[Spraguer]

I'll give it a shot, Bloggy.

There are a million ways to get you projected K, BB and HR numbers. You can use a projection system like PECOTA or some fantasy publication or you can whip up something of your own. You'll notice that K, BB and HR numbers are much less volatile from year to year for each pitcher than is ERA or any other secondary result: that is because they are a measure of skill and not circumstances/fortune. No matter what you are doing, I can't imagine doing it without Excel or Quattro Pro or some other spreadsheet program.

Let's just go with 2006 numbers, which aren't ideal because of sample size issues, but will work here for ease of explanation:

Lilly: 101.7 IP, 52 BB, 95 K, 16 HR.
Koronka: 105 IP, 34 BB, 50 K, 10 HR.

First, let's make a quick correction for park factor with the homerun totals. Texas' HR factor is 1.20, so we're going to divide half of Koronka's homerun total (5) by 1.2 and then add it back to the other half. That will give us 9.2 HR, rather than 10. Rogers Centre has a HR factor of 1.22 ... Lilly's total is going to be (then) 14.6 HR.

So, let's make our ratios using raw K and BB numbers and park-factored HR numbers:
Lilly: 8.4 K/9IP, 1.8 K/BB, 1.3 HR/9IP
Koronka: 4.3 K/9IP, 1.5 K/BB, 0.8 HR/9IP

There are lots of things that you can do with these numbers. By eye-balling it, I think its clear that Lilly has had the better season.

Those are the relevant numbers; what you do with them is up to you.

Anyhow, I weigh the three numbers according to research done by Ron Shandler, who's Baseball Forecaster books and BaseballHQ website are very good. He assigns weights of 6, 21 and 30 to K/9, K/BB and HR/9, respectively. This puts them on a roughly even scale, allowing you to add the first two and subtract the second one to give you a three true outcomes score, or BPV and Shandler calls it (his formula also makes a very small adjustment for batting average, which I've omitted here for ease of explanation; its impact is very minimal anyhow.)

Here we go - let's score Lilly.

8.4 K/9 * 6 = 50.4
1.8 K/BB * 21 = 37.8
1.3 HR/9 * 30 = 39

So we've got a dominance score (50.4), a command score (37.8) and a HR score (39). As any Jays fan can attest (I am one), these scores are not surprising.

Lilly's total score = 50.4 + 37.8 - 39 = 49.2

Now, for Koronka: 25.8 dominance score, 31.5 command score, 24 HR score.

Koronka's total score = 25.8 + 31.5 - 24 = 33.3

This shouldn't be shocking at all, that Lilly is better than Koronka.

A 49.2 score is equivalent to 4.7 runs allowed per nine innings.
A 33.3 score is equivalent to 5.3 runs allowed per nine innings.

Those equivalancies are based on two seasons worth of data from MLB pitchers (04 and 05). Making a chart of true outcomes scores to runs allowed isn't that hard; you do it once and your done.

So there you go, Bloggy. 4.7 and 5.3.

Since I've come this far, I will add the rest of the equation:

I've got the Jays bullpen at 4.12 R/9 and the Texas bullpen at 4.69 R/9. After weighing those 2/3 starter and 1/3 bullpen, you get Toronto allowing 4.51 R/9 in a Lilly start and Texas allowing 5.10 in a Koronka start.

I have Toronto's defence as neutral and Texas' defence at 1/10 of a run/game. So, the final numbers are

Toronto (Lilly) 4.51 Runs Allowed
Texas (Koronka) 5.00 Runs Allowed