysis, the really big picture, isn't how a player creates value. It's how much value the player creates. The rest are just details, and getting too wrapped up in details can make it easy to lose sight of the big picture.

Now, I'll grant that a team with too many similar players can be easily shut down by opponents who counter those particular players' strengths. For example, a lineup with Shin-Soo Choo, John Jaso, and Lucas Duda back-to-back-to-back could hypothetically be demolished in a high-leverage situation by a lefty reliever like Charlie Furbush. But in the Mariners' case, this isn't all that important of a concern. Lineup balance is something you worry about once your team is actually good; even post-Cano the Mariners are probably ten wins away from contention, and by the time they get done adding eight of those ten wins their lineup could have a totally different complexion.

Say the M's add Corey Hart as their designated hitter and set up a platoon at first base. Their lineup against left-handed pitching would then feature Hart, Mike Zunino, Robinson Cano, the RH half of the first base platoon, and Willie Bloomquist and Abraham Almonte off the bench - not unassailable, but certainly not so one-sided as to be completely shut down by a single lefty. And those additions, the ones that take the Mariners from "lefty-heavy" to "balanced", would be fairly minor. It could be pulled off with a designated hitter and a few bench pieces. There's no reason why the team should limit their search for major position player acquisitions to only right-handed hitters. Acquire talent first, worry about specific skill sets later.

Danger #3: Overdefinition Of Roles

Thinking of players in terms of player types isn't just dangerous for GMs... it's risky for managers, too. Blind attachment to the "designated closer" role probably lost the Mariners a few extra-innings games last year as their worse relievers blew leads while Danny Farquhar sat on the bench. Beyond just the bullpen, this kind of thinking can also lead managers to be inflexible with their bench players - anyone remember the "Free John Jaso" campaign? Player-type analysis is the root of a lot of evil in lineup construction; thinking of players as "experienced" in certain roles is how you wind up with Endy Chavez batting leadoff. Instead of thinking of players as fitting into defined boxes, managers should think of players as they are: individuals, each with a unique skillset. Joe Maddon is wonderful at this. Eric Wedge was... less so.

In conclusion: player-type analysis can be a useful abstraction, but it's also a very dangerous way to think. This is a mistake that the Mariners absolutely cannot afford to make. Just because Shin-Soo Choo isn't a righty doesn't mean that the team shouldn't be interested. Just because David Price has the "ace" label doesn't mean that he's as good as Felix Hernandez, or worth trading Taijuan Walker for. Just because Endy Chavez fit well into the classical "leadoff hitter" player type doesn't mean Eric Wedge was right to actually bat him first. When the Mariners get down to the details in player evaluation, they need to forget about player types and focus on players. They don't need a bat. They don't need a #2. They don't need a closer.

They need talent. It's as simple as that.