In the interest of getting a better handle on player valuation in the NFBC Online Championship, I mined my two leagues from last year to see how many innings and at-bats are generated from a typical roster.

Here are the results:

C AB is catcher at-bats. I wanted to see how many teams get from both catcher slots combined. Some of these teams had six catchers over the course of the year, some had only two. It turns out the average is 785 for all teams and 782 on the teams that finished in the top-2 in hitting in both leagues. The Titans, who won the March 29 league and finished fifth overall out of 1,788 teams, had only 757. That means when you're pricing a catcher projected for 450 at-bats, he's able to fill the entire slot all year, i.e., there are no replacement level stats to add to his numbers, and he's actually getting you a surplus even over teams that are active on the waiver wire.

I also separated innings pitched into starters and relievers. The average IP was not surprisingly higher among the top-2 pitching finishers, but it wasn't a massive margin - just 60 IP above the average, and that presumably includes teams that gave up. Moreover, the extra innings were almost entirely from starters.

The bottom line, you'll want at least 1,400 innings total, 200 of which are from relievers. The 1,200-1,250 IP should be split up among six and a half to seven starting slots, and the 200 relief innings from two and a half to three relief slots. You can buy yourself extra IP by streaming two-start pitchers aggressively while allowing for three active closers most of the time.

Ideally, we'd discover how many innings (and what numbers) to expect from each pitching slot. SP1, for example, might be worth 210 IP, SP2 205, SP3 200, etc. That doesn't mean you'll be able to roster three 200 IP pitchers - last year there were only 14 in the entire league. But you can supplement the players you draft with replacements during their DL stints and two-start streamers during their bad matchups.

That way, we can target a certain amount of IP for each slot and more accurately value pitchers who are projected to sit in, for example, an SP3 slot with 200 IP of 190 Ks and a 3.50 ERA. He might fill the entire slot, while another pitcher with better per-inning numbers and a lesser workload will only fill it partially, and a replacement value adjustment for a number of innings will be necessary. There's also the transaction (FAAB and roster space) cost of streaming for which to account too. Player valuation is imprecise, but the hope is we can get a decent estimate - close enough to understand, for example, the extent to which we need two 200-IP aces at the expense of elite hitting and at what price we're better off letting them go.