Valuing closers has always given me problems. Every year I run their Steamer projections through my formula, they always came out short of what they actually sell for in my various leagues. For example, in the article linked above, Kenley Jansen was the top closer at $14.37, assuming a 70/30 pitching split. But $14.37 equates to roughly pick No. 82 in the NFBC, and Jansen goes far higher than that in most leagues. (His NFBC ADP is 45.)

In the NFBC, closers are pushed up even more than usual because: (1) There's no trading; and (2) Roughly half your entry fee goes toward the overall prizes, at which you have virtually no shot if you tank a category. For that reason you must get saves at your draft, i.e., ignoring the category leads to almost certain ruin, so you pay what it takes even if it's far more than they're worth.

As a result, NFBC owners act as though closer is a separate position from starting pitcher and as if the format required you to start six starters, two relievers and a flex (starter or closer.) When you re-run the numbers on that assumption, the closer values make a little more sense:

But not entirely. Setting aside that these numbers are based on Steamer projections, which sometimes assign saves to the wrong players, Francisco Rodriguez's value of nine cents is based on a projection that gives him 33 saves, albeit with a 4.09 ERA and 1.32 WHIP. Similarly, A.J. Ramos ($1.05) is projected for 28 saves and Tony Watson ($5.48) for 36. Clearly, my adjustment - which builds in the necessity of starting two closers - didn't solve the problem of why lower-tier closers are so overvalued by drafters, not just in the NFBC, but everywhere.

Moreover, the closer dollar values in my piece were based on the NFBC's own budget for them! In other words, in a vacuum where you need nine pitchers of any kind, and use a 70/30 hitting/pitching split, Jansen's statistical contributions are worth only $14.37, but even when you give pitchers a bigger 34/66 share - as the NFBC does - and budget in $293 for closers - as the NFBC does - and Jansen comes in as the fifth most valuable pitcher at $26.46, Rodriguez, Ramos, Watson and other lower-tier closers are still far below where they get drafted. According to Steamer projections, if you're going to make relief pitcher a separate required starting position - as most NFBC owners do in practice - and spend $293, 9.4% of the entire budget and 28 percent of the pitching budget as they do, those lower-tier closers are still borderline worthless.

The problem, it seems, goes back to what I wrote at the top of this article:

For that reason you must get saves at your draft, i.e., ignoring the category leads to almost certain ruin, so you pay what it takes even if it's far more than they're worth. 

Even if saves, which are only one category, aren't worth the negatives relative to other relievers in Ks, ERA and WHIP, not getting them makes it virtually impossible to contend in the overall contest. As such, drafters willingly overpay to secure the category, knowing it's taking precious resources from elsewhere.

What it looks like in practice is a 12 - or 15-man game of chicken, where everyone would like to wait as long as possible without getting caught empty-handed. Moreover, because under the adjusted values, Jansen and Chapman are so much more valuable for their overall contributions, if you're going to jump, you should slightly overpay for one of them rather than vastly overpay for the lesser commodities. And while the top closers are projected for only a few more saves than the mid-tier ones, I'd have to think the certainty of their saves, i.e., their job security, is much higher.

In practice it would not matter whether the saves run came in Round 3 or Round 12 - as long as you got the same value relative to the rest of the league, you're fine. But saves runs rarely are uniform in the way they affect every team. Inevitably, some get a closer a couple rounds before the main run, others a few rounds after. The team that can hold off and grab 2-3 late-closers after the runs have stopped has a big advantage over those who paid retail (during the run.) Of course, in trying to wait, you run the risk of ruin. It might not be picking up nickels in front of a bulldozer, but perhaps $100 bills. At some point if the currency is big enough, it could be worth a gamble or two. But you can see why people don't like to do it in a contest like the NFBC.

In a regular league with trading and without an overall component, waiting on closers makes far more sense. Not only can you trade for saves or even tank them, but you can easily go nine starters for half the year and aggressively five closers for two months if need be (through trade.) Moreover, the distribution in saves in standalone leagues is usually uneven, so there are usually teams whose closers are fairly useless to them and therefore easy to acquire.

The only caveat with relying on trade partners is people are often unreasonable or charge an unfair mark-up when they realize you need a commodity. As for tanking altogether, it's theoretically possible to win a standalone league while doing so but keep in mind in many leagues, people who aren't doing well check out early and stop making moves. That often results in the few teams that try the entire time being strong across the board, making it hard to compete with them if you've punted saves. In standalone leagues where everyone's trying the whole year, like AL Tout Wars, punting saves - or another category - is entirely viable.

The bottom line, more often than not, I want to come away with enough saves to compete in the category. Ideally, I'll take at least one elite closer who has a low risk for losing the job and will help me in categories beyond saves. For my second closer, I typically play a less daunting game of chicken whereby I end getting one for a price that's good relative to what other people are paying. I time it as best I can, knowing I'm taking on some risk, but I adjust as I go with the risk being I overpay in the later rounds for two ugly lower-tier options rather than one of the mid-tier ones I was targeting at a discount. I adjust the amount of risk I'm willing to tolerate by the format - more in standalone leagues and even the NFBC 12-team format, less in the 15-team NFBC format.